Novo enables small business owners to get started fast, with free business checking and simple solutions for getting paid, paying expenses, and managing finances

Novo is the powerfully simple, secure, and award-winning banking platform built for small businesses and their unique needs. Free business checking and smart, flexible, modern tools to help any small business get an edge and grow. Powered by Middlesex Federal Savings. Novo is a fintech, not a bank. Banking services provided by Middlesex Federal Savings, F.A.; Member FDIC.

Novo

novo

A MOTH—GENUS NOVO

Img source: novoresume.com Even though you may get contacted by different recruiters via LinkedIn, chances are they will still ask for your resume. Finding you on LinkedIn was only the first part of the recruiting process. That is why it is important to also have a resume prepared that you can use. Novorésumé is a really great alternative when it comes to preparing resumes. It is a really good tool that gives you the opportunity to prepare your resume in a matter of minutes, in an easy-to-use layout that is also very user-friendly. It comes with customizable templates that you can use, interchangeable fonts, themes, and layouts. It also comes with a built-in real-time content optimizer that analyses the content, experience, and geolocation to provide immediate suggestions for the best possible content. You can add your image, contact details, short summary about yourself, education, expertise, projects and skills. Here are some of its core features, as by its CEO in ProductHunt: described - All the sections needed is already provided by our tool, so you have more time to focus on creating a quality content - Layout optimizer checks that content is readable and aligned - Live feedback will provide with tips and examples - Automated synchronization - Multiple languages — all sections easily customisable to the language preferred - Templates are approved by recruiters and employers Novorésumé is free, but has some premium features that you need to pay for if you want to use them on your resume.

Prepare your resume quickly and easily in a matter of minutes with Novorésumé

Rough Start for Novogratz: Galaxy Digital Holdings Reports Net Loss of $272.7 Million in 2018

Author:
(1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es);
(2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es);
(3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Table of Links Abstract and Introduction Related work The proposed pipeline Evaluation Conclusion & Acknowledgements References & Appendix 3. The proposed pipeline The purpose of the proposed pipeline is, given the text reviews of an Amazon product, to classify them as normal if they refer to it, and as anomalous if they describe different products or if they are so generic that they do not provide useful information to consumers. These classifications will be accompanied by a normality score and explanations that justify them. Figure 1 shows the modules that are part of the proposed pipeline to solve this task. In the first phase, we encode the text reviews of the target product using the pretrained MPNet transformer [25]. In the second phase, a DAEF [26] autoencoder is trained using these embeddings. Using the reconstruction errors issued by the network and a predefined threshold error, we can classify reviews as normal (error greater than the threshold) or anomalous (error less than the threshold). For the last phase, we propose a method based on the most frequent normal terms to generate an explanation associated with each classification. Therefore, the proposed pipeline is composed of three main modules: (1) Text encoding; (2) Anomaly detection; (3) Explainability. The following sections describe each of them in detail. 3.1. Text encoding Although there are models capable of dealing directly with images or text, machine learning models are usually designed to be trained using numerical vectors that represent the sample data (tabular data). This input format of the data is the usual one for classic anomaly detection methods. In the NLP area, there are multiple techniques to represent texts using vectors of real numbers, which are known as embeddings. These techniques allow the generation of vector spaces that try to represent the relationships and semantic similarities of the language, so that, for example, two synonymous words will be found at a shorter distance in the vector space than two unrelated words. Embeddings can be calculated independently for each word of the language (word embeddings), which led to models such as word2vec [27] or Global Vectors (GloVe) [28]. The representation of a sentence (sentence embedding) or a document (document embedding) will therefore be the sum of all the individual representations of the terms that make it up. To obtain representations of fixed length, it is usual to perform operations such as the mean. In certain cases, these operations between embeddings can worsen or even invalidate the final embedding, so specific models have been developed capable of understanding and representing a text as a whole, instead of just encoding it word by word. Among these models are those based on transformers [29], such as BERT [30], XLNet [31], GPT-3 [32] or GPT-4 [33], which have been trained over large-scale datasets and can solve different tasks, including sentence embedding. Since it inherits the advantages of the BERT and XLNet models while overcoming their limitations, we will use the pre-trained MPNet [25] model to calculate the embeddings of the reviews. MPNet combines Masked Language Modeling (MLM) and Permuted Language Modeling (PLM) to predict token dependencies, using auxiliary position information as input to enable the model to view a complete sentence and reduce position differences. Models like GPT-3 or GPT-4 are more advanced, but in addition to not being open source, they demand much higher computational resources, unaffordable for a significant part of the scientific community, as is our particular case. MPNet maps sentences and paragraphs to a 768 dimensional dense vector space e ∈ R 768×1 , providing fast and quality encodings. Computing time is a critical aspect in the area in which this work is framed, since in e-commerce platforms (and online reviews in general) we can find a huge number of products and reviews to deal with. Therefore, the proposed pipeline uses MPNet to calculate the embeddings of the reviews, which allows them to be represented as numeric vectors of fixed size. This model and others are available in the Hugging Face repository [34]. 3.2. Anomaly detection Anomaly detection is a field with a large number of algorithms that solve the problem of distinguishing between normal and anomalous instances in a wide variety of ways [4, 35]. Depending on the assumptions made and the processes they employ, we can distinguish between five main types of methods: probabilistic, distance-based, information theory-based, boundarybased, and reconstruction-based methods. Among the latter are autoencoder networks (AE) [36], one of the most widely used models. AE are a type of self-associative neural network whose output layer seeks to reproduce the data presented to the input layer after having gone through a dimensional compression phase. In this way, they manage to obtain a representation of the input data in a space with a dimension smaller than the original, learning a compact representation of the data, retaining the important information, and compressing the redundant one. One of the requirements of our proposed pipeline is to generate a normality score for each review. This score will allow us, among other things, to order the different reviews by their level of normality. When AE networks are used in anomaly detection scenarios, the classification is usually carried out based on the reconstruction error that they emit to reproduce in its output the embeddings of the reviews it receives as inputs, which represents the level of normality of the evaluated instance. This reconstruction error can be used as the normality score we are looking for. Due to the speed of its training, we have decided to use DAEF (Deep Autoencoder for Federated learning) [26] as our anomaly detection model. Unlike traditional neural networks, DAEF trains a deep AE network in a non-iterative way, which drastically reduces its training time without losing the performance of traditional (iterative) AEs. The proposed pipeline uses DAEF to calculate the embeddings of the reviews, being able to issue a normality score associated with each of the review classifications. 3.3. Explainability As black-box ML models are mainly being employed to make important predictions in critical contexts, the demand for transparency is increasing. The danger is in creating and using decisions that are not justifiable, legitimate, or that do not allow obtaining detailed explanations of their behaviour. Explanations supporting the output of a model are crucial, moreover in fields such as medicine, autonomous vehicles, security, or finance. When ML models do not meet any of the criteria imposed to declare them explainable, a separate method must be applied to the model to explain its decisions. This is the purpose of post-hoc explainability techniques [37], which aim at communicating understandable information about how an already developed model produces its predictions for any given input. Within these techniques, the so-called model-agnostic techniques are those designed to be plugged into any model with the intent of extracting some information from its prediction procedure. To generate the explanations we have implemented an approach based on a statistical analysis of the dataset. Based on the definitions of normal and anomalous review presented in Section 2, Our hypothesis assumes that normal reviews will always refer directly or indirectly to the target product so that there will be a list of terms used very frequently among normal texts. The appearance of one or more of these “normal” terms in a review would justify its classification by the system as normal. In the same way, anomalous reviews may be justified with the non-presence of said terms. The original text reviews classified as normal using the AD model will be processed (tokenization, lemmatization and stopword removal) and analyzed to obtain the list of the top-n most frequent terms. This list of terms will be compared with the lists of frequent terms of other products in order to remove the terms they have in common, so that the final list for a product does not contain generic terms. This final list of terms will be used to generate the explanations. New reviews classified as normal will be searched for the presence of any of these terms. To make this search more flexible, it will also be analyzed if there are terms in the review semantically close to any of the words on the list. To quantify the closeness of two words, the cosine similarity emitted by the MPNet transformer model itself will be used as a measure of distance. Reviews classified as anomalous will be explained by the non-occurrence of terms from the list. Figure 2 represents an overall view of the pipeline considering the product “chocolate bars” as the normal class. In summary, in this work, we propose the use of this statistical analysis of the terms to generate the explanations associated with the classification of the reviews. This paper is available on arxiv under CC 4.0 license. Author: (1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es); (2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es); (3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Author: Author: (1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es); (2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es); (3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Table of Links Abstract and Introduction Abstract and Introduction Related work Related work The proposed pipeline The proposed pipeline Evaluation Evaluation Conclusion & Acknowledgements Conclusion & Acknowledgements References & Appendix References & Appendix 3. The proposed pipeline The purpose of the proposed pipeline is, given the text reviews of an Amazon product, to classify them as normal if they refer to it, and as anomalous if they describe different products or if they are so generic that they do not provide useful information to consumers. These classifications will be accompanied by a normality score and explanations that justify them. Figure 1 shows the modules that are part of the proposed pipeline to solve this task. In the first phase, we encode the text reviews of the target product using the pretrained MPNet transformer [25]. In the second phase, a DAEF [26] autoencoder is trained using these embeddings. Using the reconstruction errors issued by the network and a predefined threshold error, we can classify reviews as normal (error greater than the threshold) or anomalous (error less than the threshold). For the last phase, we propose a method based on the most frequent normal terms to generate an explanation associated with each classification. Therefore, the proposed pipeline is composed of three main modules: (1) Text encoding; (2) Anomaly detection; (3) Explainability. The following sections describe each of them in detail. 3.1. Text encoding 3.1. Text encoding Although there are models capable of dealing directly with images or text, machine learning models are usually designed to be trained using numerical vectors that represent the sample data (tabular data). This input format of the data is the usual one for classic anomaly detection methods. In the NLP area, there are multiple techniques to represent texts using vectors of real numbers, which are known as embeddings. These techniques allow the generation of vector spaces that try to represent the relationships and semantic similarities of the language, so that, for example, two synonymous words will be found at a shorter distance in the vector space than two unrelated words. Embeddings can be calculated independently for each word of the language (word embeddings), which led to models such as word2vec [27] or Global Vectors (GloVe) [28]. The representation of a sentence (sentence embedding) or a document (document embedding) will therefore be the sum of all the individual representations of the terms that make it up. To obtain representations of fixed length, it is usual to perform operations such as the mean. In certain cases, these operations between embeddings can worsen or even invalidate the final embedding, so specific models have been developed capable of understanding and representing a text as a whole, instead of just encoding it word by word. Among these models are those based on transformers [29], such as BERT [30], XLNet [31], GPT-3 [32] or GPT-4 [33], which have been trained over large-scale datasets and can solve different tasks, including sentence embedding. Since it inherits the advantages of the BERT and XLNet models while overcoming their limitations, we will use the pre-trained MPNet [25] model to calculate the embeddings of the reviews. MPNet combines Masked Language Modeling (MLM) and Permuted Language Modeling (PLM) to predict token dependencies, using auxiliary position information as input to enable the model to view a complete sentence and reduce position differences. Models like GPT-3 or GPT-4 are more advanced, but in addition to not being open source, they demand much higher computational resources, unaffordable for a significant part of the scientific community, as is our particular case. MPNet maps sentences and paragraphs to a 768 dimensional dense vector space e ∈ R 768×1 , providing fast and quality encodings. Computing time is a critical aspect in the area in which this work is framed, since in e-commerce platforms (and online reviews in general) we can find a huge number of products and reviews to deal with. Therefore, the proposed pipeline uses MPNet to calculate the embeddings of the reviews, which allows them to be represented as numeric vectors of fixed size. This model and others are available in the Hugging Face repository [34]. 3.2. Anomaly detection 3.2. Anomaly detection Anomaly detection is a field with a large number of algorithms that solve the problem of distinguishing between normal and anomalous instances in a wide variety of ways [4, 35]. Depending on the assumptions made and the processes they employ, we can distinguish between five main types of methods: probabilistic, distance-based, information theory-based, boundarybased, and reconstruction-based methods. Among the latter are autoencoder networks (AE) [36], one of the most widely used models. AE are a type of self-associative neural network whose output layer seeks to reproduce the data presented to the input layer after having gone through a dimensional compression phase. In this way, they manage to obtain a representation of the input data in a space with a dimension smaller than the original, learning a compact representation of the data, retaining the important information, and compressing the redundant one. One of the requirements of our proposed pipeline is to generate a normality score for each review. This score will allow us, among other things, to order the different reviews by their level of normality. When AE networks are used in anomaly detection scenarios, the classification is usually carried out based on the reconstruction error that they emit to reproduce in its output the embeddings of the reviews it receives as inputs, which represents the level of normality of the evaluated instance. This reconstruction error can be used as the normality score we are looking for. Due to the speed of its training, we have decided to use DAEF (Deep Autoencoder for Federated learning) [26] as our anomaly detection model. Unlike traditional neural networks, DAEF trains a deep AE network in a non-iterative way, which drastically reduces its training time without losing the performance of traditional (iterative) AEs. The proposed pipeline uses DAEF to calculate the embeddings of the reviews, being able to issue a normality score associated with each of the review classifications. 3.3. Explainability 3.3. Explainability As black-box ML models are mainly being employed to make important predictions in critical contexts, the demand for transparency is increasing. The danger is in creating and using decisions that are not justifiable, legitimate, or that do not allow obtaining detailed explanations of their behaviour. Explanations supporting the output of a model are crucial, moreover in fields such as medicine, autonomous vehicles, security, or finance. When ML models do not meet any of the criteria imposed to declare them explainable, a separate method must be applied to the model to explain its decisions. This is the purpose of post-hoc explainability techniques [37], which aim at communicating understandable information about how an already developed model produces its predictions for any given input. Within these techniques, the so-called model-agnostic techniques are those designed to be plugged into any model with the intent of extracting some information from its prediction procedure. To generate the explanations we have implemented an approach based on a statistical analysis of the dataset. Based on the definitions of normal and anomalous review presented in Section 2, Our hypothesis assumes that normal reviews will always refer directly or indirectly to the target product so that there will be a list of terms used very frequently among normal texts. The appearance of one or more of these “normal” terms in a review would justify its classification by the system as normal. In the same way, anomalous reviews may be justified with the non-presence of said terms. The original text reviews classified as normal using the AD model will be processed (tokenization, lemmatization and stopword removal) and analyzed to obtain the list of the top-n most frequent terms. This list of terms will be compared with the lists of frequent terms of other products in order to remove the terms they have in common, so that the final list for a product does not contain generic terms. This final list of terms will be used to generate the explanations. New reviews classified as normal will be searched for the presence of any of these terms. To make this search more flexible, it will also be analyzed if there are terms in the review semantically close to any of the words on the list. To quantify the closeness of two words, the cosine similarity emitted by the MPNet transformer model itself will be used as a measure of distance. Reviews classified as anomalous will be explained by the non-occurrence of terms from the list. Figure 2 represents an overall view of the pipeline considering the product “chocolate bars” as the normal class. In summary, in this work, we propose the use of this statistical analysis of the terms to generate the explanations associated with the classification of the reviews. This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Explainable AI in Action: Generating Insights from Review Anomalies

Effective Anomaly Detection Pipeline for Amazon Reviews: References & Appendix

Author:
(1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es);
(2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es);
(3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Table of Links Abstract and Introduction Related work The proposed pipeline Evaluation Conclusion & Acknowledgements References & Appendix 5. Conclusion In this work, we have proposed a pipeline for detecting anomalous reviews associated with Amazon products, which can be directly extrapolated to other online review platforms or scenarios with similar characteristics. The representation of the reviews using MPNet embeddings has enabled the training of classical anomaly detection algorithms that have achieved a very good performance. These have been evaluated using reviews from different products and categories, and the score they emit allows us to sort the reviews based on their normality. A technique based on the occurrence of frequent terms has been proposed to generate explanations associated with the classifications of the reviews. This technique has been compared with SHAP, one of the reference post-hoc techniques in the field of explainability, and with GPT-3, due to its high power and versatility. To evaluate this aspect of the pipeline, we conducted a two-part survey in which 241 members of the university community participated. From the first part of the explainability test we can conclude that, in general terms, the effect of the explanations has not been beneficial for the users. In any case, these tests allow us to reflect on the difficulty of using explainability and evaluation techniques in borderline scenarios where subjectivity plays an important role, such as the one presented in this article or in other fields of NLP, as well as in areas such as image or audio generation. Regarding the second part of the explainability test, we have been able to conclude that respondents preferred explanations that presented a more natural and familiar appearance over more condensed and concise explanations such as those provided by SHAP, regardless of the explanation effect they provide. Explanations based on term frequency analysis have been preferred by respondents along with GPT-3, however, our approach presents a significantly lower computational costs and both its use and the explanations produced are simpler for the users. As future work, it would be interesting to evaluate GPT-3 or other large models carrying out the complete process followed by the pipeline proposed in this work, instead of being tested only in the explainability module. We have not carried out this test due to the high computational cost that would be involved in processing the thousands of reviews to be evaluated using GPT3. Another interesting possible line of future work would be to broaden the scope of the survey, both in terms of the number of products involved and the number of reviews, in order to clarify the conclusions reached at the forward simulation stage. Lastly, it would be very useful to try presenting the explanations issued by SHAP in a more familiar or natural format for the end user, so that we can see if their level of preference is increased for the general public. Acknowledgements This work was supported in part by grant Machine Learning on the Edge - Ayudas Fundación BBVA a Equipos de Investigación Científica 2019 ; the Spanish National Plan for Scientific and Technical Research and Innovation (PID2019-109238GB-C22 and TED2021-130599A-I00); the Xunta de Galicia (ED431C 2022/44) and ERDF funds. CITIC, as a Research Center of the University System of Galicia, is funded by Consellería de Educación, Universidade e Formación Profesional of the Xunta de Galicia, Spain through the European Regional Development Fund (ERDF) and the Secretaría Xeral de Universidades (Ref. ED431G 2019/01). This paper is available on arxiv under CC 4.0 license. Author: (1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es); (2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es); (3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Author: Author: (1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es); (2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es); (3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Table of Links Abstract and Introduction Abstract and Introduction Related work Related work The proposed pipeline The proposed pipeline Evaluation Evaluation Conclusion & Acknowledgements Conclusion & Acknowledgements References & Appendix References & Appendix 5. Conclusion In this work, we have proposed a pipeline for detecting anomalous reviews associated with Amazon products, which can be directly extrapolated to other online review platforms or scenarios with similar characteristics. The representation of the reviews using MPNet embeddings has enabled the training of classical anomaly detection algorithms that have achieved a very good performance. These have been evaluated using reviews from different products and categories, and the score they emit allows us to sort the reviews based on their normality. A technique based on the occurrence of frequent terms has been proposed to generate explanations associated with the classifications of the reviews. This technique has been compared with SHAP, one of the reference post-hoc techniques in the field of explainability, and with GPT-3, due to its high power and versatility. To evaluate this aspect of the pipeline, we conducted a two-part survey in which 241 members of the university community participated. From the first part of the explainability test we can conclude that, in general terms, the effect of the explanations has not been beneficial for the users. In any case, these tests allow us to reflect on the difficulty of using explainability and evaluation techniques in borderline scenarios where subjectivity plays an important role, such as the one presented in this article or in other fields of NLP, as well as in areas such as image or audio generation. Regarding the second part of the explainability test, we have been able to conclude that respondents preferred explanations that presented a more natural and familiar appearance over more condensed and concise explanations such as those provided by SHAP, regardless of the explanation effect they provide. Explanations based on term frequency analysis have been preferred by respondents along with GPT-3, however, our approach presents a significantly lower computational costs and both its use and the explanations produced are simpler for the users. As future work, it would be interesting to evaluate GPT-3 or other large models carrying out the complete process followed by the pipeline proposed in this work, instead of being tested only in the explainability module. We have not carried out this test due to the high computational cost that would be involved in processing the thousands of reviews to be evaluated using GPT3. Another interesting possible line of future work would be to broaden the scope of the survey, both in terms of the number of products involved and the number of reviews, in order to clarify the conclusions reached at the forward simulation stage. Lastly, it would be very useful to try presenting the explanations issued by SHAP in a more familiar or natural format for the end user, so that we can see if their level of preference is increased for the general public. Acknowledgements This work was supported in part by grant Machine Learning on the Edge - Ayudas Fundación BBVA a Equipos de Investigación Científica 2019 ; the Spanish National Plan for Scientific and Technical Research and Innovation (PID2019-109238GB-C22 and TED2021-130599A-I00); the Xunta de Galicia (ED431C 2022/44) and ERDF funds. CITIC, as a Research Center of the University System of Galicia, is funded by Consellería de Educación, Universidade e Formación Profesional of the Xunta de Galicia, Spain through the European Regional Development Fund (ERDF) and the Secretaría Xeral de Universidades (Ref. ED431G 2019/01). This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Effective Anomaly Detection Pipeline for Amazon Reviews: Insights and Future Directions

Real-World Evaluation of Anomaly Detection Using Amazon Reviews

Author:
(1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es);
(2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es);
(3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Table of Links Abstract and Introduction Related work The proposed pipeline Evaluation Conclusion & Acknowledgements References & Appendix 2. Related work Anomaly detection has been a consolidated research field for years. Its great utility has allowed its techniques to be applied in numerous areas: medicine [9], industrial systems [10], electronic fraud [11], cybersecurity [12], etc. However, when we talk about texts and NLP, there is no massive application of these anomaly detection techniques as in the previous cases. This may be due to the difficulty in defining the idea of an anomaly in texts. Contrary to other scenarios, such as monitoring an industrial system through its sensors, in which the anomalous class will correspond to faults in the system, when the data is text, defining the concept of an anomaly is not trivial. One of the most important lines of research related to the detection of anomalies in texts is fake reviews detection, also known as spam review detection, fake opinion detection, and spam opinion detection [13]. The main problem associated with fake review detection is classifying the review as either fake or genuine. There are generally three types of fake reviews [14]: • Type 1 (untruthful opinions): Fake reviews describing users who post negative reviews to damage a product’s reputation or post positive reviews to promote it. These reviews are called fake or deceptive reviews, and they are difficult to detect simply by reading, as real and fake reviews are similar to each other. • Type 2 (reviews on brands only): Those that do not comment on the products themselves, but talk about the brands, manufacturers, or sellers of the products. Although they can be useful, they are sometimes considered spam because they are not targeted at specific products. • Type 3 (non-reviews): Non-reviews that are irrelevant and offer no genuine opinion. In the work carried out by J. Salminen et al. [15], the authors try to distinguish genuine reviews from fake reviews on Amazon. To have a labeled dataset of fake reviews, they use GPT-2 [16] to artificially generate them. After this, they solve the task of distinguishing between genuine and fake reviews by fine-tuning a pretrained RoBERTa [17] model. They also show that his model is also capable of detecting fake reviews manually written by humans. Ş. Öztürk Birim et al. [18] proposed to use relevant information as the review length, purchase verification, sentiment score, or topic distribution as features to represent costumers reviews. Based on these features, well-known machine learning classifiers like random forests (RF) [19] are applied for fake detection. In their article, D. U. Vidanagama et al. [20] incorporate review-related features such as linguistic features, Part-of-Speech (POS) features, and sentiment analysis features using a domain ontology to detect fake reviews with a rule-based classifier. L. Ruff et al. [21] presented Context Vector Data Description (CVDD), a text-specific anomaly detection method that allows working with sequences of variable-length embeddings using self-attention mechanisms. To overcome the limitations of CVDD, J. Mu et al. [22] proposed tadnet, a textual anomaly detection network that uses an adversarial training strategy to detect anomalous texts in Social Internet of Things. In addition, thanks to the capture of the different semantic contexts of the texts, both models achieve interpretability and flexibility, allowing to detect which parts of the texts have caused the anomaly. Other authors, instead of developing text-specific AD models, make use of well-known AD and NLP techniques to design architectures that solve the problem. B. Song et al. [23] propose to analyze the accident reports of a chemical processing plant to detect anomalous conditions. In this scenario, anomalous conditions are defined as unexperienced accidents that occur in unusual conditions. The authors work directly with the original text extracting the meaningful keywords of the reports using the term frequency-inverse document frequency (TF-IDF) index. Based on this, and using the local outlier factor (LOF) algorithm, they identify anomaly accidents in terms of local density clusters, finding four major types of anomaly accidents. Working with the original texts and not with embeddings they achieve a certain interpretability in the results. In the approach presented by S. Seo et al. [24], a framework for identifying unusual but noteworthy customer responses and extracting significant words and phrases is proposed. The authors use Doc2Vec to vectorize customer responses, then LOF is applied to identify unusual responses, and based on a TF-IDF analysis and the distances in the embedding space, they can visualize useful information about the results through a network graph. In the previous works and most cases, the detection of fake reviews focuses on detecting type 1 reviews, that is, reviews that positively or negatively describe a product but whose intention is not genuine since they do not come from a real buyer. Since there are several works based on this scenario [15, 18, 20], in this work we will focus on detecting type 2 and 3 reviews, which refer to reviews that do not provide information about the product itself. In this way, the objective of this work is to design a pipeline capable of distinguishing reviews related to a specific product (normal reviews) from reviews that do not and therefore do not provide information to the users that read them (anomalous reviews), for example, because they wrongly describe other products or because they are too generic. In addition to this, the classifications carried out by the system will be explained through an analysis process based on NLP techniques. This paper is available on arxiv under CC 4.0 license. Author: (1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es); (2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es); (3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Author: Author: (1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es); (2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es); (3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Table of Links Abstract and Introduction Abstract and Introduction Related work Related work The proposed pipeline The proposed pipeline Evaluation Evaluation Conclusion & Acknowledgements Conclusion & Acknowledgements References & Appendix References & Appendix 2. Related work Anomaly detection has been a consolidated research field for years. Its great utility has allowed its techniques to be applied in numerous areas: medicine [9], industrial systems [10], electronic fraud [11], cybersecurity [12], etc. However, when we talk about texts and NLP, there is no massive application of these anomaly detection techniques as in the previous cases. This may be due to the difficulty in defining the idea of an anomaly in texts. Contrary to other scenarios, such as monitoring an industrial system through its sensors, in which the anomalous class will correspond to faults in the system, when the data is text, defining the concept of an anomaly is not trivial. One of the most important lines of research related to the detection of anomalies in texts is fake reviews detection, also known as spam review detection, fake opinion detection, and spam opinion detection [13]. The main problem associated with fake review detection is classifying the review as either fake or genuine. There are generally three types of fake reviews [14]: • Type 1 (untruthful opinions): Fake reviews describing users who post negative reviews to damage a product’s reputation or post positive reviews to promote it. These reviews are called fake or deceptive reviews, and they are difficult to detect simply by reading, as real and fake reviews are similar to each other. • Type 2 (reviews on brands only): Those that do not comment on the products themselves, but talk about the brands, manufacturers, or sellers of the products. Although they can be useful, they are sometimes considered spam because they are not targeted at specific products. • Type 3 (non-reviews): Non-reviews that are irrelevant and offer no genuine opinion. In the work carried out by J. Salminen et al. [15], the authors try to distinguish genuine reviews from fake reviews on Amazon. To have a labeled dataset of fake reviews, they use GPT-2 [16] to artificially generate them. After this, they solve the task of distinguishing between genuine and fake reviews by fine-tuning a pretrained RoBERTa [17] model. They also show that his model is also capable of detecting fake reviews manually written by humans. Ş. Öztürk Birim et al. [18] proposed to use relevant information as the review length, purchase verification, sentiment score, or topic distribution as features to represent costumers reviews. Based on these features, well-known machine learning classifiers like random forests (RF) [19] are applied for fake detection. In their article, D. U. Vidanagama et al. [20] incorporate review-related features such as linguistic features, Part-of-Speech (POS) features, and sentiment analysis features using a domain ontology to detect fake reviews with a rule-based classifier. L. Ruff et al. [21] presented Context Vector Data Description (CVDD), a text-specific anomaly detection method that allows working with sequences of variable-length embeddings using self-attention mechanisms. To overcome the limitations of CVDD, J. Mu et al. [22] proposed tadnet, a textual anomaly detection network that uses an adversarial training strategy to detect anomalous texts in Social Internet of Things. In addition, thanks to the capture of the different semantic contexts of the texts, both models achieve interpretability and flexibility, allowing to detect which parts of the texts have caused the anomaly. Other authors, instead of developing text-specific AD models, make use of well-known AD and NLP techniques to design architectures that solve the problem. B. Song et al. [23] propose to analyze the accident reports of a chemical processing plant to detect anomalous conditions. In this scenario, anomalous conditions are defined as unexperienced accidents that occur in unusual conditions. The authors work directly with the original text extracting the meaningful keywords of the reports using the term frequency-inverse document frequency (TF-IDF) index. Based on this, and using the local outlier factor (LOF) algorithm, they identify anomaly accidents in terms of local density clusters, finding four major types of anomaly accidents. Working with the original texts and not with embeddings they achieve a certain interpretability in the results. In the approach presented by S. Seo et al. [24], a framework for identifying unusual but noteworthy customer responses and extracting significant words and phrases is proposed. The authors use Doc2Vec to vectorize customer responses, then LOF is applied to identify unusual responses, and based on a TF-IDF analysis and the distances in the embedding space, they can visualize useful information about the results through a network graph. In the previous works and most cases, the detection of fake reviews focuses on detecting type 1 reviews, that is, reviews that positively or negatively describe a product but whose intention is not genuine since they do not come from a real buyer. Since there are several works based on this scenario [15, 18, 20], in this work we will focus on detecting type 2 and 3 reviews, which refer to reviews that do not provide information about the product itself. In this way, the objective of this work is to design a pipeline capable of distinguishing reviews related to a specific product (normal reviews) from reviews that do not and therefore do not provide information to the users that read them (anomalous reviews), for example, because they wrongly describe other products or because they are too generic. In addition to this, the classifications carried out by the system will be explained through an analysis process based on NLP techniques. This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Advancements in Anomaly Detection

Author:
(1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es);
(2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es);
(3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Table of Links Abstract and Introduction Related work The proposed pipeline Evaluation Conclusion & Acknowledgements References & Appendix Abstract This paper presents a pipeline to detect and explain anomalous reviews in online platforms. The pipeline is made up of three modules and allows the detection of reviews that do not generate value for users due to either worthless or malicious composition. The classifications are accompanied by a normality score and an explanation that justifies the decision made. The pipeline’s ability to solve the anomaly detection task was evaluated using different datasets created from a large Amazon database. Additionally, a study comparing three explainability techniques involving 241 participants was conducted to assess the explainability module. The study aimed to measure the impact of explanations on the respondents’ ability to reproduce the classification model and their perceived usefulness. This work can be useful to automate tasks in review online platforms, such as those for electronic commerce, and offers inspiration for addressing similar problems in the field of anomaly detection in textual data. We also consider it interesting to have carried out a human evaluation of the capacity of different explainability techniques in a real and infrequent scenario such as the detection of anomalous reviews, as well as to reflect on whether it is possible to explain tasks as humanly subjective as this one. Keywords: Anomaly Detection, Text Reviews, Transformers, Explainability 1. Introduction Nowadays more than ever in history, the user opinions about products and services have a great impact on the future of the company that offers them. In such a globalized and highly competitive world, online review platforms, such as electronic commerce (e-commerce), play a crucial role in the credibility of products and services. These user reviews usually come in the form of text reviews or numerical ratings, accompanied in some cases by images or videos. These reviews provide users with information about the product or service they are considering purchasing, which directly influences the number of sales. Most people make purchase decisions based on ratings and reviews from other users [1]. In the case of many companies, such as Amazon, each product in the store has a list of text reviews published by customers of the platform. Users can access this list of reviews (opinions) to obtain extra information about the product, being able to mark the reviews as useful, which will position those reviews with the most votes at the top of the list. In addition to this, users can report to Amazon if they feel a review is inappropriate, for example, if its content is incorrect. This procedure to rank reviews based on their usefulness and report inappropriate reviews is carried out manually by platform users. As a result, Amazon reported more than 200 million suspected fake reviews in 2020 alone [2]. This problem does not only occur on Amazon, but affects all platforms that allow their users to post reviews. For example, Tripadvisor uses an automatic system capable of distinguishing between normal, suspicious and inappropriate reviews [3]. Inappropriate ones are automatically removed (3.1% of review submissions in 2020), while those classified as suspicious are reviewed again by a human moderator (5.1% of review submissions in 2020). On the other hand, in machine learning (ML), anomaly detection (AD) is the branch that builds models capable of differentiating between normal and abnormal data [4]. At first, anomaly detection might seem like a classification problem with only two classes. However, anomalies tend to occur infrequently or are non-existent, so normal data prevails in these scenarios. Because of this, it is common for models to be trained only with normal data. The goal of these models is to represent the normal class as well as possible in order to classify new data as normal or anomalous. The technological development of recent years has allowed the construction of very powerful models for Natural Language Processing (NLP) [5]. Contrary to other tasks such as Sentiment Analysis [6] or Question Answering [7], the application of anomaly detection on texts is still at an early stage, probably due to its lower demand. For this reason, we present in this article a pipeline that, given text reviews of a product (in this case from Amazon), addresses opinion filtering as an anomaly detection problem where: • Reviews containing representative information about the product are considered as the normal class. • Reviews whose content has nothing to do with the product to be represented are considered as the anomalous class. The proposed pipeline allows us to carry out the following tasks: • Classify reviews as normal or anomalous, allowing us to locate those that do not describe characteristics of the product to which they are associated, and therefore have no value for the users of the platform. • Issue a normality score associated with each review. • Generate an explanation that justifies the classification made for each review by the system. The pipeline’s ability to solve the anomaly detection task was evaluated using different datasets created from a large Amazon database [8]. In addition, to evaluate the explainability module, a study was carried out to compare three explainability techniques in which a total of 241 people had participated. The objective of this study is both to measure the impact of the explanations on the reproducibility of the classification model by the respondents, as well as the usefulness of these explanations. We believe that this work can be useful to automate tasks such as those mentioned in online review platforms, in addition to the existing interest in the application of anomaly detection models on texts, for which it can serve as inspiration to solve similar problems. We also consider it interesting to have carried out a human evaluation of the capacity of different explainability techniques in a real and infrequent scenario such as the detection of anomalous reviews, as well as to reflect on whether it is possible to explain tasks as humanly subjective as this one. This document is structured as follows. Section 2 contains a brief review of the main anomaly detection works in texts and, more specifically, in text reviews. Section 3 describes the proposed pipeline and its operation. Section 4 collects the experimentation carried out and, finally, conclusions are drawn in Section 5. This paper is available on arxiv under CC 4.0 license. Author: (1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es); (2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es); (3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Author: Author: (1) David Novoa-Paradela, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain & Corresponding author (Email: david.novoa@udc.es); (2) Oscar Fontenla-Romero, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: oscar.fontenla@udc.es); (3) Bertha Guijarro-Berdiñas, Universidade da Coruña, CITIC, Campus de Elviña s/n, 15008, A Coruña, Spain (Email: berta.guijarro@udc.es). Table of Links Abstract and Introduction Abstract and Introduction Related work Related work The proposed pipeline The proposed pipeline Evaluation Evaluation Conclusion & Acknowledgements Conclusion & Acknowledgements References & Appendix References & Appendix Abstract This paper presents a pipeline to detect and explain anomalous reviews in online platforms. The pipeline is made up of three modules and allows the detection of reviews that do not generate value for users due to either worthless or malicious composition. The classifications are accompanied by a normality score and an explanation that justifies the decision made. The pipeline’s ability to solve the anomaly detection task was evaluated using different datasets created from a large Amazon database. Additionally, a study comparing three explainability techniques involving 241 participants was conducted to assess the explainability module. The study aimed to measure the impact of explanations on the respondents’ ability to reproduce the classification model and their perceived usefulness. This work can be useful to automate tasks in review online platforms, such as those for electronic commerce, and offers inspiration for addressing similar problems in the field of anomaly detection in textual data. We also consider it interesting to have carried out a human evaluation of the capacity of different explainability techniques in a real and infrequent scenario such as the detection of anomalous reviews, as well as to reflect on whether it is possible to explain tasks as humanly subjective as this one. Keywords: Anomaly Detection, Text Reviews, Transformers, Explainability Keywords: 1. Introduction Nowadays more than ever in history, the user opinions about products and services have a great impact on the future of the company that offers them. In such a globalized and highly competitive world, online review platforms, such as electronic commerce (e-commerce), play a crucial role in the credibility of products and services. These user reviews usually come in the form of text reviews or numerical ratings, accompanied in some cases by images or videos. These reviews provide users with information about the product or service they are considering purchasing, which directly influences the number of sales. Most people make purchase decisions based on ratings and reviews from other users [1]. In the case of many companies, such as Amazon, each product in the store has a list of text reviews published by customers of the platform. Users can access this list of reviews (opinions) to obtain extra information about the product, being able to mark the reviews as useful, which will position those reviews with the most votes at the top of the list. In addition to this, users can report to Amazon if they feel a review is inappropriate, for example, if its content is incorrect. This procedure to rank reviews based on their usefulness and report inappropriate reviews is carried out manually by platform users. As a result, Amazon reported more than 200 million suspected fake reviews in 2020 alone [2]. This problem does not only occur on Amazon, but affects all platforms that allow their users to post reviews. For example, Tripadvisor uses an automatic system capable of distinguishing between normal, suspicious and inappropriate reviews [3]. Inappropriate ones are automatically removed (3.1% of review submissions in 2020), while those classified as suspicious are reviewed again by a human moderator (5.1% of review submissions in 2020). On the other hand, in machine learning (ML), anomaly detection (AD) is the branch that builds models capable of differentiating between normal and abnormal data [4]. At first, anomaly detection might seem like a classification problem with only two classes. However, anomalies tend to occur infrequently or are non-existent, so normal data prevails in these scenarios. Because of this, it is common for models to be trained only with normal data. The goal of these models is to represent the normal class as well as possible in order to classify new data as normal or anomalous. The technological development of recent years has allowed the construction of very powerful models for Natural Language Processing (NLP) [5]. Contrary to other tasks such as Sentiment Analysis [6] or Question Answering [7], the application of anomaly detection on texts is still at an early stage, probably due to its lower demand. For this reason, we present in this article a pipeline that, given text reviews of a product (in this case from Amazon), addresses opinion filtering as an anomaly detection problem where: • Reviews containing representative information about the product are considered as the normal class. • Reviews whose content has nothing to do with the product to be represented are considered as the anomalous class. The proposed pipeline allows us to carry out the following tasks: • Classify reviews as normal or anomalous, allowing us to locate those that do not describe characteristics of the product to which they are associated, and therefore have no value for the users of the platform. • Issue a normality score associated with each review. • Generate an explanation that justifies the classification made for each review by the system. The pipeline’s ability to solve the anomaly detection task was evaluated using different datasets created from a large Amazon database [8]. In addition, to evaluate the explainability module, a study was carried out to compare three explainability techniques in which a total of 241 people had participated. The objective of this study is both to measure the impact of the explanations on the reproducibility of the classification model by the respondents, as well as the usefulness of these explanations. We believe that this work can be useful to automate tasks such as those mentioned in online review platforms, in addition to the existing interest in the application of anomaly detection models on texts, for which it can serve as inspiration to solve similar problems. We also consider it interesting to have carried out a human evaluation of the capacity of different explainability techniques in a real and infrequent scenario such as the detection of anomalous reviews, as well as to reflect on whether it is possible to explain tasks as humanly subjective as this one. This document is structured as follows. Section 2 contains a brief review of the main anomaly detection works in texts and, more specifically, in text reviews. Section 3 describes the proposed pipeline and its operation. Section 4 collects the experimentation carried out and, finally, conclusions are drawn in Section 5. This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Explained Anomaly Detection in Text Reviews: Can Subjective Scenarios Be Correctly Evaluated?

This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license.
Authors:
(1) Brisha Jain, Independent researcher India and brishajain02@gmail.com;
(2) Mainack Mondal, IIT Kharagpur India and mainack@cse.iitkgp.ac.in. Table of Links Abstract and Intro
Related Work
Data Collection
Analysis Methodology
Results
Limitations
Concluding Discussion
References REFERENCES [1] Toril Aalberg and Anders Todal Jenssen. 2017. Gender Stereotyping of Political Candidates. Nordicom Review 28, 1 (2017), 17–32. https://doi.org/10.1515/nor-2017-0198 [2] Pushkal Agarwal, Nishanth Sastry, and Edward Wood. 2019. Tweeting MPs: Digital Engagement between Citizens and Members of Parliament in the UK. Proceedings of the International AAAI Conference on Web and Social Media 13, 01. [3] Arshia Arya, Soham De, Dibyendu Mishra, Gazal Shekhawat, Ankur Sharma, Anmol Panda, Faisal M Lalani, Parantak Singh, Ramaravind Kommiya Mothilal, Rynaa Grover, Sachita Nishal, Saloni Dash, Shehla Rashid Shora, Syeda Zainab Akbar, and Joyojeet Pal. 2022. DISMISS: Database of Indian Social Media Influencers on Twitter. [4] Nichole M. Bauer. 2015. Emotional, Sensitive, and Unfit for Office? Gender Stereotype Activation and Support Female Candidates. Political Psychology 36, 6 (2015), 691–708. [5] Aditi Bhowmick, Paul Novosad, Sam Asher, Elliott Ash, Bilal Siddiqi, Christoph Goessmann, Daniel Chen, and Tanaya Devi. 2021. In-Group Bias in the Indian Judiciary: Evidence from 5.5 Million Criminal Cases. In Proceedings of the 4th ACM SIGCAS Conference on Computing and Sustainable Societies (COMPASS ’21). 47. https://doi.org/10.1145/3460112.3471943 [6] Jose Camacho-Collados, Kiamehr Rezaee, Talayeh Riahi, Asahi Ushio, Daniel Loureiro, Dimosthenis Antypas, Joanne Boisson, Luis Espinosa-Anke, Fangyu Liu, Eugenio Martínez-Cámara, et al. 2022. TweetNLP: Cutting-Edge Natural Language Processing for Social Media. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: System Demonstrations. [7] Tessa M. Ditonto, Allison J. Hamilton, and David P. Redlawsk. 2014. Gender Stereotypes, Information Search, and Voting Behavior in Political Campaigns. Political Behavior 36, 2 (2014), 335–358. https://doi.org/10.1007/s11109-013-9232-6 [8] Kathleen Dolan and Timothy Lynch. 2014. It Takes a Survey: Understanding Gender Stereotypes, Abstract Attitudes, and Voting for Women Candidates. American Politics Research 42, 4 (2014), 656–576. https://doi.org/10.1177/1532673X13503034 [9] Heather Evans. 2016. Do Women Only Talk about ‘Female Issues’? Gender and
Issue Discussion on Twitter. Online Information Review 40, 5 (2016), 660–672.
https://doi.org/10.1108/OIR-10-2015-0338 [10] Heather K. Evans and Jennifer Hayes Clark. 2016. You Tweet Like a Girl!’: How Female Candidates Campaign on Twitter. American Politics Research 44, 2 (2016), 326–352. https://doi.org/10.1177/1532673X15597747 [11] Peter A. Gloor, Joao Marcos, Patrick M. de Boer, Hauke Fuehres, Wei Lo, and Keiichi Nemoto. 2015. Cultural Anthropology through the Lens of Wikipedia: Historical Leader Networks, Gender Bias, and News-based Sentiment. arXiv:1508.00055 [cs.CY] https://doi.org/10.48550/ARXIV.1508.00055 [12] Leonie Huddy and Nayda Terkildsen. 1993. Gender Stereotypes and the Perception of Male and Female Candidates. American Journal of Political Science 37, 1 (1993), 119. https://doi.org/10.2307/2111526 [13] Berrin Beasley John H. Parmelee, Nataliya Roman and Stephynie C. Perkins. 2019. Gender and Generational Differences in Political Reporters’ Interactivity on Twitter. Journalism Studies 20, 2 (2019), 232–247. https://doi.org/10.1080/1461670X.2017.1364140 [14] Ashiqur R. KhudaBukhsh Kunal Khadilkar and Tom M. Mitchell. 2022. Gender Bias, Social Bias, and Representation in Bollywood and Hollywood. Patterns 3, 4 (2022), 100486. https://doi.org/10.1016/j.patter.2022.100486 [15] MingYu (Ethen) Liu. 2023. Topic Modeling. https://ethen8181.github.io/machine-learning/clustering/topic_model/LDA.html. [16] J. Nathan Matias, Sarah Szalavitz, and Ethan Zuckerman. 2017. FollowBias: Supporting Behavior Change toward Gender Equality by Networked Gatekeepers on Social Media (CSCW ’17). 1082–1095. https://doi.org/10.1145/2998181.2998287 [17] Lindsey Meeks. 2016. Gendered Styles, Gendered Differences: Candidates’ Use of Personalization and Interactivity on Twitter. Journal of Information Technology & Politics 13, 4 (2016), 295–310. https://doi.org/10.1080/19331681.2016.1160268 [18] Armin Mertens, Franziska Pradel, Ayjeren Rozyjumayeva, and Jens Wäckerle. 2019. As the Tweet, so the Reply? Gender Bias in Digital Communication with Politicians. In Proceedings of the 10th ACM Conference on Web Science (WebSci ’19). 193–201. [19] Sebo P. 2021. Using genderize.io to infer the gender of first names: how to improve the accuracy of the inference. J Med Libr Assoc. (2021). [20] Anmol Panda, A’ndre Gonawela, Sreangsu Acharyya, Dibyendu Mishra, Mugdha Mohapatra, Ramgopal Chandrasekaran, and Joyojeet Pal. 2020. NivaDuck - A Scalable Pipeline to Build a Database of Political Twitter Handles for India and the United States. In International Conference on Social Media and Society. 200–209. [21] John H. Parmelee. 2013. Political Journalists and Twitter: Influences on Norms and Practices. Journal of Media Practice 14, 4 (2013), 291–305. https://doi.org/10.1386/jmpr.14.4.291_1 [22] Pallavi Prasad. 2020. One in Every Seven Tweets About Indian Women Politicians Is Abusive: Study. https://www.theswaddle.com/women-politicians-india-twitter-trolls. [23] Gazal Shekhawat, Parantak Singh, Rynaa Grover, Saloni Dash, Himani Negi, Ankur Sharma, Somya Sagarika, and Joyojeet Pal. 2021. Twitter Use by Politicians Across States in India. [24] Carole Spary. 2007. Female Political Leadership in India. Commonwealth & Comparative Politics 45, 3 (2007), 253–277. https://doi.org/10.1080/14662040701516821 [25] Genderize team. 2023. Genderize. https://genderize.io/. [26] Jesse Holcomb Usher, Nikki and Justin Littman. 2018. Twitter Makes It Worse: Political Journalists, Gendered Echo Chambers, and the Amplification of Gender Bias. The International Journal of Press/Politics 23, 3 (2018), 324–344. https://doi.org/10.1177/1940161218781254 [27] Jason Gainous Wagner, Kevin M. and Mirya R. Holman. 2017. I Am Woman, Hear Me Tweet! Gender Differences in Twitter Use among Congressional Candidates. Journal of Women, Politics & Policy 38, 4 (2017), 430–455. https://doi.org/10.1080/1554477X.2016.1268871 This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. Authors: (1) Brisha Jain, Independent researcher India and brishajain02@gmail.com; (2) Mainack Mondal, IIT Kharagpur India and mainack@cse.iitkgp.ac.in. This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. Authors: Authors: (1) Brisha Jain, Independent researcher India and brishajain02@gmail.com; (2) Mainack Mondal, IIT Kharagpur India and mainack@cse.iitkgp.ac.in. Table of Links Abstract and Intro Related Work Data Collection Analysis Methodology Results Limitations Concluding Discussion References Abstract and Intro Abstract and Intro Related Work Related Work Data Collection Data Collection Analysis Methodology Analysis Methodology Results Results Limitations Limitations Concluding Discussion Concluding Discussion References References REFERENCES [1] Toril Aalberg and Anders Todal Jenssen. 2017. Gender Stereotyping of Political Candidates. Nordicom Review 28, 1 (2017), 17–32. https://doi.org/10.1515/nor-2017-0198 [2] Pushkal Agarwal, Nishanth Sastry, and Edward Wood. 2019. Tweeting MPs: Digital Engagement between Citizens and Members of Parliament in the UK. Proceedings of the International AAAI Conference on Web and Social Media 13, 01. [3] Arshia Arya, Soham De, Dibyendu Mishra, Gazal Shekhawat, Ankur Sharma, Anmol Panda, Faisal M Lalani, Parantak Singh, Ramaravind Kommiya Mothilal, Rynaa Grover, Sachita Nishal, Saloni Dash, Shehla Rashid Shora, Syeda Zainab Akbar, and Joyojeet Pal. 2022. DISMISS: Database of Indian Social Media Influencers on Twitter. [4] Nichole M. Bauer. 2015. Emotional, Sensitive, and Unfit for Office? Gender Stereotype Activation and Support Female Candidates. Political Psychology 36, 6 (2015), 691–708. [5] Aditi Bhowmick, Paul Novosad, Sam Asher, Elliott Ash, Bilal Siddiqi, Christoph Goessmann, Daniel Chen, and Tanaya Devi. 2021. In-Group Bias in the Indian Judiciary: Evidence from 5.5 Million Criminal Cases. In Proceedings of the 4th ACM SIGCAS Conference on Computing and Sustainable Societies (COMPASS ’21). 47. https://doi.org/10.1145/3460112.3471943 https://doi.org/10.1145/3460112.3471943 [6] Jose Camacho-Collados, Kiamehr Rezaee, Talayeh Riahi, Asahi Ushio, Daniel Loureiro, Dimosthenis Antypas, Joanne Boisson, Luis Espinosa-Anke, Fangyu Liu, Eugenio Martínez-Cámara, et al. 2022. TweetNLP: Cutting-Edge Natural Language Processing for Social Media. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: System Demonstrations. [7] Tessa M. Ditonto, Allison J. Hamilton, and David P. Redlawsk. 2014. Gender Stereotypes, Information Search, and Voting Behavior in Political Campaigns. Political Behavior 36, 2 (2014), 335–358. https://doi.org/10.1007/s11109-013-9232-6 https://doi.org/10.1007/s11109-013-9232-6 [8] Kathleen Dolan and Timothy Lynch. 2014. It Takes a Survey: Understanding Gender Stereotypes, Abstract Attitudes, and Voting for Women Candidates. American Politics Research 42, 4 (2014), 656–576. https://doi.org/10.1177/1532673X13503034 https://doi.org/10.1177/1532673X13503034 [9] Heather Evans. 2016. Do Women Only Talk about ‘Female Issues’? Gender and
Issue Discussion on Twitter. Online Information Review 40, 5 (2016), 660–672. https://doi.org/10.1108/OIR-10-2015-0338 https://doi.org/10.1108/OIR-10-2015-0338 [10] Heather K. Evans and Jennifer Hayes Clark. 2016. You Tweet Like a Girl!’: How Female Candidates Campaign on Twitter. American Politics Research 44, 2 (2016), 326–352. https://doi.org/10.1177/1532673X15597747 https://doi.org/10.1177/1532673X15597747 [11] Peter A. Gloor, Joao Marcos, Patrick M. de Boer, Hauke Fuehres, Wei Lo, and Keiichi Nemoto. 2015. Cultural Anthropology through the Lens of Wikipedia: Historical Leader Networks, Gender Bias, and News-based Sentiment. arXiv:1508.00055 [ cs.CY ] https://doi.org/10.48550/ARXIV.1508.00055 cs.CY https://doi.org/10.48550/ARXIV.1508.00055 [12] Leonie Huddy and Nayda Terkildsen. 1993. Gender Stereotypes and the Perception of Male and Female Candidates. American Journal of Political Science 37, 1 (1993), 119. https://doi.org/10.2307/2111526 https://doi.org/10.2307/2111526 [13] Berrin Beasley John H. Parmelee, Nataliya Roman and Stephynie C. Perkins. 2019. Gender and Generational Differences in Political Reporters’ Interactivity on Twitter. Journalism Studies 20, 2 (2019), 232–247. https://doi.org/10.1080/1461670X.2017.1364140 https://doi.org/10.1080/1461670X.2017.1364140 [14] Ashiqur R. KhudaBukhsh Kunal Khadilkar and Tom M. Mitchell. 2022. Gender Bias, Social Bias, and Representation in Bollywood and Hollywood. Patterns 3, 4 (2022), 100486. https://doi.org/10.1016/j.patter.2022.100486 https://doi.org/10.1016/j.patter.2022.100486 [15] MingYu (Ethen) Liu. 2023. Topic Modeling. https://ethen8181.github.io/machine-learning/clustering/topic_model/LDA.html . https://ethen8181.github.io/machine-learning/clustering/topic_model/LDA.html [16] J. Nathan Matias, Sarah Szalavitz, and Ethan Zuckerman. 2017. FollowBias: Supporting Behavior Change toward Gender Equality by Networked Gatekeepers on Social Media (CSCW ’17). 1082–1095. https://doi.org/10.1145/2998181.2998287 https://doi.org/10.1145/2998181.2998287 [17] Lindsey Meeks. 2016. Gendered Styles, Gendered Differences: Candidates’ Use of Personalization and Interactivity on Twitter. Journal of Information Technology & Politics 13, 4 (2016), 295–310. https://doi.org/10.1080/19331681.2016.1160268 https://doi.org/10.1080/19331681.2016.1160268 [18] Armin Mertens, Franziska Pradel, Ayjeren Rozyjumayeva, and Jens Wäckerle. 2019. As the Tweet, so the Reply? Gender Bias in Digital Communication with Politicians. In Proceedings of the 10th ACM Conference on Web Science (WebSci ’19). 193–201. [19] Sebo P. 2021. Using genderize.io to infer the gender of first names: how to improve the accuracy of the inference. J Med Libr Assoc. (2021). genderize.io [20] Anmol Panda, A’ndre Gonawela, Sreangsu Acharyya, Dibyendu Mishra, Mugdha Mohapatra, Ramgopal Chandrasekaran, and Joyojeet Pal. 2020. NivaDuck - A Scalable Pipeline to Build a Database of Political Twitter Handles for India and the United States. In International Conference on Social Media and Society. 200–209. [21] John H. Parmelee. 2013. Political Journalists and Twitter: Influences on Norms and Practices. Journal of Media Practice 14, 4 (2013), 291–305. https://doi.org/10.1386/jmpr.14.4.291_1 https://doi.org/10.1386/jmpr.14.4.291_1 [22] Pallavi Prasad. 2020. One in Every Seven Tweets About Indian Women Politicians Is Abusive: Study. https://www.theswaddle.com/women-politicians-india-twitter-trolls . https://www.theswaddle.com/women-politicians-india-twitter-trolls [23] Gazal Shekhawat, Parantak Singh, Rynaa Grover, Saloni Dash, Himani Negi, Ankur Sharma, Somya Sagarika, and Joyojeet Pal. 2021. Twitter Use by Politicians Across States in India. [24] Carole Spary. 2007. Female Political Leadership in India. Commonwealth & Comparative Politics 45, 3 (2007), 253–277. https://doi.org/10.1080/14662040701516821 https://doi.org/10.1080/14662040701516821 [25] Genderize team. 2023. Genderize. https://genderize.io/ . https://genderize.io/ [26] Jesse Holcomb Usher, Nikki and Justin Littman. 2018. Twitter Makes It Worse: Political Journalists, Gendered Echo Chambers, and the Amplification of Gender Bias. The International Journal of Press/Politics 23, 3 (2018), 324–344. https://doi.org/10.1177/1940161218781254 https://doi.org/10.1177/1940161218781254 [27] Jason Gainous Wagner, Kevin M. and Mirya R. Holman. 2017. I Am Woman, Hear Me Tweet! Gender Differences in Twitter Use among Congressional Candidates. Journal of Women, Politics & Policy 38, 4 (2017), 430–455. https://doi.org/10.1080/1554477X.2016.1268871 https://doi.org/10.1080/1554477X.2016.1268871

Uncovering Gender Bias within Journalist-Politician Interaction in Indian Twitter: References

ভারতীয় টুইটারে সাংবাদিক-রাজনীতিবিদ ইন্টারঅ্যাকশনের মধ্যে জেন্ডার বায়াস উন্মোচন: রেফারেন্স

Aufdeckung geschlechtsspezifischer Voreingenommenheit in der Interaktion zwischen Journalisten und Politikern im indischen Twitter: Referenzen

Descubriendo el sesgo de género en la interacción entre periodistas y políticos en Twitter indio: referencias

Découvrir les préjugés sexistes dans les interactions journalistes-hommes politiques sur Twitter indien : références

भारतीय ट्विटर पर पत्रकार-राजनेता बातचीत में लैंगिक पूर्वाग्रह को उजागर करना: संदर्भ

インドのツイッターにおけるジャーナリストと政治家の交流におけるジェンダーバイアスの解明：参考文献

인도 트위터의 언론인-정치인 상호작용에서 성 편견 발견: 참고자료

Descobrindo o preconceito de gênero na interação jornalista-político no Twitter indiano: referências

Выявление гендерных предубеждений во взаимодействии журналистов и политиков в индийском Твиттере: ссылки

Hint Twitter'ında Gazeteci-Siyasetçi Etkileşiminde Cinsiyet Önyargısını Ortaya Çıkarmak: Referanslar

Khám phá sự thiên vị giới tính trong tương tác giữa nhà báo và chính trị gia trên Twitter của Ấn Độ: Tài liệu tham khảo

揭示印度推特上记者与政客互动中的性别偏见：参考文献

The Food and Drug Administration’s (FDA’s) recent approval of the loss drug Wegovy means it’ll soon sweep the nation. However, as its popularity and price have surged, it’s become less accessible to the average American. Can wearables provide a solution to this problem? The Problem With Wegovy and How Wearables Can Fix It The weight loss market is at an all-time high — and indicators suggest it’ll continue expanding. According to one estimate, its value will surpass $399.3 billion by 2032, achieving a CAGR of 10.3%. This growth has driven demand for drugs like Wegovy. Wegovy is a treatment for chronic weight management. Its recent FDA approval means it will likely become one of the leading weight loss drugs in the United States. That said, despite the country’s high obesity rate most — 41.9% of adults over 20 are obese — Americans won’t benefit from it. Most people don’t have an extra $1,300 to spend each month — that’s Wegovy’s list price. Although one report found its average net price was $700 monthly between 2022 to 2023, it’s still more than most can afford. What’s worse is that Wegovy isn’t just a weight loss drug — new evidence suggests it can also lower the risk of cardiovascular issues, prolonging people’s lives. A 17,000-patient study discovered those taking it had a 20% lower risk of experiencing a cardiac-related health event than those taking a placebo. Skyrocketing demand, high research and development costs, the limited availability of clinical trials, and technology gaps might make it almost impossible for Novo Nordisk — the drugmaker behind Wegovy — to improve the accessibility of its lifesaving, fat-burning drug. That’s where wearable tech comes in. Wearables could improve Wegovy’s treatment efficacy, increasing its ability to accomplish desired outcomes. Their unique data collection capabilities could also accelerate research, potentially lowering costs and increasing accessibility. How Wearables Can Track and Enhance Treatment Efficacy Wearable tech has emerged as the next horizon in health care. It has proven exceptionally valuable in clinical trials because it can gather patient data 24/7 regardless of participants’ location — specialized facilities and invasive observation technology can move aside. Medical professionals can track treatment efficacy by monitoring the real-time patient data internet-connected wearables constantly send. Being able to view data-driven results immediately instead of waiting for testing or in-person visits streamlines their processes. Eventually, clinical trial teams can gather enough information to segment participants by demographics, physical states, or preexisting conditions. This way, they can tailor factors like dosage or medication frequency to each group, increasing the drug’s effectiveness. Unlike other sensor technologies, wearable tech often pairs with a supplementary app. Drugmakers and clinical trial leaders could use this unique functionality to send helpful push notifications or help people track their progress. Wearable Tech Tracks Patients’ Treatment Responses Wearable tech can monitor physiological responses like glucose levels and heart rate without invasive procedures like blood drawing. It can track metrics like blood pressure, blood oxygen level, or daily calories burned the same way. Non-invasive sensor tech — meaning no needles, swallowable capsules, or electrodes — could help people feel better about participating in clinical trials or taking new drugs. There’s no longer any need for repetitive, painful, or time-consuming testing. How Wearable Tech Can Promote Self-Efficacy One overlooked opportunity for wearables involves self-efficacy — a person’s belief in their ability to reach a specific goal. When people wear something that tracks their activity and health, they feel more inclined to improve their habits. While a medication can help people slim down, weight loss usually requires various efforts. In a 68-week clinical trial, Wegovy lowered participants’ mean body weight by 14.9% by targeting hormones responsible for appetite and food intake — a noticeable difference from the placebo’s 2.4% reduction. Notably, the loss is within the 5%–15% range recommended to improve obesity-related conditions. Although research shows Wegovy outperforms similar drugs, its long-term outcomes remain unproven — the 68-week clinical trial saw participants adopt healthy habits and take the drug to achieve clinically relevant body weight reduction. In other words, the self-efficacy wearables could prompt lifestyle changes, increasing the drug’s effectiveness. Research suggests wearables increase self-efficacy in weight loss trial participants by encouraging physical activity, motivating and enhancing awareness. While there’s no clear link between diet, exercise, and this technology, researchers have made some connections. Wearables could go a long way in making Wegovy more accessible to the average American seeking to shed a few pounds. Typically, people gain weight quickly after stopping these kinds of drugs, but most won’t be able to afford the $700 monthly bill. Pairing the medication with lifestyle changes could enhance its effectiveness, increasing long-term success. Integration Obstacles That Need Overcoming One of the largest obstacles drugmakers and clinical trial leaders must overcome is data accuracy. Although 24/7 at-home monitoring makes processes easier, it can also mess them up. Countless variables could affect results since participants aren’t inside a controlled environment. Wegovy isn’t supposed to be used with other semaglutide-containing products for safety reasons. This factor is simple enough to control in a lab. However, it’s easy for someone at home to pick up an over-the-counter prescription or use an herbal remedy without reading the label to check for semaglutide. Privacy concerns are another obstacle since wearable tech tracks excess information. While the abundance of measurements is excellent for researchers, patients might hesitate to share so much. There’s also the issue of data breaches and leaks. If the weight loss market plans to integrate wearables into health care — a sector notorious for its late adoption of technology due to several legal, moral, and regulatory hurdles — they’ll need to consider the upfront cost. Do clinical trial participants pay for the tech? Who pays when someone breaks theirs? Decision-makers should also consider long-term effectiveness. While this tech can inspire lifestyle changes, people might be unwilling to stick to their new fitness routines. Unless they ensure participants stay dedicated, it’ll look like their decreasing use correlates to diminishing efficacy. The Power of Wearable Tech for Treatment Efficacy If clinical trial leaders and medical professionals give people wearables when they hand out Wegovy, there’s a good chance they’ll enhance treatment success, leading to better long-term patient outcomes. Eventually, these improvements could lead to a more affordable drug and higher demand — a win-win for both the drugmaker and those looking to lose weight. The Food and Drug Administration’s (FDA’s) recent approval of the loss drug Wegovy means it’ll soon sweep the nation. However, as its popularity and price have surged, it’s become less accessible to the average American. Can wearables provide a solution to this problem? The Problem With Wegovy and How Wearables Can Fix It The Problem With Wegovy and How Wearables Can Fix It The weight loss market is at an all-time high — and indicators suggest it’ll continue expanding. According to one estimate, its value will surpass $399.3 billion by 2032, achieving a CAGR of 10.3%. This growth has driven demand for drugs like Wegovy. will surpass $399.3 billion will surpass $399.3 billion Wegovy is a treatment for chronic weight management. Its recent FDA approval means it will likely become one of the leading weight loss drugs in the United States. That said, despite the country’s high obesity rate most — 41.9% of adults over 20 are obese — Americans won’t benefit from it. 41.9% of adults over 20 41.9% of adults over 20 Most people don’t have an extra $1,300 to spend each month — that’s Wegovy’s list price. Although one report found its average net price was $700 monthly between 2022 to 2023, it’s still more than most can afford. net price was $700 monthly net price was $700 monthly What’s worse is that Wegovy isn’t just a weight loss drug — new evidence suggests it can also lower the risk of cardiovascular issues, prolonging people’s lives. A 17,000-patient study discovered those taking it had a 20% lower risk of experiencing a cardiac-related health event than those taking a placebo. had a 20% lower risk had a 20% lower risk Skyrocketing demand, high research and development costs, the limited availability of clinical trials, and technology gaps might make it almost impossible for Novo Nordisk — the drugmaker behind Wegovy — to improve the accessibility of its lifesaving, fat-burning drug. That’s where wearable tech comes in. Wearables could improve Wegovy’s treatment efficacy, increasing its ability to accomplish desired outcomes. Their unique data collection capabilities could also accelerate research, potentially lowering costs and increasing accessibility. How Wearables Can Track and Enhance Treatment Efficacy Wearable tech has emerged as the next horizon in health care. It has proven exceptionally valuable in clinical trials because it can gather patient data 24/7 regardless of participants’ location — specialized facilities and invasive observation technology can move aside. has proven exceptionally valuable has proven exceptionally valuable Medical professionals can track treatment efficacy by monitoring the real-time patient data internet-connected wearables constantly send. Being able to view data-driven results immediately instead of waiting for testing or in-person visits streamlines their processes. Eventually, clinical trial teams can gather enough information to segment participants by demographics, physical states, or preexisting conditions. This way, they can tailor factors like dosage or medication frequency to each group, increasing the drug’s effectiveness. Unlike other sensor technologies, wearable tech often pairs with a supplementary app. Drugmakers and clinical trial leaders could use this unique functionality to send helpful push notifications or help people track their progress. Wearable Tech Tracks Patients’ Treatment Responses Wearable tech can monitor physiological responses like glucose levels and heart rate without invasive procedures like blood drawing. It can track metrics like blood pressure, blood oxygen level, or daily calories burned the same way. Non-invasive sensor tech — meaning no needles, swallowable capsules, or electrodes — could help people feel better about participating in clinical trials or taking new drugs. There’s no longer any need for repetitive, painful, or time-consuming testing. How Wearable Tech Can Promote Self-Efficacy One overlooked opportunity for wearables involves self-efficacy — a person’s belief in their ability to reach a specific goal. When people wear something that tracks their activity and health, they feel more inclined to improve their habits. While a medication can help people slim down, weight loss usually requires various efforts. In a 68-week clinical trial, Wegovy lowered participants’ mean body weight by 14.9% by targeting hormones responsible for appetite and food intake — a noticeable difference from the placebo’s 2.4% reduction. Notably, the loss is within the 5%–15% range recommended to improve obesity-related conditions. mean body weight by 14.9% mean body weight by 14.9% Although research shows Wegovy outperforms similar drugs, its long-term outcomes remain unproven — the 68-week clinical trial saw participants adopt healthy habits and take the drug to achieve clinically relevant body weight reduction. In other words, the self-efficacy wearables could prompt lifestyle changes, increasing the drug’s effectiveness. Research suggests wearables increase self-efficacy in weight loss trial participants by encouraging physical activity, motivating and enhancing awareness. While there’s no clear link between diet, exercise, and this technology, researchers have made some connections. wearables increase self-efficacy in weight loss wearables increase self-efficacy in weight loss Wearables could go a long way in making Wegovy more accessible to the average American seeking to shed a few pounds. Typically, people gain weight quickly after stopping these kinds of drugs, but most won’t be able to afford the $700 monthly bill. Pairing the medication with lifestyle changes could enhance its effectiveness, increasing long-term success. Integration Obstacles That Need Overcoming One of the largest obstacles drugmakers and clinical trial leaders must overcome is data accuracy. Although 24/7 at-home monitoring makes processes easier, it can also mess them up. Countless variables could affect results since participants aren’t inside a controlled environment. Wegovy isn’t supposed to be used with other semaglutide-containing products for safety reasons. This factor is simple enough to control in a lab. However, it’s easy for someone at home to pick up an over-the-counter prescription or use an herbal remedy without reading the label to check for semaglutide. with other semaglutide-containing products with other semaglutide-containing products Privacy concerns are another obstacle since wearable tech tracks excess information. While the abundance of measurements is excellent for researchers, patients might hesitate to share so much. There’s also the issue of data breaches and leaks. If the weight loss market plans to integrate wearables into health care — a sector notorious for its late adoption of technology due to several legal, moral, and regulatory hurdles — they’ll need to consider the upfront cost. Do clinical trial participants pay for the tech? Who pays when someone breaks theirs? Decision-makers should also consider long-term effectiveness. While this tech can inspire lifestyle changes, people might be unwilling to stick to their new fitness routines. Unless they ensure participants stay dedicated, it’ll look like their decreasing use correlates to diminishing efficacy. The Power of Wearable Tech for Treatment Efficacy If clinical trial leaders and medical professionals give people wearables when they hand out Wegovy, there’s a good chance they’ll enhance treatment success, leading to better long-term patient outcomes. Eventually, these improvements could lead to a more affordable drug and higher demand — a win-win for both the drugmaker and those looking to lose weight.

Wearable Tech and Its Potential to Monitor Wegovy Treatment Efficacy

**BERLIN, Germany, March 27th, 2024/Chainwire/--**On March 13th, 2024, Casino.Guide for Canadian game-of-chance enthusiasts. That is a selection of gaming hubs licensed by international regulators, whose games get tested by third-party independent randomness verification agencies like iTech Labs, GLI, and eCOGRA. published its selection of the Internet’s highest-quality gambling platforms The number one spot on this list gets reserved for Stake.com, what many consider the world’s most famous crypto casino, licensed by Curacao Gaming. It is a remote casino/sportsbook that works with various sports franchises, like English Premiership side Everton FC, the UFC, and the Alfa Romeo F1 racing team. Stake has also made the news cycle in recent years for its , who is a massive online gambling aficionado. affiliation with Canadian musical superstar Drake Casino.Guide followed up Stake.com with Lucky Days, Cloudbet, InstaSpin, and BC.Game, completing its top five rankings with these five sites. All four of these, like Stake, operate from Curacao, accepting virtual currencies as viable payment methods while offering welcome bonuses for new users. BC.Game is also a notable interactive gaming brand that many online gamblers have gotten familiar with thanks to its affiliation/sponsorship with the Cloud9 eSports team, which has squads that compete in various video games like Counter Strike Global Offensive, League of Legends, Rocket League, Halo Infinite, StarCraft, and Overwatch. BC.Game is also the official betting partner of the Argentinian national football team. Stake is the wagering partner of the UFC. And it is of its former middleweight champion, Israel Adesanya. a long-time sponsor About Stake It comes as little surprise that Casino.Guide places Stake.com as . It debuted in 2017 and quickly established itself as a credible gaming brand. It gets operated by Medium Rare N.V., a company run by Ed Craven, the face of the Stake brand. the number one Canadian online casino In August 2022, Stake.com entered the United States through a social casino offshoot, and the company’s staff manages its operations from offices in Cyprus, Australia, and Serbia. The site itself gets renowned for its vast collection of games, advanced sportsbook page, and wide range of accepted cryptos. How Casino.Guide Rates Online Casino Sites. Casino.Guide claims that it is the go-to place for objective reviews and ratings, listing the most trustworthy gaming sites for Canadian players. It also supplies info on the latest casino streamer trends and everything anyone would want to know about crypto gambling. The site staff evaluates casinos based on their registration promos, game selections, payment options, payout timeframes, and the competence of their customer service staff. Casino.Guide specialists diligently seek to analyze these main areas and others to conclude if a specific website meets their quality control standards. If it does, they rank it accordingly, stacking it up with other crypto-centric competitors that allow Canadians to join their user pools. A Casino.Guide rating between 95% and 100% means that a platform comes highly recommended. A 94% to 90% rating gets reserved for recommended brands. An 89% to 85% rating is for passable websites, and those that fall under the 85% rating get labeled as casinos in need of improvement. There is also a Best-in-Class group, the crème de la crème of online gaming operators for Canadian remote casino lovers. Casino.Guide’s Team The Casino.Guide website got established in 2020, founded by a group of iGaming experts that total over three decades of online gambling sector experience. The pack gets headlined by Vienne Garcia, a resident of Malta, Europe’s prime iGaming hotspot. Garcia has spent much time in British Columbia, so she knows about what Canadians desire concerning gaming, and she looks to put her decade’s worth of experience to good use to cater to their needs. Christian Webber is the in-house mobile gaming expert at Casino.Guide. He looks to keep up to date with the latest developments on the smartphone gambling side of things. Christian lives in Germany and has spent almost ten years working , Austria’s land-based machine gambling juggernaut. Peter Schmidt is the final part of the Casino.Guide puzzle, supplying his unmatched knowledge of all things related to casino play over the Web. He is the chief site analyst, primarily focusing on reviewed gaming platforms. as a software tester for Novomatic Contact Teamlead Outreach Kush Sharma impulsQ GmbH ks@impulsq.de This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here.

Casino.Guide Posts Its List Of The Best Canadian Online Casinos For March 2024

Casino.Guide মার্চ 2024 এর জন্য সেরা কানাডিয়ান অনলাইন ক্যাসিনোগুলির তালিকা পোস্ট করে

Casino.Guide veröffentlicht seine Liste der besten kanadischen Online-Casinos für März 2024

Casino.Guide publica su lista de los mejores casinos en línea canadienses para marzo de 2024

Casino.Guide publie sa liste des meilleurs casinos en ligne canadiens pour mars 2024

Casino.Guide ने मार्च 2024 के लिए सर्वश्रेष्ठ कनाडाई ऑनलाइन कैसीनो की अपनी सूची पोस्ट की

Casino.Guide が 2024 年 3 月のカナダのベストオンラインカジノのリストを掲載

Casino.Guide가 2024년 3월 최고의 캐나다 온라인 카지노 목록을 게시했습니다.

Casino.Guide publica sua lista dos melhores cassinos online canadenses para março de 2024

Casino.Guide опубликовал список лучших канадских онлайн-казино на март 2024 года

Casino.Guide Mart 2024 İçin En İyi Kanada Çevrimiçi Casinolarının Listesini Yayınladı

Casino.Guide đăng danh sách các sòng bạc trực tuyến tốt nhất ở Canada vào tháng 3 năm 2024

Casino.Guide 发布 2024 年 3 月加拿大最佳在线赌场名单

This paper is available on arxiv under CC 4.0 license. Authors: (1) Matu´s Medo, Department for BioMedical Research, Inselspital, Bern University Hospital, University of Bern, Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern and matus.medo@unifr.ch; (2)  Michaela Medova, Department for BioMedical Research, Inselspital, Bern University Hospital, University of Bern, and Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern. Table of Links Abstract & Introduction Results Discussion Materials and methods Acknowledgement & References Supporting Information Acknowledgement We thank Charlotte Ng and Peter Degen for their useful feedback. References Alexandrov, L.B., Kim, J., Haradhvala, N.J., Huang, M.N., Tian Ng, A.W., Wu, Y., Boot, A., Covington, K.R., Gordenin, D.A., Bergstrom, E.N., et al., 2020. The repertoire of mutational signatures in human cancer. Nature 578, 94–101. Alexandrov, L.B., Nik-Zainal, S., Wedge, D.C., Aparicio, S.A., Behjati, S., Biankin, A.V., Bignell, G.R., Bolli, N., Borg, A., Børresen-Dale, A.L., et al., 2013. Signatures of mutational processes in human cancer. Nature 500, 415–421. Berger, B., Yu, Y.W., 2022. Navigating bottlenecks and trade-offs in genomic data analysis. Nature Reviews Genetics , 1–16. Blokzijl, F., Janssen, R., van Boxtel, R., Cuppen, E., 2018. MutationalPatterns: Comprehensive genome-wide analysis of mutational processes. Genome Medicine 10, 1–11. Brady, S.W., Gout, A.M., Zhang, J., 2022. Therapeutic and prognostic insights from the analysis of cancer mutational signatures. Trends in Genetics 38, 194–208. Cannataro, V.L., Mandell, J.D., Townsend, J.P., 2022. Attribution of cancer origins to endogenous, exogenous, and preventable mutational processes. Molecular Biology and Evolution 39, msac084. Degasperi, A., Zou, X., Dias Amarante, T., Martinez-Martinez, A., Koh, G.C.C., Dias, J.M., Heskin, L., Chmelova, L., Rinaldi, G., Wang, V.Y.W., et al., 2022. Substitution mutational signatures in whole-genome–sequenced cancers in the UK population. Science 376, abl9283. Diaz-Gay, M., Vangara, R., Barnes, M., Wang, X., Islam, S.A., Vermes, I., Narasimman, N.B., Yang, T., Jiang, Z., Moody, S., et al., 2023. Assigning mutational signatures to individual samples and individual somatic mutations with SigProfilerAssignment. bioRxiv , 548264doi:10. 1101/2023.07.10.548264. bioRxiv 2023.07.10.548264. Gori, K., Baez-Ortega, A., 2020. sigfit: Flexible Bayesian inference of mutational signatures. bioRxiv , 372896doi:10.1101/372896. bioRxiv, 372896. Gulhan, D.C., Lee, J.J.K., Melloni, G.E., Cortes-Ciriano, I., Park, P.J., 2019. Detecting the mutational signature of homologous recombination deficiency in clinical samples. Nature Genetics 51, 912–919. Hubschmann, D., Jopp-Saile, L., Andresen, C., Kr ¨ amer, S., Gu, Z., Heilig, C.E., Kreutzfeldt, S., Teleanu, V., Fr ¨ ohling, S., Eils, R., et al., 2021. ¨ Analysis of mutational signatures with yet another package for signature analysis. Genes, Chromosomes and Cancer 60, 314–331. Islam, S.A., D´ıaz-Gay, M., Wu, Y., Barnes, M., Vangara, R., Bergstrom, E.N., He, Y., Vella, M., Wang, J., Teague, J.W., et al., 2022. Uncovering novel mutational signatures by de novo extraction with SigProfilerExtractor. Cell Genomics 2, 100179. Kim, Y.A., Leiserson, M.D., Moorjani, P., Sharan, R., Wojtowicz, D., Przytycka, T.M., 2021. Mutational signatures: From methods to mechanisms. Annual Review of Biomedical Data Science 4, 189–206. Koh, G., Degasperi, A., Zou, X., Momen, S., Nik-Zainal, S., 2021. Mutational signatures: Emerging concepts, caveats and clinical applications. Nature Reviews Cancer 21, 619–637. Li, S., 2020. sigLASSO: Optimizing cancer mutation signatures jointly with sampling likelihood. URL: https://github.com/gersteinlab/ siglasso. R package version 1.1. Li, S., Crawford, F.W., Gerstein, M.B., 2020. Using sigLASSO to optimize cancer mutation signatures jointly with sampling likelihood. Nature Communications 11, 3575. Maura, F., Degasperi, A., Nadeu, F., Leongamornlert, D., Davies, H., Moore, L., Royo, R., Ziccheddu, B., Puente, X.S., Avet-Loiseau, H., Campbell, P.J., Nik-Zainal, S., Campo, E., Munshi, N., Bolli, N., 2019. A practical guide for mutational signature analysis in hematological malignancies. Nature Communications 10, 1–12. Missa, O., Dytham, C., Morlon, H., 2016. Understanding how biodiversity unfolds through time under neutral theory. Philosophical Transactions of the Royal Society B: Biological Sciences 371, 20150226. Nik-Zainal, S., Alexandrov, L.B., Wedge, D.C., Van Loo, P., Greenman, C.D., Raine, K., Jones, D., Hinton, J., Marshall, J., Stebbings, L.A., et al., 2012. Mutational processes molding the genomes of 21 breast cancers. Cell 149, 979–993. Nik-Zainal, S., Davies, H., Staaf, J., Ramakrishna, M., Glodzik, D., Zou, X., Martincorena, I., Alexandrov, L.B., Martin, S., Wedge, D.C., et al., 2016. Landscape of somatic mutations in 560 breast cancer whole-genome sequences. Nature 534, 47–54. Omichessan, H., Severi, G., Perduca, V., 2019. Computational tools to detect signatures of mutational processes in DNA from tumours: A review and empirical comparison of performance. PLoS ONE 14, e0221235. Pandey, P., Arora, S., Rosen, G.L., 2022. MetaMutationalSigs: Comparison of mutational signature refitting results made easy. Bioinformatics 38, 2344–2347. Riva, L., Pandiri, A.R., Li, Y.R., Droop, A., Hewinson, J., Quail, M.A., Iyer, V., Shepherd, R., Herbert, R.A., Campbell, P.J., et al., 2020. The mutational signature profile of known and suspected human carcinogens in mice. Nature Genetics 52, 1189–1197. Rosenthal, R., McGranahan, N., Herrero, J., Taylor, B.S., Swanton, C., 2016. DeconstructSigs: Delineating mutational processes in single tumors distinguishes DNA repair deficiencies and patterns of carcinoma evolution. Genome Biology 17, 1–11. Rustad, E.H., Nadeu, F., Angelopoulos, N., Ziccheddu, B., Bolli, N., Puente, X.S., Campo, E., Landgren, O., Maura, F., 2021. mmsig: A fitting approach to accurately identify somatic mutational signatures in hematological malignancies. Communications Biology 4, 1–12. Schumann, F., Blanc, E., Messerschmidt, C., Blankenstein, T., Busse, A., Beule, D., 2019. SigsPack, A package for cancer mutational signatures. BMC Bioinformatics 20, 1–9. Van Hoeck, A., Tjoonk, N.H., van Boxtel, R., Cuppen, E., 2019. Portrait of a cancer: Mutational signature analyses for cancer diagnostics. BMC Cancer 19, 1–14. Wang, S., Li, H., Song, M., Tao, Z., Wu, T., He, Z., Zhao, X., Wu, K., Liu, X.S., 2021. Copy number signature analysis tool and its application in prostate cancer reveals distinct mutational processes and clinical outcomes. PLoS Genetics 17, e1009557. Wu, Y., Chua, E.H.Z., Ng, A.W.T., Boot, A., Rozen, S.G., 2022. Accuracy of mutational signature software on correlated signatures. Scientific Reports 12, 1–12. This paper is available on under CC 4.0 license. Arxiv

The People We Have to Thank In Our Comparison of Tools for Fitting Mutational Signatures

This paper is available on arxiv under CC 4.0 license. Authors: (1) Matu´s Medo, Department for BioMedical Research, Inselspital, Bern University Hospital, University of Bern, Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern and matus.medo@unifr.ch; (2)  Michaela Medova, Department for BioMedical Research, Inselspital, Bern University Hospital, University of Bern, and Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern. Table of Links Abstract & Introduction Results Discussion Materials and methods Acknowledgement & References Supporting Information 4. Materials and methods 4.1. Reference signatures We used all signatures from version 3.0 of the COSMIC catalog (COSMICv3, https://cancer.sanger.ac. uk/signatures/sbs/), in particular the SBS signatures for human genome assembly GRCh38. COSMICv3 contains 67 mutational signatures (18 are artifact signatures) defined in 96 mutational contexts. While newer COSMIC versions (the latest version is 3.3.1 from June 2022) added more signatures, profiles of the COSMICv3 signatures have not been altered (all absolute weight differences are below 10−7 ). To measure the diversity of signature profiles, we compute their exponentiated Shannon index which is a common measure of diversity (Missa et al., 2016) that can be understood as the effective number of active mutational contexts. For a signature whose all weight is concentrated in one context, the exponentiated Shannon index is one. For a signature with the same weight 1/96 of all contexts, the exponentiated Shannon index is 96. For COSMICv3 signatures, the exponentiated Shannon index ranges from 2.7 (SBS48) to 80.3 (SBS3). 4.2. Synthetic mutational catalogs Absolute contributions of signatures to sequenced tissues from various cancer types were downloaded from the Catalogue Of Somatic Mutations In Cancer (COSMIC, https://cancer.sanger.ac.uk/signatures/sbs/, catalog version 3.2) on November 14, 2021 for all 46 non-artifact SBS signatures that have tissue contribution data available. These data included only samples with reconstruction accuracy above 0.90 and at least 10 mutations attributed to the specific SBS signature. Using the provided unique sample identifiers, we used these files to compute the relative signature contributions in individual WGS-sequenced samples that are stratified by the cancer type. We chose eight different cancer types for further analysis: Hepatocellular carcinoma (Liver-HCC, n = 422), stomach adenocarcinoma (Stomach-AdenoCA, n = 170), head and neck squamous cell carcinoma (Head-SCC, n = 57), colorectal carcinoma (ColoRect-AdenoCA, n = 72), lung adenocarcinoma (Lung-AdenoCA, n = 62), cutaneous melanoma (Skin-Melanoma, n = 280), non-Hodgkin Lymphoma (Lymph-BNHL, n = 117), and glioblastoma (CNS-GBM, n = 63). The approach described above allows us to reproduce empirical signature weights in previously analyzed samples without resorting to assumptions such as a log-normal distribution of the number of mutations due to a given signature (Alexandrov et al., 2020, Islam et al., 2022) or adding additional zeros to the Poisson distribution to reproduce signatures that are not active in a sample (Omichessan et al., 2019). The code for generating synthetic mutational catalogs and evaluating the signature fitting tools is available at https://github.com/8medom/SigFitTest. For Figure 6, we created synthetic mutational catalogs with a systematic difference in the activity of signature SBS40. After generating sample weights as described in the previous paragraph (assuming the CNS-GBM cancer type), the relative weights of SBS40 were multiplied by 1.3 in the first half of the samples and divided by 1.3 in the second half of the samples. The relative weights of all signatures in each sample were then normalized to one. Note that signature SBS40 is active in nearly all CNS-GBM samples (Fig. SF9 in SI) and its median relative weight is close to 50%. By the described process, the median weights in the two groups of samples become 42% and 55%, respectively. 4.3. Tools for signature fitting The task of fitting known mutational signatures to a mutational catalog consists of finding the combination of signatures from a reference set that “best” matches the given catalog. Denoting the matrix with reference signatures as S, where S cs is the relative weight of context c for signature s, and the given normalized mutational catalog as m, where mci is the fraction of mutations in context c in sample i, this amounts to solving where wsi is the relative weight of signature s in sample i. Mathematically, this is an over-determined system as the number of reference signatures is smaller than the number of mutational contexts (96 contexts for the common SBS signatures). Several fitting tools are therefore based on minimizing the difference between m and Sw through nonnegative least squares (as the signature weights cannot be negative) or quadratic programming. We evaluated several tools that belong to this class: MutationalPatterns v3.4.1 (Blokzijl et al., 2018), YAPSA v1.20.1 (Hubschmann et al. ¨ , 2021), SigsPack v1.8.0 (Schumann et al., 2019), and sigminer v2.2.0 (Wang et al., 2021) which has three separate methods based on quadratic programming, non-linear least squares, and simulated annealing. We find that all these tools produce similar results. Other tools use various iterative processes by which the provided set of reference signatures is gradually reduced by removing the signatures, for example, whose inclusion does not considerably improve the match between the observed and reconstructed mutational catalogs (or, opposite, signatures are gradually added as long as the reconstruction accuracy sufficiently improves). We evaluated deconstructSigs v1.8.0 (Rosenthal et al., 2016), SigProfilerSingleSample v0.0.0.27 (Alexandrov et al., 2020), SigProfilerAssignment v0.0.30 (Diaz-Gay et al., 2023), mmsig v0.0.0.9000 (Rustad et al., 2021), and signature.tools.lib v2.2.0 (Degasperi et al., 2022), that all belong to this category. The newest tool, SigProfilerAssignment, combines backward and forward iterative adjustment of the reference catalog and these steps are repeated until convergence. Finally, SigLASSO v1.1 combines the data likelihood in a generative model with L1 regularization and prior knowledge (Li, 2020, Li et al., 2020). To allow for a fair comparison with other tools, we did not use prior knowledge when evaluating sigLASSO. In our experience, this tool sometimes fails to halt but starting it again with the same input data resolves the issue. A similar Bayesian framework is used by sigfit v2.2 (Gori and Baez-Ortega, 2020). The authors of sigfit recommend setting all signatures whose lower bounds of the estimated 95% confidence intervals are below 0.01 to zero. The authors of deconstructSigs recommend setting all signatures whose relative weights are below 0.06 to zero. All results shown for sigfit and deconstructSigs follow these recommendations. 4.4. Trimming reference signatures It has been argued that the removal of irrelevant reference signatures can improve the fitting results (Kim et al., 2021). To assess this hypothesis, we use a two-step process where: (1) We fit signatures using all COSMICv3 signatures as a reference and (2) keep only the signatures whose relative weight exceeds threshold w0 for at least 5 samples in our cohorts with 100 samples. We use thresholds w0 = 0.05 (m = 100), w0 = 0.03 (m = 2, 000) and w0 = 0.01 (m = 50, 000) which correspond to absolute signature contributions 5, 60, and 500, respectively. This pruning of reference signatures is beneficial when the number of mutations per sample is small (see m = 100 in Fig. S17 in SI). However, the converse is true when the number of mutations per sample is high, in particular for tools that perform well with COSMICv3 as a reference (m = 50, 000 in Fig. S17 in SI). To overcome this problem, we have also tested a less strict pruning where we keep all signatures whose relative weight exceeds w0 for at least one sample (see Fig. S18 in SI for the results). We have also tested the approach based on de novo extraction of signatures as suggested in (Maura et al., 2019). For signature extraction, we used SignatureAnalyzed v0.0.7 (https://github.com/getzlab/SignatureAnalyzer) with L1 prior on both W and H and a Poisson objective function for optimization. As recommended in (Maura et al., 2019), each extracted signature was then compared with individual signatures from COSMICv3 as well as linear combinations of two signatures from COSMICv3. The signature (or a pair of signatures) that yielded the smallest cosine distance was then added to the list of trimmed reference signatures for the given input data. When the number of mutations is small (m = 100), less than five signatures are selected, on average. When the number of mutations is large (m = 50, 000), 89% of the active signatures are selected, on average. A comparison of the results obtained by the two reference-trimming schemes is shown in Fig. S19 in SI. 4.5. Evaluation metrics Recent literature often classifies fitting results as true positives, false positives, true negatives, and false negatives, which allows computing evaluation metrics such as precision and specificity Omichessan et al. (2019), Islam et al. (2022), Diaz-Gay et al. (2023). We use instead quantitative metrics that measure the differences between the true relative signature weights (which are used to generate the input mutational catalogs) and the estimated relative signature weights. For tools that estimate absolute signature weights, those are first normalized by the number of mutations in each sample. and average it over all samples in the cohort to obtain fitting error. The lowest fitting error, 0, is achieved when the estimated signature weights are exact for all signatures and all samples. The highest fitting error, 1, is achieved when the estimated signature weights sum to one for each sample but they are all false positives. For example, when a sample has 40% contribution of SBS1 and 60% of SBS4 but a tool estimates 20% contribution of SBS2 and 80% contribution of SBS3, the fitting error is 1. There are some tools whose signature weight estimates do not sum up to one for each sample, typically as a result of the uncertainty implied by a small number of mutations in samples. The fitting error of those tools can be inflated due to these tools’ conservativeness. However, we probed whether additionally normalizing the estimated weights to one improves the fitting error achieved by those tools and found that this is not the case. We can thus conclude that these tools are not disfavored by measuring their fitting error. Another way to quantify the agreement between the true and estimated signature weights is by computing their Pearson correlation. For each sample where at least three signatures had positive either true or estimated weights, we computed the Pearson correlation coefficient between these two weight vectors whilst excluding all signatures that are zero for both of them. The obtained values were then averaged over all samples in the cohort. We finally computed false positive weight which quantifies how much weight is assigned to the signatures that are not active in a given sample, and then this quantity is averaged over all samples. This quantity too ranges from 0 (when no weight is assigned to inactive signatures) to 1 (when all weight is assigned to inactive signatures). This paper is available on under CC 4.0 license. arxiv

The Materials and Methods That We Used to Compare Tools for Fitting Mutational Signatures

This paper is available on arxiv under CC 4.0 license. Authors: (1) Matu´s Medo, Department for BioMedical Research, Inselspital, Bern University Hospital, University of Bern, Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern and matus.medo@unifr.ch; (2)  Michaela Medova, Department for BioMedical Research, Inselspital, Bern University Hospital, University of Bern, and Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern. Table of Links Abstract & Introduction Results Discussion Materials and methods Acknowledgement & References Supporting Information 3. Discussion The broad range of tools for fitting mutational signatures makes it difficult to understand which tool to choose for a given project. In this work, we provide a comprehensive assessment of eleven different tools on synthetic mutational catalogs. Using mutational catalogs where only one signature is active allows us to quantify the differences in the fitting difficulty of individual signatures. We find that flat signatures whose average similarity with other signatures is high are the most difficult to fit. To assess fitting tools, we use synthetic mutational catalogs whose signature activities are modeled based on eight distinct cancer types. We find that when the number of mutations is small (100 mutations per sample), SigProfilerSingleSample is the best tool to use for all cancer types. As the number of mutations increases, SigProfilerAssignment becomes the best tool for all cancer types with mmsig best for some cancer types for an intermediate number of mutations (2,000 mutations per sample). The results change little when Pearson correlation is used instead of fitting error for the evaluation (Figs. S22 and S23 in SI). The risk of overfitting the data by including too many signatures in the reference catalog is often discussed in the literature (Maura et al., 2019, Koh et al., 2021, Degasperi et al., 2022). We show that three methods (SigProfilerSingleSample, SigProfilerExtractor, and mmsig) are robust to such overfitting when the number of mutations per sample is not too small (Figs. S12 and S13 in SI). Crucially, we find that the common practice of excluding signatures from the reference catalog because they do not seem to be active (or are little active) in the analyzed data is little effective or even harmful, unless the number of mutations per sample is small (Fig. S16 in SI). In most cases it is preferable to let the built-in statistical methods of these tools to decide using the complete COSMIC catalog as a reference. While our work gives clear recommendations on which tools to consider and which to avoid, many issues can be addressed in the future to further improve the fitting of mutational signatures. First, a similar assessment of fitting tools can be done for other types of signatures: doublet base substitutions, small insertions and deletions, and copy number variations. Then, some tools (e.g., sigfit and mmsig) can compute confidence intervals for their estimates. For sigfit, we used these estimates in the way recommended by the authors: When the lower bound of the confidence interval for the relative signature weight is below 0.01, the signature is marked as absent in the sample (its relative weight is set to zero). Our results show that this recommended practice indeed improves the results achieved by sigfit. By bootstrapping (resampling the original mutational catalogs with replacement), confidence intervals could be computed also for the tools that do not compute them by themselves. It would be worth investigating whether thus-determined confidence intervals could also improve the performance of other fitting tools. To fit mutational signatures, it is commonly required that each sample has at least 50 (Riva et al., 2020) or 200 (Blokzijl et al., 2018) single base substitutions. However, different signatures are differently difficult to fit (Figure 1), so it is unlikely that such a universal threshold is meaningful. Based on the simulation framework established here, it would be possible to study the minimum necessary number of mutations for different signatures of interest and different cancer types. To best match the needs of practitioners, it would be possible to extend the simulation framework so that it finds the best-performing fitting tool for a given cancer type, a list of signatures of interest, and a given distribution of sample mutation counts. When an active signature is missing from the reference catalog because it is new or is falsely deemed inactive in a given set of samples, most fitting tools cannot cope with this situation and distribute all (or nearly all) mutations among signatures from the reference set (Fig. S15 in SI). Three tools do not suffer from this drawback—deconstructSigs, signature.tools.lib, and sigfit—but their performance is generally weak. One could aim to enhance one of the bestperforming tools, such as SigProfilerAssignment, with an additional step that would quantify whether given signatures are indeed likely to produce given mutational catalogs. Another approach to this problem is offered in (Maura et al., 2019) where the authors suggest a multistep process to simultaneously avoid under- and over-fitting. Mutational signatures are first extracted de novo, then they are assigned to existing signatures from a reference catalog, and thusdetermined active signatures are finally used to fit the original input data. While there are well-defined statistical principles for the first and third steps of this process, the second assignment step is largely ad hoc. A more principled approach could improve the outcome. More generally, COSMIC signatures are the result of analyzing tumor tissues from many different organs, so they can be viewed as an average across them (Degasperi et al., 2022). It might be more appropriate to construct organ-specific reference catalogs that would better reflect mutational processes that occur in these biologically diverse systems. The fast-growing number of sequenced samples would make such an effort possible. This paper is available on under CC 4.0 license. Arxiv

Discussing Tools for Fitting Mutational Signatures: A Comprehensive Comparison

Comparing Tools for Fitting Mutational Signatures: The Results of Our Work

This paper is available on arxiv under CC 4.0 license. Authors: (1) Matu´s Medo, Department for BioMedical Research, Inselspital, Bern University Hospital, University of Bern, Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern and matus.medo@unifr.ch; (2)  Michaela Medova, Department for BioMedical Research, Inselspital, Bern University Hospital, University of Bern, and Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern. Table of Links Abstract & Introduction Results Discussion Materials and methods Acknowledgement & References Supporting Information Abstract Mutational signatures connect characteristic mutational patterns in the genome with biological processes that take place in the tumor tissues. Analysis of mutational signatures can help elucidate tumor evolution, prognosis, and therapeutic strategies. Although tools for extracting mutational signatures de novo have been extensively benchmarked, a similar effort is lacking for tools that fit known mutational signatures to a given catalog of mutations. We fill this gap by comprehensively evaluating eleven signature fitting tools (well-established as well as recent) on synthetic input data. To create realistic input data, we use empirical signature weights in tumor tissue samples from the COSMIC database. The study design allows us to assess the effects of the number of mutations, type of cancer, and the catalog of reference signatures on the results obtained with various fitting tools. We find substantial performance differences between the evaluated tools. Averaged over 120,000 simulated mutational catalogs corresponding to eight different cancer types, SigProfilerSingleSample and SigProfilerAssignment perform best for small and large numbers of mutations per sample, respectively. We further show that ad hoc constraining the list of reference signatures is likely to produce inferior results and that noisy estimates of signature weights in samples with as few as 100 mutations can still be useful in downstream analysis. Cancer genomics, statistical methods, mutational signatures, computational models Keywords: 1. Introduction Since their introduction a decade ago by Nik-Zainal et al. (2012), Alexandrov et al. (2013), mutational signatures have become a widely used tool in genomics (Maura et al., 2019, Koh et al., 2021). They allow researchers to move from individual mutations in the genome to biological processes that take place in living tissues (Kim et al., 2021). In (Cannataro et al., 2022), for example, signature activities have been used to attribute mutations to endogenous, exogenous, and preventable mutational processes. The activity of various mutational signatures can also serve as prognostic or therapeutic biomarkers (Van Hoeck et al., 2019, Brady et al., 2022). Homologous recombination defficiency leads to the accumulation of DNA damage and manifests itself in a specific mutational signature (signature SBS3 from the COSMIC catalog) (Nik-Zainal et al., 2016, Gulhan et al., 2019). Mutational signatures can be introduced for single base substitutions (SBS), doublet base substitutions (DBS), small insertions and deletions (ID), as well as for copy number variations (CN). We focus here on SBS-based mutational signatures which are most commonly used in the literature. Current SBS signatures are defined using 6 possible classes of substitutions (C>A, C>G, C>T, T>A, T>C, and T>G) together with their two immediate neighboring bases, thus giving rise to 6 × 4 × 4 = 96 different nucleotide contexts into which all SBS mutations in a given sample are classified. De novo extraction of signatures from somatic mutations in sequenced samples has been used to gradually map the landscape of mutational signatures in cancer tissues. Over time, the initial catalogue of 22 SBS-based mutational signatures in the first version of the Catalogue Of Somatic Mutations In Cancer (COSMIC) released in August 2013 has expanded to 80 signatures in the COSMICv3.3 version released in June 2022. This expansion was possible owing to the increased availability of WES and WGS sequencing data as well as improved tools for signature extraction. Extensive benchmarking of extraction tools on synthetic data has recently shown that SigProfilerExtractor outperforms other methods in terms of sensitivity, precision, and false discovery rate, particularly in cohorts with more than 20 active signatures (Islam et al., 2022). A two-step process consisting of first extracting common signatures and then rare signatures has been recently recommended in (Degasperi et al., 2022). Nevertheless, with the existing signatures extracted from cohorts including thousands of sequenced samples (Alexandrov et al., 2020, Islam et al., 2022, Degasperi et al., 2022), the authors of smaller studies are unlikely to discover new signatures. For example, the analysis of more than 23,000 WGS and WES cancers in (Islam et al., 2022) has only discovered four new signatures. The much more relevant task is thus that of fitting existing signatures to given sequenced samples. In this process, the catalogs of somatic mutations are used to determine the signature contributions to each individual sample. Many different tools have been developed for this task (Koh et al., 2021) (see Methods for their description and classification). However, while tools for extracting mutational signatures have been recently extensively benchmarked on synthetic data by various studies (Omichessan et al., 2019, Alexandrov et al., 2020, Islam et al., 2022, Wu et al., 2022), a similar quantitative comparison is lacking for tools for fitting mutational signatures. This need is further exacerbated by substantial variations between results obtained by different methods (Pandey et al., 2022). Furthermore, newly introduced tools for fitting mutational signatures are commonly compared with only a few existing tools, rarely the most recent ones, and a standardized comparison methodology is lacking. In this study, our aim is to fill this gap and provide a comprehensive evaluation of a broad range of fitting tools on synthetic data motivated by various types of cancer. We constrain on fitting SBS signatures as they are the most widely used signature type. Many of the included tools can be nevertheless used for other types of signatures as their fitting is mathematically not different from the fitting of SBS signatures. We generate two classes of synthetic mutational catalogs. In the first one, only one mutational signature is responsible for all mutations (single-signature cohorts). In the second one, signature activities in each sample are modeled after the activities found in real tumor samples from various cancers (heterogeneous cohorts). We have collected eleven tools for fitting mutational signatures, from earlier tools such as deconstructSigs to very recent ones such as SigProfilerAssignment. We assess their performance by comparing the known true signature activities in the created catalogs with results obtained by various fitting tools. We find that there is no single fitting tool that performs best regardless of how many mutations are in the samples and which cancer type is chosen to model signature activities. Averaged across all cancer types, two related tools (SigProfilerSingleSample and SigProfilerAssignment) perform best when the number of mutations per sample is small (below 1000, roughly) and large (above 1000), respectively. We also compare the tools by how prone they are to overfitting caused by increasing the size of the reference signature catalog and evaluate whether it is beneficial to constrain the reference catalog based on which signatures seem to be absent or little active in the analyzed samples. We close with a discussion of open problems in fitting mutational signatures. Figure 1: (a, b) The bars show the 95th percentile range of the observed fraction of mutations in synthetic data for SBS mutational contexts (horizontal axis) and different total mutation counts (100, 1,000 and 10,000 represented with red, blue and gray bars, respectively). The effect of the number of mutations in single-signature cohorts. Panels a and b show signature SBS1 with four distinct C>T peaks and signature SBS5 with a flat mutational spectrum, respectively. (c) The fitting errors of the evaluated fitting tools for synthetic data corresponding to 100 samples with 100% contributions of a single signature; the catalog of all 67 COSMICv3 signatures was used for fitting. We show results for ten common signatures (horizontal axis) and various mutation counts (vertical axis). The tools are ordered by the error averaged over the displayed signatures achieved for 51,200 mutations. This paper is available on under CC 4.0 license. arxiv

A Comprehensive Comparison of Tools for Fitting Mutational Signatures

This paper is available on arxiv under CC BY-NC-SA 4.0 DEED license. Authors: (1) Pascal Baumann, Max Planck Institute for Software Systems (MPI-SWS), Germany; (2) Rupak Majumdar, Max Planck Institute for Software Systems (MPI-SWS), Germany; (3) Ramanathan S. Thinniyam, Max Planck Institute for Software Systems (MPI-SWS), Germany; (4) Georg Zetzsche, Max Planck Institute for Software Systems (MPI-SWS), Germany. Table of Links Introduction Dynamic Networks of Concurret Pushdown Systems (DCPS) Warm-up: Non-Termination Fair Non-Termination From Progressive Runs for VASSB to Reachability Starvation Conclusion, Acknowledgment & References A. Strengthening Fairness to Progressive Runs B. Proofs from Section 4.1 C. Proofs for Section 5 D. Proofs for Section 5.3 E. Proofs for Section 6 7 CONCLUSION ACKNOWLEDGMENTS This research was sponsored in part by the Deutsche Forschungsgemeinschaft project 389792660 TRR 248–CPEC and by the European Research Council under the Grant Agreement 610150 (http://www.impact-erc.eu/) (ERC Synergy Grant ImPACT). REFERENCES Parosh Aziz Abdulla, Karlis Čer ¯ ans, Bengt Jonsson, and Yih-Kuen Tsay. 1996. General decidability theo ¯ rems for infinitestate systems. In Proceedings of the Eleventh Annual Symposium on Logic in Computer Science. IEEE Computer Society Press, 313–321. Krzysztof R. Apt and Ernst-Rüdiger Olderog. 1991. Verification of Sequential and Concurrent Programs. Springer-Verlag. Mohamed Faouzi Atig, Ahmed Bouajjani, Michael Emmi, and Akash Lal. 2012a. Detecting Fair Non-termination in Multithreaded Programs. In Proceedings of CAV 2012. 210–226. https://doi.org/10.1007/978-3-642-31424-7_19 Mohamed Faouzi Atig, Ahmed Bouajjani, K. Narayan Kumar, and Prakash Saivasan. 2012b. Linear-Time Model-Checking for Multithreaded Programs under Scope-Bounding. In Automated Technology for Verification and Analysis - 10th International Symposium, ATVA 2012, Thiruvananthapuram, India, October 3-6, 2012. Proceedings (Lecture Notes in Computer Science), Vol. 7561. Springer, 152–166. Mohamed Faouzi Atig, Ahmed Bouajjani, K. Narayan Kumar, and Prakash Saivasan. 2017. Parity Games on Bounded Phase Multi-pushdown Systems. In Networked Systems - 5th International Conference, NETYS 2017, Marrakech, Morocco, May 17-19, 2017, Proceedings (Lecture Notes in Computer Science), Vol. 10299. Springer, 272–287. Mohamed Faouzi Atig, Ahmed Bouajjani, and Shaz Qadeer. 2009. Context-Bounded Analysis for Concurrent Programs with Dynamic Creation of Threads. In Proceedings of TACAS 2009. 107–123. Mohamed Faouzi Atig, Ahmed Bouajjani, and Shaz Qadeer. 2011. Context-Bounded Analysis For Concurrent Programs With Dynamic Creation of Threads. Log. Methods Comput. Sci. 7, 4 (2011). https://doi.org/10.2168/LMCS-7(4:4)2011 Georg Bachmeier, Michael Luttenberger, and Maximilian Schlund. 2015. Finite Automata for the Sub- and Superword Closure of CFLs: Descriptional and Computational Complexity. In 9th International Conference on Language and Automata Theory and Applications, LATA 2015, Nice, France, March 2-6, 2015, Proceedings. Springer, 473–485. Pascal Baumann, Rupak Majumdar, Ramanathan S. Thinniyam, and Georg Zetzsche. 2020. The Complexity of Bounded Context Switching with Dynamic Thread Creation. In 47th International Colloquium on Automata, Languages, and Programming, ICALP 2020, July 8-11, 2020, Saarbrücken, Germany (Virtual Conference) (LIPIcs), Vol. 168. Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 111:1–111:16. Julius Richard Büchi. 1964. Regular canonical systems. Archiv für mathematische Logik und Grundlagenforschung 6, 3-4 (1964), 91–111. Heino Carstensen. 1987. Decidability questions for fairness in Petri nets. In Proceedings of STACS 1987. Springer, 396–407. Byron Cook, Andreas Podelski, and Andrey Rybalchenko. 2007. Proving thread termination. In Proceedings of the ACM SIGPLAN 2007 Conference on Programming Language Design and Implementation, San Diego, California, USA, June 10-13, 2007. ACM, 320–330. Byron Cook, Andreas Podelski, and Andrey Rybalchenko. 2011. Proving program termination. Commun. ACM 54, 5 (2011), 88–98. https://doi.org/10.1145/1941487.1941509 Bruno Courcelle. 1991. On construction obstruction sets of words. EATCS 44 (1991), 178–185. Wojciech Czerwinski, Slawomir Lasota, Ranko Lazic, Jérôme Leroux, and Filip Mazowiecki. 2019. The reachability problem for Petri nets is not elementary. In Proceedings of STOC 2019. 24–33. Antoine Durand-Gasselin, Javier Esparza, Pierre Ganty, and Rupak Majumdar. 2017. Model checking parameterized asynchronous shared-memory systems. Formal Methods Syst. Des. 50, 2-3 (2017), 140–167. https://doi.org/10.1007/s10703-016-0258-3 Paul Erdős and Richard Rado. 1952. Combinatorial Theorems on Classifications of Subsets of a Given Set. Proceedings of the London Mathematical Society s3-2, 1 (01 1952), 417–439. https://doi.org/10.1112/plms/s3-2.1.417 Azadeh Farzan, Zachary Kincaid, and Andreas Podelski. 2016. Proving Liveness of Parameterized Programs. In Proceedings of the 31st Annual ACM/IEEE Symposium on Logic in Computer Science, LICS ’16, New York, NY, USA, July 5-8, 2016. ACM, 185–196. Alain Finkel and Philippe Schnoebelen. 2001. Well-structured transition systems everywhere! Theor. Comput. Sci. 256, 1-2 (2001), 63–92. Marie Fortin, Anca Muscholl, and Igor Walukiewicz. 2017. Model-Checking Linear-Time Properties of Parametrized Asynchronous Shared-Memory Pushdown Systems. In Computer Aided Verification - 29th International Conference, CAV 2017, Heidelberg, Germany, July 24-28, 2017, Proceedings, Part II (Lecture Notes in Computer Science), Vol. 10427. Springer, 155– 175. Pierre Ganty and Rupak Majumdar. 2012. Algorithmic verification of asynchronous programs. ACM Transactions on Programming Languages and Systems (TOPLAS) 34, 1 (2012), 6. Leonard H Haines. 1969. On free monoids partially ordered by embedding. Journal of Combinatorial Theory 6, 1 (1969), 94–98. David Harel. 1986. Effective transformations on infinite trees, with applications to high undecidability, dominoes, and fairness. J. ACM 33 (1986), 224–248. Rodney R Howell, Louis E Rosier, and Hsu-Chun Yen. 1991. A taxonomy of fairness and temporal logic problems for Petri nets. Theoretical Computer Science 82, 2 (1991), 341–372. Petr Jančar. 1990. Decidability of a temporal logic problem for Petri nets. Theoretical Computer Science 74, 1 (1990), 71–93. Vineet Kahlon. 2008. Parameterization as Abstraction: A Tractable Approach to the Dataflow Analysis of Concurrent Programs. IEEE Computer Society, 181–192. Sambasiva Rao Kosaraju. 1982. Decidability of Reachability in Vector Addition Systems (Preliminary Version). In STOC ’82: Proc. of 14th ACM symp. on Theory of Computing. ACM, 267–281. Bernhard Kragl, Constantin Enea, Thomas A. Henzinger, Suha Orhun Mutluergil, and Shaz Qadeer. 2020. Inductive sequentialization of asynchronous programs. In Proceedings of the 41st ACM SIGPLAN International Conference on Programming Language Design and Implementation, PLDI 2020, London, UK, June 15-20, 2020. ACM, 227–242. Akash Lal and Thomas W. Reps. 2009. Reducing concurrent analysis under a context bound to sequential analysis. Formal Methods in System Design 35, 1 (2009), 73–97. https://doi.org/10.1007/s10703-009-0078-9 Akash Lal, Tayssir Touili, Nicholas Kidd, and Thomas W. Reps. 2008. Interprocedural Analysis of Concurrent Programs Under a Context Bound. In Tools and Algorithms for the Construction and Analysis of Systems, 14th International Conference, TACAS 2008, Held as Part of the Joint European Conferences on Theory and Practice of Software, ETAPS 2008, Budapest, Hungary, March 29-April 6, 2008. Proceedings (Lecture Notes in Computer Science), Vol. 4963. Springer, 282–298. Jérôme Leroux, Grégoire Sutre, and Patrick Totzke. 2015. On the coverability problem for pushdown vector addition systems in one dimension. In ICALP 2015, Vol. 9135. 324–336. https://doi.org/10.1007/978-3-662-47666-6_26 Irina A. Lomazova and Philippe Schnoebelen. 1999. Some Decidability Results for Nested Petri Nets. In Perspectives of System Informatics, Third International Andrei Ershov Memorial Conference, PSI’99, Akademgorodok, Novosibirsk, Russia, July 6-9, 1999, Proceedings (Lecture Notes in Computer Science), Vol. 1755. Springer, 208–220. Ernst W. Mayr. 1981. An Algorithm for the General Petri Net Reachability Problem. In Proceedings of STOC 1981. 238–246. Anca Muscholl, Helmut Seidl, and Igor Walukiewicz. 2017. Reachability for Dynamic Parametric Processes. In Verification, Model Checking, and Abstract Interpretation - 18th International Conference, VMCAI 2017, Paris, France, January 15-17, 2017, Proceedings (Lecture Notes in Computer Science), Vol. 10145. Springer, 424–441. Madanlal Musuvathi and Shaz Qadeer. 2007. Iterative context bounding for systematic testing of multithreaded programs. In Proceedings of the ACM SIGPLAN 2007 Conference on Programming Language Design and Implementation, PLDI 2007, San Diego, CA, USA, June 10-13, 2007. ACM, 446–455. Oded Padon, Jochen Hoenicke, Giuliano Losa, Andreas Podelski, Mooly Sagiv, and Sharon Shoham. 2018. Reducing liveness to safety in first-order logic. Proc. ACM Program. Lang. 2, POPL (2018), 26:1–26:33. Rohit J. Parikh. 1966. On Context-Free Languages. J. ACM 13, 4 (1966), 570–581. Shaz Qadeer and Jakob Rehof. 2005. Context-Bounded Model Checking of Concurrent Software. In Tools and Algorithms for the Construction and Analysis of Systems, 11th International Conference, TACAS 2005, Held as Part of the Joint European Conferences on Theory and Practice of Software, ETAPS 2005, Edinburgh, UK, April 4-8, 2005, Proceedings (Lecture Notes in Computer Science), Vol. 3440. Springer, 93–107. Charles Rackoff. 1978. The covering and boundedness problems for vector addition systems. Theoretical Computer Science 6, 2 (1978), 223–231. Ganesan Ramalingam. 2000. Context-sensitive synchronization-sensitive analysis is undecidable. ACM TOPLAS 22(2) (2000), 416–430. Frank Plumpton Ramsey. 1930. On a Problem of Formal Logic. Proceedings of the London Mathematical Society s2-30, 1 (1930), 264–286. https://doi.org/10.1112/plms/s2-30.1.264 Salvatore La Torre, Margherita Napoli, and Gennaro Parlato. 2016. Scope-Bounded Pushdown Languages. Int. J. Found. Comput. Sci. 27, 2 (2016), 215–234. https://doi.org/10.1142/S0129054116400074 Moshe Y. Vardi. 1991. Verification of concurrent programs—the automata-theoretic framework. Annals of Pure and Applied Logic 51 (1991), 79–98. Kumar Neeraj Verma and Jean Goubault-Larrecq. 2005. Karp-Miller Trees for a Branching Extension of VASS. Discrete Mathematics & Theoretical Computer Science Vol. 7 (2005). Georg Zetzsche. 2013. Silent Transitions in Automata with Storage. (2013). arXiv:1302.3798 Full version of an article in Proceedings of ICALP 2013.

Stories, News & Mentions about NOVO

Novo’s Stories on HackerNoon