Text Processing and Sentiment Analysis of Twitter Data

A complete guide to text processing using Twitter data and R. Why Text Processing using R? With the increasing importance of computational text analysis in research , many researchers face the challenge of learning how to use advanced software that enables this text analysis. Currently, one of the most popular environments for computational methods and the emerging field of “data science” is the R statistical software. However, for researchers that are not well-versed in programming, learning how to use R can be a challenge, and performing text analysis in particular can seem daunting. In this step by step guide one can learn that performing text analysis in R is not as hard as some might fear. This is an introduction into the use of common techniques, with the aim of helping researchers get acquainted with computational text analysis in general, as well as getting a start at performing advanced text analysis studies in R. Why R? R was specifically designed for statistical analysis, which makes it highly suitable for data science applications. Although the learning curve for programming with R can be steep, especially for people without prior programming experience, the tools now available for carrying out text analysis in R make it easy to perform powerful, cutting-edge text analytics using only a few simple commands. One of the keys to R’s explosive growth has been its densely populated collection of extension software libraries, known in R terminology as packages, supplied and maintained by R’s extensive user community. Each package extends the functionality of the base R language and core packages, and in addition to functions and data must include documentation and examples, often in the form of vignettes demonstrating the use of the package. The best-known package repository, the Comprehensive R Archive Network (CRAN), currently has over 10,000 packages that are published. Text analysis in particular has become well established in R. There is a vast collection of dedicated text processing and text analysis packages, from low-level string operations to advanced text modeling techniques such as fitting Latent Dirichlet Allocation models, R provides it all. One of the main advantages of performing text analysis in R is that it is often possible, and relatively easy, to switch between different packages or to combine them. Recent efforts among the R text analysis developers’ community are designed to promote this interoperability to maximize flexibility and choice among users.4 As a result, learning the basics for text analysis in R provides access to a wide range of advanced text analysis features. Why Twitter Data? Twitter is an online microblogging tool that disseminates more than 400 million messages per day, including vast amounts of information about almost all industries from entertainment to sports, health to business etc. One of the best things about Twitter — indeed, perhaps its greatest appeal — is in its accessibility. It’s easy to use both for sharing information and for collecting it. Twitter provides unprecedented access to our lawmakers and to our celebrities, as well as to news as it’s happening. Twitter represents an important data source for the business models of huge companies as well. All the above characteristics make twitter a best place to collect real time and latest data to analyse and do any sought of research for real life situations. If you need other datasets, you can download pre-exiting datasets of various use cases like cancer detection to Q&A dataset to sports comments to chatbots. Here is various text classification datasets. Or if you have your unique use case, you can create your very own dataset for it. You can download images from the web and to make a big dataset in no time, use an annotation tool like , where you upload the raw data and tag manually in a ziffy. Dataturks Dataset: In order to extract data of Twitter we need to create a Twitter Application. STEPS TO CREATE A TWITTER APPLICATION 1.Navigate to My Applications. ( ) Link 2. Since I already have this app created, it appears on my page. Click on “create New App”. Fill in all the details of the application. 3. Once all the details are filled in and verified you will be granted the customer and access keys. INSTALL AND LOAD R PACKAGES: R comes with a standard set of packages. A number of other packages are and installation. For the purpose of this post, we will need the following packages: available for download ROAuth: Provides an interface to the OAuth 1.0 specification, allowing users to authenticate via OAuth to the server of their choice. Twitter: Provides an interface to the Twitter web API. Let’s start by installing and loading all the required packages. install.packages( ) install.packages( ) ( , lib.loc= ) ( , lib.loc= ) ( , lib.loc= ) ( , lib.loc= ) ( , lib.loc= ) ( , lib.loc= ) ( ) ( , lib.loc= ) ( ) ( ) "twitteR" "ROAuth" library "NLP" "~/R/win-library/3.3" library "twitteR" "~/R/win-library/3.3" library "syuzhet" "~/R/win-library/3.3" library "tm" "~/R/win-library/3.3" library "SnowballC" "~/R/win-library/3.3" library "stringi" "~/R/win-library/3.3" library "topicmodels" library "syuzhet" "~/R/win-library/3.3" library "twitteR" library "ROAuth" Connect your twitter account to R, in order to extract the required tweets. consumer_key <- consumer_secret <- access_token <- access_secret <- setup_twitter_oauth(consumer_key, consumer_secret, access_token, access_secret) 'YOUR CONSUMER KEY' 'YOUR CONSUMER SECRET KEY' 'YOUR ACCESS TOKEN' 'YOUR ACCESS SECRET TOKEN' Extracting tweets using a particular hashtag: tweets_g <- searchTwitter( , n= ,lang = ) tweets_a <- searchTwitter( , n= ,lang = ) tweets_f <- searchTwitter( , n= ,lang = ) tweets_tech <- searchTwitter( , n= ,lang = ) "#google" 1000 "en" "#amazon" 1000 "en" "#facebook" 1000 "en" "#technology" 1000 "en" Convert this extracted data to a dataframe which makes it more readable and easier to work with. amazon_tweets <- twListToDF(tweets_a) google_tweets <- twListToDF(tweets_g) facebook_tweets <- twListToDF(tweets_f) tech_tweets <- twListToDF(tweets_tech) View(amazon_tweets) View(google_tweets) View(facebook_tweets) View(tech_tweets) Here is how the data would look like: It is very clear that in the text section of the data, which is what we need to process there are a lot of special characters and unnecessary data which we would not require. Hence it becomes extremely important to pre-process this data and then we can continue with out analysis. Below is a code to pre-process the data and remove tabs, blank spaces, links etc. This section can be modified according to one’s requirements. google_text<- google_tweets$text amazon_text<- amazon_tweets$text facebook_text<- facebook_tweets$text tech_text<- tech_tweets$text google_text<- tolower(google_text) amazon_text<- tolower(amazon_text) facebook_text<- tolower(facebook_text) tech_text<- tolower(tech_text) google_text <- gsub( , , google_text) amazon_text <- gsub( , , amazon_text) facebook_text <- gsub( , , facebook_text) tech_text <- gsub( , , tech_text) google_text <- gsub( , , google_text) amazon_text <- gsub( , , amazon_text) facebook_text <- gsub( , , facebook_text) tech_text <- gsub( , , tech_text) google_text <- gsub( , , google_text) amazon_text <- gsub( , , amazon_text) facebook_text <- gsub( , , facebook_text) tech_text <- gsub( , , tech_text) google_text <- gsub( , , google_text) amazon_text <- gsub( , , amazon_text) facebook_text <- gsub( , , facebook_text) tech_text <- gsub( , , tech_text) google_text <- gsub( , , google_text) amazon_text <- gsub( , , amazon_text) facebook_text <- gsub( , , facebook_text) tech_text <- gsub( , , tech_text) google_text <- gsub( , , google_text) amazon_text <- gsub( , , amazon_text) facebook_text <- gsub( , , facebook_text) tech_text <- gsub( , , tech_text) google_text <- gsub( , , google_text) amazon_text <- gsub( , , amazon_text) facebook_text <- gsub( , , facebook_text) tech_text <- gsub( , , tech_text) #convert all text to lower case # Replace blank space (“rt”) "rt" "" "rt" "" "rt" "" "rt" "" # Replace @UserName "@\\w+" "" "@\\w+" "" "@\\w+" "" "@\\w+" "" # Remove punctuation "[[:punct:]]" "" "[[:punct:]]" "" "[[:punct:]]" "" "[[:punct:]]" "" # Remove links "http\\w+" "" "http\\w+" "" "http\\w+" "" "http\\w+" "" # Remove tabs "[ |\t]{2,}" "" "[ |\t]{2,}" "" "[ |\t]{2,}" "" "[ |\t]{2,}" "" # Remove blank spaces at the beginning "^ " "" "^ " "" "^ " "" "^ " "" # Remove blank spaces at the end " $" "" " $" "" " $" "" " $" "" A little more pre-processing — Removal of stop words! What are Stop Words? When working with text mining applications, we often hear of the term “stop words” or “stop word list” or even “stop list”. Stop words are basically a set of commonly used words in any language, not just English. The reason why stop words are critical to many applications is that, if we remove the words that are very commonly used in a given language, we can focus on the important words instead. Stop words are generally thought to be a . It really can mean different things to different applications. For example, in some applications removing all stop words right from determiners (e.g. the, a, an) to prepositions (e.g. above, across, before) to some adjectives (e.g. good, nice) can be an appropriate stop word list. To some applications however, this can be detrimental. For instance, in sentiment analysis removing adjective terms such as ‘good’ and ‘nice’ as well as negations such as ‘not’ can throw algorithms off their tracks. In such cases, one can choose to use a minimal stop list consisting of just determiners or determiners with prepositions or just coordinating conjunctions depending on the needs of the application. “single set of words” google_tweets.text.corpus <- tm_map(google_tweets.text.corpus, (x)removeWords(x,stopwords())) amazon_tweets.text.corpus <- tm_map(amazon_tweets.text.corpus, (x)removeWords(x,stopwords())) facebook_tweets.text.corpus <- tm_map(facebook_tweets.text.corpus, (x)removeWords(x,stopwords())) tech_tweets.text.corpus <- tm_map(tech_tweets.text.corpus, (x)removeWords(x,stopwords())) #clean up by removing stop words function function function function We are done pre-processing our data, and are ready to do some analysis. Data visualizations (like charts, graphs, infographics, and more) give analysts a valuable way to communicate important information at a glance. If you want a stunning visualization format to highlight important textual data points, using a word cloud can make dull data sizzle and immediately convey crucial information. What are Word Clouds? Word clouds (also known as text clouds or tag clouds) work in a simple way: the more a specific word appears in a source of textual data (such as a speech, blog post, or database), the bigger and bolder it appears in the word cloud. So let’s generate some word clouds and find out some of the frequent and important terms being used in the tweets we have extracted. ( ) wordcloud(google_tweets.text.corpus,min.freq = ,colors=brewer.pal( , ),random.color = ,max.words = ) wordcloud(amazon_tweets.text.corpus,min.freq = ,colors=brewer.pal( , ),random.color = ,max.words = ) wordcloud(facebook_tweets.text.corpus,min.freq = ,colors=brewer.pal( , ),random.color = ,max.words = ) wordcloud(tech_tweets.text.corpus,min.freq = ,colors=brewer.pal( , ),random.color = ,max.words = ) library "wordcloud" #generate wordcloud 10 8 "Dark2" TRUE 500 10 8 "Dark2" TRUE 500 10 8 "Dark2" TRUE 500 10 8 "Dark2" TRUE 500 Below are the word clouds for the data I have: Next I will be doing sentiment analysis on the tweets. Introduction to Sentiment Analysis What is Sentiment Analysis? Sentiment essentially relates to feelings; attitudes, emotions and opinions. Sentiment Analysis refers to the practice of applying Natural Language Processing and Text Analysis techniques to identify and extract subjective information from a piece of text. A person’s opinion or feelings are for the most part subjective and not facts. Which means to accurately analyze an individual’s opinion or mood from a piece of text can be extremely difficult. With Sentiment Analysis from a text analytics point of view, we are essentially looking to get an understanding of the attitude of a writer with respect to a topic in a piece of text and its polarity; whether it’s positive, negative or neutral. In recent years there has been a steady increase in interest from brands, companies and researchers in Sentiment Analysis and its application to business analytics. The business world today, as is the case in many data analytics streams, are looking for “business insight.” In relation to sentiment analysis, I am talking about insights into consumer behavior, what customers want, what are customers like and dislike about the products, what their buying signals are, what their decision process looks like etc because in the end of the its the customers for whose satisfaction these businesses work for. I have used the inbuilt sentiment analyzer in R, which uses the NRC sentiment dictionary to calculate the presence of eight different emotions and their corresponding valence in a text. mysentiment_google<-get_nrc_sentiment((google_text)) mysentiment_amazon<-get_nrc_sentiment((amazon_text)) mysentiment_facebook<-get_nrc_sentiment((facebook_text)) mysentiment_tech<-get_nrc_sentiment((tech_text)) Sentimentscores_google<-data.frame(colSums(mysentiment_google[,])) Sentimentscores_amazon<-data.frame(colSums(mysentiment_amazon[,])) Sentimentscores_facebook<-data.frame(colSums(mysentiment_facebook[,])) Sentimentscores_tech<-data.frame(colSums(mysentiment_tech[,])) names(Sentimentscores_google)<- Sentimentscores_google<-cbind( =rownames(Sentimentscores_google),Sentimentscores_google) rownames(Sentimentscores_google)<- names(Sentimentscores_amazon)<- Sentimentscores_amazon<-cbind( =rownames(Sentimentscores_amazon),Sentimentscores_amazon) rownames(Sentimentscores_amazon)<- names(Sentimentscores_facebook)<- Sentimentscores_facebook<-cbind( =rownames(Sentimentscores_facebook),Sentimentscores_facebook) rownames(Sentimentscores_facebook)<- names(Sentimentscores_tech)<- Sentimentscores_tech<-cbind( =rownames(Sentimentscores_tech),Sentimentscores_tech) rownames(Sentimentscores_tech)<- ggplot(data=Sentimentscores_google,aes(x=sentiment,y=Score))+geom_bar(aes(fill=sentiment),stat = )+ theme(legend.position= )+ xlab( )+ylab( )+ggtitle( ) ggplot(data=Sentimentscores_amazon,aes(x=sentiment,y=Score))+geom_bar(aes(fill=sentiment),stat = )+ theme(legend.position= )+ xlab( )+ylab( )+ggtitle( ) ggplot(data=Sentimentscores_facebook,aes(x=sentiment,y=Score))+geom_bar(aes(fill=sentiment),stat = )+ theme(legend.position= )+ xlab( )+ylab( )+ggtitle( ) ggplot(data=Sentimentscores_tech,aes(x=sentiment,y=Score))+geom_bar(aes(fill=sentiment),stat = )+ theme(legend.position= )+ xlab( )+ylab( )+ggtitle( ) #getting emotions using in-built function #calculationg total score for each sentiment "Score" "sentiment" NULL "Score" "sentiment" NULL "Score" "sentiment" NULL "Score" "sentiment" NULL #************************************************************************************* #plotting the sentiments with scores "identity" "none" "Sentiments" "scores" "Sentiments of people behind the tweets on tech giant GOOGLE" "identity" "none" "Sentiments" "scores" "Sentiments of people behind the tweets on ecomerce giant AMAZON" "identity" "none" "Sentiments" "scores" "Sentiments of people behind the tweets on Social Netwoking site FACEBOOK" "identity" "none" "Sentiments" "scores" "Sentiments of people behind the tweets on tech as a whole" From the following plot of sentiment analysis on tweets with #Google, its observable that Google has a positive impact on the consumers and is extremely trusted by them. There have lately been quite a good number of anticipations maybe on their products or the company employee’s or any other possibility. From the following plot of sentiment analysis on tweets with #Amazon, its observable that the E-Commerce giant, has maximum tweets being positive and similar to google has won the trust of consumers. Apart from these there is a notable amount of joy among their customers. The tweets with #Facebook show a little different trend. Lately, there seems to be a lot of negative sentiment associated with the tweets on facebook. But along with that there is also a trust which the company has maintained. In general, the tweets on technology have a lot of positivity and trust among the people, across the globe. These are only a few very obvious sentiments which I have described. You can always look at the visualizations and draw multiple conclusions about the data. Conclusion Text Processing and Sentiment analysis emerges as a challenging field with lots of obstacles as it involves natural language processing. It has a wide variety of applications that could benefit from its results, such as news analytics, marketing, question answering, readers do. Getting important insights from opinions expressed on the internet especially from social media blogs is vital for many companies and institutions, whether it is in terms of product feedback, public mood, or investors opinions. Sentiment analysis is a difficult technology to get right. However, when you do, the benefits are great. Look for a tool that has uses Natural Language Processing technology and ideally with machine learning capabilities. Look for a vendor that treats sentiment analysis seriously and shows advancements and updates in their sentiment analysis technology. Complete code.