A company’s unstructured data is often overlooked and definitely underutilized. But it doesn’t have to be that way. What if companies could instead apply the power of AI, Machine Learning, Natural Language Understanding (NLU), and Natural Language Processing (NLP) to automatically sift through this unstructured data and extract meaningful information?
NLP/NLU tools can help make this happen, helping companies achieve crucial actionable insights not accomplishable with human analysis alone. One of these NLU/NLP tools is Topic Detection.
Topic Detection APIs locate and label hundreds of topics in bodies of texts, typically in accordance with the standardized IAB Taxonomy. This article examines the best APIs on the market for performing Topic Detection in 2022.
AssemblyAI creates industry-leading
Its Topic Detection API, part of its Audio Intelligence offering, returns a topic and relevance key for audio and video files transcribed with its Speech-to-Text API. Pricing for Topic Detection starts at $.000583 per second of transcribed audio, in addition to its Core Transcription pricing.
__TextRazor __offers a competitive Topic Detection API called Topic Tagging. Trained on Wikipedia pages, Topic Tagging automatically identifies and categorizes topics in unstructured texts.
Users can utilize TextRazor’s Topic Tagging feature for free for up to 500 requests per day, with additional usage starting at $200/month.
The Topic Extraction API is free to use–users can follow the guides set out in thedocumentation to get started.
uClassify is free to use for up to 500 calls per day, with additional paid plans available as needed.
Pricing to use Amazon Comprehend varies widely based on which service is required and at what usage rate.
Finally, Microsoft Azure’s __Cognitive Services__offers Text Analytics, a Topic Detection feature for unstructured, static texts. Azure is one of the only APIs to also offer this advanced analysis for medical texts and terminology as well.
Like AWS, the pricing for Text Analytics varies greatly based on the type of API needed and the individual usage requirements.