Lecture: Topic Modeling in Natural Language Processing (NLP)

How can you understand the underlying theme in a collection of millions of documents? This is a common problem: examining through an organization’s e-mails, understanding a decade worth of newspaper articles, or classifying a scientific field’s research methodologies and findings.

Topic models are a statistical framework that helps users understand large document collections. Using topic modeling, you do not just find individual documents, but also understand the general themes present in the collection.

This session describes the use of topic modeling and constructing meaningful topic models that present an overview of the documents and the underlying sentiment attached to the documents.