In this session, you will learn how to use topic modeling to discover trends in large bodies of text. Topic models treat texts as collections of words about a mixture of topics. When given a set of texts, they can automatically discover topics that the texts discuss, and report what percentage of each text is devoted to each topic.

Because they discover topics without any additional human input, their output must be interpreted with care. This workshop will give you the background knowledge you'll need to make sense of topic model output. It will also show you ways to track the rise and fall of topics' popularity in a corpus over time, and to visualize the relationships between topics.