EVENT DETAILS

An Introduction to Topic Modeling

Topic modeling is a computational method of exploring what is sometimes described as "hidden" thematic or semantic structures in collections of texts. It is used by Digital Humanities researchers to explore textual data at scale and to offer a different perspective on data from which to gain new insights about the materials they are studying. This workshop will introduce the principles of topic modeling, along with the tools used for topic model construction in the Humanities and Humanities-inflected disciplines. Discussion will include the interpretation of topic models and the tools being created by the Mellon-funded WhatEvery1Says project to make topic modeling workflows more accessible to scholars and students in the Humanities.

The workshop will last two hours, with a third hour for those who wish to gain some hands-on experience in implementing topic models with their data. The workshop is geared for beginners, including students and faculty. No prior knowledge is assumed or required, but it is recommended that those who stay for the third hour bring along a laptop and, optionally, their own digital collections (some test collections will also be available). Workshop materials will be available here.

Scott Kleinman is one of the principal investigators for the WhatEvery1Says project, which uses topic modeling to study public discourse about the Humanities. A professor of English at California State University, Northridge, Kleinman is also project lead for the Lexomics project, which produces the online text-analysis tool Lexos.



Brain image courtesy of Alan Liu. Image source: Medical News Today.