Many datasets have substantial amounts of textual content, such as transcripts, archives, and blogs. The challenge of identifying the subjects or topics hidden in the textual bodies can be daunting. This workshop will teach the participants how to use the Stanford Topic Modeling Toolbox (TMT) to perform analysis on text collections, and will cover issues such as tool installation, input preparation, parameter tuning, and topic slicing. The goal of the workshop is to introduce not only a computer tool, but also the workflow of discovering topics in text collections. TMT is a text mining tool used by researchers in social sciences and humanities, but the workshop should also be interesting to people in other fields who want to analyze large amounts of text.