Allison Chaney will present her Pre-FPO on Wednesday, February 24, 2016 at 1:30pm in Room CS 302. The members of her committee are:  David Blei (advisor, reader), Barbara Engelhardt (reader), Brandon Stewart (Sociology), and Elad Hazan Everyone is invited to attend her talk.  The talk title and abstract follow below: Title: A Bayesian Model for Detecting and Characterizing Events Social scientists, such as historians and political scientists, commonly read large quantities of text to construct an accurate picture of who, what, when, and where an event happened.  In this talk, I will present a Bayesian model for analyzing documents to identify and characterize events of potential significance, where events are characterized by interactions between entities (e.g., countries, organizations, individuals) that deviate from typical interaction patterns. Specifically, the model builds on topic modeling to distinguish between topics that describe “business-as-usual” and topics that deviate from these patterns.  I will also describe details of approximate posterior inference for this model as well as visualization and evaluation of the results.  This talk includes ongoing work.