Evaluating syntactic topic models

• Grant number: 13/15353-6
• Support Opportunities: Scholarships abroad - Research Internship - Doctorate
• Start date: November 11, 2013
• End date: May 10, 2014
• Field of knowledge: Physical Sciences and Mathematics - Computer Science
• Principal Investigator: Alneu de Andrade Lopes
• Grantee: Thiago de Paulo Faleiros
• Supervisor: Jordan Boyd-Graber
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Institution abroad: University of Maryland, College Park, United States
• Associated to the scholarship: 11/23689-9 - Propagation in bipartite graphs for Topic Extraction in Data Streams, BP.DR

Abstract

We intend to evaluate the syntactic plausibility of Syntactic Topic Model (STM). The STMis a non-parametric Bayesian model of parsed documents. The STM generates words thatare both thematically and syntactically constrained, which combines the semantic insightsof topic models with the syntactic information available from parse tress. The expertise thatwe wish to obtain from STM study include the understand the non-parametric Bayesianmodelling, techniques of approximate posterior inference of probabilistic topics models,and evaluation methodology. As all this knowledge is important for the full developmentof the student doctorate project research. (AU)