file.pdf (305.59 kB)
Download file

Learning Topics and Positions from Debatepedia

Download (305.59 kB)
journal contribution
posted on 01.10.2013, 00:00 by Swapna Gottipati, Minghui Qui, Yanchuan Sim, Jing Jiang, Noah A. Smith

We explore Debatepedia, a communityauthored encyclopedia of sociopolitical debates, as evidence for inferring a lowdimensional, human-interpretable representation in the domain of issues and positions. We introduce a generative model positing latent topics and cross-cutting positions that gives special treatment to person mentions and opinion words. We evaluate the resulting representation’s usefulness in attaching opinionated documents to arguments and its consistency with human judgments about positions.

History

Publisher Statement

Copyright 2013 Association for Computational Linguistics

Date

01/10/2013

Usage metrics

Exports