Abstract / Description of output
Datasets annotated with semantic roles are an important prerequisite to developing high-performance role labeling systems. Unfortunately, the reliance on manual annotations, which are both difficult and highly expensive to produce, presents a major obstacle to the widespread application of these systems across different languages and text genres. In this paper we describe a method for inducing the semantic roles of verbal arguments directly from unannotated text. We formulate the role induction problem as one of detecting alternations and finding a canonical syntactic form for them. Both steps are implemented in a novel probabilistic model, a latent-variable variant of the logistic classifier. Our method increases the purity of the induced role clusters by a wide margin over a strong baseline.
Original language | English |
---|---|
Title of host publication | Human Language Technologies: Conference of the North American Chapter of the Association of Computational Linguistics, Proceedings, June 2-4, 2010, Los Angeles, California, USA |
Publisher | Association for Computational Linguistics |
Pages | 939-947 |
Number of pages | 9 |
Publication status | Published - 2010 |