Paper: Learning Grammar with Explicit Annotations for Subordinating Conjunctions

ACL ID P14-3007
Title Learning Grammar with Explicit Annotations for Subordinating Conjunctions
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2014
Authors

Data-driven approach for parsing may suf- fer from data sparsity when entirely un- supervised. External knowledge has been shown to be an effective way to alleviate this problem. Subordinating conjunctions impose important constraints on Chinese syntactic structures. This paper proposes a method to develop a grammar with hierar- chical category knowledge of subordinat- ing conjunctions as explicit annotations. Firstly, each part-of-speech tag of the sub- ordinating conjunctions is annotated with the most general category in the hierar- chical knowledge. Those categories are human-defined to represent distinct syn- tactic constraints, and provide an appropri- ate starting point for splitting. Secondly, based on the data-driven state-split ap- proach, we establish a mapping from each automat...