A Cascaded Syntactic and Semantic Dependency Parsing System

We describe our CoNLL 2008 Shared Task system in this paper. The system includes two cascaded components: a syntactic and a semantic dependency parsers. A firstorder projective MSTParser is used as our syntactic dependency parser. In order to overcome the shortcoming of the MSTParser, that it cannot model more global information, we add a relabeling stage after the parsing to distinguish some confusable labels, such as ADV, TMP, and LOC. Besides adding a predicate identification and a classification stages, our semantic dependency parsing simplifies the traditional four stages semantic role labeling into two: a maximum entropy based argument classification and an ILP-based post inference. Finally, we gain the overall labeled macro F1 = 82.66, which ranked the second position in the closed challenge. 1 System Architecture Our CoNLL 2008 Shared Task (Surdeanu et al., 2008) participating system includes two cascaded components: a syntactic and a semantic dependency parsers. They are described in Section 2 and 3 respectively. Their experimental results are shown in Section 4. Section 5 gives our conclusion and future work. 2 Syntactic Dependency Parsing MSTParser (McDonald, 2006) is selected as our basic syntactic dependency parser. It views the c © 2008. Licensed under the Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported license (http://creativecommons.org/licenses/by-nc-sa/3.0/). Some rights reserved. syntactic dependency parsing as a problem of finding maximum spanning trees (MST) in directed graphs. MSTParser provides the state-ofthe-art performance for both projective and nonprojective tree banks.