Semantic Role Labeling by Tagging Syntactic Chunks

Semantic Role Labeling by Tagging Syntactic Chunks∗ Kadri Hacioglu1, Sameer Pradhan1, Wayne Ward1, James H. Martin1, Daniel Jurafsky2 1University of Colorado at Boulder, 2Stanford University {hacioglu,spradhan,whw}@cslr.colorado.edu, martin@cs.colorado.edu, jurafsky@stanford.edu Abstract In this paper, we present a semantic role la- beler (or chunker) that groups syntactic chunks (i.e. base phrases) into the arguments of a pred- icate. This is accomplished by casting the se- mantic labeling as the classification of syntactic chunks (e.g. NP-chunk, PP-chunk) into one of several classes such as the beginning of an ar- gument (B-ARG), inside an argument (I-ARG) and outside an argument (O). This amounts to tagging syntactic chunks with semantic labels using the IOB representation. The chunker is realized using support vector machines as one- versus-all classifiers. We describe the represen- tation of data and information used to accom- plish the task. We participate in the “closed challenge” of the CoNLL-2004 shared task and report results on both development and test sets. 1 Introduction In semantic role labeling the goal is to group sequences of words together and classify them by using semantic la- bels. For meaning representation the predicate-argument structure that exists in most languages is used. In this structure a word (most frequently a verb) is specified as a predicate, and a number of word groups are considered as arguments accompanying the word (or predicate). In this paper, we select support vector machines (SVMs) (Vapnik, 1995; Burges, 1998) to implement the semantic role classifiers, due to their ability to han- dle an extremely large number of (overlapping) features with quite strong generalization properties. Support vec- tor machines for semantic role chunking were first used ∗This research was partially supported by the ARDA AQUAINT program via contract OCG4423B and by the NSF via grant IIS-9978025 in (Hacioglu and Ward, 2003) as word-by-word (W-by- W) classifiers. The system was then applied to the constituent-by-constituent (C-by-C) classification in (Ha- cioglu et al., 2003). In (Pradhan et al., 2003; Prad- han et al., 2004), several extensions to the basic system have been proposed, extensively studied and systemati- cally compared to other systems. In this paper, we imple- ment a system that classifies syntactic chunks (i.e. base phrases) instead of words or the constituents derived from syntactic trees. This system is referred to as the phrase- by-phrase (P-by-P) semantic role classifier. We partici- pate in the “closed challenge” of the CoNLL-2004 shared task and report results on both development and test sets. A detailed description of the task, data and related work can be found in (Carreras and M`arquez, 2004). 2 System Description 2.1 Data Representation In this paper, we change the representation of the original data as follows: • Bracketed representation of roles is converted into IOB2 representation (Ramhsaw and Marcus, 1995; Sang and Veenstra, 1995) • Word tokens are collapsed into base phrase (BP) to- kens. Since the semantic annotation in the PropBank corpus does not have any embedded structure there is no loss of information in the first change. However, this results in a simpler representation …

Content truncated to 500 words