Abstract

Unknown lexical items present a major obstacle to the development of broadcoverage semantic role labeling systems. We address this problem with a semisupervised learning approach which acquires training instances for unseen verbs from an unlabeled corpus. Our method relies on the hypothesis that unknown lexical items will be structurally and semantically similar to known items for which annotations are available. Accordingly, we represent known and unknown sentences as graphs, formalize the search for the most similar verb as a graph alignment problem and solve the optimization using integer linear programming. Experimental results show that role labeling performance for unknown lexical items improves with training data produced automatically by our method. 1

Citations