Classifying the Semantic Relations in Noun Compounds via a Domain-Specific Lexical Hierarchy

We are developing corpus-based techniques for identifying semantic relations at an intermediate level of description (more specific than those used in case frames, but more general than those used in traditional knowledge representation systems). In this paper we describe a classification algorithm for identifying relationships between two-word noun compounds. We find that a very simple approach using machine learning algorithms and a domain-specific lexical hierarchy successfully generalizes from training instances, performing better than a baseline consisting of training on the words themselves.