Automatic Acquisition of GL Resources, Using an Explanatory, Symbolic Technique

This chapter presents a symbolic machine learning method that automatically infers, from descriptions of noun-verb pairs found in a corpus in which the verb plays (or not) one of the qualia roles of the noun, corpus-specific morpho-syntactic and semantic patterns that convey qualia relations. The patterns are explanatory and linguistically motivated, and can be applied to a corpus to efficiently extract GL resources and populate Generative Lexicons. The linguistic relevance of these patterns is examined, and the N-V qualia pairs that they can detect or not is discussed. Comparisons to other methods for corpus-based qualia couple extraction are also presented.