Automatic Acquisition of Hyponyms from Large Text Corpora

This paper describes a method for the automatic acquisition of the hyponymy lexical relation from unrestricted text, motivated by two goals: avoiding the need for pre-encoded knowledge and being applicable across a wide range of text. It identifies a set of lexico-syntactic patterns that are easily recognizable, occur frequently and across text genre boundaries, and indisputably indicate the lexical relation of interest, and it describes a method for discovering these patterns, suggesting that other lexical relations could also be acquired in this way.

A subset of the acquisition algorithm is implemented, and its results are used to augment and critique the structure of a large hand-built thesaurus. The author suggests extensions and applications to areas such as information retrieval. By relying on frequently occurring, genre-independent patterns, the approach avoids dependence on pre-encoded knowledge while remaining broadly applicable across text.