Discovery of Definition Patterns by Compressing Dictionary Sentences

This paper proposes an automatic method to discover definition patterns from an ordinary dictionary. There are frequent patterns to describe words and concepts in a ordinary dictionary. Each definition pattern gives a set of similar words and can be used as a template to clarify distinctions among them. To discover these definition patterns, we convert definition sentences into tree structures, and compress them using the MDL principle. The experiment on a Japanese children dictionary is reported, showing the effectiveness of our method.