Medical vocabulary mining using distributional semantics on Japanese patient blogs

要旨

Random indexing has previously been successfully used for medical vocabulary expansion for Germanic languages. In this study, we used this approach to ex- tract medical terms from a Japanese pa- tient blog corpus. The corpus was seg- mented into semantic units by a semantic role labeller, and different pre-processing and parameter settings were then evalu- ated. The evaluation showed that simi- lar settings are suitable for Japanese as for previously explored Germanic languages, and that distributional semantics is equally useful for semi-automatic expansion of Japanese medical vocabularies as for med- ical vocabularies in Germanic languages.

主題