Thursday, July 16, 2009

Digital Data and Japanese Language Research

I listened to this lecture by Tanomura last weekend. He stressed the importance of using a large corpus for studying linguistic phenomena, as a large sampling of data allows us to see patterns in language change more accurately. A corpus is also useful for seeing patterns in collocations. Search engines are not suitable for use as corpora, because the results they give are too variable.

Corpus:

日本語コーパス

No comments: