Corpus Annotation/Management Tools for the Project: Balanced Corpus of Contemporary Written Japanese

摘要

This paper introduces our activities on corpus annotation and management tool development in the Japanese government funded project, Balanced Corpus of Contemporary Written Japanese. We are investigating various levels of text annotation that covers morphological and POS tagging, syntactic dependency parsing, predicate-argument analysis, and coreference analysis. Since automatic annotation is not perfect, we need annotated corpus management tools that facilitate corpus browsing and error correction. We especially take up our corpus management tool ChaKi, explains its functions, and discuss how we are trying to maintain consistency of corpus annotation.