漢文の形態素解析・依存文法解析・直接構成鎖解析

Abstract

In this paper, the author describes syntactical analysis for classical Chinese texts : tokenization, part-of-speech tagging, dependency parsing, sentencization, and immediate catena analysis. MeCab, a morphological analyzer using Conditional Random Fields, is utilized for tokenization and part-of-speech tagging. UDPipe, a natural language processor using Universal Dependencies, is utilized for dependency parsing and sentencization. The author and his colleagues are now investigating an algorithm for immediate catena analysis.