Chinese Partial Parsing with Modal Finite-State Approach

Abstract

A novel finite-state approach with the practical use of modal operators for Chinese partial parsing is presented in this paper. The traditional rule-based partial parsing approaches make use of regular grammar to approximate the context free or simplified context sensitive grammar with some loss of descriptive ability. Until now, how to solve the non-categorical problem within rule-based paradigm has not been reported, except for adding probability to grammars. In the approach presented in this paper, basic modal operators, necessity and possibility, are introduced into classic regular grammar. The technique of using such grammar to partially parse the free Chinese text is also discussed in detail. The experimental results suggest that the proposed approach can achieve Chinese partial parsing well and give a considerable performance improvement compared to traditional finite-state approach. A new application in natural language processing of modal logic is revealed through this paper.