Improved Chinese Word Segmentation Disambiguation Model Based on Conditional Random Fields

摘要

This paper proposes an improved model that can eliminate the sense ambiguity of Chinese word segmentation based on conditional random fields (CRFs). First, this model segments words based on a bidirectional maximum matching algorithm and extracts the ambiguous part. Then it resolves ambiguity based on the conditional random field algorithm for segmentation ambiguity and outputs a more accurate result for the segmentation. The test results show that this model can reduce the error rate of segmentation caused by the ambiguity of word segmentation.