An Improved Forward Maximum Matching Algorithm for Chinese Word Segmentation

摘要：

With the original Forward Maximum Matching algorithm (FMM), the initial value of the maximum word length is a constant As a result, long Chinese words cannot be segmented correctly and matched repeatedly. This paper presents an improved way for FMM to dynamically adjust the maximum word length. In addition, a new Chinese word segmentation lexicon is introduced which would work corporately with the improved algorithm. Compare with the original FMM, the improved algorithm dramatically reduces the number of times each segment is being matched. Furthermore, analysis has shown obvious improvement in both speed and efficiency for Chinese word segmentation.

关键词： forward maximum matching Chinese word segmentation lexicon structure Hash table

作者: Jing Luan Ruilei Wang Xiupei Lu

作者单位: College of the Computer Science and Technology Xinjiang Normal University 102 Xinyi Road, Urumqi, Ch College of the Computer Science and Technology Xinjiang Normal University 102 Xinyi Road,Urumqi,Chin

会议类型: 国际会议

会议名称: 2010 International Conference on Future Information Technology(2010年未来信息技术国际会议 ICFIT 2010)

会议地点: 长沙

会议语种:英文

页码: 991-994

在线出版日期: 2010-12-14（万方平台首次上网日期，不代表论文的发表时间）