网络新词识别算法研究  

Research on algorithm for networks new words identification

作　　者：刘哲[1] 黄永峰[1] 罗芳[2] 陈跻[2] 王丙坤[1] 

机构地区：[1]清华大学电子工程系,北京100084 [2]华中科技大学计算机学院,湖北武汉430074

出　　处：《计算机工程与科学》2013年第9期141-145,共5页Computer Engineering & Science

基　　金：国家863计划资助项目(2012AA011004);清华大学自主科研项目基金(20111081023)

摘　　要：针对社交网络新词识别过程中"旧词新义"所引起的语义模糊问题,提出了网络新词识别算法。通过检测词语频度变化、共现词语分布一致性、情感倾向性迁移三项指标综合分析判断网络新词产生变化的规律特点,从而设计一种网络新词识别算法。最后以实验验证了该算法对提高现有系统网络新词识别准确率的可行性和有效性。Abstract：Aiming at the problem of ambiguity caused by Neologism in new words identification in so- cial networks, an algorithm for networks new words identification is proposed. The algorithm analyzes the changing patterns and features of networks new words by three indicators： frequency gain, co-occur- rence distribution consistency and emotional tendencies transference. Finally, real experiments validate the feasibility and effectiveness of the algorithm improving existing systems of new words identification.

关 键 词：社交网络 新词识别 准确率 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]