基于本体的产品命名实体识别研究  

Product Named Entity Recognition Based on Ontology

作　　者：罗芳[1] 熊前兴[1] 肖敏[1] 

机构地区：[1]武汉理工大学计算机科学与技术学院,湖北武汉430063

出　　处：《武汉理工大学学报（信息与管理工程版）》2011年第6期948-952,共5页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：中央高校基本科研业务费专项基金资助项目(2010-IV-023)

摘　　要：论述了近年来国内外在常规命名实体识别方面研究工作的进展状况,针对其中最为关键的产品命名实体识别技术,考虑到领域本体对产品命名实体识别的支持,提出了将本体特征融入到统计模型中,结合词性特征、上下文特征,以及本体特征的多特征模型进行产品命名实体识别实验,实验结果证明,该方法能有效地提高产品命名实体识别的性能。In recent years,there are a lot of research achievements in the field of NER （name entity recognition）, while PNER （product named entity recognition） is just starting. With the rapid development of e -commerce, PNER becomes one of the key techniques of intelligent e - commerce. In the view of supporting domain ontology to PNER, a method of PNER was proposed, which combined CRF（conditional random field） model with domain ontology as a semantic feature besides word and part -of- speech features. Experiments were conducted to compare the two kinds of feature templates, and results indicate that the method can effectively improve the performance of PNER.

关 键 词：产品命名实体识别 领域本体 条件随机域 特征选择 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]