在数字资源日益丰富的环境下,人名歧义现象为数据检索带来了很多不确定性,降低了数据检索的准确度.运用人名实体的个人信息特征和作者文献话题等特征,采用多特征的融合方法充分挖掘与实体相关联的信息,实现人名消歧.实验使用爬取的某数据库文献数据,进行了人名消歧特征的选取和参数的确定等实验,结果表明提出的方法具有较好的性能.%Under the environment of digital resource increasingly rich,personal name entity ambiguity phenomenon has brought a lot of uncertainty for data retrieval,and reduces the accuracy of the data retrieval.This paper used personal characteristics of personal name entity and subject characteristics of the literature,fully mined the information associated with the entity based on many characteristics fusion method,to achieve the personal name entity disambiguation.The experiments used the crawled literature data of a database,and researched on the feature selection of personal name entity disambiguation and determination of the parameters.The results show that the proposed method has good feasibility.
展开▼