摘要
提出了一种利用维基百科作为中介层的两阶段查询主题分类模型,对其实现方法做了详细的分析与设计。通过扩展查询获取与查询词主题相关的语义特征,并将查询与分类标签转移桥接在维基目录中以语义相关度判别分类标签,为实际应用提供了一条良好的设计与实现思路。实验结果表明,该方法能够有效解决查询分类中的数据稀疏问题,具有较好的分类准确率与召回率。
A query topic classifier with two-stage based on Wikipedia is analyzed and designed. The theme-related semantic features with query words are acquired by query expansion, and query are labeled by semantic relatedness when user queries and classification labels are transferred and bridged to wikipedia directories. A design and realization in practical application are provided. The experiment results show that the proposed method can be used to solve the data sparseness problem, and performs well on precision rate and recall rate in query classification.
出处
《情报科学》
CSSCI
北大核心
2014年第5期131-136,共6页
Information Science
关键词
查询分类
主题分类
维基百科
语义相关度
query classification
topic classification
wikipedia
semantic relatedness
