文档介绍：
短文本查询扩展中扩展词间的关联性挖掘
刘悦,徐蔚然**
(北京邮电大学信息与通信工程学院,北京 100876)
5
10
15
20
25
30
35
40
摘要:本文针对经典查询扩展算法展开深入调研,分析了现阶段查询扩展方法所存在的缺陷。
提出了一种基于词激活力模型的扩展词间关联性挖掘算法。利用词激活力模型中词间亲密
度,计算扩展词间的关联性,得到扩展词对,并利用扩展词对进行查询重构。实验数据说明,
词激活力模型可以很好的对于词间关系建模,同时扩展词对可以有效的减少因扩展词引起的
信息偏移,同时提升检索系统的整体性能。
关键词:信息检索;查询扩展;扩展词对;词激活力模型;查询重构
中图分类号:TP391
Relevance exploration between short text query expansion
words
LIU Yue, XU Weiran
(School of Information munication Engineering,Beijing University of Posts and
munications, Beijing 100876)
Abstract: This paper mainly focuses on the relationship between expansion words. Based on the
analysis of traditional query expansion models, we find that query expansion sometimes give rise
to information shifting. In order to improve this situation, we propose a method to explore the
relevance between expansion words using word activation forces model (WAF). We generate
expansion-words-pairs by calculating affinity of different expansion words and reconstruct query
with these pairs of words. According to experiment data, it’s a wise choice using WAF modeling
the relationship between words and pairs of expanded words can effectively reduce shift caused by
expansion terms while enhancing the overall performance of the retrieval system.
Key words: Information Retrieval; Query Expansion; Expansion-words-pairs; Word Activation
Forces; Query Reconstruction
0