针对信息检索中存在的词不匹配问题,提出了基于频繁项集和负关联规则挖掘的局部反馈查询扩展模型及其算法.该算法对前列n篇初检文档挖掘频繁项集和非频繁项集,并从频繁项集中提取关联词；从频繁项集和非频繁项集中挖掘负关联规则,提取负关联规则后件作为负关联词,计算负关联词与整个原查询词的相关性；根据相关性删除关联词库中与负关联词相同的词项,将余下的关联词项作为最终扩展词,并与原查询组合成新查询,实现查询扩展.实验结果表明,该算法能发现虚假的负关联词,有效地提高和改善信息检索性能.%Aiming at the term mismatch issues of existing information retrieval system, a novel query expansion model and its algorithm of local feedback is proposed based on frequent itemsets and negative association rules mining. Firstly, the frequent itemsets and non-frequent itemsets are mined synchronously in the top-ranked n chapter retrieved local documents. On one hand, the association terms are extracted from the frequent itemsets, on the other hand, negative association rules are mined in frequent itemsets and non-frequent itemsets and the consequents of negative association rules are extracted to make into negative association term. And then, final negative association terms are obtained according to the correlation of each negative association term and the entire original query. Finally, the terms the same as negative association terms are removed from association terms database and the rest of the terms of the association terms database are combined with original query for query expansion. The experimental results show that the proposed algorithm can not only detect those false negative association terms but also effectively improve and enhance the information retrieval performance.
展开▼
