基于相关性-兴趣度架构的关联规则挖掘的查询扩展

Query Expansion Based on Association Rules Mining Using Correlation-Interest Measure Framework

针对情报检索系统中存在的词不匹配问题,提出一种基于相关性-兴趣度架构的关联规则挖掘的局部反馈查询扩展算法,并论述查询扩展基本思想、扩展算法模型以及扩展词权值的计算方法。该算法主要特点是采用支持度-置信度-相关性-兴趣度框架衡量关联规则,避免产生负相关的、虚假的和无兴趣的规则,提高来自于关联规则的扩展词的质量。实验结果表明,该算法能有效地改善和提高信息检索性能,有很高的实际应用价值和推广前景。

Aiming at the term mismatch issues of existing information retrieval system, a novel query expansion algorithm of local feedback is proposed based on association rules mining under ton＇elation -interest measure framework. Its basic conception and algo- rithm as well as model are expounded, and a new computing method for weights of expansion terms is also expatiated. The framework of support-confidence-relevance-interest measure is used to judge association rules in the algorithm and negative-related and false as well as no interest association rules are avoided, to improve the quality of the expansion terms from the association rules. The results of the experiment show that the proposed algorithm is effective and improves the performance of irfformation retrieval with superior applied value of practice as well as popularizing prospect.