为提高XML信息检索的查询准确率,提出一种基于词项扩展的XML信息检索反馈技术。利用词项所在节点的语义权重、词项与查询词间的相邻频度、共现程度,评估词项权重并排序,取权重较大的词项对初始检索词进行扩展,给出各因子的计算方法。在Wikipedia2009数据集上的实验结果表明,扩展后的查询准确率较高。
In order to improve the precision rate of XML information retrieval,this paper proposes a feedback technique for XML information retrieval based on term expansion.It considers lexical semantic weights of nodes,adjacency frequency between lexical items and queries,and local co-occurrence between terms and keywords in computing term weight.Experimental results on Wikipedia2009 data set show that the proposed technique can improve the precision rate.