基于词汇时间分布的微博查询扩展

发布时间：2018-07-04 16:13

本文选题：微博检索 + 查询扩展　；参考：《计算机学报》2016年10期

【摘要】：该文提出了一种面向微博检索的基于词汇时间分布的查询扩展方法.该方法利用扩展词与查询词的时间分布的相似性来度量扩展词与查询词之间的相关度,建立了基于词汇时间分布的查询模型.具体而言,该文在提出词汇时间分布的定义和估计方法的基础上,给出了查询词与扩展词的时间分布相似性的度量,以此作为它们的相关度,完成扩展词的选择和查询模型的重估.该文方法利用时间信息而不是内容来扩展查询,避免了基于内容的查询扩展方法因微博内容短而无法准确估计扩展词的不足.由TREC 2011和TREC 2012微博检索评测数据上的实验结果表明,基于词汇时间分布的查询扩展模型有效地提高了微博检索的性能,不仅显著优于经典的基于内容的查询扩展模型,而且优于其他利用时间进行查询扩展的方法.
[Abstract]:In this paper, a query extension method based on word time distribution for Weibo retrieval is proposed. This method uses the similarity of time distribution between extended words and query words to measure the correlation between extended words and query words, and establishes a query model based on word time distribution. Specifically, on the basis of the definition and estimation method of lexical time distribution, this paper gives a measure of the similarity of temporal distribution between query words and extended words, which is regarded as their correlation degree. The selection of extended words and the revaluation of query model are completed. In this paper, time information rather than content is used to extend the query, which avoids the shortage of the method of query expansion based on content because of the short content of Weibo, which can not accurately estimate the extension words. The experimental results of TREC 2011 and TREC 2012 Weibo retrieval and evaluation data show that the query expansion model based on lexical time distribution can effectively improve the performance of Weibo retrieval, and it is not only superior to the classical content-based query expansion model, but also better than that of the classical content-based query extension model. Moreover, it is superior to other methods of query expansion using time.
【作者单位】：哈尔滨工业大学计算机科学与技术学院;黑龙江工程学院计算机科学与技术学院;
【基金】：国家自然科学基金(61370170,61402134,61173074) 国家社科基金(14CTQ032)资助~~
【分类号】：TP391.3

【相似文献】