基于主题Hub值的元搜索

Topic Hub Based Meta Search Engine

为了提高元搜索引擎排序结果的质量,提出了成员引擎特征的主题Hub值表示和基于主题Hub值的结果排序算法.特征学习算法利用一组主题关联词对成员引擎的特征进行学习,并表示为主题Hub值的形式.排序算法根据主题Hub值计算结果的全局相关度对结果进行排序.实验结果表明,该模型取得了更好的排序质量.

To improve the ranking result quality of meta search engine,the authors propose and design a ranking algorithm based on the component engines by topic hub values.The algorithm uses a set of topic relating words to learn the feature of the component engines and denotes them as topic hub values.The ranking algorithm calculates the global similarity of the results according to the topic hub values.The experiments show that the model can get a better ranking quality.