Abstract:
To improve the satisfaction degree of Web page ranking, this paper proposes a new algorithm based on the feature selection technology. This algorithm utilizes multiple feature selection to choose a subset of a paper’s feature set, and selects those features with largest total importance scores and smallest total similarity scores. Experimental results on the public dataset in the region of information retrieval show that this algorithm has higher performance than traditional ranking algorithms.
Key words:
ranking,
feature selection,
information retrieval
摘要: 为提高网页排名满意度,基于特征选择技术提出一种新的页面排名算法。该算法利用多特征选择技术对页面的特征子集进行筛选,寻找一组具有最大权值与最小相似性的特征集合。在通用信息检索数据集上进行特征词选择测试,结果表明该算法的性能优于传统排名 算法。
关键词:
排名,
特征选择,
信息检索
CLC Number:
LI Hui, LI Cun-Hua, WANG Xia. Web Page Ranking Algorithm Based on Feature Selection[J]. Computer Engineering, 2010, 36(13): 37-39.
李慧, 李存华, 王霞. 基于特征选择的网页排名算法[J]. 计算机工程, 2010, 36(13): 37-39.