基于云计算的改进余弦向量度量法索引项权值算法

Improved Cosine Vector Measuring Method Indexing Term Weight Algorithm Based on Cloud Computing

针对用户对索引项重要程度无区分以及普通计算平台承载能力受限提出基于云计算的改进余弦向量度量法索引项权值算法(ICVMMITWCC算法),该算法通过从包含全部索引项的文本集中计算索引项平均权值对经典的余弦向量度量法索引项权值算法(CCVMMITW算法)修改求得相对权值;通过实验对比ICVMMITWCC算法与CCVMMITW算法下文本的排序效率,说明ICVMMITWCC算法更贴近用户查询需求.

Aiming the importance degree of indexing terms hadn＇t distinguish to users and the common computing platform carrying capacity limited proposed the improved cosine vector measuring method indexing term weight algorithm based on cloud computing （ICVMMITWCC algorithm）, the algorithm corrected the classic cosine vector measuring method indexing term weight algorithm （CCVMMITW algorithm） to acquire relative weight by computing indexing term average weight from texts set that included all indexing terms; comparing the efficiency of sorting of texts in ICVMMITWCC algorithm and CCVMMITW algorithm by experiment, indicating ICVMMITWCC algorithm is closer to the query requirement of users.