W eb数据挖掘及其在微博话题检测中的应用研究

摘要

For the real⁃time characteristic of microblog content,the Single⁃Pass algorithm of topic clustering was improved. The time parameter is added to the method of microblog topic similarity detection,and the processing steps of the algorithm are given according to the processing method of the microblog forwarding characteristic. The simulation results show that the algo⁃rithm has the advantages of simple logic and high execution efficiency,and can effectively improve the accuracy of microblog topic detection based on Web data mining.%针对微博内容实时性的特点，对话题聚类的Single⁃Pass算法进行了改进，提出将时间参数添加到微博话题相似度检测方法中，并针对微博转发特性的处理方法给出了算法的处理步骤。仿真实验表明，该算法的优点是逻辑简单、算法执行效率高，通过算法的应用有效提高了基于Web数据挖掘的微博话题检测准确度。