基于云计算的微博舆情流式快速自聚类方法研究

发布时间：2018-01-07 07:37

本文关键词：基于云计算的微博舆情流式快速自聚类方法研究　出处：《情报科学》2017年08期 　论文类型：期刊论文

【摘要】：【目的/意义】当前微博已成为重要的舆论场,针对海量微博信息的舆情难以快速获取的问题,提出一种基于云计算的微博舆情流式快速自聚类方法。【方法/过程】该方法首先设计舆情最小距离聚类算法,包括构建舆情相似度计算模型,及构建舆情最佳聚类阈值确定方法;然后构建舆情流式自聚类模型,该模型利用云计算和最小距离聚类算法在横、纵两个方向聚类舆情信息,得到各主题的舆情集合。在横向上,以云计算的多个计算节点为聚类起始,同步并行聚类分配到其上的舆情信息。在纵向上,多个计算节点协同、流水线式聚类同一起始节点的舆情信息;最后在纵向上聚类舆情集合,将同主题舆情集合聚为一类。【结果/结论】实验结果表明:该方法能有效加快微博舆情获取速度,且具有较高的舆情获取准确率。
[Abstract]:[Objective] the significance / micro-blog has become an important field of public opinion, public opinion for the massive micro-blog information is difficult to quickly obtain the problem, propose a micro-blog public cloud based streaming fast clustering method. [method] the first design / process method of minimum distance clustering algorithm including the construction of public opinion, public opinion similarity calculation model and, to build public opinion the best clustering threshold determination method; and then construct public opinion flow from clustering model, the model uses the cloud computing and the minimum distance clustering algorithm in horizontal, longitudinal direction of the two clustering information of public opinion, public opinion has each topic set. On the other hand, many computing nodes in cloud computing for initial clustering, synchronization parallel clustering assigned to public opinion information on it. In the vertical direction, a plurality of collaborative computing nodes, in line with public opinion information clustering starting node; finally set vertically in the clustering of public opinion, The same topic public opinion collection is clustered into a class. [result / Conclusion] experimental results show that this method can effectively speed up the acquisition of micro-blog public opinion and has high accuracy of public opinion acquisition.

【作者单位】：河南牧业经济学院图书馆信息部;
【基金】：教育部人文社会科学研究项目(15YJC870004)
【分类号】：G206;TP391.1
【正文快照】： 微博是一种重要的社交网络平台,其以便捷、友好的交互优势广受大众青睐。目前,我国的微博注册用户量超过6亿,月活跃用户量超过2亿。由于允许快捷地编发、评论和分享微博信息,微博平台每天都有上百万条信息产生,已成为舆情产生和传播的重要源头,是我国最重要的网络舆论场之一【

【相似文献】