Design And Implementation Of Weibo Public Opinion Analysis Based On Single-Pass Algorithm

As the Internet has become a necessity of people's life,online public opinions,especially those on weibo,have acquired a strong social influence and a very fast social communication speed.For the promotion of the government's ruling ability,completes the microblogging public opinion analysis is of great significance,it can make the government more easily to understand the current event hot spot of social time,mastering the public opinion and development trends of time,help topic to guide and control for all kinds of network,builds a brisk Internet environment,also collected the decision-making for the government to make relevant public opinion and other basic information.This paper carries out a detailed research on sina weibo-related network public opinion analysis technology,and on this basis designs and implements a micro-blog public opinion analysis system based on Single-Pass algorithm,including micro-blog collection,text preprocessing,micro-blog clustering,micro-blog cluster summary and public opinion information display and other modules.The main work of this paper includes:1.Collect and preprocess the microblog,customize the topic keywords through the microblog search box,and conduct targeted crawling on the search situation page;The NLPIR word segmentation system was used for Chinese word segmentation,word stopping and part of speech tagging.Then the space vector representation is transformed into computer-recognizable text.2.In the microblog clustering and cluster summary,the improved Single-Pass algorithm and tf-idf algorithm were selected for the clustering of microblog articles and the acquisition of the characteristic items and weights of microblog clusters,and an appropriate microblog was selected as the summary summary of microblog clusters.3.By demonstrating a topic cycle on the release of the amount of charts,according to the total number of similar information descending TOP5 arrangement and cluster,the related topic publisher fans more than 20000 users TOP5 in descending order effect,the microblogging public opinion on information more intuitive to present to the user,thus achieve the microblogging public opinion information monitoring and analysis of the final result.