| In recent years,micro-blog users show explosive growth,and micro-blog's guiding role in social public opinion is also increasingly significant.In order to ensure the users have a green and healthy network environment,micro-blog public opinion monitoring is indispensable,but with the expansion of the user group,the information of micro-blog is increasing exponentially.The traditional method of artificial summary analysis can not meet the actual needs,and the new method of visualization of micro-blog public opinion based on large data has come into being.Data visualization is an important way to help people to excavate the features contained in large-scale data.The significance of it is to communicate and communicate clearly and effectively with the help of graphical means.According to the characteristics of micro-blog data,this paper develops a visualization system for analysis of micro-blog public opinion,and designs a visual form suitable for displaying the situation of micro-blog public opinion.The main contents are as follows:First,micro-blog text preprocessing,using the simple word segmentation method to Chinese micro-blog text segmentation,combined with vector space model to express micro-blog text,transform the micro-blog text information into the mathematical form which can be processed by the computer and eliminate invalid information,thus further research.Secondly,the hot topic of micro-blog is extracted.According to the characteristics of micro-blog data,the Single-pass algorithm is improved from two aspects of the input sequence and the topic center,so as to achieve a fast and efficient extraction of the hot topic of micro-blog.Again,the analysis of micro-blog text emotion,this paper presents the emotional analysis of the introduction of expression,based on the traditional dictionary based affective analysis method,the expression is also used as a reference index for emotional analysis,in order to improve the accuracy of analysis.Finally,the visualization system of micro-blog public opinion is realized,and the visual display of micro-blog public opinion is carried out from many angles,such as users,topics and communication routes,and the validity of the proposed algorithm is verified. |