Font Size: a A A

The Algorithm Analysis And System Realization Of Microblog Public Opinion

Posted on:2021-02-06Degree:MasterType:Thesis
Country:ChinaCandidate:L YangFull Text:PDF
GTID:2428330611451374Subject:Software engineering
Abstract/Summary:PDF Full Text Request
The rapid development of the network makes it more and more convenient for people to view various social platforms.At the same time,the amount of data generated every day is also considerable,and how to make good use of this data has become a research hotspot of scholars.Among many social platforms,microblog has gradually become one of the most popular social platforms in China due to its advantages of being more free and faster.Combined with various analysis methods based on text,this paper designs and implements a micro-blog public opinion analysis system built on B/S.The system can quickly show the dynamics of public opinions on the microblog platform,and obviously improve the timeliness and convenience of monitoring and understanding the development trend of the target public opinions.The detailed work mainly includes the following aspects:1.Data acquisition level.For the data collection of the system,this paper designed a distributed network crawler to collect the comments made by users under each microblog information.Secondly,it improved the anti-crawling mechanism and incremental crawling of data,which greatly improved the collection efficiency.2.System business logic layer,use two machine learning methods,namely the SVM algorithm,naive bayesian algorithm,and the four deep learning methods,namely,CNN,LSTM,CNN + LSTM,BiLSTM algorithm,has carried on the emotional polarity to users of public opinion text analysis,and on this basis to realize the keyword extraction and text summary,based on TextRank algorithm DBSCAN clustering algorithm and target function such as early warning.3.System implementation level.The micro-blog public opinion analysis system is designed at four levels,which are data acquisition layer,data storage layer,analysis layer and presentation layer.Compared with the traditional public opinion analysis system,the structure of this system is clearer,and the function division is clearer and easier to understand.4.At the level of system guarantee,the integrity and effectiveness of system function are verified from three aspects of system function,performance and availability,and the high efficiency and stability of the system are guaranteed.
Keywords/Search Tags:Microblog Public Opinion, Text Analysis, Deep Learning, Data Collection
PDF Full Text Request
Related items