Research And Implementation Of Network Public Opinion Recognition System Based On Game Theory

The generation and dissemination of public opinion information based on online networks has the following three characteristics.First,the entry barrier of online networks is low,and the time and space costs required for users to produce public opinion information are relatively low;second,the characteristics of online networks determinepublic opinion information can be quickly disseminated on the network;third,the wide range of user groups have uneven identification capabilities for false public opinion content,which makes false public opinion information spread rapidly on the network platform.Therefore,studying the credibility recognition of public opinion information in online networks has important practical significance for effectively controlling the generation and dissemination of false public opinion information.In this thesis,the content data of the blockchain-based social network Steemit and the traditional social network Tumblr were obtained from the network by using the Python crawler tool.The social network analysis method was used to compare Steemit and Tumblr from the degree centrality,closeness centrality,eigenvector centrality and betweenness centrality.Through comparative analysis,this thesis finds out that the generation and transmission of public opinion information can be better controlled in Steemit.Therefore,this thesis selects the design mechanism of Steemit for the identification and analysis of online public opinion.Through the analysis of Steemit,this thesis finds two problems existing in Steemit: one is the unreasonable shareholding mechanism of user agent;Second,the lack of constraints on user voting rights,resulting in the abuse of the incentive mechanism.Therefore,this thesis improves the reward and punishment mechanism,and proves that the improved reward and punishment mechanism is effective in practical application through game theory.Then,the improved reward and punishment mechanism is coded and realized in the system.The network public opinion recognition system in this article is based on the mainstream technology framework SSH.Struts is used in the control layer.Spring is responsible for the centralized management of all classes in the project.The Hibernate framework interacts with the database and is responsible for data persistence.In the actual coding,a hierarchical design idea is adopted,Action is the logic layer;Bean is the entity layer,and the database corresponds to as many specific classes as there are entities;Dao layer is used as SQL statements,responsible for writing basic operations of functional design;Service layer for specific functional implementation methods.This article uses a lightweight database My Sql for data storage.The main users of this system are administrators and ordinary users,and the administrator is responsible for the management and maintenance of system affairs.Ordinary users have two roles.They can not only publish public opinion content in the system as content publishers,but also can vote on content in the system as content receivers.Constrain user behavior through incentive mechanisms,and encourage users to maximize their own interests.As a content publisher,only publish trusted public opinion content,and as a receiver,he will only correctly judge the credibility of public opinion content.Only credible and high-quality public opinion content exists in the public opinion recognition system.