网络舆情监控算法分析与研究

Analysis and Research of Network Public Opinion Monitoring Algorithms

摘要:

对现有的网络舆情监控算法进行梳理，将其归纳为经典算法和拓展算法，分析这些算法的优缺点和常见算法的性能对比，整理部分专家学者对这些监控算法的优化改进研究情况并对其研究成果进行分析。从技术、管理、应用等三个角度评价现有网络舆情监控算法的优化改进环节和取得的成效，并指出现有网络舆情监控算法存在的问题主要集中在只基于文本、针对单一数据类型、没有考虑事件和用户的差异化特点和动态变化情况、缺乏综合监控体系思维和管理机制等方面。进而从网络舆情的特点、发展规律、驱动因素、现有监控算法不足、监控效果期望等角度探讨未来网络舆情监控算法的发展趋势。

Abstract:

This paper reorganizes the current network public opinion monitoring algorithms, summarizes them into classical algorithms and extended algorithms, compares the advantages and disadvantages of these algorithms, also the capabilities of the famous algorithms, and reorganizes these monitoring algorithms’ improvements from some experts and scholars, and then analyzes their research reports. It evaluates the optimizations and improvements of these monitoring algorithms from the perspectives of technology, management and application. It also points out that the problems existing in these monitoring algorithms are mainly only for text and single data type, not considering the different characteristics and dynamic changes of events and users, lack of comprehensive monitoring system thought and management mechanism. It points out the tendencies of future network public opinion monitoring algorithms from the perspectives of network public opinions’ characteristics, development pattern, driving factors, the deficiencies of existing monitoring algorithms and expected effects.