智慧信息服务大数据框架下网络舆情信息集成应用研究

The Integrated Application of Internet Public Opinion Information under the Big Data Framework of Smart Information Service

【目的/意义】随着互联网及大数据的飞速发展,网络信息海量涌现,对网络舆情信息集成与分析成为及时应对用户网络行为规律以及优化政府决策的重要手段。【方法/过程】在分析了网络舆情信息集成的发展现状、研究意义以及主要研究方向的基础上,辨析了大数据及网络舆情的概念及特征。基于此,本文以智慧信息服务大数据框架下网络舆情信息为研究对象,结合日志分析法和统计分析法,采用Awstats软件大数据集成技术等对智慧信息服务大数据框架下网络舆情的主要特征进行全面分析;进一步利用回归系数显著性检验方法,获得网络用户的行为规律,并将其应用到优化政府决策中。【结果/结论】实证结果表明,描述性分析的回归系数显著性检验结果低于0.05,推断性分析的回归系数显著性检验结果低于0.1,相关性和聚类分析的回归系数显著性检验值皆低于0.001,说明可根据描述性、推断性、相关性和聚类分析,获得有效的网络舆情集成分析结果,可将其应用到优化政府决策中。【创新/局限】但由于条件有限,本文设计的集成方法在集成效率方面还有所不足,未来的研究方向可以从提高集成速度入手,加强舆论信息集成的效果。

【Purpose/significance】With the rapid development of the Internet and big data,massive amounts of network information have emerged.The integration and analysis of network public opinion information has become an important means to respond to the laws of user network behavior in a timely manner and optimize government decision-making.【Method/process】To this end,this article analyzes the concept and characteristics of big data and network public opinion based on the analysis of the current development,research significance and main research directions of online public opinion information integration.Based on this,the article takes the network public opinion information under the smart information service big data framework as the research object,combines log analysis and statistical analysis,and uses Awstats software big data integration technology to comprehensively analyze the main characteristics of the relevant network public opinion;Further use the significance test method of regression coefficients to obtain the behavior patterns of network users and apply them to the optimization of government decision-making.【Result/conclusion】the empirical results show that the significance test result of the regression coefficient of descriptive analysis is lower than 0.05,the significance test result of the regression coefficient of inferential analysis is lower than 0.1,and the significance test value of the regression coefficient of correlation and cluster analysis is lower than 0.001,indicating that effective integrated analysis results of online public opinion can be obtained according to descriptive,inferential,correlation and cluster analysis,which can be applied to optimize government decisionmaking.【Innovation/limitation】However,due to limited conditions,the integration method designed in this article is still insufficient in terms of integration efficiency.The future research direction can start from increasing the integration speed and enhancing the effect of public opinion information integration.