情报科学

• • 上一篇    下一篇

大数据情境下网络舆情分众化演变趋势分析

  

  1. 南京农业大学 工学院

Analysis of the evolution trend of network public opinion segmentation in the context of big data

  1. College of Engineering, Nanjing Agricultural University

摘要:

[目的/意义] 为有效判断网络舆情演变趋势分析网络舆情传播和形成的规律,研究网络舆情分众化演变的状态和特征,本文基于AEMIPO方法提出大数据情境下网络舆情分众化演变趋势分析方法,以期为网络舆情传播引导策略提供优化参考[方法/过程]通过对网络舆情分众化演变过程的自相似性、周期性和平稳性等统计特性进行动态跟踪,选取ARMAARIMASARIMAFARIMA模型对上述统计特性进行描述,构建备选模型库,从备选模型库中根据选择规则选择模型对网络舆情分众化演变趋势进行建模,并在大数据情境下预测网络舆情分众化演变趋势。[创新/局限]由于本文针对一个实证案例进行分析,存在一定研究局限性,因此在日后研究中需结合更多案例进行验证,并对该方法进行不断优化,从而全方面提升其有效性和准确度。[结果/结论]山东金矿爆炸救援的微博数据为实例对所提方法进行验证分析,结果反映该方法预测准确率高达80%,表明其可在大数据情境下对网络舆情分众化演变趋势进行准确预测和分析。

关键词:

大数据, 网络舆情, 演变趋势, 统计特征, 实例分析

Abstract:

[Purpose/Meaning] In order to effectively judge the evolution trend of online public opinion, analyze the law of the spread and formation of online public opinion, and study the status and characteristics of the evolution of online public opinion segmentation, this paper proposes the evolution of online public opinion segmentation in the context of big data based on the AEMIPO method Trend analysis method, so as to provide an optimized reference for the guidance strategy of network public opinion dissemination. [Method/Process] By dynamically tracking the statistical characteristics of the evolution process of online public opinion segmentation, such as self-similarity, periodicity and stability, select ARMA, ARIMA, SARIMA, FARIMA models to describe the above statistical characteristics, and construct alternative models Library; Select the corresponding model from the candidate model library according to the selection rules to model the evolution trend of online public opinion segmentation, and predict the evolution trend of online public opinion segmentation in the context of big data. [Innovation/Limitations] Due to the analysis of an empirical case in this article, there are certain research limitations. Therefore, it is necessary to combine more cases for verification in future research and continue to optimize the method to improve its effectiveness and accuracy in all aspects degree. [Results/Conclusions] The proposed method was validated and analyzed with the Weibo data of the "Shandong Gold Mine Explosion Rescue Case" as an example. The results showed that the prediction accuracy rate of the method was as high as 80%, indicating that the method can be used in the context of big data. The evolution trend of public opinion segmentation is accurately predicted and analyzed.

Key words:

Big Data, Internet public opinion, Evolution trend, Statistical Features, Case Analysis