网络舆情观点提取的LDA主题模型方法

目的/意义]无处不在的网络舆情信息深深影响甚至误导网络受众,探讨揭示网络舆情观点的方法,旨在拓展用户的认知深度和广度,提高大众对舆论的辨识能力。方法/过程]从技术上对比分析观点提取方法间的差异,从认知上阐释网络舆论平台的群体智慧和受众个体的认知过程,进而明确LDA主题模型提取舆情观点的优势及路径。结果/结论]结合舆论主题和情感因素,基于LDA的网络舆情观点提取,可从海量评论中判定深度评论,摘取主要观点,借助群众智慧,有效拓展个体思想和认知,为从大规模舆情中有序呈现受众观点提供新路径,也为舆情监测与疏导提供切实的依据。

Extraction Method of Network Public Opinion Based on LDA Topic Model

Purpose/significance]The pervasive network public opinion information deeply affects and even misleads the network audience. This paper explores ways to reveal the network public opinion points, to expand the users' depth and breadth to cognize, and improve the public's ability to distinguish.Method/process]The differences between two methods are analyzed from the view of technology, and the cognitive process of the masses and the audience is interpreted from the perspective of cognition, and then the advantages and path of the LDA topic model are described.Result/conclusion]Combined with the public opinion topic and emotional factors and extracting the network public opinion points based on LDA model, this paper determines the depth comments from mass comments and extracts the main opinions, and effectively expands the individual thought and cognition with the wisdom of crowds, to explore a new path to present audience ideas, and provide the practical basis for public opinion monitoring and counseling.