Research on Portrait of Online Public Opinion Subject Based on Big Data of Public Opinion—A Case Study of Notre Dame

• Published in: Journal of physics. Conference series Vol. 1861; no. 1; p. 12023
• Main Authors: Jia, Nan, Xia, Yixue, Jia, Ling
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.03.2021
• Subjects: Algorithms; Big Data; Machine learning; Oversampling; Physics; Public opinion; Social networks
• ISSN: 1742-6588; 1742-6596
• DOI: 10.1088/1742-6596/1861/1/012023

[Purpose/Significance] As online public opinion gradually becomes an important carrier of public opinion expression, multi-dimensional portrayal of each subject of public opinion based on public opinion big data helps to understand the characteristics of each dimension in an all-round way, thereby enhancing the pertinence of network social governance. This article aims to actively explore the subject of online public opinion subject based on public opinion big data and machine learning algorithms. [ Method/Process ]Use web crawler technology to crawl the microblog and comment data of the “Notre Dame” event, and describe it from the three dimensions of the basic attributes of microblog users, the microblog comment stand and the topic of microblog comments. In the basic attribute dimension, the reputation index of each user is calculated using information such as the number of fans and followers; in the Weibo comment stand dimension, machine learning algorithms are used to classify the comment stand, and the oversampling technique is used to oversample the data set Processing to improve the classification effect; in the microblog comment subject dimension, the LDA model is used to divide the subject of the review data; finally, a fusion portrait is developed based on the above portrait dimension.[ Results/Conclusion ]Through the integration of portraits, comprehensive use of image information such as microblog user influence, comment stances, comment topics, comment time, etc., to propose guidance strategies for different scenarios, helps to improve the targeted and effectiveness of network social governance.