Abstract
The Internet is full of rumor posts, the spread of rumors will have a negative impact on social harmony and stability, affecting the healthy development of network information ecology. The uncertainty, timeliness, subjectivity and other characteristics of rumors make them different from general false network information. Social network rumor detection is a hot issue in the research field of social network and information transmission, which helps to further improve the efficiency and effect of rumor governance and purify the network environment. The social network rumor was defined as a social network transmission and unproven, or has been officially confirmed as false, and in the social network of information, the traditional rumors detection based on characteristics of the research focuses on text messages, publish static flat characteristics of users, the respect such as transmission, ignoring the message transmission evolution reaction of structure and transmission groups. Scholars at home and abroad have carried out many relevant studies on how to improve the efficiency of rumor detection. 138 Chinese literatures and 331 English literatures were retrieved from CNKI and Web of Science databases (retrieval date: August 2, 2021), and irrelevant literatures and news reports were removed, totaling 127 Chinese literatures and 238 English literatures. CiteSpace software was used for bibliometric analysis. By analyzing the publication time of relevant literatures, it is found that rumor detection research ushered in an outbreak period after 2019, and the number of literatures at home and abroad increased significantly. By analyzing the authors and research institutions, it is found that the maximum number of articles published by a single author at home and abroad is no more than three. The overall cooperative relationship between research institutions is not very close, and there are many independent research institutions. Through keyword analysis and existing literature review, it is found that the technologies and methods of rumor detection in recent years mainly involve “deep learning”, “attention mechanism”, “semi-supervised learning” and other technologies. The basic process of rumor detection is to combine the selected content features and social context features effectively, and then use advanced technologies such as Natural Language Processing (NLP), machine learning and deep learning to predict whether the information to be tested is false. With the deepening of the research in this field, the detection method of hybrid model is more and more popular. Rumor detection applications mainly focus on “online rumor”, “social media”, “public opinion monitoring” and other issues. The main research hotspots can be summarized as rumor detection algorithms, characteristics and propagation models, as well as research on different kinds of rumor identification and related tasks. The current research trends are the recognition of rumormonger, feature extraction of rumor corpus and multimodal rumor recognition.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.