Research status of deep learning methods for rumor detection

Abstract

To manage the rumors in social media to reduce the harm of rumors in society. Many studies used methods of deep learning to detect rumors in open networks. To comprehensively sort out the research status of rumor detection from multiple perspectives, this paper analyzes the highly focused work from three perspectives: Feature Selection, Model Structure, and Research Methods. From the perspective of feature selection, we divide methods into content feature, social feature, and propagation structure feature of the rumors. Then, this work divides deep learning models of rumor detection into CNN, RNN, GNN, Transformer based on the model structure, which is convenient for comparison. Besides, this work summarizes 30 works into 7 rumor detection methods such as propagation trees, adversarial learning, cross-domain methods, multi-task learning, unsupervised and semi-supervised methods, based knowledge graph, and other methods for the first time. And compare the advantages of different methods to detect rumors. In addition, this review enumerate datasets available and discusses the potential issues and future work to help researchers advance the development of field.

Research on domain ontology construction based on the content features of online rumors

“…Huang et al 11 conducted an empirical study based on WeChat rumors, and found that the "confirmation" means of rumors include data corroboration and specific information, hot events and authoritative release; Butt et al 12 analyzed the psycholinguistic features of rumors, and extracted four features from the rumor data set: LIWC, readability, senticnet and emotions. Zhou et al 13 analyzed the semantic features of fake news content in theme and emotion, and found that the distribution of fake news and real news is different in theme features, and the overall mood, negative mood and anger of fake news are higher; Tan et al 14 divided the content characteristics of rumors into content characteristics with certain emotional tendency and social characteristics that affect credibility; Damstra et al 15 identified the elements as a consistent indicator of intentionally deceptive news content, including negative emotions causing anger or fear, lengthy sensational headlines, using informal language or swearing, etc. Lai et al 16 put forward that emotional rumors can make the rumor audience have similar positive and negative emotions through emotional contagion; Yuan et al 17 found that multimedia evidence form and topic shaping are important means to create rumors, which mostly convey negative emotions of fear and anger, and the provision of information sources is related to the popularity and duration of rumors; Ruan et al 18 analyzed the content types, emotional types and discourse focus of Weibo's rumor samples, and found that the proportion of social life rumors was the highest, and the emotional types were mainly hostile and fearful, with the focus on the general public and the personnel of the party, government and military institutions.…”

RumorLLM: A Rumor Large Language Model-Based Fake-News-Detection Data-Augmentation Approach

“…In the field of fake news detection, supervised learning is currently the mainstream method. It is mainly divided into two main categories: methods based on traditional machine-learning methods [16] and methods based on deep-learning methods [17,18]. In previous studies, text and user information were mainly extracted using statistical machine learning or neural networks to extract textual features [19].…”

KG-MFEND: an efficient knowledge graph-based model for multi-domain fake news detection

“…Although user information-based and propagation-based methods have had some success, they have the following limitations when it comes to acquiring additional users and propagation structure information [ 22 ]: (1) Data ethics issues: User information is personal and subject to relevant data protection laws; it must be obtained in reasonable and legal ways [ 38 ]. (2) Data acquisition difficulty: Researchers must strictly follow their propagation paths to capture relevant information.…”