Finalists Announced for the Eighth Communication Data Mining Competition: 12 Teams to Compete in Final Round
School of Journalism and Communication, South China University of Technology
(The following article was originally published by the WeHot Research Institute.)
On August 24, the finalist list for the Eighth Communication Data Mining Competition was released. Guided by the Intelligent and Computational Communication Committee of the Chinese Association for History of Journalism and Communication and jointly organized by the School of Journalism and Communication of South China University of Technology and the WeHot Research Institute, the competition will see 12 teams compete in the final round. This also marks the entry of this year’s competition into the final stage.
This year’s competition attracted 536 student applicants from 92 universities at home and abroad, including Peking University, Tsinghua University, Zhejiang University, Shanghai Jiao Tong University, Fudan University, Nanjing University, University of Science and Technology of China, Wuhan University, Huazhong University of Science and Technology, Xi’an Jiaotong University, Renmin University of China, Minzu University of China, Beijing Normal University, The Chinese University of Hong Kong, City University of Hong Kong, the London School of Economics and Political Science, and the University of Bristol. A total of 147 teams, comprising 485 participants, passed the eligibility review and selected one of four topics for competition.
During the preliminary round, 56 teams completed and submitted their entries on time. The judging panel consisted of six experts from academia and industry in intelligent and computational communication: Ye Wu, Director of the Computational Communication Research Center at Beijing Normal University; Lu Guan, Associate Professor at the School of Journalism, Fudan University; Yuanhang Lu, Associate Professor at the School of Journalism and Communication, South China University of Technology; Quanyi Zou, Associate Professor and doctoral supervisor at the School of Journalism and Communication, South China University of Technology; Chengchen Wang, General Manager of the Intelligent Retrieval Division of Midu Technology Co., Ltd.; and Gen Li, Head of the Natural Language Processing Laboratory at the WeHot Research Institute.
The judging panel conducted a comprehensive evaluation based on the topic and content, research methods and value of each entry. This year, submitted entries were required to include a research report, a dataset that could be used to reproduce the results, and a description of data processing. The research report was mandatory, and failure to submit it was regarded as withdrawal from the competition. Points were deducted from the total score for missing submission materials. One point was deducted for each missing dataset or data processing description, with a maximum deduction of two points. Teams were ranked independently under each topic, and the top three teams by total score for each topic entered the final round.
After fierce competition, the final-round teams are listed below.
Finalist list of the preliminary round of the Eighth Communication Data Mining Competition
Topic 1: Fact Consistency Detection and Risk Warning for Multimodal Content
Rank | Team | Work Title | Score |
1 | Tou Nao Te Gong Dui | Fact Consistency and Communication Risk Governance of Multimodal Content: A Computational Communication Study Based on Text-OCR Alignment and Credible Evidence Chains | 8.43 |
2 | Shuo Wan Mu Sheng | MMFact: A Multimodal Fact Consistency Detection and Risk Warning Model | 8.23 |
3 | Ji Chuan Dui | Research on Fact Verification and Risk Warning of Online Information Based on Multi-Source Evidence and Multimodal Consistency Fusion | 7.72 |
Topic 2: Intelligent Modeling of Communication Strategies and Multidimensional Effect Prediction
Rank | Team | Work Title | Score |
1 | Mythbuster | Circle-Based Communication and Agent Simulation of Family Affairs Judicial Information on Social Media: A Multi-Method Study Based on BERTopic, Machine Learning and OASIS | 7.98 |
2 | Xiao Xiao De Hua Yuan Li Wa Ya Wa Ya Wa | Intelligent Modeling, Multidimensional Effect Prediction and Dynamic Optimization of Cross-Platform Policy Communication Strategies: A Multi-Agent Study Based on the Policy Event of Installing Elevators in Residential Buildings of Four or More Floors | 7.87 |
3 | Guang Zong Yao Zu | How Far Can Prediction Go? The Predictability of Public Issues on Digital Platforms and Its Cross-Domain Limits | 7.83 |
Topic 3: Identification of AIGC Content Communication Features and Assessment of Social Impact
Rank | Team | Work Title | Score |
1 | realscope | From Source to Ripple: Research on the Communication Influence Mechanism and Risk Governance of AIGC Content | 7.85 |
2 | Token is burning | When “Generation” Enters the Public Opinion Field: Dual-Model Identification, Communication Features and Risks of Suspected AIGC Content in 2025 Weibo Hotspot Events | 7.78 |
3 | Whatever you do is ‘right’ | Broad but Cold: Residual Penetration and Interaction Marginalization of AIGC Content in the Weibo Information Ecosystem: Identification of Communication Features and Assessment of Social Impact Based on 375 Events and 13.10 Million Weibo Posts | 7.70 |
Topic 4: User Guidance Strategies and Effect Prediction for Social Media Circles
Rank | Team | Work Title | Score |
1 | Xiang Fa Hen Dui | Entertainment Enters the Circle, Risks Leave Traces: Risk Identification, Guidance Strategies and Effect Prediction for Circle-Based Communication of Entertainment Issues on Social Media | 8.82 |
2 | Little genius | Who Is the Real Victim? Narrative Features and Diffusion Mechanisms of Gender-Opposition Information on Social Media | 8.70 |
3 | Mei Fen Quan Kai Dui | User Circle Identification and Differentiated Communication Strategies for the Issue of AI & Employment | 8.34 |
(Team name translations are for reference only; please refer to official sources for the definitive versions.)
The online final of this year’s competition will be held in late September. Each finalist team is required to prepare its analysis report as a presentation and deliver a 10-minute presentation and explanation. The judging panel will ask questions and score the teams online. Finalist teams are encouraged to prepare their final entries as soon as possible. Best wishes to all participants for excellent results.
