Unsupervised Fake News Detection on Social Media: A Generative Approach
Authors: Shuo Yang, Kai Shu, Suhang Wang, Renjie Gu, Fan Wu, Huan Liu5644-5651
AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiment results on two datasets show that the proposed method significantly outperforms the compared unsupervised methods. We conduct experiments on two real-world social media datasets, and the experiment results demonstrate the effectiveness of the proposed framework for fake news detection on social media. |
| Researcher Affiliation | Academia | Shuo Yang, Kai Shu, Suhang Wang, Renjie Gu, Fan Wu, Huan Liu Department of Computer Science and Engineering, Shanghai Jiao Tong University, China School of Computing, Informatics and Decision Systems Engineering, Arizona State University, USA College of Information Sciences and Technology, Pennsylvania State University, USA {wnmmxy, grj165, wu-fan}@sjtu.edu.cn; {kai.shu, huan.liu}@asu.edu; szw494@psu.edu |
| Pseudocode | Yes | Algorithm 1 shows the pseudo-code of the algorithm. |
| Open Source Code | No | The paper does not provide any specific links to source code or explicitly state that the code for their methodology is open-source or publicly available. |
| Open Datasets | Yes | In the experiment, we use two public datasets, i.e., LIAR (Wang 2017) and Buzz Feed News4 to evaluate the performance of our algorithm. 4https://github.com/BuzzFeedNews/2016-10-facebook-fact-check/blob/master/data |
| Dataset Splits | No | The paper describes the datasets used and their statistics but does not explicitly provide training/validation/test dataset splits. While the method is unsupervised and thus doesn't have a traditional training phase, no specific data partitioning details are given for reproducibility in terms of data splits for evaluation. |
| Hardware Specification | No | The paper does not specify the hardware used for running its experiments, such as particular CPU or GPU models, or cloud computing instance types. |
| Software Dependencies | No | The paper mentions using 'Twitter s advanced search API' and 'off-the-shelf sentiment analysis' but does not provide specific software names with version numbers (e.g., Python 3.x, specific library versions) that would be needed for reproducibility. |
| Experiment Setup | Yes | Parameter Settings: In the experiment, we set uniform priors for news count, i.e., r=(5, 5) so that each news has an equal chance of being true or fake. We set prior for sensitivity as α0 =(7, 3) and prior for 1-specificity as α1 =(3, 7) to plug in the assumption that verified users are usually reliable and do not have high false positive or false negative rates. As for unverified users, for each pair of (u, v) {0, 1}2, we set βu,v = (1, 9) indicating the observation that most of the unverified users reveal positive opinions. As for the Gibbs sampling algorithm, the number of iterations is set to 100. The burn-in period and thinning are set to 20 and 4, respectively. |