Cross-Topic Rumor Detection using Topic-Mixtures

Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics, pages 1534–1538 April 19 - 23, 2021. ©2021 Association for Computational Linguistics 1534 Cross-Topic Rumor Detection using Topic-Mixtures Xiaoying Ren, Jing Jiang∗ Singapore Management University 80 Stamford Road Singapore 178902 lemontreejying@gmail.com jingjiang@smu.edu.sg Ling Min Serena Khoo, Hai Leong Chieu DSO National Laboratories 12 Science Park Drive Singapore 118225 {klingmin,chaileon}@dso.org.sg Abstract There has been much interest in rumor de- tection using deep learning models in recent years. A well-known limitation of deep learn- ing models is that they tend to learn superfi- cial patterns, which restricts their generaliza- tion ability. We find that this is also true for cross-topic rumor detection. In this paper, we propose a method inspired by the “mixture of experts” paradigm. We assume that the predic- tion of the rumor class label given an instance is dependent on the topic distribution of the in- stance. After deriving a vector representation for each topic, given an instance, we derive a “topic mixture” vector for the instance based on its topic distribution. This topic mixture is combined with the vector representation of the instance itself to make rumor predictions. Our experiments show that our proposed method can outperform two baseline debiasing meth- ods in a cross-topic setting. In a synthetic set- ting when we removed topic-specific words, our method also works better than the base- lines, showing that our method does not rely on superficial features. 1 Introduction Recently, there has been much interest in detect- ing online false information such as rumors and fake news. Existing work has explored different features including network structures (Ma et al., 2019a), propagation paths (Liu and Wu, 2018), user credibility (Castillo et al., 2011) and the fusion of heterogeneous data such as image and text (Wang et al., 2018). However, these proposed algorithms still cannot be easily deployed for real-world appli- cations, and one of the key reasons is that, just like many other NLP problems, rumor or fake news de- tection models may easily overfit the training data and thus cannot perform well on new data. The ∗Corresponding author problem can be more serious with deep learning so- lutions, because deep neural networks tend to learn superficial patterns that are specific to the training data but do not always generalize well (Wang et al., 2018). In this work, we study the task of rumor detec- tion and focus on the problem of adapting a rumor detection model trained on a set of source topics to a target topic, which we refer to as cross-topic rumor detection. In a recent study by Khoo et al. (2020), the authors compared the performance of rumor detection in an in-topic setting and an out- of-topic setting. They found that their model could achieve 77.4% macro F-score on the in-topic test- ing data but the performance of the same classifier dropped to 39.5% when applied to out-of-topic test- ing data, which describe events different from the training events. In this paper, …

Content truncated to 500 words