Enhancing group polarity of temporal patterns for rumour detection on Twitter
- Publisher:
- Institute of Electrical and Electronics Engineers (IEEE)
- Publication Type:
- Conference Proceeding
- Citation:
- Proceedings of the 2022 IEEE International Conference on Behavioural and Social Computing, BESC 2022, 2022, 00, pp. 1-4
- Issue Date:
- 2022-01-01
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| Enhancing_group_polarity_of_temporal_patterns_for_rumour_detection_on_Twitter.pdf | Published version | 1.41 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
Rumour detection on social media is an urgent issue to be solved in our age. Current rumour detection methods focus on capturing the discriminative patterns from multiple modalities and detecting anomalies temporally or semantically. They cannot detect the rumours showing less maliciousness and suspiciousness. We utilise BiLSTM to capture the group semantic polarity and temporal patterns of diffusion to detect this kind of rumours. We also leverage text CNN to capture the semantic patterns of source tweets. Finally, the concatenated vector of enhanced temporal diffusion vector and semantic source vector is fed into the classifier to get the authenticity label. The experimental results show the effectiveness of retweeting sequence of user description to detect the group semantic polarity and the superiority of our idea to detect rumours on Twitter.
Please use this identifier to cite or link to this item: