Sentiment Aware Fake News Detection on Online Social Networks
The DOI matches its Crossref record; 24 of the 27 references it cites were found in a registry or an index.
Abstract
Messages posted to online social networks (OSN) causes a recent stir due to the intended spread of fake news or rumor. This work aims to understand and analyse the characteristics of fake news especially in relation to sentiments, for the automatic detection of fake news and rumors. Based on empirical observations, we propose a hypothesis that there exists a relation between fake messages or rumours and sentiments of the texts posted online. We verify our hypothesis by comparing with the state-of-the-art baseline text-only fake news detection methods that do not consider sentiments. We performed experiments on standard Twitter fake news dataset and show good improvements in detecting fake news or rumor posts.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“For instance, fake videos created by fraudsters can easily deceive the general public [1].”
From A Comprehensive Review of DeepFake Detection Using Advanced Machine Learning and Fusion Methods (2023) · cites this paper as [1]
Citations
Cites
Glove: Global Vectors for Word Representation
Pennington, J., Socher, R., & Manning, C. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1162 · Cited by 34,071 · cited 1×
“We used the ones proposed by [23].”
(2003) · Journal of Machine Learning Research · doi:10.5555/944919.944937 · Cited by 32,337
WordNet: a lexical database for English
Miller, G. A. (1995) · Communications of the ACM · doi:10.1145/219717.219748 · Cited by 14,015 · cited 1×
“Some sentiment dictionaries exist to help in the achievement of this task such as [17] and [18].”
BackgroundBlei, D. M., Ng, A. Y., & Jordan, M. I. (2002) · Advances in Neural Information Processing Systems 14 · doi:10.7551/mitpress/1120.003.0082 · Cited by 12,223
The spread of true and false news online
Vosoughi, S., Roy, D., & Aral, S. (2018) · Science · doi:10.1126/science.aap9559 · Cited by 8,847 · cited 1×
“[11] has so far created the largest rumour dataset of 126,000 messages spread by almost 3 million people and found that fake news diffused up to 100,000 people while the truth only reached 1,000 people.”
BackgroundMining and summarizing customer reviews
Hu, M., & Liu, B. (2004) · Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining · doi:10.1145/1014052.1014073 · Cited by 7,793 · cited 2×
“Some sentiment dictionaries exist to help in the achievement of this task such as [17] and [18].”
BackgroundMethodThe Psychological Meaning of Words: LIWC and Computerized Text Analysis Methods
Tausczik, Y. R., & Pennebaker, J. W. (2010) · Journal of Language and Social Psychology · doi:10.1177/0261927x09351676 · Cited by 5,774 · cited 1×
“Therefore a sentiment analysis is proposed to be performed on each of the event corpus with a focus on the sentiment scoring function using Linguistic Word Count application’s (LIWC) [19] psychological and linguistic analytic capabilities.”
MethodBlei, D. M. (2012) · Communications of the ACM · doi:10.1145/2133806.2133826 · Cited by 5,516 · cited 1×
“Topic models enable the identification of most relevant words and concepts within a text corpus [20].”
BackgroundSentiWordNet 3.0: An Enhanced Lexical Resource for Sentiment Analysis and Opinion Mining
Baccianella, S., Esuli, A., & Sebastiani, F. (2010) · Language Resources and Evaluation Conference (LREC) · doi:10.63317/2jktnrykui9s · Cited by 2,666 · cited 1×
“Similar tools and methods were employed by [15] that used a weak supervised, semisupervised and random-walk step to create lexicons and bagof-words sentiments.”
MethodInformation credibility on twitter
Castillo, C., Mendoza, M., & Poblete, B. (2011) · Proceedings of the 20th international conference on World wide web · doi:10.1145/1963405.1963500 · Cited by 2,593 · cited 1×
“These could be broadly categorised into: (a) Content-based: Text (linguistics [5]); Media (images [6], GIFs and video) and URLs, (b) User-based: activity tracking (bots and spam [7]); bio information (registration age [8]); opposing views of other online users [9] and (c) Metadata-based: GPS Geotags, device source, Fo…”
BackgroundProbabilistic Latent Semantic Analysis
Hofmann, T. (2013) · arXiv (Cornell University) · doi:10.48550/arxiv.1301.6705 · Cited by 2,047 · cited 2×
“We have used two models in extracting emotion scores: a) Latent Semantic Analysis (LSA) [21] and b) Latent Dirichlet Allocation (LDA) [22] as described below: Latent Semantic Analysis: Given an unobserved event topic t, a tweet corpus d containing a word wi are conditionally independent. p (d,wi) = p(d) ∑ t p (wi|t) p…”
MethodFerrara, E., Varol, O., Davis, C., Menczer, F., & Flammini, A. (2016) · Communications of the ACM · doi:10.1145/2818717 · Cited by 1,973 · cited 1×
“These could be broadly categorised into: (a) Content-based: Text (linguistics [5]); Media (images [6], GIFs and video) and URLs, (b) User-based: activity tracking (bots and spam [7]); bio information (registration age [8]); opposing views of other online users [9] and (c) Metadata-based: GPS Geotags, device source, Fo…”
BackgroundLinguistic styles: language use as an individual difference
Pennebaker, J. W., & King, L. A. (1999) · Journal of Personality and Social Psychology · doi:10.1037/0022-3514.77.6.1296 · Cited by 1,961 · cited 1×
“Previous work has shown that deception and false statements can be detected from the writing style of the authors or linguistics and sometimes be used to infer their personalities [2].”
BackgroundFrom Tweets to Polls: Linking Text Sentiment to Public Opinion Time Series
O'Connor, B., Balasubramanyan, R., Routledge, B., & Smith, N. (2010) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v4i1.14031 · Cited by 1,923
Lying Words: Predicting Deception from Linguistic Styles
Newman, M. L., Pennebaker, J. W., Berry, D. S., & Richards, J. M. (2003) · Personality and Social Psychology Bulletin · doi:10.1177/0146167203029005010 · Cited by 1,439
CSI: A Hybrid Deep Model for Fake News Detection
Ruchansky, N., Seo, S., & Liu, Y. (2017) · Proceedings of the 2017 ACM on Conference on Information and Knowledge Management · doi:10.1145/3132847.3132877 · Cited by 1,039 · cited 1×
“[14] also used a text-based approach for fake news detection but considered the test, response and clustering of user features determined by support vector decomposition and integrated into a hybrid model.”
MethodWho says what to whom on twitter
Wu, S., Hofman, J. M., Mason, W. A., & Watts, D. J. (2011) · Proceedings of the 20th international conference on World wide web · doi:10.1145/1963405.1963504 · Cited by 957 · cited 1×
“A Twitter message has been shown to have a lifespan of as little as less than one day and up to a 70 day span depending on the type of content and URL being shared [1].”
BackgroundGupta, A., Lamba, H., Kumaraguru, P., & Joshi, A. (2013) · Proceedings of the 22nd International Conference on World Wide Web Companion · doi:10.1145/2487788.2488033 · Cited by 573 · cited 1×
“These could be broadly categorised into: (a) Content-based: Text (linguistics [5]); Media (images [6], GIFs and video) and URLs, (b) User-based: activity tracking (bots and spam [7]); bio information (registration age [8]); opposing views of other online users [9] and (c) Metadata-based: GPS Geotags, device source, Fo…”
BackgroundNews Verification by Exploiting Conflicting Social Viewpoints in Microblogs
Jin, Z., Cao, J., Zhang, Y., & Luo, J. (2016) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v30i1.10382 · Cited by 364 · cited 1×
“These could be broadly categorised into: (a) Content-based: Text (linguistics [5]); Media (images [6], GIFs and video) and URLs, (b) User-based: activity tracking (bots and spam [7]); bio information (registration age [8]); opposing views of other online users [9] and (c) Metadata-based: GPS Geotags, device source, Fo…”
BackgroundFake News identification on Twitter with Hybrid CNN and RNN Models
Ajao, O., Bhowmik, D., & Zargari, S. A. (2018) · e-space (Manchester Metropolitan University) · Cited by 318
Self-citationFake News Identification on Twitter with Hybrid CNN and RNN Models
Ajao, O., Bhowmik, D., & Zargari, S. (2018) · Proceedings of the 9th International Conference on Social Media and Society · doi:10.1145/3217804.3217917 · Cited by 253
Self-citationLearning Reporting Dynamics during Breaking News for Rumour Detection in Social Media
Zubiaga, A., Liakata, M., & Procter, R. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1610.07363 · Cited by 149 · cited 4×
“We used the PHEME [13] labeled Twitter dataset, from which images were also retrieved for testing on the deep learning model.”
MethodResultInfluentialSome Like it Hoax: Automated Fake News Detection in Social Networks
Tacchini, E., Ballarin, G., Vedova, M. L. D., Moret, S., & de Alfaro, L. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1704.07506 · Cited by 112 · cited 1×
“These could be broadly categorised into: (a) Content-based: Text (linguistics [5]); Media (images [6], GIFs and video) and URLs, (b) User-based: activity tracking (bots and spam [7]); bio information (registration age [8]); opposing views of other online users [9] and (c) Metadata-based: GPS Geotags, device source, Fo…”
BackgroundMIS2: Misinformation and Misbehavior Mining on the Web
Kumar, S., Jiang, M., Jung, T., Luo, R. J., & Leskovec, J. (2018) · ACM International Conference on Web Search and Data Mining (WSDM) · doi:10.1145/3159652.3160597 · Cited by 18 · cited 1×
“[12] identified that ‘lone wolves’ spread their message faster by creating fake accounts which express the same opinion in multiple ways to help propagate their message faster.”
Background
Research works printed but not linked: 1
- blei, 2003, Latent dirich-let allocation, Journal of Machine Learning Research, 3, 993
Cited by
A Comprehensive Review on Fake News Detection With Deep Learning
Mridha, M. F., Keya, A. J., Hamid, M. A., Monowar, M. M., & Rahman, M. S. (2021) · IEEE Access · doi:10.1109/access.2021.3129329 · Cited by 281
“Early research concentrated on leveraging textual information derived from the article’s content, such as statistical text features [20] and emotional information [21]–[23].”
BackgroundFANG: Leveraging Social Context for Fake News Detection Using Graph Representation
Nguyen, V.-H., Sugiyama, K., Nakov, P., & Kan, M.-Y. (2020) · Proceedings of the 29th ACM International Conference on Information & Knowledge Management · doi:10.1145/3340531.3412046 · Cited by 270
“, [28]), whereas the two types of support —neutral support and negative support— are based on reported correlation between news factuality and invoked sentiment [1].”
Mining Dual Emotion for Fake News Detection
Zhang, X., Cao, J., Li, X., Sheng, Q., Zhong, L., & Shu, K. (2021) · Proceedings of the Web Conference 2021 · doi:10.1145/3442381.3450004 · Cited by 251
“The authors of [1] point that there exists a relationship between fake news’ contents and their emotional words, and they construct an emotion feature (the ratio of the count of negative and positive words) to help detect fake news.”
BackgroundInfluentialMDFEND: Multi-domain Fake News Detection
Nan, Q., Cao, J., Zhu, Y., Wang, Y., & Li, J. (2021) · Proceedings of the 30th ACM International Conference on Information & Knowledge Management · doi:10.1145/3459637.3482139 · Cited by 248
“Earlier studies used hand-craft features [5, 6, 10, 18].”
Fake News in Sheep's Clothing: Robust Fake News Detection Against LLM-Empowered Style Attacks
Wu, J., Guo, J., & Hooi, B. (2024) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3637528.3671977 · Cited by 244
“Building upon this, prior advances have extracted sentiment features [1, 51] to complement and enhance the detector, and highlighted the significance of writing style between hyperpartisan news and well-balanced mainstream reporting [35].”
Background
Metadata
- Authors
- Ajao, O., Bhowmik, D., & Zargari, S.
- Affiliations
- Sheffield Hallam University (GB)
- University of Stirling (GB)
- Venue
- ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
- Publisher
- IEEE
- Published
- 2019-05
- Pages
- 2507-2511
- ISSN
- 1520-6149 (ISSN-L)
- OpenAlex
- W2911668186
- Cited by
- 228
- References
- 27
- Citations by year
- 2026: 18 · 2025: 34 · 2024: 47 · 2023: 50 · 2022: 30 · 2021: 32 · 2020: 13 · 2019: 3
- FWCI
- 55.59
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE International Conference on Acoustics Speech and Signal Processing (publishedVersion)
- SHURA (Sheffield Hallam University Research Archive) (Sheffield Hallam University) (acceptedVersion)
- Stirling Online Research Repository (University of Stirling) (submittedVersion, public-domain)
- Topics
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Physical Sciences › Computer Science › Information Systems › Spam and Phishing Detection
- Physical Sciences › Computer Science › Artificial Intelligence › Hate Speech and Cyberbullying Detection
- Keywords
- fake news detection, rumor detection, sentiment analysis, social networking sites, Twitter, text classification, misinformation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.