Negative Deceptive Opinion Spam

By Myle Ott, Claire Cardie, Jeffrey T. Hancock. North American Chapter of the Association for Computational Linguistics, 2013

444 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/dcaade47c361eaff61c99590ca3d5b5a709fb034

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Mohit Bansal
2. James Thorne
3. Andreas Vlachos
4. Myle Ott
5. Jiwei Li
6. Jianfeng Gao
7. Zhe Gan
8. Claire Cardie

Citations in the corpus

• Towards a General Rule for Identifying Deceptive Opinion Spam (2014, methodology): “Moreover, by soliciting deceptive opinion spam in an arti-ficial environment, we are endorsing the deception, which may influence the cues that we observe (Feeley and others, 1998; Frank and Ekman, 1997; Newman et al., 2003; Ott, 2013).”
• Automatic deception detection: Methods for finding fake news (2015, background): “The classification of sentiment (Pang & Lee, 2008; Ott et al., 2013) is based on the underlying intuition that deceivers use unintended emotional communication, judgment or evaluation of affective state (Hancock, Woodworth, & Porter, 2011).”
• On Human Predictions with Explanations and Predictions of Machine Learning Models: A Case Study on Deception Detection (2018, methodology): “ce the machine performance with linear SVM in Ott et al. [60] already surpasses humans (∼50%) by a wide margin and linear classifiers are generally considered more interpretable, we follow Ott et al. [59] and use linear SVM with bag-of-words features as our machine learning model. The linear SVM cla”
• Automatic Detection of Fake News (2017, methodology): “Previous work on verbal deception detection showed that LIWC is a valuable tool for the deception detection in various contexts (e.g., genuine and fake hotel reviews, (Ott et al., 2011b, 2013); prisoners’ lies (Bond and Lee, 2005)).”
• False Information on Web and Social Media: A Survey (2018, background): “In a followup work [69], negative sentiment deceptive reviews were studied, with additional 400 negative reviews from AMT.”
• Automated Fact Checking: Task Formulations, Methods and Future Directions (2018, background): “This is corroborated by the study of sentiment in deceptive texts (Ott et al., 2013) which also identified that texts with negative sentiment were more likely to be deceptive.”
• Description Based Text Classification with Reinforcement Learning (2020, background): “…Pontiki et al., 2016; Sun et al., 2019a), topic classification (Schwartz et al., 1997; Quercia et al., 2012; Wang and Manning, 2012), spam detection (Ott et al., 2011, 2013; Li et al., 2014), etc. Standardly, text classification is divided into the following two steps: (1) text feature extraction…”
• Identifying the sentiment styles of YouTube’s vloggers (2018, background): “Applications range from the automated detection of fake reviews (Ott et al., 2013), fake news (Thorne et al., 2018; Pérez-Rosas et al., 2018) and lies and truths (Mihalcea and Strapparava, 2009; Kleinberg et al., 2018) to detecting political sentiments (Cambria, 2016).”
• Combating Fake News (2019, methodology): “e linguistic analysis based methods that are applied to fake news detection. 4.2.1 N-gram Approach. The most effective linguistic analysis method applied to fake news detection is the n-gram approach [67, 76, 77]. n-grams are sequences of n contiguous words in a text, constituting words (unigrams) a”