Deception Detection with Feature-Augmentation by Soft Domain Transfer

Shahriar, Sadat; Mukherjee, Arjun; Gnawali, Omprakash

doi:10.1007/978-3-031-19097-1_23

Cited by 3 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The goal of the conversation is to persuade the readers by using several persuasion techniques (Cialdini and Cialdini, 2007;Gragg, 2003;Stajano and Wilson, 2011). Utilization of such techniques are observed in fake review detection, and phishing email detection (Munzel, 2016;Shahriar et al, 2022). Table 2 shows some examples of strategies used in a collusion scam.…”

Section: Persuasion Strategymentioning

confidence: 99%

Tackling the Myriads of Collusion Scams on YouTube Comments of Cryptocurrency Videos

Shahriar,

Mukherjee

2023

Proceedings of the Conference Recent Advances in Natural Language Processing - Large Language Models for Natural Language Proce

View full text Add to dashboard Cite

Despite repeated measures, YouTube's comment section has been a fertile ground for scammers. With the growth of the cryptocurrency market and obscurity around it, a new form of scam, namely "Collusion Scam" has emerged as a dominant force within YouTube's comment space. Unlike typical scams and spams, collusion scams employ a cunning persuasion strategy, using the facade of genuine social interactions within comment threads to create an aura of trust and success to entrap innocent users. In this research, we collect 1,174 such collusion scam threads and perform a detailed analysis, which is tailored towards the successful detection of these scams. We find that utilization of the collusion dynamics can provide an accuracy of 96.67% and an F1-score of 93.04%. Furthermore, we demonstrate the robust predictive power of metadata associated with these threads and user channels, which act as compelling indicators of collusion scams. Finally, we show that modern LLM, like chatGPT, can effectively detect collusion scams without the need for any training.

show abstract

Section: Persuasion Strategymentioning

confidence: 99%

Tackling the Myriads of Collusion Scams on YouTube Comments of Cryptocurrency Videos

Shahriar,

Mukherjee

2023

Proceedings of the Conference Recent Advances in Natural Language Processing - Large Language Models for Natural Language Proce

View full text Add to dashboard Cite

show abstract

“…The intuition behind this approach is that by obtaining the feature representation in different domains' highlevel latent space, the deceptive text may obtain richer information to detect deception than in its own domain only. The training strategy is adopted from Shahriar et al 2022.…”

Section: Intermediate Layer Concatenation (Ilc)mentioning

confidence: 99%

“…While this approach provides a domain-agnostic system, it is possible that the intrinsic variations between deception domains mean that a single network cannot give an effective solution. A feature-augmentationbased soft domain transfer approach using the last layer of learned models was proposed in (Shahriar et al, 2022). However, the last layers are prone to capturing the domain-specific noise which may have adverse effects in deceptive transfer.…”

Section: Introductionmentioning

confidence: 99%

“…To address this problem, we use a Multi-Task Learning (MTL) approach with the LSTM and BERT model, where the part of the model is shared across all domains to capture the generic information, and multiple branches downstream to account for domain-specific information. We compare our approach with an All For One (OFA) mode of deception generalization and Intermediate Layer Concatenation (ILC) mode of domain transfer (Shahriar et al, 2021(Shahriar et al, , 2022.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Exploring Deceptive Domain Transfer Strategies: Mitigating the Differences among Deceptive Domains

Shahriar,

Mukherjee,

Gnawali

2023

Proceedings of the Conference Recent Advances in Natural Language Processing - Large Language Models for Natural Language Proce

View full text Add to dashboard Cite

Deceptive text poses a significant threat to users, resulting in widespread misinformation and disorder. While researchers have created numerous cutting-edge techniques for detecting deception in domain-specific settings, whether there is a generic deception pattern so that deception-related knowledge in one domain can be transferred to the other remains mostly unexplored. Moreover, the disparities in textual expression across these many mediums pose an additional obstacle for generalization.To this end, we present a Multi-Task Learning (MTL) based deception generalization strategy to reduce the domain-specific noise and facilitate a better understanding of deception via a generalized training. As deceptive domains, we use News (fake news), Tweets (rumors), and Reviews (fake reviews) and employ LSTM and BERT models to incorporate domain transfer techniques. Our proposed architecture for the combined approach of domain-independent and domain-specific training improves the deception detection performance by up to 5.28% in F1-score.

show abstract