Sarcasm Detection with Commonsense Knowledge

Li, Jiangnan; Pan, Hongliang; Lin, Zheng; Fu, Peng; Wang, Weiping

doi:10.1109/taslp.2021.3120601

Cited by 23 publications

(27 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lou et al (2021) design a GCN-based model combining SenticNet (Cambria et al, 2020), dependency tree and LSTM with GCN (Kipf and Welling, 2017) together, which achieves promising performance. Similar to previous studies, to better understand sarcasm, many approaches are able to utilize external information such as emoji expressions (Felbo et al, 2017), affective knowledge (Babanejad et al, 2020) and commonsense (Li et al, 2021). Joshi et al (2017) provide a more comprehensive survey.…”

Section: Neural Modelsmentioning

confidence: 99%

A Dual-Channel Framework for Sarcasm Recognition by Detecting Sentiment Conflict

Liu¹,

Wang²,

Sun³

et al. 2022

Findings of the Association for Computational Linguistics: NAACL 2022

View full text Add to dashboard Cite

Sarcasm employs ambivalence, where one says something positive but actually means negative, and vice versa. The essence of sarcasm, which is also a sufficient and necessary condition, is the conflict between literal and implied sentiments expressed in one sentence. However, it is difficult to recognize such sentiment conflict because the sentiments are mixed or even implicit. As a result, the recognition of sophisticated and obscure sentiment brings in a great challenge to sarcasm detection. In this paper, we propose a Dual-Channel Framework by modeling both literal and implied sentiments separately. Based on this dual-channel framework, we design the Dual-Channel Network (DC-Net) to recognize sentiment conflict. Experiments on political debates (i.e., IAC-V1 and IAC-V2) and Twitter datasets show that our proposed DC-Net achieves state-of-the-art performance on sarcasm recognition. Our code is released to support research 1 .

show abstract

Section: Neural Modelsmentioning

confidence: 99%

A Dual-Channel Framework for Sarcasm Recognition by Detecting Sentiment Conflict

Liu¹,

Wang²,

Sun³

et al. 2022

Findings of the Association for Computational Linguistics: NAACL 2022

View full text Add to dashboard Cite

show abstract

“…Similarly, in 2022, the researchers in [66] introduced an enhancement to the BERT model by fine-tuning it to related intermediate tasks before applying it to the target task. The authors in [67] applied the pre-trained COMET model to generate relevant commonsense knowledge. The experiment was conducted on three datasets, including Ghosh and Ptácek from Twitter and SARC-Pol from Reddit [35], [65], and [68].…”

Section: E Transformer-based Approachesmentioning

confidence: 99%

Text-based Sarcasm Detection on Social Networks: A Systematic Review

Alqahtani¹,

Alhenaki²,

Alsheddi³

2023

IJACSA

View full text Add to dashboard Cite

Sarcasm is a sophisticated phenomenon used for conveying a meaning that differs from what is being said, and it is usually used to express displeasure or ridicule others. Sentiment analysis is a process of uncovering the subjective information from a text. Detecting figurative language such as irony or sarcasm, is a focused challenging research field of sentiment analysis. Detecting and understanding the use of sarcasm in social networks could provide businesses and politicians with significant insight, since it reflects people's opinions about certain topics, news, and products. This has especially become relevant recently because sarcastic texts have been trending on social networks and are being posted by millions of active users. As a result of this situation, there is now an increasing amount of research on the detection of sarcasm in social network posts. Many works have been published on sarcasm detection, and they include a wide variety of techniques based on rules, lexicons, traditional machine learning, deep learning, and transformers. However, sarcasm detection is a challenging task due to the ambiguity and non-straightforward nature of sarcastic text. In addition, very few reviews have been conducted on the research in this area. Therefore, this systematic review mainly aims at exploring the newly published sarcasm detection articles on social networks in the years between 2019 and 2022. Several databases were extensively searched, and 30 articles that met the criteria were included. The selected articles were reviewed based on their approaches, datasets, and evaluation metrics. The findings emphasized that deep learning is the most commonly used technique for sarcasm detection in recent literature, and Twitter and F-measure are the most used source and performance metric, respectively. Finally, this article presents a brief discussion regarding the challenges in sarcasm detection and future research directions.

show abstract

“…While using text-image pair can benefit sarcasm detection compared with only using a single modality , recent works have shown that it might be still challenging to detect sarcasm solely from a textimage pair (Li et al, 2021a;Veale and Hao, 2010). To this end, we explore the effect of fusing various external knowledge extracted from an image for sarcasm detection.…”

Section: Knowledge Enhancementmentioning

confidence: 99%

“…Moreover, as figurativeness and subtlety inherent in sarcasm utterances may bring a negative impact to sarcasm detection, some works (Li et al, 2021a;Veale and Hao, 2010) found that the identification of sarcasm also relies on the external knowledge of the world beyond the input texts and images as new contextual information. What's more, it has drawn increasing research interest in how to incorporate knowledge to boost many machine learning algorithms such as recommendation system (Sun et al, 2021) and relation extraction (Sun et al, 2022).…”

Section: Introductionmentioning

confidence: 99%

Towards Multi-Modal Sarcasm Detection via Hierarchical Congruity Modeling with Knowledge Enhancement

Liu¹,

Wang²,

Li³

2022

Preprint

View full text Add to dashboard Cite

Sarcasm is a linguistic phenomenon indicating a discrepancy between literal meanings and implied intentions. Due to its sophisticated nature, it is usually challenging to be detected from the text itself. As a result, multimodal sarcasm detection has received more attention in both academia and industries. However, most existing techniques only modeled the atomic-level inconsistencies between the text input and its accompanying image, ignoring more complex compositions for both modalities. Moreover, they neglected the rich information contained in external knowledge, e.g., image captions. In this paper, we propose a novel hierarchical framework for sarcasm detection by exploring both the atomiclevel congruity based on multi-head cross attention mechanism and the composition-level congruity based on graph neural networks, where a post with low congruity can be identified as sarcasm. In addition, we exploit the effect of various knowledge resources for sarcasm detection. Evaluation results on a public multi-modal sarcasm detection dataset based on Twitter demonstrate the superiority of our proposed model.

show abstract

Sarcasm Detection with Commonsense Knowledge

Cited by 23 publications

References 26 publications

A Dual-Channel Framework for Sarcasm Recognition by Detecting Sentiment Conflict

A Dual-Channel Framework for Sarcasm Recognition by Detecting Sentiment Conflict

Text-based Sarcasm Detection on Social Networks: A Systematic Review

Towards Multi-Modal Sarcasm Detection via Hierarchical Congruity Modeling with Knowledge Enhancement

Contact Info

Product

Resources

About