scispace - formally typeset
Search or ask a question
Topic

Sarcasm

About: Sarcasm is a research topic. Over the lifetime, 1130 publications have been published within this topic receiving 25388 citations.


Papers
More filters
Journal ArticleDOI
TL;DR: Directives to hearers can be expressed in a variety of syntactic forms as mentioned in this paper, and the social distribution of such forms shows them to occur systematically, according to familiarity, rank, territorial location, difficulty of task, whether or not a duty is normally expected, whether compliance is likely, or not non-compliance is likely.
Abstract: Directives to hearers can be expressed in a variety of syntactic forms. The social distribution of such forms shows them to occur systematically, according to familiarity, rank, territorial location, difficulty of task, whether or not a duty is normally expected, whether or not non-compliance is likely. Except for some hints and questions not mentioning what is desired, directives do not require inference from a prior literal interpretation to be understood. Indeed, misunderstandings and puns imply that the interpretation of many directives is not likely to include a literal phase. On the contrary, where knowledge of obligations and prohibitions is shared, simple interpretation rules suffice, allowing prompt understanding. To interpret the affective significance of directives, one must compare the expected and realized forms, and recognize the social features that the difference implies. Deference, solidarity, coldness, sarcasm, rudeness, and qualitatively specified compliments or insults can be communicated systematically by such departures. (Pragmatics, directives, requests, politeness, conversational analysis, performatives, US English.)

609 citations

Proceedings ArticleDOI
01 Aug 2017
TL;DR: This paper shows that by extending the distant supervision to a more diverse set of noisy labels, the models can learn richer representations and obtain state-of-the-art performance on 8 benchmark datasets within emotion, sentiment and sarcasm detection using a single pretrained model.
Abstract: NLP tasks are often limited by scarcity of manually annotated data. In social media sentiment analysis and related tasks, researchers have therefore used binarized emoticons and specific hashtags as forms of distant supervision. Our paper shows that by extending the distant supervision to a more diverse set of noisy labels, the models can learn richer representations. Through emoji prediction on a dataset of 1246 million tweets containing one of 64 common emojis we obtain state-of-the-art performance on 8 benchmark datasets within emotion, sentiment and sarcasm detection using a single pretrained model. Our analyses confirm that the diversity of our emotional labels yield a performance improvement over previous distant supervision approaches.

592 citations

Proceedings Article
19 Jun 2011
TL;DR: This work reports on a method for constructing a corpus of sarcastic Twitter messages in which determination of the sarcasm of each message has been made by its author and uses this reliable corpus to compare sarcastic utterances in Twitter to utterances that express positive or negative attitudes without sarcasm.
Abstract: Sarcasm transforms the polarity of an apparently positive or negative utterance into its opposite. We report on a method for constructing a corpus of sarcastic Twitter messages in which determination of the sarcasm of each message has been made by its author. We use this reliable corpus to compare sarcastic utterances in Twitter to utterances that express positive or negative attitudes without sarcasm. We investigate the impact of lexical and pragmatic factors on machine learning effectiveness for identifying sarcastic utterances and we compare the performance of machine learning techniques and human judges on this task. Perhaps unsurprisingly, neither the human judges nor the machine learning techniques perform very well.

592 citations

Journal ArticleDOI
TL;DR: TASIT is straightforward for people with a normal range of social skills while being sensitive to social perception deficits after traumatic brain injury, with some influence from both education and intelligence.
Abstract: Objective To develop a clinically sensitive test of social perception for people with traumatic brain injury (TBI). Design An assessment tool comprising videotaped vignettes and response probes was developed in successive stages and tested on both normal participants and those with TBI. Subjects A total of 169 normal adults and 7 adults with severe TBI (pilot studies), 283 normal adults, and 12 people with severe TBI (main studies). Main outcome measures "The Awareness of Social Inference Test" (TASIT) comprises videotaped vignettes of everyday social interactions and has three parts, each with alternate forms. The Emotion Evaluation Test (EET) assesses recognition of spontaneous emotional expression (happy, surprised, sad, anxious, angry, disgusted, and neutral). The Social Inference-Minimal (SI-M) test assesses comprehension of sincere versus sarcastic exchanges, whereas the Social Inference-Enriched test (SI-E) assesses lies versus sarcasm. In both SI-M and SI-E speaker demeanor (voice, facial expression) indicate the intended meaning of the exchange. In addition, the SI-E vignettes have other contextual clues that reveal the speakers' intentions. Performance on SI-E and SI-E is assessed via four standard questions per item probing for understanding of the emotions, intentions, beliefs, and meanings of the speakers and their exchanges. Results Groups taken from the pool of 283 normal adults achieved a high level of performance on all aspects of the test with some influence from both education and intelligence. The 12 people with TBI were poorer at judging emotions than were matched controls, with particular difficulties recognizing neutral items, fear, and disgust. They were as capable as matched controls when understanding sincere exchanges and lies but had difficulty with sarcasm. Conclusions TASIT is straightforward for people with a normal range of social skills while being sensitive to social perception deficits after traumatic brain injury.

562 citations

Proceedings Article
01 Oct 2013
TL;DR: This work develops a sarcasm recognizer that automatically learns lists of positive sentiment phrases and negative situation phrases from sarcastic tweets and shows that identifying contrasting contexts using the phrases learned through bootstrapping yields improved recall for sarcasm recognition.
Abstract: A common form of sarcasm on Twitter consists of a positive sentiment contrasted with a negative situation. For example, many sarcastic tweets include a positive sentiment, such as “love” or “enjoy”, followed by an expression that describes an undesirable activity or state (e.g., “taking exams” or “being ignored”). We have developed a sarcasm recognizer to identify this type of sarcasm in tweets. We present a novel bootstrapping algorithm that automatically learns lists of positive sentiment phrases and negative situation phrases from sarcastic tweets. We show that identifying contrasting contexts using the phrases learned through bootstrapping yields improved recall for sarcasm recognition.

503 citations


Network Information
Related Topics (5)
Sentence
41.2K papers, 929.6K citations
81% related
Natural language
31.1K papers, 806.8K citations
80% related
Vocabulary
44.6K papers, 941.5K citations
78% related
Narrative
64.2K papers, 1.1M citations
77% related
Recall
23.6K papers, 989.7K citations
76% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
2023173
2022408
2021180
2020154
2019107
2018114