ICWSM - A Great Catchy Name: Semi-Supervised Recognition of Sarcastic Sentences in Online Product Reviews

By Oren Tsur, Dmitry Davidov, Ari Rappoport. International Conference on Web and Social Media, 2010

377 citations recorded by the source corpus.

Sarcasm is a sophisticated form of speech act widely used in online communities. Automatic recognition of sarcasm is, however, a novel task. Sarcasm recognition could contribute to the performance of review summarization and ranking systems. This paper presents SASI, a novel Semi-supervised Algorithm for Sarcasm Identification that recognizes sarcastic sentences in product reviews. SASI has two stages: semi-supervised pattern acquisition, and sarcasm classification. We experimented on a data set of about 66000 Amazon reviews for various books and products. Using a gold standard in which each sentence was tagged by 3 annotators, we obtained precision of 77% and recall of 83.1% for identifying sarcastic sentences. We found some strong features that characterize sarcastic utterances. However, a combination of more subtle pattern-based features proved more promising in identifying the various facets of sarcasm. We also speculate on the motivation for using sarcasm in online communities and social networks.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/9ce8e449151d4cb9047168f3a90c257035768847

How researchers see this paper

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

In Artificial Intelligence

Citations in the corpus

A Deeper Look into Sarcastic Tweets Using Deep Convolutional Neural Networks (2016, methodology): “An early work in this field was done by (Tsur et al., 2010) on a dataset of 6,600 manually annotated Amazon reviews using a kNN-classifier over punctuation-based and pattern-based features, i.”
From humor recognition to irony detection: The figurative language of social media (2012, background): “For instance, the research works focused on automatic humor generation [7, 36] and humor recognition [26, 25, 31]; as well as the ones focused on irony detection [39, 40, 10], satire detection [9], and sarcasm detection [38], have shown the feasibility of computationally dealing with figurative lang”
Who cares about Sarcastic Tweets? Investigating the Impact of Sarcasm on Sentiment Analysis. (2014, background): “Tsur et al. (Tsur et al., 2010) use a semi-supervised approach to classify sentences in online product reviews into different sarcastic classes, and report an F-measure of 82.7% on the binary sarcasm detection task (although Precision is much higher than Recall).”
Sarcasm as Contrast between a Positive Sentiment and Negative Situation (2013, background): “Tsur et al. (2010) presented a semi-supervised learning framework that exploits syntactic and pattern based features in sar-castic sentences of Amazon product reviews.”
The Effect of Different Writing Tasks on Linguistic Style: A Case Study of the ROC Story Cloze Task (2017, background): “Finally, writing style can also change from one sentence to another, for instance between positive and negative text (Davidov et al., 2010) or when writing sarcastic text (Tsur et al., 2010).”
Enhanced Sentiment Learning Using Twitter Hashtags and Smileys (2010, background): “This pattern based framework was proven efﬁ-cient for sarcasm detection in (Tsur et al., 2010; Davidov et al., 2010).”
Sentiment Analysis: An Overview from Linguistics (2016, positive): “Indeed, the following examples, all titles of reviews from Tsur et al (2010), would be difficult to interpret without some knowledge of context and, most importantly, without world knowledge (e.g., the fact that mentioning a book’s cover as the main positive feature of the book implies a negative…”
Bridging CNNs, RNNs, and Weighted Finite-State Machines (2018, methodology): “Following Tsur et al. (2010), we replace low frequency words with a special wildcard symbol.”
A multidimensional approach for detecting irony in Twitter (2013, background): “For instance, Tsur et al. (2010) address the problem of finding linguistic elements that mark the use of sarcasm in online product reviews.”
A survey of sentiment analysis in social media (2018, background): “Tsur et al. [124] presented a novel semi-supervised algorithm called SASI for sarcasm identiﬁcation that recognizes sarcastic sentences in product reviews.”
Sarcasm Detection on Twitter: A Behavioral Modeling Approach (2015, background): “Current research on sarcasm detection on Twitter [38, 10, 20, 29] has primarily focused on obtaining information from the text of the tweets.”
Humans Require Context to Infer Ironic Intent (so Computers Probably do, too) (2014, background): “…source used to experiment with irony detection systems has been Twitter (Reyes et al., 2012; Gonz´alez-Ib´a˜nez et al., 2011; Davidov et al., 2010), though Amazon product reviews have been used experimentally as well (Tsur et al., 2010; Davidov et al., 2010; Reyes et al., 2012; Filatova, 2012).”
Harnessing Context Incongruity for Sarcasm Detection (2015, methodology): “…Barbieri et al., 2014), (b) extraction of common patterns, such as hashtag-based sentiment (Maynard and Green-wood, 2014; Liebrecht et al., 2013), a positive verb being followed by a negative situation (Riloff et al., 2013), or discriminative n-grams (Tsur et al., 2010a; Davidov et al., 2010).”
Story Cloze Task: UW NLP System (2017, background): “Following Tsur et al. (2010) and Schwartz et al. (2013), we distinguish between high frequency and low frequency words.”
Natural Language Processing to the Rescue? Extracting "Situational Awareness" Tweets During Mass Emergency (2011, background): “However, subjectivity classification is useful in its own right, for discovery of sarcasm in text (Tsur et al., 2010), or as in our data, for assessing the amount of emotion or opinion content a user expresses in tweets, and the utility of subjective tweets in information extraction or assessment of”
Fracking Sarcasm using Neural Network (2016, methodology): “Since our We evaluated our system with two publicly available datasets (Tsur et al., 2010; Riloff et al., 2013).”
Implicitly Abusive Language – What does it actually look like and why are we not getting there? (2021, background): “Although the automatic detection of sarcasm has been investigated (Tsur et al., 2010; Riloff et al., 2013), the classiﬁcation performance is still fairly limited.”
Beneath the Tip of the Iceberg: Current Challenges and New Directions in Sentiment Analysis Research (2020, background): “, 2011b), stylistic patterns (Tsur et al., 2010), incongruity (Joshi et al.”