Recognizing Contextual Polarity in Phrase-Level Sentiment Analysis

By Theresa Wilson, Janyce Wiebe, Paul Hoffmann. 2005

3,727 citations recorded by the source corpus.

This paper presents a new approach to phrase-level sentiment analysis that first determines whether an expression is neutral or polar and then disambiguates the polarity of the polar expressions. With this approach, the system is able to automatically identify the contextual polarity for a large subset of sentiment expressions, achieving results that are significantly better than baseline.

Subfields: Artificial Intelligence, Natural Language Processing

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

In Natural Language Processing

Citations in the corpus

Refining Language Models with Compositional Explanations (2021, methodology): “For sentiment labels, we use the subjectivity lexicon [49] to decide if a word is positive, negative or neutral.”
Toward Controlled Generation of Text (2017, background): “The lexicon from (Wilson et al., 2005) contains 2700 words with sentiment labels.”
Understanding Neural Networks through Representation Erasure (2016, methodology): “• Sentiment: We use the MPQA subjectivity lexicon list (Deng and Wiebe, 2015; Wilson et al., 2005), which consists roughly 8,000 lexicons.”
Text Understanding from Scratch (2015, background): “Many researchers have shown that with some random text, the inter-rater consensus on polarity is only about 60% - 80%(Gamon & Aue, 2005)(Kim & Hovy, 2004)(Strappa-rava & Mihalcea, 2008)(Viera et al., 2005)(Wiebe et al., 2001)(Wilson et al., 2005).”
LSDSem 2017 Shared Task: The Story Cloze Test (2017, methodology): “One difference is that these more successful approaches used sentiment lexicons to score words and sentences, whereas Mostafazadeh et al. used 2017) showcase results with alternative sentiment dictionaries such as MPQA (Wilson et al., 2005).”
Twitter mood predicts the stock market (2010, methodology): “We adopt OF’s subjective lexicon that has been established upon previous work [37], [38], [24].”
Political Ideology Detection Using Recursive Neural Networks (2014, background): “A growing NLP subﬁeld detects private states such as opinions, sentiment, and beliefs (Wilson et al., 2005; Pang and Lee, 2008) from text.”
When does a compliment become sexist? Analysis and classification of ambivalent sexism using twitter data (2017, methodology): “To detect contextual polarity using phrase-level sentiment analysis, Wilson et al. (2005) identify whether a phrase is neutral or polar.”
Automated Fact Checking: Task Formulations, Methods and Future Directions (2018, background): “…classiﬁers augmented with lexicons including the Linguistic In-quiry and Word Count (LIWC) (Pennebaker et al., 2015), a sentiment lexicon (Wilson et al., 2005), a hedging lexicon (Hyland, 2015), and a novel ‘dramatic’ language lexicon to identify emotive, subjective and sensational…”
DynaSent: A Dynamic Benchmark for Sentiment Analysis (2020, background): “Other important examples are Opin-ionFinder (Wilson et al., 2005), LIWC (Pennebaker et al., 2007), MPQA, Opinion Lexicon (Liu, 2011), SentiWordNet (Baccianella et al., 2010), and the valence and arousal ratings of Warriner et al. (2013).”
Harnessing Twitter "Big Data" for Automatic Emotion Identification (2012, neutral): “MPQA Lexicon: MPQA (Multi-Perspective Question Answering) Subjectivity Lexicon [17] provides prior polarities (i.”
Dependency Tree-based Sentiment Classification using CRFs with Hidden Variables (2010, methodology): “As discussed in Section 1, Wilson et al. (2005) studied a bag-of-features based statistical sentiment classiﬁcation method incorporating head-modiﬁer relation.”
NRC-Canada-2014: Detecting Aspects and Sentiment in Customer Reviews (2014, background): “In addition, we employed existing out-ofdomain sentiment lexicons: (1) large-coverage automatic tweet sentiment lexicons, Hashtag Sentiment lexicons and Sentiment140 lexicons (Kiritchenko et al., 2014), and (2) three manually created sentiment lexicons, NRC Emotion Lexicon (Mohammad and Turney, 2010”
Aspect-based sentiment analysis of movie reviews on discussion boards (2010, methodology): “The approach makes use of both a domain-specific lexicon and a generic opinion lexicon, derived from SentiWordNet [6] and the subjectivity lexicon of [7], to assign a prior sentiment score to each word in a sentence.”
SemEval-2016 Task 6: Detecting Stance in Tweets (2016, background): “…nineteen submissions made extensive use of publicly-available sentiment and emotion lexicons such as the NRC Emotion Lexicon (Moham-mad and Turney, 2010), Hu and Liu Lexicon (Hu and Liu, 2004), MPQA Subjectivity Lexicon (Wilson et al., 2005), and NRC Hashtag Lexicons (Kir-itchenko et al., 2014b).”
United we Stand: Improving Sentiment Analysis by Joining Machine Learning and Rule Based Methods (2010, methodology): “• SL+: This is the subjectivity lexicon(Wilson et al., 2005) with 54 added valence operators.”