Computing Word-Pair Antonymy

By Saif M. Mohammad, Bonnie J. Dorr, Graeme Hirst. 2008

109 citations recorded by the source corpus.

Knowing the degree of antonymy between words has widespread applications in natural language processing. Manually-created lexicons have limited coverage and do not include most semantically contrasting word pairs. We present a new automatic and empirical measure of antonymy that combines corpus statistics with the structure of a published thesaurus. The approach is evaluated on a set of closest-opposite questions, obtaining a precision of over 80%. Along the way, we discuss what humans consider antonymous and how antonymy manifests itself in utterances.

Subfields: Artificial Intelligence, Language and Linguistics

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 Language and Linguistics

Citations in the corpus

Multi-Relational Latent Semantic Analysis (2013, methodology): “For instance, morphological variations discovered from the Google n -gram corpus have been combined with information from thesauri and vector-based word relatedness models for detecting antonyms (Mohammad et al., 2008).”
Revisiting Word Embedding for Contrasting Meaning (2015, methodology): “…0.41 0.58 Roget lookup (Zhang et al., 2014) 1.00 0.35 0.52 0.99 0.31 0.47 Roget lookup 1.00 0.32 0.49 0.97 0.29 0.44 W&R lookup (Zhang et al., 2014) 1.00 0.48 0.64 0.98 0.45 0.62 W&R lookup 0.98 0.52 0.68 0.97 0.52 0.68 (Mohammad et al., 2008) connected with these two types of edges respectively.”
Emotions Evoked by Common Words and Phrases: Using Mechanical Turk to Create an Emotion Lexicon (2010, background): “Natural symmetry apart, we believe that prior work on automatically computing word–pair antonymy (Lin et al., 2003; Mohammad et al., 2008; Lobanova et al., 2010) can now be leveraged in automatic emotion detection.”
Polarity Inducing Latent Semantic Analysis (2012, methodology): “In comparison, words with high cosine similarity scores are typically semantically related, which includes both syn-onyms and antonyms, as contrasting words often co-occur (Murphy and Andrew, 1993; Mohammad et al., 2008).”
Word Semantic Representations using Bayesian Probabilistic Tensor Factorization (2014, methodology): “(Mohammad et al., 2008; Mohammad et al., 2013) proposed empirical approaches considering corpus co-occurrence statistics and the structure of a published thesaurus.”
UMBC_EBIQUITY-CORE: Semantic Textual Similarity Systems (2013, methodology): “We also added to it more than 2,000 verb phrases extracted from WordNet.”
Counter-fitting Word Vectors to Linguistic Constraints (2016, background): “Most prior work focuses on extracting antonym pairs from text rather than exploiting them (Lin et al., 2003; Mohammad et al., 2008; Turney, 2008; Hashimoto et al., 2012; Mohammad et al., 2013).”
Computational Approaches to Sentence Completion (2012, methodology): “Among the narrower tasks, the identification of synonyms and antonyms has been studied by (Landauer and Dumais, 1997; Mohammed et al., 2008; Mohammed et al., 2011; Turney et al., 2003; Turney, 2008), who used questions from the Test of English as a Foreign Language (TOEFL), Graduate Record Exams (GR”