Word sense disambiguation: A survey
By Roberto Navigli. 2009
2,219 citations recorded by the source corpus.
Word sense disambiguation (WSD) is the ability to identify the meaning of words in context in a computational manner. WSD is considered an AI-complete problem, that is, a task whose solution is at least as hard as the most difficult problems in artificial intelligence. We introduce the reader to the motivations for solving the ambiguity of words and provide a description of the task. We overview supervised, unsupervised, and knowledge-based approaches. The assessment of WSD systems is discussed in the context of the Senseval/Semeval campaigns, aiming at the objective evaluation of systems participating in several different disambiguation tasks. Finally, applications, open problems, and future directions are discussed.
Subfields: Artificial Intelligence, Natural Language Processing, Computational Linguistics & Semantics
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
- Mohammad Taher Pilehvar
- Mohit Bansal
- José Camacho-Collados
- Luke S. Zettlemoyer
- Mirella Lapata
- Yejin Choi
- Percy Liang
- Jianfeng Gao
In Natural Language Processing
- Mohammad Taher Pilehvar
- José Camacho-Collados
- Roberto Navigli
- Mirella Lapata
- K. Erk
- Chris Biemann
- Anna Korhonen
- Alessandro Raganato
In Computational Linguistics & Semantics
- Mirella Lapata
- José Camacho-Collados
- Mohammad Taher Pilehvar
- Lasha Abzianidze
- Sebastian Padó
- Peter D. Turney
- Alessandro Raganato
- K. Erk
Citations in the corpus
- SimLex-999: Evaluating Semantic Models With (Genuine) Similarity Estimation (2014, background): “Models of association, on the other hand, may be better suited to tasks such as word-sense disambiguation (Navigli, 2009), and applications such as text classification (Phan et al., 2008) in which the target classes correspond to topical domains such as agriculture or sport (Rose et al., 2002).”
- FlauBERT: Unsupervised Language Model Pre-training for French (2019, methodology): “First, WordNet is a resource that is entirely manually verified, and widely used in WSD research (Navigli, 2009).”
- Entity Linking meets Word Sense Disambiguation: a Unified Approach (2014, background): “Word Sense Disambiguation (WSD) (Navigli, 2009; Navigli, 2012) is a historical task aimed at assigning meanings to single-word and multi-word occurrences within text, a task which is more alive than ever in the research community.”
- Natural Language Questions for the Web of Data (2012, methodology): “Our joint disambiguation method was inspired by recent work on NED (Milne and Witten, 2008; Kulkarni et al., 2009; Hoffart et al., 2011a) and WSD (Navigli, 2009).”
- Neural Sequence Learning Models for Word Sense Disambiguation (2017, background): “While in its classical formulation (Navigli, 2009) WSD is viewed as a classification problem for a given word w in context, with word senses of w being the class labels, here we consider a variable-length sequence of input symbols ~x = h x 1 , ..., x T i and we aim at predicting a sequence of output…”
- Embeddings for Word Sense Disambiguation: An Evaluation Study (2016, background): “The task of automatically assigning pre-defined meanings to words in contexts, known as Word Sense Disambiguation, is a fundamental task in computational lexical semantics (Navigli, 2009).”
- Unsupervised Does Not Mean Uninterpretable: The Case for Word Sense Induction and Disambiguation (2017, methodology): “Despite significant progress in methodology during the two last decades (Ide and V´eronis, 1998; Agirre and Edmonds, 2007; Moro and Nav-igli, 2015), WSD is still not widespread in applications (Navigli, 2009), which indicates the need for further progress.”
- Constrained Semi-supervised Learning in the Presence of Unanticipated Classes (2016, background): “Glosses are an important resource of word meanings, which has been used by word sense disambiguation (WSD) [96] algorithms for decades.”
- SensEmbed: Learning Sense Embeddings for Word and Relational Similarity (2015, background): “com/p/word2vec/ We follow Navigli (2009) and show the n sense of the word with part of speech x as wordn.”
- Align, Disambiguate and Walk: A Unified Approach for Measuring Semantic Similarity (2013, neutral): “…retrieval (Hliaoutakis et al., 2006), while identifying similar words enables tasks such as paraphrasing (Glickman and Dagan, 2003), lexical substitution (McCarthy and Navigli, 2009), lexical simplifica-tion (Biran et al., 2011), and Web search result clustering (Di Marco and Navigli, 2013).”
- An automatic approach to identify word sense changes in text media across timescales (2015, background): “Automatic discovery and disambiguation of word senses from a given text is an important and challenging problem that has been extensively studied in the literature (Spärk-Jones 1986; Ide and Veronis 1998; Schütze 1998; Navigli 2009).”
- Context Dependent Semantic Parsing: A Survey (2020, background): “However, semantic parsing literature largely neglects word sense disambiguation, which is regarded as an AI complete problem (Navigli, 2009).”
- An Experimental Study of Graph Connectivity for Unsupervised Word Sense Disambiguation (2010, methodology): “Throughout this paper, we will use the term unsupervised to refer to knowledge-based WSD methods that employ an existing sense inventory but no labeled data (see also [13], [14] for a more detailed discussion of these issues).”
- Machine Learning-Based Sentiment Analysis for Twitter Accounts (2018, methodology): “In addition to that, W-WSD [31,61] has the ability to detect the correct word sense within a specified context.”
- The communicative function of ambiguity in language. (2011, positive): “Nearly all frequent words have multiple senses but word senses can be disambiguated reasonably well by computational models (Navigli, 2009), even using very simple knowledge of context or heuristics (Gale, Church, & Yarowsky, 1992; Yarowsky, 1993).”
- Biases in Large Language Models: Origins, Inventory, and Discussion (2023, background): “For example, SemCor [90] is the de facto training corpus for WordNet-based Word Sense Disambiguation (WSD) – the task of automatically assigning the most appropriate sense to a word in context [16, 97] – but is based on the Brown Corpus, the majority of whose text is from the 1960s (e.”