Align, Disambiguate and Walk: A Unified Approach for Measuring Semantic Similarity

By Mohammad Taher Pilehvar, David Jurgens, Roberto Navigli. 2013

196 citations recorded by the source corpus.

How researchers see this paper

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

In Natural Language Processing

Citations in the corpus

WiC: the Word-in-Context Dataset for Evaluating Context-Sensitive Meaning Representations (2018, background): “distinction (Snow et al., 2007; Pilehvar et al., 2013; Mancini et al., 2017).”
Robust Entity Linking via Random Walks (2014, background): “It has been shown to be effective in the word similarity measurement [1, 14], and word sense disambiguation [22].”
Set-Theoretic Alignment for Comparable Corpora (2016, methodology): “Jaccard similarity, a core component of the approach we describe, has been standardly used as a text similarity measure in information retrieval and text summarisation tasks, or to compute semantic similarity (Pilehvar et al., 2013).”
Evolution of Semantic Similarity—A Survey (2020, methodology): “The similarity between these vectors is calculated using the measure called Weighted Overlap[75] as, where O denotes the overlapping terms in each vector and rank ( d , (cid:174) v i ) represent the rank of the term d in the vector v i .”
SensEmbed: Learning Sense Embeddings for Word and Relational Similarity (2015, methodology): “For instance, basing their work on the original sense discrimination approach of Reisinger and Mooney (2010), Huang et al.”
Robust named entity disambiguation with random walks (2018, background): “It has been shown to be effective in the word similarity measurement [1,16], and word sense disambiguation [30].”
Generating with Confidence: Uncertainty Quantification for Black-box Large Language Models (2023, methodology): “We observe that with the default generation configs, the choice of similarity measure ( a NLI,contra , a NLI,entail , a Jaccard ) seems to play a bigger role than the construction of U and C measures. a NLI,entail consistently performs the best, especially on nq , which is likely due to the nature o”