Word Sense Disambiguation Using Wikipedia

Abstract

We present three approaches to word sense disambiguation that use Wikipedia as a source of sense annotations. Starting from a basic monolingual approach, we develop two multilingual systems: one that uses a machine translation system to create multilingual features, and one where multilingual features are extracted primarily through the interlingual links available in Wikipedia. Experiments on four languages confirm that the Wikipedia sense annotations are reliable and can be used to construct accurate monolingual sense classifiers. The experiments also show that the multilingual systems obtain on average a substantial relative error reduction when compared to the monolingual systems.

NASARI: a Novel Approach to a Semantically-Aware Representation of Items

“…On the other hand, there are highly similar pairs in the dataset with relatively low gold scores. Examples include school-center 5 and term-life 6 with the respective gold similarity scores of 3.4 and 4.5, whereas 5 School and center have a pair of highly similar senses in WordNet 3.0: center 3 n : "a building dedicated to a particular activity" and school Dandala et al (2013).…”

“…Wikipedia can be considered as a sense inventory wherein the different meanings of a word are denoted by the articles listed in its disambiguation page (Mihalcea and Csomai, 2007). Given the high granularity of this inventory, clustering of senses can be highly beneficial to tasks that take this encyclopedic resource as their sense inventory (Hovy et al, 2013), such as Wikipedia-based Word Sense Disambiguation Dandala et al, 2013).…”

“…For the sense clustering task, we take as our benchmark the two datasets created by Dandala et al (2013). In these datasets, clustering has been viewed as a binary classification problem in which all possible pairings of senses of a word are annotated whether they ought to be clustered or not.…”

“…Therefore, most sense modeling techniques tend to base their computation on the knowledge obtained from various lexical resources. However, these techniques mainly utilize the knowledge derived from either WordNet (Banerjee and Pedersen, 2002;Budanitsky and Hirst, 2006;Pilehvar et al, 2013) or Wikipedia (Medelyan et al, 2009;Dandala et al, 2013;Gabrilovich and Markovitch, 2007;Strube and Ponzetto, 2006), which are, respectively, the most widely-used lexicographic and encyclopedic resources in lexical semantics (Hovy et al, 2013). This restriction to a single resource brings about two main limitations: (1) the sense modeling does not benefit from the complementary knowledge of different resources, and (2) the obtained representations are resource-specific and cannot be used across settings.…”

SupWSD: A Flexible Toolkit for Supervised Word Sense Disambiguation

“…Since scaling up to multiple languages is considerably easier for knowledgebased systems, as they do not require senseannotated training data, various efforts have been made towards the automatic construction of highquality sense-annotated corpora for multiple languages (Otegi et al, 2016;Delli Bovi et al, 2017), aimed at overcoming the so-called knowledge acquisition bottleneck of supervised models (Pilehvar and Navigli, 2014). These efforts include the use of Wikipedia, which can be considered a full-fledged, manually sense-annotated resource for numerous languages, and hence exploited as training data (Dandala et al, 2013).…”