Lexical Hierarchy

Originally posted as part of the CMPE58H course by Susan Üsküdarlı in the fall semester of 2022 at https://medium.com/@furkanakkurt7642/lexical-hierarchy-85943cd4c0b6.


WordNet is a lexical database of English. It groups its lexemes by synonymy and calls the groups synsets. There are more than a hundred thousand synsets.1 A synset also has some kind of a definition and optionally an example usage. It serves for many people as a dictionary and a thesaurus together.2 The structure of the grouping resembles a hierarchy where a word of a distinct conceptual idea has mostly a small set of connections.1 It seems to me even a single semantic conception of a surface form of a word should have more connections, it definitely does for a human brain; however, probably it’s not possible to create such a web just now.

Some third parties also consider WordNet itself as an ontology that’s used in computer science.2 In another projectP.S., its database has been converted to RDF and its data has been restructured according to Lexicon Model for Ontologies of W3C, making it a more accurate ontology.

I wonder if WordNet can be used to semanticize the web. Language learning in a semantic web would be helpful as there would be applications that would show a word’s translation in the context of the paragraph, not just the most immediate usage. Even translations of whole sentences could be done, as semantic information would definitely help a machine translation tool to more accurately do its job. I believe the convenience it would provide would make a lot of people interested in its applications. Many people can’t find time to learn a language, even though they want to for their career, as a heritage language, or just as a hobby. Duolingo just taps into this and helps many people learn a language in small bites. Recently, I started using it “actively” again.

As an example application that could use WordNet, I know of an extension, Toucan, that changes the words in a text to a desired language. Semanticization of web would make this extension a whole lot better, while making the process of adding other languages to the extension a whole lot easier.

References

  1. WordNet
  2. WordNet - Wikipedia

P.S. The link http://wordnet-rdf.princeton.edu/about does not work.