A simplified taxonomy of offensive language (SOL) for computational applications Cover Image

A simplified taxonomy of offensive language (SOL) for computational applications
A simplified taxonomy of offensive language (SOL) for computational applications

Author(s): Barbara Lewandowska-Tomaszczyk
Subject(s): Semantics
Published by: Akademia Nauk Stosowanych w Koninie
Keywords: annotation; computational applications; conceptual fuzziness; con- ceptual indeterminacy; offensive language; simplified taxonomy;

Summary/Abstract: The focus of the paper is a discussion of problems connected with the analysis and semantic categorization of concepts reflecting meanings of types of linguistic offense in English. The first part of the paper is related to general definitional and cross-boundary categorization problems of classes of words using the ex- ample of English offensive lexicon with the identification of main approaches to it and solutions proposed in linguistic and computational literature. The issue of computational identification of classes of offensive language is particularly em- phasized in the further sections with the discussion of extended models of of- fensive language. Insights from relevant lexical embeddings and further annotation exercise lead in conclusion to positing a simplified semantic taxonomy of offensive language (SOL) in the final section of the paper.

  • Issue Year: 10/2022
  • Issue No: 3
  • Page Range: 213-227
  • Page Count: 15
  • Language: English