La dimensione Testuale del Videogioco

Classificazione dei transcript dei videogiochi basata sul lessico

Autori

  • Alessandro Maisto Dipartimento di Scienze Politiche e della Comunicazione, Università degli Studi di Salerno, Fisciano (SA), Italia.

DOI:

https://doi.org/10.57574/596540765

Parole chiave:

Videogame transcript corpus, Automatic videogame rating, Bad language, Slang dictionary, Text analysis

Abstract

In this work, we explore the textual dimension of video games. Despite their pronounced visual and interctive characteristics, video games can be regarded as documents due to their narrative and communicative elements. Our research delves into this textual dimension to automatically generate rating tags associated with offensive language, violence, and the presence of drugs. We utilized a dictionary of English slang, compiled from various online sources and manually annotated with four categories: Slang, Violence, Drugs, and Discrimination. The resulting electronic dictionary facilitated the automatic assignment of the three rating tags with high precision. It has also been employed to classify video games based on their lexical content. The two classification tasks – by rating tags and by lexical dimension – could pave the way for an automatic warning system capable of nalyzing the full textual dimension of a video game.

##submission.downloads##

Pubblicato

30-06-2024