The ACUMEN Project

Most studies based on critical discourse analysis employ corpus linguistics techniques to perform quantitative and qualitative discourse analyses. Although integrating corpus linguistics into critical discourse analysis has provided research results with greater objectivity, since they are empirically based on a larger amount of textual data, the computational methods that have been used are restricted to statistical metrics that determine the relevance of words or to the exploration of lexical patterns through concordances. This project proposes integrating methods and techniques from artificial intelligence into critical discourse analysis, where the synergy between these disciplines results in an emerging research field. To demonstrate the usefulness of our methodology, we focus on comments on YouTube videos dealing with various specific types of problems in the domain of social inequality. For this purpose, a corpus of YouTube comments will be constructed and annotated with an extensive inventory of lexico-grammatical features based on the Systemic Functional Linguistics model. The choice of YouTube as a text source is motivated by the fact that this video platform has become a prominent online news source, where users often reflect their ideological profile in the comments they post. The project has a strong multidisciplinary character, requiring knowledge from linguistics (e.g., theoretical linguistics, corpus linguistics, and critical discourse analysis) and artificial intelligence, where the latter is put at the service of linguistic research.