Application of AI tools as methodology for the analysis of toxicity in social media: A case study of Spanish politics on Twitter

Abstract

Introduction: A new artificial intelligence (AI) methodology is analyzed with the understanding that communication is one of the most important fields of work for its application. In addition to the content collection and production phases, other areas within the world of communication such as distribution, and specifically the moderation of comments (on social networks and in the media) are also experiencing a period of innovation, but in a less obvious way for the audience. Methodology: To find out how various AI tools can measure the quality of the conversation and combat toxicity in communicative spaces, we analyzed 43,165 tweets published from October 18th to October 24th 2021 corresponding to seven Spanish politicians and the cascade of user responses. Results: The most significant outcome reveals insults as the predominant toxic category in the comments, regardless of ideology. In addition, the conversations have an average of 21% of bots. Discussion: Therefore, this research shows how new AI methodologies can account for a hitherto qualitative term such as toxicity and contradicts previous findings on bots as its spreaders, with real users generating the most. Conclusions: In the specific study of politics, there is a perceived difference in behaviors between horizontal conversation among peers and vertical conversations with politicians. Therefore, these tools help to make visible new realities such as toxicity, with the ultimate aim of eradicating it and cleaning up the online debate.