“Thou Shalt Not Take the Lord’s Name in Vain”: A Methodological Proposal to Identify Religious Hate Content on Digital Social Networks

Abstract

This study explores biblical terms in contexts of hate speech dissemination on digital social networks and proposes a method to filter hateful terms on Facebook. The objective is to identify and index words and expressions of religious intolerance and violence in Portuguese. Specifically, it aims to identify the biblical terms used in the context of the dissemination of hateful content on Facebook and, based on the results, build a block list that expands the analytical capacity of response, investigation, and timely intervention methods to abusive, fundamentalist, and extremist content. The methodological approach is based on text mining and content analysis procedures, combined with the application of Zipf's law, with the adoption of Goffman’s transition point (T). As a result, we collected 4,214,699 comments from the official page of a neo-Pentecostal pastor who served as federal deputy between 2013 and 2016, recognized for intolerant and radical statements against nonpractitioners of his religion, non-Christians, agnostics, atheists, and secularists. In addition, the method determines 108 terms distributed in 3,614 comments. The content analysis verifies 25 terms used in the context of hate and systematizes them about the forms of speech (intolerant, fundamentalist, extremist, curse, or praise), as well as different recipients of the message found. The findings support the validation of the methodological proposal and provide the creation of a controlled vocabulary that organizes and monitors the hateful comments, the speech form, and the message recipient.