Identification of Unsuitable Content for Children in Video Gaming Forums

Abstract

Online forums for videogames fans may represent a threat of exposure to topics and language non-suitable for their age. The following case study shows how the analysis of topics threads based on interactions of micro messages extracted from videogames forums allowed us to identify the challenges associated to the processing of this particular type of content. An original technique for segmentation of groups of interactions was evaluated with the objective of applying algorhythms for topics detection. Despite the high volume of micro messages that were highly disperse, overloaded with emojis and which contained slang associated with every community and videogame; it was possible to detect inappropriate topics according to the age associated to the category of the videogame.