Prototype to Perform Audit in Social Networks to Determine Cyberbullying

摘要

It was analyzed information on references to cyberbullying in social networks, text analysis, harassment detection, factors and behavior. The problem is the lack of auditing models on social networks that generate reliable information to determine cyberbullying. The objective of this research is to develop a prototype to perform audits on social networks to determine cyberbullying. It was used the deductive method and exploratory research to identify proposals that detect cyberbullying on social networks. It resulted in an Architecture prototype to perform audits on social networks to determine cyberbullying and an Algorithm prototype to perform audits on social networks to determine cyberbullying expressed in a flow chart. It was concluded that the algorithm prototype obtains valid information from social networks in a clear and precise way through the classification and weight of the offensive text; the prototype obtained 93.80% of average performance.