Negobot: A Conversational Agent Based on Game Theory for the Detection of Paedophile Behaviour

摘要: Children have been increasingly becoming active users of the Internet and, although any segment population is susceptible to falling victim existing risks, they in particular are one most vulnerable. Thus, some major scourges this cyber-society paedophile behaviours on Internet, child pornography or sexual exploitation children. In light background, Negobot a conversational agent posing as child, chats, social networks and other channels suffering from behaviour. As agent, Negobot, has strong technical base Natural Language Processing information retrieval, well Artificial Intelligence Machine Learning. However, innovative proposal consider conversation itself game, applying game theory. context, proposes, first, competitive which system identifies best strategies for achieving its goal, obtain that leads us infer if subject involved with tendencies, while our actions do not bring alleged offender leave due suspicious behaviour agent.