Although the use of the Internet by children is usually safe and enriching, there are also risks associated with the use of especially social networking sites (SNS). Several initiatives were launched to protect children against these potential threats through prevention and intervene when things go wrong, but given the sheer abundance of information, it is impossible for the party to any potentially harmful situations (such as cyberbullying, suicidal behavior or grooming by trace pedophiles). The automatic tracking of harmful content and harmful conduct requires a cross-media filtering approach, which can detect immediate risk, which critical situations indicates moderators of social networking and provide the accurate quantitative information that longitudinal and policies supporting information is available online child safety. This filtering will be done taking into account the expectations and experiences of adolescents and children online and also using an approach that deals in a respectful and caring way with privacy-sensitive information. To achieve this will bring Amica Project advances in state-of-the-art from different disciplines