Harassment Detection Using Machine Learning and Fuzzy Logic Techniques

Proceedings

卷:31

JoséA. Concepción-Sánchez¹  Jezabel Molina-Gil¹  Pino Caballero-Gil¹ 

[1] Department of Computer Engineering and Systems, University of La Laguna, 38271 La Laguna, Tenerife, Spain;

关键词: cyberbullying;    data procession;    fuzzy logic;    machine learning;    real time;   

DOI  :  10.3390/proceedings2019031027

来源: DOAJ

【 摘 要 】

Social networks, instant messaging applications, smartphones and the Internet are the main technological tools used by adolescents for communication. While they can benefit from those tools, they can also be used as a weapon for harassment. Cyberbullying is the name used for a current global social problem derived from harassment that uses offensive messages, which is severely affecting the youngest. Different types of software to identify and filter offensive contents have been developed in the last years. However, most of them are time consuming, not scalable and focused on very specific environments. To address this problem, we propose a mobile application for smartphones that provides a potential offensive content detection in order to determine whether a cyberbullying attack exists or not. In particular, we have developed an application that combines data pre-processing, fuzzy logic and machine learning to predict cyberbullying content. The main idea is to install a mobile application on the smartphone of a possible victim, so that it runs in the background. The system analyzes all received messages and notifications using data processing and decision-making algorithms. Finally, a fuzzy logic technique helps the system to reach a conclusion under a certain degree of imprecision.

【 授权许可】

Unknown