Etiqa'a: An Android Mobile Application for Monitoring Teen's Private Messages on WhatsApp to Detect Harmful/Inappropriate Words in Arabic using Machine Learning
Faiza Mohammed Usman Baran, Lama Saleh Abdullah Alzughaybi, Manar Ahmed Saeed Bajafar, Maram Nasser Muslih Alsaedi, Thraa Freed Hassan Serdar, Olfat M. Mirza · Engineering Technology & Applied Science Research · 2023
In today's world, social networks, such as WhatsApp, have become essential to daily life. An increasing number of Arab children use WhatsApp to communicate with others on a local and global scale, which has led to several negative consequences in their lives, including those associated with being bullied and harassed online. This study presents Etiqa'a, an application aiming to minimize risks and keep threats against minors from becoming a reality. Etiqa'a scans received WhatsApp messages which are then analyzed, and classified using a Logistic Regression (LR) machine learning model. The test results showed an accuracy of 81% in classifying messages as appropriate or inappropriate based on the text of the message. In the case of the latter, the application sends a detailed alert to parents.