Abstract:
Millions of people use social media platforms to communicate, share information, and interact with others, and it has become an essential part of our lives. However, incidents of violence, vulgarity, and child labour do occur on social media platforms. Monitoring and identifying these incidents is crucial to preventing and reducing their occurrence because they have serious repercussions. The Social Media Violence Vulgar and Child Labour Analyzer comes into play in this situation. A machine learning-based tool called the Social Media Violence Vulgar and Child labour Analyzer was created to recognise and classify images and videos into various categories, including child labour, violence, vulgar, and clean. The tool uses three models for prediction: the MainModel with four labels, the SecondModel with three labels, and the ThirdModel with three labels. It is powered by Python. Based on their content, these models are able to classify images and videos into various categories. The tool is integrated with the Angular and d3.js frameworks for the frontend and visualisation, respectively, and the Flask microframework for the backend. The database has been built using Postgres SQL. With the aid of the tool, incidents involving violence, vulgarity, and child labour can be prevented and reduced by being able to identify and monitor images and videos that contain such content. In conclusion, social media has developed into a potent tool for sharing information and facilitating communication. However, it is crucial to guarantee that social media platforms are secure for all users, especially kids. An effective tool for locating and tracking instances of violence, vulgarity, and child labour on social media platforms is the Social Media Violence, Vulgarity, and Child Labour Analyzer. Enforcement of community rules, monitoring of illegal content, and reporting of incidents pertaining to these categories are just a few of its numerous significant use cases. We can all use this tool to make social media a safer place.
