ADVANCED SOCIAL MEDIA ANALYTICS FOR REAL-TIME CRIME DETECTION AND PREVENTION: A MULTI-DIMENSIONAL APPROACH
Suresh V Reddy · International Journal of Apllied Mathematics · 2025
Social networking sites produce enormous volumes of data in real time, which might provide important information on illegal activity. This study combines computer vision, geographic clustering, and natural language processing (NLP) to provide a unique method for real-time crime detection using social media analytics. The suggested model classifies text using Long Short-Term Memory (LSTM) networks, images using Convolutional Neural Networks (CNN), and geographical data using k-means clustering. A multi-modal dataset with more than a million posts and photos from social media sites including Facebook, Instagram, and Twitter is used to test the architecture. The usefulness of the suggested system in real-time crime detection is demonstrated by the testing findings, which indicate an overall accuracy of 94.5% with a detection delay of less than 3 seconds.