Social Media Data Analysis For Disaster Response

Abstract

Natural disasters such as earthquakes, floods, wildfires, and cyclones cause widespread damage and require timely identification and response to minimize their impact. The proposed system introduces a multimodal disaster detection framework capable of analyzing both text and image inputs. For text-based disaster detection, the system employs machine learning algorithms to classify the type of disaster based on user-provided information. For image-based detection, the system utilizes Convolutional Neural Networks (CNN) and Transfer Learning models such as VGG16, InceptionV3, and ResNet to accurately recognize the disaster type from input images. The models are compared to identify the most effective one for each data type. Once the disaster category is determined from either text or image input, the system automatically provides relevant helpline numbers, safety precautions, preventive measures, and alerts. Developed using the Flask framework, this system serves as a reliable, user-friendly platform to support disaster awareness, preparedness, and rapid response.