Image Tweet Classification for Crisis Informative Task

In the realm of crisis events, the overflow of visual data across social media platforms necessitates the development of advanced techniques for thorough analysis and categorization. This research paper tackles the crucial task of classifying images based on informativeness within the CrisisMMD dataset, with the objective of identifying and categorizing informative content within images related to crisis. The central challenge lies in estab-lishing precise and efficient classification methods to differentiate between informative and non-informative visual content during emergency situations. Given the pivotal role of social media as a primary source for real-time information dissemination our proposed research aims to enrich our comprehension of crisis-related images, thereby contributing to more effective disaster response strategies. This study extends upon the existing body of literature that has explored various facets of crisis-related social media data, with a particular emphasis on the CrisisMMD dataset. Focusing specifically on the image modality, our research leverages advanced artificial intelligence and machine learning techniques emphasizing the application of deep multi-modal neural networks including the VGG-16 model. The significance of this research lies in its potential to significantly enhance the accuracy and efficiency of classifying informative content within crisis-related images, thereby facilitating more targeted and rapid disaster response efforts. Through meticulous experimentation and evaluation, the findings of this study are anticipated to provide valuable insights, propelling forward the field of image classification for informativeness on social media in times of crisis.