Spectral Clustering of Events in Social Media Flood Images based on Multimodal Analysis

In Data Science, Clustering is one of the most popular techniques. It has wide potential application in data analysis, market research, pattern recognition, image processing, etc. Spectral Clustering is a clustering algorithm shows improved performance than the conventional clustering algorithms. To handle Social media multimodal data in Event Clustering and for better understanding of the context of the post, Spectral Clustering of Events in Flood Images Based on Multimodal Analysis (SCEFIMA) approach is proposed. This approach makes use of Post's Image and textual content (Multimodal data) in Clustering the events. Also considering Spatial and Temporal Information in understanding the post. The existing approaches lacks on multimodal analysis in social media data. Use of Multimodal data gives the better understanding of the context of the post in event clustering. In addition to this, Using Multimodal data in Event Clustering significantly boost performance than the existing methods. Detecting and Clustering events from multimodal data in social networks helpful in monitoring events by public organizations and authorities. Also helpful in early planning for preventive measures.