Analysis of Social Media Data using Multimodal Deep Learning for Disaster Response
This work addresses the need for more effective disaster response tools by leveraging multimodal data, though it is incremental as it builds on existing deep learning techniques.
The paper tackled the problem of analyzing social media data for disaster response by proposing a multimodal deep learning architecture that uses both text and image modalities to learn a joint representation, resulting in better performance than single-modality models as shown in experiments on real-world datasets.
Multimedia content in social media platforms provides significant information during disaster events. The types of information shared include reports of injured or deceased people, infrastructure damage, and missing or found people, among others. Although many studies have shown the usefulness of both text and image content for disaster response purposes, the research has been mostly focused on analyzing only the text modality in the past. In this paper, we propose to use both text and image modalities of social media data to learn a joint representation using state-of-the-art deep learning techniques. Specifically, we utilize convolutional neural networks to define a multimodal deep learning architecture with a modality-agnostic shared representation. Extensive experiments on real-world disaster datasets show that the proposed multimodal architecture yields better performance than models trained using a single modality (e.g., either text or image).