Building the Multi-Modal Storytelling of Urban Emergency Events Based on Crowdsensing of Social Media Analytics
Zheng Xu,Yunhuai Liu,Hui Zhang,Xiangfeng Luo,Lin Mei,Chuanping Hu
DOI: https://doi.org/10.1007/s11036-016-0789-2
2016-01-01
Mobile Networks and Applications
Abstract:With the development of Web 2.0, ubiquitous computing, and corresponding technologies, social media has the ability to provide the concepts of information contribution, diffusion, and exchange. Different from the permitting the general public to issue the user-generated information, social media has enabled them to avoid the need to use centralized, authoritative agencies. One of the important functions of Weibo is to monitor real time urban emergency events, such as fire, explosion, traffic jam, etc. Weibo user can be seen as social sensors and Weibo can be seen as the sensor platform. In this paper, the proposed method focuses on the step for storytelling of urban emergency events: given the Weibo posts related to a detected urban emergency event, the proposed method targets at mining the multi-modal information (e.g., images, videos, and texts), as well as storytelling the event precisely and concisely. To sum up, we propose a novel urban emergency event storytelling method to generate multi-modal summary from Weibo. Specifically, the proposed method consists of three stages: irrelevant Weibo post filtering, mining multimodal information and storytelling generation. We conduct extensive case studies on real-world microblog datasets to demonstrate the superiority of the proposed framework.