Special Issue on Using AI and Social Media for Disaster Response and Management
摘要截稿:
全文截稿: 2019-03-01
影响因子: 4.787
期刊难度:
CCF分类: B类
中科院JCR分区:
• 大类 : 计算机科学 - 1区
• 小类 : 计算机:信息系统 - 1区
• 小类 : 图书情报与档案管理 - 1区
Overview
At the onset of a disaster event, victims, bystanders, and general public increasingly use social media platforms (e.g., Twitter and Facebook) to post situational updates such as reports of injured or dead people, infrastructure damage, requests of urgent needs, and so on. This online information on social media is available in different forms such as textual messages, images, and videos. Several research studies have shown that social media information is useful for disaster response and management, if processed timely and effectively.
Encouraged by these findings, humanitarian organizations have already started considering to incorporate more information from non-traditional data sources into their workflows. However, there are still a number of challenges that prevent these organizations using social media information for response efforts. These challenges include near real-time information processing, information overload, information extraction, summarization, and verification among others.
The aim of this special issue is to bring together diverse research communities such as information retrieval, data mining and machine learning, natural language processing, computer vision, computational social science, and human computer interaction, to potentially contribute towards building AI-based next-generation Information Processing Systems for an effective utilization of social media data for disaster response and management.
Topics of interest include but are not limited to:
Robust data mining, natural language processing, computer vision, and machine learning techniques to process social media data in real time
Indexing algorithms and technical challenges of handling big social media data during disasters
Extracting situational and actionable insights from social media text messages and/or images
Reducing information overload for better situational awareness
Aggregating multiple data sources for better information extraction
Automatic geo-inference from social media text messages and/or images
Transfer learning and domain adaptation techniques that exploit existing data and models from past disasters to deal with current events
Methods for dealing with task ambiguity, noise, bias, and long tails in social media data
Webly-supervised/weakly-supervised learning algorithms on raw user-provided content
Multimodal/crossmodal learning for robust classification of social media data
Social media information credibility, veracity, and misinformation