基于群体无意识协作的社会事件地理信息推断及空间关联

项目名称： 基于群体无意识协作的社会事件地理信息推断及空间关联

项目编号： No.61332004

项目类型： 重点项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 曹建农

作者单位： 香港理工大学深圳研究院

项目金额： 300万元

中文摘要： 近年来随着社会热点事件的规模日益增大，其对政府的正常运转和人民的日常生活影响日益明显。因此迫切需要大规模的、随时随地、无所不在的复杂感知能力，来实现对这些社会事件的及时发现与监测。现有探测社会事件的技术严重依赖专业人员和设备（如交通监测系统），需耗费大量人力物力，代价巨大。针对这一问题，本项目拟对群智计算理论进行研究，利用群智的思想来实现对社会事件的发现和监测，即利用人们的移动设备（智能手机、平板电脑等）自动收集群体日常生活中的信息，通过挖掘其空间关联来发现和监测社会事件。本项目将利用信号指纹空间、序列空间、多维标度、时空扫描统计等方法，探索城市规模群体的无意识协作机制，研究信号指纹-位置关联、室内地图的自动化生成、城市交通感知、城市热点事件发现等关键技术，形成一系列基于群智计算来推断社会事件地理信息的理论与方法，为指导相关群智计算系统的设计与分析提供坚实的理论和技术支撑。

中文关键词： 社会事件地理信息推断;社会事件空间关联;无意识协作;众包;参与式感知

英文摘要： For the past few years, different kinds of social events start to show increasingly larger scale and more significant impact on government operation and civic life than ever before. Therefore, there is an imperative need of large-scale sensing capability for discovering these social events at anywhere and anytime. Current state-of-the-art approaches rely heavily on specialized personnel and professional equipment (e.g. road traffic monitoring systems), thus are either too costly to be practical or inefficient. To solve this problem, we plan to conduct research on crowd intelligence and propose to develop a series of key techniques to detect these social events in a much cheaper while more efficient way. Through the usage of fingerprint space, sequential space, multi-dimensional scaling, space-time scan statistic, etc., we plan to develop metropolis-scale unconscious crowd-collaboration mechanism and geo-information based social event inference and spatial correlation. Our study includes the fingerprint-location correlation, indoor map generation, road and traffic sensing, and city-scale social event discovery, etc. Based on our research, we expect to develop a set of theories, algorithms, methods and systems that can utilize the mobility and geo-information of the crowd to deduce the geographic information of social events. This set of theories will facilitate crowd-computing based social incident discovery and surveillance, and provide a solid theoretical support to the design and analysis of related crowd computing systems.

英文关键词： Social events geoinfo inferenc;Social events spatial correlat;Unconsciously Collaboration;Crowdsourcing;Participatory Sensing