Urban emergency events are still difficult to be detected for their real-time feature, the limited physical sensors such as surveillance cameras in a city are still not spread, pervasive, and reliable enough for this goal.
Recently, social media feeds are rapidly emerging as a novel platform for providing and dissemination of information that is often geographic. In fact, the content from social media often includes (explicit or implicit) references to urban emergency events occurring at, or affecting specific locations.
Various studies have shown the potential of using social network streams data to monitor the public health status of a population.
Users of social networks can be seen as an hybrid form of a sensor system that allows for the identification and localization of the impact area of emergency events.
However, despite the widespread use of social media in various domains, its use in the scope of security and/or emergency management applications is still in its infancy.
RELATED RESEARCH
Social media such as Twitter and Facebook have been popularly used as everyday communication tools. Millions of people use "tweets" or Facebook "statuses" to inform family, friends, colleagues or any others about information, opinion and emotions about events just happening, leading to the great potential of using social media for monitoring and rescue purposes.
Usage statistics during recent disasters (e.g. the 20 million tweets regarding "Sandy" storm and a throughput of 10 Instagram photos per second) suggest that social media information might be efficiently exploited for the understanding and management of real-world events by civil protection and law enforcement agencies.
Microblogging social network, such as Twitter, are evaluated in terms of content and metadata. The main challenges are related to high throughput of Tweet stream and the slang of textual content.
Most relevant homepages and groups (e.g. related to civil defense) are monitored for event confirmation and latter for statistical evaluations. User-defined search against public users posts are supported.
Metadata and spatial information of multimedia contents (i.e. pictures) are analyzed for event detection. Eventually valuable contents are attached to event description. Image processing is not implemented yet.
INSTAGRAM, FLICKR
EYESHADE provides and experimental support to any kind of data-source accessible as RSS or Geo-RSS feed. The feed channel must be properly configured by user with few a-priori classification data which is exploited during the analysis (e.g. credibility).
RSS Feeds
Training sets and labeled data are a valuable goods in todays digital word.
In fact, most datasets (e.g. Twitter) are not allowed to be distributed without prior approval.
EYESHADE developer team likes to share the following free online data-sources which have been considered during testing and validation of the platform, they includes: knowledge base, labeled data, language data models.
