SBIR Phase I: Enhancing the Viability of Emergency Alerting over Social Media via a Collaborative Computer-Aided System for Handling Incoming Citizen Messages

这一创新是一个软件引擎，使应急管理人员能够分析和处理大量传入的市民消息(例如，查询、求助等)。在应急经理的社交媒体网站(S)(例如，脸书、推特)上，通常在一个或多个社交媒体网站上发布紧急警报后经历。互联网连接软件引擎的核心由两个算法组成：多元匹配算法和自然语言处理算法。匹配算法基于多个参数最优地选择能够与受影响司法管辖区的应急管理者协作处理传入消息的其他应急管理者。采用自然语言处理算法，基于智能关键字驱动的优先级排序方法对传入消息进行重新排序。重新确定优先顺序是必要的，因为某些信息比其他信息更需要立即关注。这项创新的更广泛/更广泛的商业影响将是增强国家紧急通知能力和有效性的社会效益，这反过来又加强了国家的应急准备和响应努力，有可能在危及生命的紧急情况下拯救生命。目前使用社交媒体进行紧急警报的应急管理人员将证明其作为警报工具的力量，能够快速、廉价地向目标公民提供详细信息；同时立即与公民建立双向沟通，以便接收援助请求以及对局势事态发展的目击者和第一手描述。然而，同样的应急经理会警告说，随着越来越多的公民加入他们的社交媒体账户，已经很大数量的传入消息正在增长，这最终将使应急经理手动处理所有传入消息变得不可行/不可持续，特别是在警报发布后的一段时间内，当时他们的工作重点是应对紧急事件。拟议的活动旨在开发一个基于智能协作的系统，使社交媒体消息处理工作量在紧急响应期间保持可管理。这项服务以保险为灵感的商业模式确保了应急经理的负担能力和业务的商业成功。

The innovation is a software engine that empowers emergency managers to analyze and process the high volume of incoming citizen messages (e.g., inquiries, calls for help, etc.) on the emergency manager's social media site(s) (e.g., Facebook, Twitter), typically experienced after an emergency alert is issued over one or more social media sites. The core of the Internet-connected software engine consists of two algorithms: a multivariate matching algorithm and a natural language processing algorithm. The matching algorithm optimally selects, based on a number of parameters, other emergency managers who can collaborate with the affected jurisdiction's emergency manager on handling the incoming messages. The natural language processing algorithm is employed to reorder incoming messages based on an intelligent keyword-driven prioritization method. The reprioritization is necessary since certain messages warrant more immediate attention than others.The broader/commercial impact of the innovation will be the societal benefit of enhancing the nation's emergency notification capability and effectiveness, which in turn bolsters the nation's emergency preparedness and response efforts, potentially saving lives during life-threatening emergencies. Emergency managers currently using social media for emergency alerting will attest to its power as an alerting tool in terms of reaching targeted citizens quickly, inexpensively, and with detailed information; while immediately establishing two-way communication with citizens for the purpose of receiving requests for assistance as well as eyewitness and first-hand accounts of situational developments. Yet, the same emergency managers will warn that the already large number of incoming messages is growing as more citizens join their social media accounts, which will eventually make it unviable/unsustainable for emergency managers to manually process all the incoming messages, especially during period after the alert is issued, when their efforts are focused on responding to the emergency incident. The proposed activity is aimed at developing an intelligent collaboration-based system that will keep the social media message-processing workload manageable during emergency response periods. The insurance-inspired business model for the service ensures affordability to emergency managers and commercial success for the business.