Humanitarian Assistance/Disaster Relief (HA/DR) Social Media Analytics Tools Navy SBIR FY2012.1 Sol No.: Navy SBIR FY2012.1 Topic No.: N121-092 Topic Title: Humanitarian Assistance/Disaster Relief (HA/DR) Social Media Analytics Tools Proposal No.: N121-092-0447 Firm: Progeny Systems Corporation 9500 Innovation Drive Manassas, Virginia 20110 Contact: Gary Sikora Phone: (703) 368-6107 Web Site: www.progeny.net Abstract: We participated in the FEMA National Level Exercise 2011 which staged a New Madrid fault earthquake event. Our role was to support a Crisis Management Decision Support System (DSS) used by NORTHCOM to manage FEMA Mission Assignments. A primary finding was the need to "integrate social media monitoring and reporting into the overall operational picture" akin to this topic's objectives. The plan is to leverage our Clinical Surveillance DSS that employs innovated Ontology-based NLP, crowdsourcing and cloud services technologies for syndromic surveillance use-cases - the initial prototype will address health risks caused by disasters. There are hundreds of social media analytics within academia, and the open and commercial marketplace that will be researched and evaluated, with openness being a strong criterion. A mashup type open architecture will be developed to allow end users to seamlessly plug in new or improved text analytics algorithms. Processed tweets will be integrated collaborative, social mapping applications and technologies. Social engineering techniques and gradient scale methodology will be used to assess information trust. The fundamental approach is initiate tweets as untrusted, whereby trust levels are increased based on various measures such as information source/mention, circumstances around tweet flow and linked conversations, and secondary social media channels. Benefits: Natural and human-caused disasters lead to human suffering and create needs that the victims cannot alleviate without assistance. Examples of disasters include hurricanes, tornadoes, floods, earthquakes, drought, blizzards, famine, war, fire, volcanic eruption, a building collapse, or a transportation wreck. A variety of international organizations offer a range of humanitarian assistance, civil affairs, human security, reconstruction, and disaster relief when a disaster event occurs. While these disaster relief organizations vary in different objectives, expertise, and resources, they all can benefit from real-time information discovery and intelligence that a Social Media Analytic Toolkit (SMAT) would provide. SMAT can be shared with the public and other governments and nongovernment organizations to improve collaboration and enhance human security. Enabled by real-time situational awareness, SMAT assist in developing improved interventions in crisis, improving civil affairs and humanitarian operations abroad and at home. Operationally SMAT will enable the creation of multilayered, real-time, intelligent maps and mashups within widely used web mapping service application and technologies. Specific to Office of Naval Research (ONR's) Humanitarian Assistance/Disaster Relief (HA/DR) Tools, Maps and Models Program development of new capabilities to assist the military in their homeland mission as well as abroad, SMAT will satisfy the following capabilities: Improve situational awareness with the public, local authorities, first responders, nongovernment agencies and other governments Mechanisms for data mining social media in a high-tempo, rapid-onset, massive disaster capable of generating thousands of "tweets" per hour for first responder purposes and needs Multilayered maps with shareable and secure layers designed to keep sensitive information secure while allowing broader distribution of less sensitive information to appropriate partners Improved means to enable first responders to collaborate using inputs from the public via crisis maps and other social media inputs We have two on-going projects that will benefit from SMAT's social media access, filtering, discovery and organization into actionable information, visualization on collaborative mapping applications and technologies, and the social engineering part of information trust, validation, or verification: one is in the disaster relief domain involving a Crisis Management Decision Support System (CM-DSS) for DoD; and the other in the Syndromic Surveillance domain involving Clinical Surveillance Decision Support System (CS-DSS) for DoD. Crisis Management Decision Support System (DSS) The DoD Crisis Management DSS variant we developed and support is called, "Department of Defense, Defense Support to Civil Authority, Automated Support System" (DDASS), which provides information and workflows across agencies aiding in Mission Assignment assessment, decision, support, and reconciliation. There are many well known reasons why real-time information with respect to survivors, rescues, secondary crisis risks, infrastructure status, life sustainment supplies, sanitary state, health trends, etc., is needed for DDASS and other Crisis Management DSS variants. crowdsourcing techniques can significantly improve the breadth, thoroughness and timelines over traditional methods used today - e.g. Haiti earthquake on January 12, 2010 used a crowdsourcing application that extracted information from multiple channels including SMS, email, Twitter and the web called Ushahidi. We participated in the FEMA National Level Exercise 2011 (NLE 2011) that took place in May 2011. The purpose of the exercise was to prepare and coordinate a multiple-jurisdictional integrated response to a national catastrophic event. Our role and responsibility was to support the DoD Crisis Management DSS variant, DDASS, across several regional DoD Coordination Elements (DCEs)/FEMA regions. Results published in the NLE 2011 Quick Look Report (QLR), June 14, 2011, supports the need solicited here to use social media, and discovering and enhancing information, for use in planning, monitoring, and pro-active execution of humanitarian missions - some applicable excerpts are sited below: There was not a national strategy or supporting technologies to address the significant, anticipated need for damage assessment. Although there was a large amount of reporting on damaged or destroyed hospitals, schools, and facilities, insufficient information was available regarding alternate operational facilities, their locations, and available resources for mass care and sheltering needs. Planning decisions did not always take into account key gaps and shortages in resources. The integration of social media monitoring and reporting into the overall COP needs further refinement if it is to be a primary method of survivor communication and assistance requests. Clinical Surveillance Decision Support System (CS-DSS) The DoD Clinical Surveillance DSS variant we our developing leverages emerging crowdsourcing and cloud services technologies into a framework for syndromic surveillance use-cases for DoD's Armed Forces Health Longitudinal Technology Application (AHLTA). Today public health researchers dedicate considerable resources to population surveillance, which requires clinical encounters with health professionals. Crowdsourcing is a low cost alternative source for tracking public health trends. Cloud services enables integration of robust, processing heavy, capabilities such as language translation, geoname transliteration, medical classification, and symbolic knowledgebase services into light-weight, mobile device applications. To illustrate, Twitter can be used for healthcare related topics, such as "I got fever 102.5 I got sore eyes my throat hurts taking Tylenol". NLP, NLU and International Classification of Diseases 10th Revision (ICD-10) medical classification cloud services can be used to extract and code the ailment (flu), the associated symptoms (fever, etc.) and treatments (Tylenol), and represent the result in a symbolic knowledgebase were mobile devices can further interact with the results for other personal, clinical, and surveillance use-cases. According to US Centers for Disease Control and Prevention (CDC), the term 'syndromic surveillance' applies to surveillance using health-related data that precede diagnosis and signal a sufficient probability of a case or an outbreak to warrant further public health response. Though historically syndromic surveillance has been utilized to target investigation of potential cases, its utility for detecting outbreaks associated with bioterrorism is increasingly being explored by public health officials. Our product core technology incorporates new methods for extracting public health information from millions of health related social network messages such as Twitter tweets. Key to this capability is a generalized medical corpus that organizes health terms into ailments, including associated symptoms and treatments, using explicit knowledge of symptoms and treatments to separate out coherent ailment groups from more general topics. This generalized medical corpus approach key capabilities includes: discovery of larger number of coherent ailments; generation of detailed ailment information (symptoms/treatments); and consistency tracking of disease rates with published government statistics. Our commercialization/transition plan involves a phased military to public deployment approach. The military deployment phase involves the AHLTA, the military's Electronic Health Record (EHR) for the Military Health System (MHS). This includes both current and new mobile product and information collaboration and sharing process. The public deployment phase involves integrating syndromic surveillance process and results with the CDC, a United States federal agen Return