Joint Interagency Field Experiment (JIFX) 14-1 VIRTUAL Event: Using Social Media for Enhanced Situational Awareness and Decision Support 19-20 November 2013 High-impact and high-visibility disasters have revealed the proliferation and widespread use of mobile devices, social media, photos, videos, and other sensory data and channels as information sources. This can be helpful in planning for, responding to, and recovering from disasters and emergencies. The amount and speed of available information, however, in addition to a lack in ability to identify, authenticate, coordinate, aggregate, and contextualize, leaves data and information often unused and un-actionable. Government agencies and response partners are responsible for identification of specific, pre-existing information requirements. These information requirements are often satisfied through traditional channels and methods. Social media, if explored strategically and in relation to the mission-specific pre- existing information requirements, or “essential elements of information”, may serve as an additional means to satisfy these requirements, providing actionable, contextualized information in real-time to both enhance existing methods, and for verification to and from the field. EVENT FOCUS. In order to facilitate this experiment, JIFX will host a 2 day virtual event aimed at achieving the following outcomes: Development of algorithms for answering pre-existing information requirements and essential elements of information associated within the transportation sector from social media data; Development of data formats and structures for social media once identified as applicable to pre- existing information requirements and essential elements of information that is technology and platform-agnostic; Development of methods for integration of structured social media data within operational environments, including integration with other transportation-related data layers, to achieve situational awareness of complex events and dynamic data sources; and Development of methods for visualization of integrated social media and other data layers. PARTICIPATION. NPS is seeking participation from organizations and individuals that may contribute to the following (through personal knowledge/skillset or developed software tool): Aggregation and Search with the ability to search based on geographic location, keyword, and/or a set of scenario-specific parameters, using natural language processing (NLP) and applying inferred context based on pre-determined, (and machine-learning) domain-specific ontologies; identify and establish baseline monitoring and detect events and applicable trends, based on user-generated thresholds and mission-specific operational requirements; infer and assign relationships for aggregation purposes and to assign location based on inference or other method; and identify and assign meaning and context to aggregated content (versus original), including consideration for temporal lapse in publication of information as well as distance and time from point of event. The following assumptions are made with regards to aggregation and search: o Information collected from social media can help to inform and enhance operational decision- making, resource allocation, and general situational awareness; o Information collected from social media can be integrated with pre-existing data layers and sensor data in order to produce contextual awareness of events, identify needs, and to inform and enhance operational decision-making, resource allocation, and general situational awareness; o Information collected from social media can be stripped of required personally identifiable information, without removing essential elements of information needed to provide situational awareness about target and specific information requirements. Authentication and Filtering to integrate crowdsourcing efforts and to provide a means for manual verification and/or comparison of crowd-sourcing results; simple GUI to enable user-generated filtering parameters; and to filter and remove publically identifiable information. Technologies that perform analysis with the ability to integrate results with pre-existing data sets and sensor data, to establish meaningful relationships and context between social media data and other information sources (automated and user-generated); predict and model potential outcomes based on relationships identified through integration of social media and other data points (automated and user-generated); and assign tags or metadata (or similar solution), and to produce notification and/or alerts, for the purposes of routing verified information, based on mission objectives and responsibilities, to the appropriate entity. Additional capabilities being considered also include the ability to share results of aggregation, filtering, and/or analysis across third party platforms and technologies, regardless of format; and produce visualization that is meaningful and applicable to mission objectives, as identified by end user; and integrate within external visualization environments, regardless of format. EXPERIMENTATION QUESTIONS: Can standing information requirements be answered or verified via social media sources? How are the answers to standing information requirements represented via social media and can they be identified in a repeatable way? What are the decision points connected to each information requirement, once answered via social media, and how can additional needs identification produce or answer additional questions as they arise? Can information identified via social media sources be tagged with standardized terminology so as to identify the type or category of the identified information? To what pre-existing information structure should tags be assigned (e.g. Emergency Support Functions (ESF), National Information Exchange Model (NEIM), National Core Capabilities, other, etc.)? Can publically identifiable information (PII) be successfully removed while maintaining useful information applicable to standing information requirements? Can the geo-location of information identified be ascertained and if so, what is the best way? Can information identified via social media sources be authenticated through verification against additional sources (e.g. existing data)? If so – how? Can meaning be established in the relationships between information identified via social media sources and pre-existing data? If so – how, and what does this look like? Interested parties should request to participate by completing the online registration. You will be notified of your acceptance.