Anticipating SurpriseKey research areas for OAS include forecasting events related to science and technology (S&T); social, political, and economic crises; epidemiology and biosecurity; counterintelligence; and cybersecurity.Office for Anticipating SurpriseOAS_feaca570-0bbf-11e5-9927-d7c0363d1fb3The Office for Anticipating Surprise (OAS) focuses on characterizing and reducing uncertainty through anticipatory intelligence. The Office executes scientific research programs that develop new capabilities to deliver timely and accurate forecasts for a range of events relevant to national security.IARPADr. Jason MathenyAssociate Office DirectorDr. Dewey MurdickAssociate Office DirectorUncertainty is characterized and reduced through anticipatory intelligence_feaca782-0bbf-11e5-9927-d7c0363d1fb3To execute scientific research programs that develop new capabilities to deliver timely and accurate forecasts for a range of events relevant to national security._feaca872-0bbf-11e5-9927-d7c0363d1fb3Technical DiversityOur programs are technically diverse, but each program:ForecastingDevelops technologies to generate timely forecasts for well-defined events and their characteristics (e.g., who, what, when, where, and how).TimelinessTestingUses a rigorous, open and ongoing test and evaluation process.EvaluationRigorOpennessMetricsHas metrics that include lead time, accuracy, false positive and false negative rates, and are calculated by comparing forecasts to real-world events.AccuracyCommunicationCommunicates forecasts in context.ContextACEEnhance the accuracy, precision, and timeliness of intelligence forecasts for a broad range of event types._feaca91c-0bbf-11e5-9927-d7c0363d1fb31Steven RieberProgram ManagerIntelligence AnalystsForeSTRelated ProgramFUSERelated ProgramOSIRelated ProgramForecasting, human judgment, machine learning, logic and critical thinking -- The goal of the ACE Program is to dramatically enhance the accuracy, precision, and timeliness of intelligence forecasts for a broad range of event types, through the development of advanced techniques that elicit, weight, and combine the judgments of many intelligence analysts. The ACE Program seeks technical innovations in the following areas: (a) efficient elicitation of probabilistic judgments, including conditional probabilities for contingent events; (b) mathematical aggregation of judgments by many individuals, based on factors that may include: past performance, expertise, cognitive style, metaknowledge, and other attributes predictive of accuracy; and (c) effective representation of aggregated probabilistic forecasts and their distributions. The ACE Program will build upon technical achievements of past research and on state-of-the-art systems used today for generating probabilistic forecasts from widely-dispersed experts. The program will involve empirical testing of forecasting accuracy against real events._feaca9bc-0bbf-11e5-9927-d7c0363d1fb3CAUSEDevelop cyber-attack forecasting methods and detect emerging cyber phenomena._feacaa70-0bbf-11e5-9927-d7c0363d1fb32Robert RahmerProgram ManagerCyber DefendersCybersecurity, cyber-event forecasting, cyber-actor behavior and cultural understanding, threat intelligence, threat modeling, cyber-event coding, cyber-kinetic event detection -- Cyber-attack Automated Unconventional Sensor Environment (CAUSE) -- The Intelligence Advanced Research Projects Activity (IARPA) will host a Proposers' Day Conference for the Cyber-attack Automated Unconventional Sensor Environment (CAUSE) Program on January 21, 2015, in anticipation of the release of a new solicitation in support of the Program. The Conference will be held from 9:00 AM to 4:00 PM EST in the Washington, DC metropolitan area. The purpose of the Conference will be to provide introductory information on CAUSE and the research problems that the Program aims to address, to respond to questions from potential proposers, and to provide a forum for potential proposers to present their capabilities and identify potential team partners. Approaches to cyber defense typically focus on post-mortem analysis of the various attack vectors utilized by adversaries. As attacks have evolved and increased over the years, established approaches (e.g., signature-based detection, anomaly detection) have not adequately enabled cybersecurity practitioners to get ahead of these threats. This has led to an industry that has invested heavily in analyzing the effects of cyber-attacks instead of analyzing and mitigating the “cause” of cyber-attacks. The CAUSE Program seeks to develop cyber-attack forecasting methods and detect emerging cyber phenomena to assist cyber defenders with the earliest detection of a cyber-attack (e.g., Distributed Denial of Service (DDoS), successful spearphishing, successful drive-by, remote exploitation, unauthorized access, reconnaissance). The CAUSE Program aims to develop and validate unconventional multi-disciplined sensor technology (e.g., actor behavior models, black market sales) that will forecast cyber-attacks and complement existing advanced intrusion detection capabilities. Anticipated innovations include: methods to manage and extract huge amounts of streaming and batch data, the application and introduction of new and existing features from other disciplines to the cyber domain, and the development of models to generate probabilistic warnings for future cyber events. Successful proposers will combine cutting-edge research with the ability to develop robust forecasting capabilities from multiple sensors not typically used in the cyber domain. The CAUSE Program will consist of both unclassified and optional classified research activities and expects to draw upon the strengths of academia and industry through collaborative teaming. It is anticipated that teams will be multidisciplinary and might include computer scientists, data scientists, social and behavioral scientists, mathematicians, statisticians, content extraction experts, information theorists, and cyber-security subject matter experts having applied experience with cyber capabilities._feacab10-0bbf-11e5-9927-d7c0363d1fb3CREATEImprove analytic thinking by combining structured reasoning techniques with crowdsourcing._feacabce-0bbf-11e5-9927-d7c0363d1fb33Steven RieberProgram ManagerForecasting, logic and critical thinking, human judgment -- Crowdsourcing Evidence, Argumentation, Thinking and Evaluation (CREATE) -- The Intelligence Advanced Research Projects Activity (IARPA) will host a Proposers' Day conference for the Crowdsourcing Evidence, Argumentation, Thinking and Evaluation (CREATE) program on June 30, 2015, in anticipation of the release of a new solicitation in support of the program. The conference will be held from 9:00 AM to 4:00 PM EDT in the Washington, DC metropolitan area. The purpose of the conference will be to provide introductory information on CREATE and the research problems that the program aims to address, to respond to questions from potential proposers and to provide a forum for potential proposers to present their capabilities and identify potential team partners. This announcement serves as a pre-solicitation notice and is issued solely for information and planning purposes. The Proposers' Day conference does not constitute a formal solicitation for proposals or proposal abstracts. Conference attendance is voluntary and is not required to propose to future solicitations (if any) associated with this program. IARPA will not provide reimbursement for any costs incurred to participate in this Proposers' Day. Background and Program Goals -- CREATE aims to improve analytic thinking by combining structured reasoning techniques with crowdsourcing. CREATE will develop and test methods to help dispersed groups of individuals identify and evaluate the structure and contents of arguments in relation to alternative hypotheses. Intelligence analysts, along with professionals in other fields, assess competing hypotheses in light of multiple considerations, including reasons, evidence and assumptions. Reasons and evidence often differ in credibility and diagnosticity. CREATE will develop (1) structured methods to elicit and aggregate the elements of an argument and (2) ways to crowdsource the use of these methods, so that many individuals can collectively develop and refine an argument. The methods will be capable of treating reasoning involving quantitative and qualitative information. The CREATE program expects to draw upon the strengths of academia and industry through collaborative teaming. It is anticipated that teams will be multidisciplinary and might include social and behavioral scientists, experts in informal logic and computer scientists ArgumentsDevelop structured methods to elicit and aggregate the elements of an argument._feacac82-0bbf-11e5-9927-d7c0363d1fb33.1CrowdsourcingCrowdsource the use of these methods, so that many individuals can collectively develop and refine an argument._feacad40-0bbf-11e5-9927-d7c0363d1fb33.2ForeSTDevelop and test methods for generating accurate forecasts for significant science and technology (S&T) milestones._feacae12-0bbf-11e5-9927-d7c0363d1fb34Jason MathenyProgram ManagerS&T ExpertsACERelated ProgramFUSERelated ProgramOSIRelated ProgramForecasting, human judgment, machine learning, technical emergence, text analytics, big data, natural language processing -- Forecasting Science & Technology (ForeST) -- The goal of the ForeST Program is to develop and test methods for generating accurate forecasts for significant science and technology (S&T) milestones, by combining the judgments of many experts. The ForeST Program seeks technical innovations in the following areas: a) efficient elicitation and aggregation of judgments using combinatorial prediction markets; b) generation of S&T forecasting questions from indicators within the scientific and patent literatures; c) methods for crowdsourcing question development and resolution with over 10,000 scientists and engineers, globally. As part of this program, ForeST-funded researchers manage the world’s largest S&T forecasting tournament, www.SciCast.org, generating public forecasts for hundreds of real-world S&T events. The ForeST Program directly leverages the programmatic and technical achievements of IARPA's ACE and FUSE programs._feacaeda-0bbf-11e5-9927-d7c0363d1fb3FUSEDevelop automated methods that aid in the systematic, continuous, and comprehensive assessment of technical emergence using information found in published scientific, technical, and patent literature._feacafa2-0bbf-11e5-9927-d7c0363d1fb35Dewey MurdickProgram ManagerTechnical emergence, text analytics, knowledge discovery, big data, social network analysis, natural language processing, forecasting, machine learning -- Foresight and Understanding from Scientific Exposition (FUSE) -- Today, the identification and assessment of emerging technical capabilities is a time-consuming, domain-specific, and expert-intensive process. This demanding process is often carried out under severe time constraints on either too much or too little data, with limited reproducible auditing and bias controls, and with limited systematic validation against real world activities. furthermore, the increasing globalization of science and technology raises the potential for high-impact technical capabilities to emerge in increasingly diverse technical, socio-economic, and geographic areas. Analysts and subject-matter experts need a reliable, evidence-based capability that allows them to dramatically accelerate the horizon-scanning process and reduce the labor involved to identify specific technical areas for in-depth review. It is essential that an automated capability can nominate both known and novel technical areas based on quantified indications of technical emergence with sufficient supporting evidence and arguments for that nomination. It is anticipated that FUSE technology will provide new analytic tools to help analysts maintain technical vigilance, across all disciplines and multiple languages, in the face of the exponentially growing flood of textual content. The FUSE program seeks to develop automated methods that aid in the systematic, continuous, and comprehensive assessment of technical emergence using information found in published scientific, technical, and patent literature. A fundamental hypothesis of the FUSE program is that real-world processes of technical emergence leaves discernible traces in the public scientific and patent literature. FUSE is creating a system that can (1) process the massive, multi-discipline, growing, noisy, and multilingual body of scientific and patent literature from around the world; (2) automatically generate and prioritize technical terms within emerging technical areas, nominate those that exhibit technical emergence, and provide compelling evidence for the emergence; and (3) provide this capability for literature in the English and Chinese languages. Technology developed from the FUSE program would automatically nominate both known and novel technical terms based on quantified indicators of technical emergence with sufficient supporting evidence and arguments for that nomination. The FUSE program also addresses the vital challenge of validating such a system using real-world data._feacb0b0-0bbf-11e5-9927-d7c0363d1fb3MercuryDevelop methods for continuous, automated analysis of SIGINT in order to anticipate and/or detect political crises, disease outbreaks, terrorist activity, and military actions._feacb182-0bbf-11e5-9927-d7c0363d1fb36Kristen JordanProgram ManagerSIGINT analytics, event forecasting, machine learning, streaming data, data fusion, weapons of mass destruction, chemical/biological warfare, human biomarkers, emerging biotechnologies -- Program Description and Goals -- Past research has found that publicly available data can be used to accurately forecast events such as political crises and disease outbreaks. However, in many cases, relevant data are not available, have significant lag times, or lack accuracy. Little research has examined whether data from foreign Signals Intelligence (SIGINT) can be used to improve forecasting accuracy in these cases. The Mercury Program seeks to develop methods for continuous, automated analysis of SIGINT in order to anticipate and/or detect political crises, disease outbreaks, terrorist activity, and military actions. Anticipated innovations include: development of empirically driven sociological models for population-level behavior change in anticipation of, and response to, these events; processing and analysis of streaming data that represent those population behavior changes; development of data extraction techniques that focus on volume, rather than depth, by identifying shallow features of streaming SIGINT data that correlate with events; and development of models to generate probabilistic forecasts of future events. Successful proposers will combine cutting-edge research with the ability to develop robust forecasting capabilities from SIGINT data. Mercury will not fund research on U.S. events, or on the identification or movement of specific individuals, and will only leverage existing foreign SIGINT data for research purposes. The Mercury Program will consist of both unclassified and classified research activities and expects to draw upon the strengths of academia and industry through collaborative teaming. It is anticipated that teams will be multidisciplinary, and might include social scientists, mathematicians, statisticians, computer scientists, content extraction experts, information theorists, and SIGINT subject matter experts with applied experience in the U.S. SIGINT System.Sociological ModelsDevelop sociological models for population-level behavior change in anticipation of, and response to, these events._feacb25e-0bbf-11e5-9927-d7c0363d1fb36.1Data Processing & AnalysisProcess and analyze streaming data that represent those population behavior changes._feacb358-0bbf-11e5-9927-d7c0363d1fb36.2Data ExtractionDevelop data extraction techniques that focus on volume, rather than depth, by identifying shallow features of streaming SIGINT data that correlate with events. _feacb43e-0bbf-11e5-9927-d7c0363d1fb36.3Probabilistic ForecastsDevelop models to generate probabilistic forecasts of future events._feacb524-0bbf-11e5-9927-d7c0363d1fb36.4OSIDevelop methods for continuous, automated analysis of publicly available data in order to anticipate and/or detect significant societal events._feacb628-0bbf-11e5-9927-d7c0363d1fb37Jason MathenyProgram ManagerACERelated ProgramForeSTRelated ProgramFUSERelated ProgramLarge/errorful data sets, forecasting, public health, machine learning -- Open Source Indicators (OSI) -- The OSI Program aims to develop methods for continuous, automated analysis of publicly available data in order to anticipate and/or detect significant societal events, such as political crises, humanitarian crises, mass violence, riots, mass migrations, disease outbreaks, economic instability, resource shortages, and responses to natural disasters. Performers will be evaluated on the basis of warnings that they deliver about real-world events. If successful, OSI methods will "beat the news" by fusing early indicators of events from multiple publicly available data sources and types._feacb70e-0bbf-11e5-9927-d7c0363d1fb3SCITEDevelop and test methods to detect insider threats._feacb7fe-0bbf-11e5-9927-d7c0363d1fb38Paul LehnerProgram ManagerEngineering enterprises that detect low probability events with low accuracy sensors, innovative research methods to evaluate analytic and forecasting tradecraft, innovative statistical methods to estimate performance of systems addressing complex analysis and forecasting problems, scientific research on organizational lessons-learned methods, evidence-based forecasting methods, inductive logic, probabilistic reasoning and its application to analytic tradecraft -- Scientific advances to Continuous Insider Threat Evaluation (SCITE) -- The Intelligence Advanced Research Projects Activity (IARPA) will host a Proposers' Day conference for the Scientific advances to Continuous Insider Threat Evaluation (SCITE) program on 16 April 2015, in anticipation of the release of a new solicitation in support of the program. The conference will be held from 9:00 AM to 4:00 PM EDT in the Washington, DC metropolitan area. The purpose of the conference will be to provide introductory information on SCITE and the research problems that the program aims to address, to respond to questions from potential proposers, and to provide a forum for potential proposers to present their capabilities and identify potential team partners. This announcement serves as a pre-solicitation notice and is issued solely for information and planning purposes. The Proposers' Day conference does not constitute a formal solicitation for proposals or proposal abstracts. Conference attendance is voluntary and is not required to propose to future solicitations (if any) associated with this program. IARPA will not provide reimbursement for any costs incurred to participate in this Proposers' Day. Background and Program Goals -- Insider threats are individuals with privileged access within an organization who are, or intend to be, engaged in malicious behaviors such as espionage, sabotage or violence. The SCITE program seeks to develop and test methods to detect insider threats, through two separate research tracks. The first track of research will develop a new class of indicators, called active indicators, and associated automated detection tools. The SCITE program will develop and rigorously test a diverse array of potential active indicators. The second track of research will develop Inference Enterprise Models (IEM)—models of enterprises organized around detecting insider threats. An IEM forecasts the accuracy of an enterprise in detecting potential threats. SCITE research will develop flexible IEM approaches that can be used to forecast performance of specified subsets of an enterprise (e.g., forecast the impact of adding a new tool to find instances of a specific behavior) or complete enterprise models (e.g., forecast performance of enterprises that employ diverse tools). The SCITE program expects to draw upon the strengths of academia and industry through collaborative teaming. It is anticipated that teams will be multidisciplinary and might include computer scientists, data scientists, social and behavioral scientists, mathematicians, statisticians, and subject matter experts having applied experience with personnel security and insider threat detection.Active IndicatorsDevelop a new class of indicators, called active indicators, and associated automated detection tools._feacb902-0bbf-11e5-9927-d7c0363d1fb38.1Insider ThreatsDevelop Inference Enterprise Models (IEM) -- models of enterprises organized around detecting insider threats._feacba06-0bbf-11e5-9927-d7c0363d1fb38.22015-06-05http://www.iarpa.gov/index.php/about-iarpa/anticipating-surpriseOwenAmburOwen.Ambur@verizon.net