Myriam Abramson

GRACE: an autonomous robot for the AAAI Robot challenge

In an attempt to solve as much of the AAAI Robot Challenge as possible, five research institutions representing academia, industry, and government integrated their research into a single robot named GRACE. This article describes this first-year effort by the GRACE team, including not only the various techniques each participant brought to GRACE but also the difficult integration effort itself.

Multi-agent systems in mobile ad hoc networks

A number of technologies are evolving that will help formulate more adaptive and robust network architectures intended to operate in dynamic, mobile environments. One technology area, mobile ad hoc networking (MANET) enables self-organizing, multi-hop heterogeneous network routing services and organization. Such technology is important in future DoD networking, especially in the forward edge of the battlespace where self-organizing, robust networking is needed. A second technology area, multi-agent systems (MAS) can enable autonomous, team-based problem solving under varying environmental conditions. Previous work done in MAS has assumed relatively benign wired network behavior and inter-agent communications characteristics that may not be well supported in MANET environments. In addition, the resource costs associated with performing inter-agent communications have a more profound impact in a mobile wireless environment. The combined operation of these technology areas, including cross-layer design considerations, has largely been unexplored to date. This paper describes ongoing research to improve the ability of these technologies to work in concert. An outline of various design and system architecture issues is first presented. We then describe models, agent systems, MANET protocols, and additional components that are being applied in our research. We present an analysis method to measure agent effectiveness and early evaluations of working prototypes within MANET environments. We conclude by outlining some open issues and areas of further work

The Human Factor in Cybersecurity: Robust & Intelligent Defense

In this chapter, we review the pervasiveness of cyber threats and the roles of both attackers and cyber users (i.e. the targets of the attackers); the lack of awareness of cyber-threats by users; the complexity of the new cyber environment, including cyber risks; engineering approaches and tools to mitigate cyber threats; and current research to identify proactive steps that users and groups can take to reduce cyber-threats. In addition, we review the research needed on the psychology of users that poses risks to users from cyber-attacks. For the latter, we review the available theory at the individual and group levels that may help individual users, groups and organizations take actions against cyber threats. We end with future research needs and conclusions. In our discussion, we first agreed that cyber threats are making cyber environments more complex and uncomfortable for average users; second, we concluded that various factors are important (e.g., timely actions are often necessary in cyber space to counter the threats of the attacks that commonly occur at internet speeds, but also the ‘slow and low’ attacks that are difficult to detect, threats that occur only after pre-specified conditions have been satisfied that trigger an unsuspecting attack). Third, we concluded that advanced persistent threats (APTs) pose a risk to users but also to national security (viz., the persistent threats posed by other Nations). Fourth, we contend that using “red” teams to search cyber defenses for vulnerabilities encourages users and organizations to better defend themselves. Fifth, the current state of theory leaves many questions unanswered that researchers must pursue to mitigate or neutralize present and future threats. Lastly, we agree with the literature that cyber space has had a dramatic impact on American life and that the cyber domain is a breeding ground for disorder. However, we also believe that actions by users and researchers can be taken to stay safe and ahead of existing and future threats.

Cooperative Multi-Agent Systems in Mobile AD HOC Networks

Two important enabling but evolving technologies supporting future DoD network-centric systems at the tactical edge are, mobile ad hoc networking (MANET) and multi-agent systems (MAS). Despite their value in enabling more autonomous network system operation scenarios, open research and engineering questions remain regarding robust interoperation, standardization, and design of these two technologies. We describe recent research and development that is helping to better understand crosslayer design issues within both MAS and MANET. We describe the problem area and the open software components developed to support our research. We summarize recent modeling and simulation advances in a mixed MAS and MANET scenario environment. MANET multicast approaches for interagent communications are discussed and described. Some early analysis of MAS performance is presented using a variety of interagent MANET communication models. The behavior of MAS autonomous cooperative teamwork and role allocation within disruptive and dynamic MANET environments is examined. We conclude by outlining open issues and areas of further work

Toward the attribution of Web behavior

As more people browse the Web to gather information, recognizing Web browsing behavior signatures can replace or complement keystroke authentication where authentication is defined as the capability of identifying an individual within a set of individuals. We claim that recurring temporal patterns of Web site visits can help identify an individual of interest and, more generally, categorize Web browsing behavior. Furthermore, just like keystroke authentication, attribution of Web behavior is not obtrusive and has applications in cyberwarfare as a new biometric technique. In this paper we describe some exploratory work and preliminary comparative results of machine learning techniques applicable to the attribution of Web browsing behavior problem.

Learning and coordination: An overview

Adaptive learning techniques can automate the large-scale coordination of multi-agent systems and enhance their robustness in dynamic environments. This paper surveys several learning approaches that have been developed to address three different aspects of coordination, namely, learning coordination behavior, team learning, and the integrated learning of trust and reputation in order to facilitate coordination in open systems including collaborative systems where artificial agents and humans interact. Although convergence in multi-agent learning is still an open research question, several applications have emerged using some of the learning techniques presented.

Coordination in Disaster Management and Response: A Unified Approach

Natural, technological and man-made disasters are typically followed by chaos that results from an inadequate overall response. Three separate levels of coordination are addressed in the mitigation and preparedness phase of disaster management where environmental conditions are slowly changing: (1) communication and transportation infrastructure, (2) monitoring and assessment tools, (3) collaborative tools and services for information sharing. However, the nature of emergencies is to be unpredictable. Toward that end, a fourth level of coordination --- distributed resource/role allocation algorithms of first responders, mobile workers, aid supplies and victims --- addresses the dynamic environmental conditions of the response phase during an emergency. A tiered peer-to-peer system architecture could combine those different levels of coordination to address the changing needs of disaster management. We describe in this paper the architecture of such a tiered peer-to-peer agent-based coordination decision support system for disaster management and response and the applicable coordination algorithms including ATF, a novel, self-organized algorithm for adaptive team formation.

Coordination Challenges and Issues in Stability, Security, Transition and Reconstruction (SSTR) and Cooperative Unmanned Aerial Vehicles

SSTR operations are becoming a priority for the Department of Defense (DoD) and are being given the same stature as the more traditional combat operations. The near-term goal of SSTR is to provide the local populace with security, restore essential services, and meet humanitarian needs, while in the longer term ensure a stable infrastructure. Large scale disasters are an example where SSTR operations can provide value. In such scenarios, local governments and non-governmental institutions are under great stress in order to respond in a timely manner to provide basic relief to the affected communities, and there may be outburst of aid from the local groups. All of this can lead to a less than desirable outcome if poorly coordinated! The use of unmanned aerial vehicles (UAVs) to support intelligence, surveillance and reconnaissance (ISR) is becoming increasingly important. These assets can enable the collection of needed information for the execution of a given set of tasks. In large scale operations, however, the ability for the UAVs to self-coordinate may be needed as it will be difficult for human operators to effectively control large teams of UAVs. This paper will begin by introducing some of the key aspects of multiagent coordination, with a focus on the operational challenges with regard to SSTR such as disaster management response as well as UAV coordination. We will then discuss the coordination challenges and gaps in order to motivate an adaptive, multiagent based approach to coordination as well as additional opportunities for research. We will conclude with a brief summary

Modeling User Behaviors to Enable Context-Aware Proactive Decision Support

The problem of automatically recognizing a user’s operational context, the implications of its shifting properties, and reacting in a dynamic manner are at the core of mission intelligence and decision making. Environments such as the OZONE Widget Framework (http://www.owfgoss.org) (OWF) provide the foundation for capturing the objectives, actions, and activities of both the mission analyst and the decision maker. By utilizing a “context container” that envelops an OZONE Application, we hypothesize that both user action and intent can be used to characterize user context with respect to operational modality (strategic, tactical, opportunistic, or random). As the analyst moves from one operational modality to another, we propose that information visualization techniques should adapt and present data and analysis pertinent to the new modality and to the trend of the shift. As a system captures the analyst’s actions and decisions in response to the new visualizations, the context container has the opportunity to assess the analyst’s perception of the information value, risk, uncertainty, prioritization, projection, and insight with respect to the current context stage. This paper will describe a conceptual architecture for an adaptive work environment for inferring user behavior and interaction within the OZONE framework, in order to provide the decision maker with context relevant information. We then bridge from our more conceptual OWF discussion to specific examples describing the role of context in decision making. Our first concrete example demonstrates how the Web analytics of a user’s browsing behavior can be used to authenticate users. The second example briefly examines the role of context in cyber security. Our third example illustrates how to capture the behavior of expert analysts in exploratory data analysis, which coupled with a recommender system, advises domain experts of “standard” analytical operations in order to suggest operations novel to the domain, but consistent with analytical goals. Finally, our fourth example discusses the role of context in a supervisory control problem when managing multiple autonomous systems.

Sequence Classification with Neural Conditional Random Fields

The proliferation of sensor devices monitoring human activity generates voluminous amount of temporal sequences needing to be interpreted and categorized. Moreover, complex behavior detection requires the personalization of multi-sensor fusion algorithms. Conditional random fields (CRFs) are commonly used in structured prediction tasks such as part-of-speech tagging in natural language processing. Conditional probabilities guide the choice of each tag/label in the sequence conflating the structured prediction task with the sequence classification task where different models provide different categorization of the same sequence. The claim of this paper is that CRF models also provide discriminative models to distinguish between types of sequence regardless of the accuracy of the labels obtained if we calibrate the class membership estimate of the sequence. We introduce and compare different neural network based linear-chain CRFs and we present experiments on two complex sequence classification and structured prediction tasks to support this claim.