We explore self-organizing strategies for role assignment in a foraging task carried out by a colony of artificial agents. Our strategies are inspired by various mechanisms of division of labor (polyethism) observed in eusocial insects like ants, termites, or bees. Specifically we instantiate models of caste polyethism and age or temporal polyethism to evaluated the benefits to foraging in a dynamic environment. Our experiment is directly related to the exploration/exploitation trade of in machine learning.

In large systems, it is important for agents to learn to act effectively, but sophisticated multi-agent learning algorithms generally do not scale. An alternative approach is to find restricted classes of games where simple, efficient algorithms converge. It is shown that stage learning efficiently converges to Nash equilibria in large anonymous games if best-reply dynamics converge. Two features are identified that improve convergence. First, rather than making learning more difficult, more agents are actually beneficial in many settings. Second, providing agents with statistical information about the behavior of others can significantly reduce the number of observations needed.

While computers assist humans with tasks such as navigation that involve spatial aspects, agents that can interact in a meaningful way in this context are still in their infancy. One core issue is the mismatch in the representation of spatial information a computer-based system is likely to use, and the one a human is likely to use. Computers are better suited for quantitative schemes such as maps or diagrams that rely on measurable distances between entities. Humans frequently use higher-level, domain-specific conceptual representations such as buildings, rooms, or streets for orientation purposes. Combined with the person-centric world view that we often assume when we refer to spatial information, it is challenging for agents to convert statements using spatial references into assertions that match their own internal representation. In this paper, we discuss an approach that uses natural language processing and information extraction tool kits to identify entities and statements about their spatial relations. These extractions are then processed by a spatial reasoner to convert them from the human conceptual space into the quantitative space used by the computer-based agent.

Humans naturally use multiple modes of instruction while teaching one another. We would like our robots and artificial agents to be instructed in the same way, rather than programmed. In this paper, we review prior work on human instruction of autonomous agents and present observations from two exploratory pilot studies and the results of a full study investigating how multiple instruction modes are used by humans. We describe our Bootstrapped Learning User Interface, a prototype multiinstruction interface informed by our human-user studies.

It has been demonstrated that people have a tendency to adapt both their linguistic representations and physical Communication is often required between agents as they actions in response to those they are interacting with, i.e., attempt to solve collaborative multi-agent tasks. This is they tend to formulate behavior and speech that will be particularly true in conditions in which an agent is working salient and sensible to a collaborating partner (Whittaker alongside a human--clearly, conventional electronic 2003). Collaboration in humans occurs via a process in communication is not feasible in this scenario; rather, these which people align their linguistic representations of the agents, including humans, must take advantage of physical environment allowing for more effective communicative communication in the shared context to confer necessary behavior. This alignment is achieved via a process in information. As an agent observes the actions of the others, which local alignment of environmental representations, it must modify its own behavior accordingly.

Trust is an approach to managing the uncertainty about autonomous entities and the information they store, and so can play an important role in any decentralized system. As a result, trust has been widely studied in multiagent systems and related fields such as the semantic web. Here we introduce a simple approach to reasoning about trust with logi

The past decade has witnessed an increasing interest in the use of virtual coaches in healthcare. This paper describes a virtual coach to provide mindfulness meditation training, and the coaching support necessary to begin a regular practice. The coach is implemented as an embodied conversational character, and provides mindfulness training and coaching support via a web-based application. The coach is represented as a female character, capable of showing a variety of affective and conversational expressions, and interacts with the user via a mixed-initiative, text-based, natural language dialogue. The coach adapts both its facial expressions and the dialogue content to the user’s learning needs and motivational state. Findings from a pilot evaluation study indicate that the coach-based training is more effective in helping users establish a regular practice than self-administered training via written and audio materials. The paper concludes with an analysis of the coach features that contribute to these results, discussion of key challenges in affect-adaptive coaching, and plans for future work.

Formations are often seen in nature, and bring many benefits for the group as a whole. They can allow a group to explore a large area more effectively, can ease movement of the group through the environment, and can increase group perceptual coverage and increase defensive capabilities, for example. The benefits of any particular formation vary and are obtained from the structure the formation provides. Robotic formations can have similar applications. To date, the techniques used and formations employed in robotic applications are significantly simpler than those seen in nature. Current techniques often require some level of global knowledge, central processing or other unrealistic assumptions. We seek to develop a formation control technique that has as few of these limitations as possible. Each agent under our approach has only local knowledge of the environment, uses no broadcast communication, and can communicate only over a limited range. Formations are achieved by organizing agents into a graph structure, where agents occupying the vertices take on the role of maintaining an appropriate number of agents on each edge, thus preserving the formation's shape and scale. We do not assume a known or static population: the evolving formation acts as a physical data structure to assist in placing and rearranging agents as the population changes. This approach does not require a global coordinate system, fixed positions within the formation, or any single lead agent. All agents within our approach are peers, and any can adopt any role within the formation.

If computational social science is to find practical application in informing policy decisions and proportionately analyzing courses of action, then it will have to make progress in the area of composition of social models.  Since a single simulation cannot hold a world of information, policy makers need to switch in and out modules in federations of simulations to test policies against all possible social environments.  Voting processes as they occur in nature, both in the form of cognition in a human mind of disparate world views, and in the form of equilibria seeking coevolution of species, inform how to combine model results externally and deeply, respectively.  These algorithms, which use the same principles of soft computation found in nature, enable any models to mesh together, even if they have different ontologies, or their data conflict, regardless of the degree they overlap.  A whiteboard architecture in which models report in their own ontologies how other models may inform them and what they have to offer other models, is a framework for the arbitrary meshing of social models.

This paper describes a framework for an agent to learn verb-phrase meanings from human teachers and combine these models with environmental dynamics so the agent can enact verb commands from the human teacher. This style of human/agent interaction allows the human teacher to issue natural-language commands and demonstrate ground actions, thereby alleviating the need for advanced teaching interfaces or difficult goal encodings. The framework extends prior work in apprenticeship learning and builds off of recent advancements in learning to recognize activities and modeling domains with multiple objects. In our studies, we show how to both learn a verb model and turn it into reward and heuristic functions that can then be composed with a dynamics model. The resulting "combined model" can then be efficiently searched by a sample-based planner which determines a policy for enacting a verb command in a given environment. Our experiments with a simulated robot domain show this framework can be used to quickly teach verb commands that the agent can then enact in new environments.