This is the 8th instantiation of this international workshop on cooperative games and multiagent systems, traditionally held as part of the AAMAS conference. The workshop is intended to focus on topics in cooperation in multiagent systems, cooperative game theory, cooperative solution concepts, coalition formation, and applications. The use of cooperative game theory to study how agents should cooperate and collaborate, along with the related topic of coalition formation, has received growing attention from the multiagent systems, game theory, and electronic commerce communities. The focus of much of the current work in this area has been on exploring methods by which agents can form coalitions so as to solve problems of joint interest, make group decisions, and distribute gains arising from such cooperation.
The domain of cooperative game theory is characterized by several unique features. First, the problem of finding good ways for agents to collaborate may require solving complex combinatorial problems, and is by itself an important research area. Second, in practice, collaborating agents are typically required to make collective decisions, coordinate and come up with joint plans. This requires designing appropriate methods for such interaction. Furthermore, in many such domains, agents are self-interested, i.e. each agent may have its own agenda or goal. Achieving stable and lasting collaboration in such environments requires careful planning. For example, gains from collaboration must be divided amongst the agents in certain ways, or the coalition would break apart. Finally, in certain such settings, agents must deal with human participants, who must be modeled properly in order to achieve good outcomes.
As a consequence of the above features, solving problems in this domain requires designing sophisticated algorithms and mechanisms, bringing together state of the art methods from several fields. In many cases, to solve such problems one needs to take into account both computational and game-theoretic considerations, and occasionally also apply a heuristic approach.
Cooperation is, and always has been, at the very core of multiagent systems research. Papers on cooperative games have been present in past AAMAS conferences (in sessions such as Game Theory, Coordination and Cooperation, Coalitions). One goal of the workshop is to gather together the researchers from the community to discuss current work in progress. The CoopMAS-2017 workshop will be an ideal environment to allow open discussion and facilitate interdisciplinary collaboration. We believe such an exchange of ideas between researchers from various fields, such as game theory, artificial intelligence, combinatorial optimization and knowledge representation will make it possible to advance the state of the art in this field from both theoretical and practical perspective. We hope that the workshop’s informal atmosphere will encourage the participants to offer a wide range of methods to address problems of collaboration between agents, and will allow achieving significant progress in this area.