next up previous
Next: Maintaining Data Repositories Up: Mechanisms of Cooperation Previous: Refining Service Requests



Facilitation plays a central role in OAA. At its core, our notion of facilitation is similar to that proposed by Genesereth [[Genesereth and Singh1993]] and others. In short, a facilitator maintains a knowledge base that records the capabilities of a collection of agents, and uses that knowledge to assist requesters and providers of services in making contact. But our notion of facilitation is also considerably stronger in four respects.

First, it encompasses a very general notion of transparent delegation, which means that a requesting agent can generate a request, and a facilitator can manage the satisfaction of that request, without the requester needing to have any knowledge of the identities or locations of the satisfying agents. In some cases, such as when the request is a data query, the requesting agent may also be oblivious to the number of agents involved in satisfying a request. Transparent delegation is possible because agents' capabilities (solvables) are treated as an abstract description of a service, rather than as an entry point into a library or body of code.

Second, an OAA facilitator is distinguished by its handling of compound goals (introduced in Section 5.3.1). This involves three types of processing: delegation, that is, determination of who (which specific agents) will execute a compound goal and how (combination and routing of results from subgoals); optimization of the completed goal, including parallelization where appropriate; and interpretation of the optimized goal. The delegation step results in a goal that is unambiguous as to its meaning and as to the agents that will participate in satisfying it. Completing the addressing of a goal involves the selection of one or more agents to handle each of its subgoals (that is, each subgoal for which this selection has not been specified by the requester). In doing this, the facilitator uses its knowledge of the capabilities of its client agents (and possibly of other facilitators, in a multifacilitator system). It may also use strategies or advice specified by the requester, as explained below. The optimization step results in a goal whose interpretation will require as few exchanges as possible, between the facilitator and the satisfying agents, and can exploit parallel efforts of the satisfying agents, wherever this does not affect the goal's meaning. The interpretation of a goal involves the coordination of requests to the satisfying agents, and assembling their responses into a coherent whole, for return to the requester.

The third respect in which OAA facilitation extends the basic concept of facilitation is that the facilitator can employ strategies and advice given by the requesting agent, thus resulting in a variety of interaction patterns that may be instantiated in the satisfaction of a request. Some of these strategies are mentioned in Section 5.4, and additional possibilities under consideration are mentioned in Section 11.

Finally, the OAA concept of facilitation has been generalized so as to handle the distribution of both data update requests and requests for installation of triggers, using some of the same strategies that are employed in the delegation of service requests. (Triggers and data maintenance mechanisms are discussed in sections 7 and 6 respectively.)

It should be noted that the reliance on facilitation is not absolute; that is, there is no hard requirement that requests and services be matched up by the facilitator, or that interagent communications go through the facilitator. (Indeed, as mentioned elsewhere, there is support in the agent library for explicit addressing of requests, and planned support for peer-to-peer communications.) However, OAA has been designed so as to encourage developers to employ the paradigm of community, and to minimize their development effort in doing so, by taking advantage of the facilitator's provision of transparent delegation and handling of compound goals.

In summary, we stress that a facilitator is always viewed as a coordinator, not a controller, of cooperative task completion. The facilitator never initiates an activity, but rather responds to requests to manage the satisfaction of some goal, the update of some data repository, or the installation of a trigger by the appropriate agent or agents. This approach makes it possible for all agents to take advantage of the facilitator's expertise in delegation, and its up-to-date knowledge about the current membership of a dynamic community. In addition, in many situations, the facilitator's coordination services allows the developer to lessen the complexity of individual agents, resulting in a more manageable software development process, and enabling the creation of lightweight agents.

next up previous
Next: Maintaining Data Repositories Up: Mechanisms of Cooperation Previous: Refining Service Requests

Adam Cheyer
Mon Oct 19 17:14:26 PDT 1998