Daniel D. Corkill

AAMAS Conference 2007 Conference Paper

Determining Confidence When Integrating Contributions from Multiple Agents

• Raphen Becker
• Daniel D. Corkill

Integrating contributions received from other agents is an essential activity in multi-agent systems (MASs). Not only must related contributions be integrated together, but the confidence in each integrated contribution must be determined. In this paper we look specifically at the issue of confidence determination and its effect on developing "principled, " highly collaborating MASs. Confidence determination is often masked by ad hoc contribution-integration techniques, viewed as being addressed by agent trust and reputation models, or simply assumed away. We present a domain-independent analysis model that can be used to measure the sensitivity of a collaborative problem-solving system to potentially incorrect confidence-integration assumptions. In analyses performed using our model, we focus on the typical assumption of independence among contributions and the effect that unaccounted-for dependencies have on the expected error in the confidence that the answers produced by the MAS are correct. We then demonstrate how the analysis model can be used to determine confidence bounds on integrated contributions and to identify where efforts to improve contribution-dependency estimates lead to the greatest improvement in solution-confidence accuracy.

IJCAI Conference 1989 Conference Paper

Focus of Control Through Goal Relationships

• Victor R. Lesser
• Daniel D. Corkill
• Robert C. Whitehair
• Joseph A. Hernandez

Goal relationships resulting from the initial data and subsequent processing can he used to dynamically construct a partial topology of the solution space based on what appear to be feasible solutions. This structure can be used to make control decisions that significantly reduce the amount of search required to solve a problem in a complex domain. We examine the utility of this approach in the context of a multi-level, cooperative knowledge source model of problem solving. We present a taxonomy of goal relationships for constructing partial topologies of the solution space and show that mechanisms using this information can be built as natural extensions of an integrated data-directed and goal directed archi tecture. Examples and performance results demonstrating how these additions improve the system's ability to evaluate potential activities are provided.

AAAI Conference 1988 Conference Paper

Tuning a Blackboard-Based Application: A Case Study: Using GBB

• Daniel D. Corkill

AAAI Conference 1987 Conference Paper

Achieving Flexibility, Efficiency, and Generality in Blackboard Architectures

• Daniel D. Corkill

Achieving flexibility and efficiency in blackboardbased AI applications are often conflicting goals. Flexibility, the ability to easily change the blackboard representation and retrieval machinery, can be achieved by using a general purpose blackboard database implementation, at the cost of efficient performance for a particular application. Conversely, a customized blackboard database implementation, while efficient, leads to strong interdependencies between the application code (knowledge sources) and the blackboard database implementation. Both flexibility and efficiency can be achieved by maintaining a sufficient level of data abstraction between the application code and the blackboard implementation. The abstraction techniques we present are a crucial aspect of the generic blackboard development system GBB. Applied in concert, these techniques simultaneously provide flexibility, efficiency, and sufficient generality to make GBB an appropriate blackboard development tool for a wide range of applications.

AAAI Conference 1986 Conference Paper

GBB: A Generic Blackboard Development System

• Daniel D. Corkill
• Kelly F

This paper describes a generic blackboard development system (GBB) that unifies many characteristics of the blackboard systems constructed to date. The goal of GBB is to provide flexibility, ease of implementation, and efficient execution of the resulting application system. Efficient insertion/retrieval of blackboard objects is achieved using a language for specifying the detailed structure of the blackboard as well as how that structure is to be implemented for a specific application. These specifications are used to generate a blackboard database kernel tailored to the application. GBB consists of two distinct subsystems: a blackboard database development subsystem and a control shell. This paper focuses on the database support and pattern matching capabilities of GBB, and presents the concepts and functionality used in providing an efficient blackboard database development subsystem.

IJCAI Conference 1985 Conference Paper

Increasing Coherence in a Distributed Problem-Solving Network

• Edmund H. Durfee
• Victor R. Lesser
• Daniel D. Corkill

IJCAI Conference 1983 Conference Paper

The Use of Meta-Level Control for Coordination in a Distributed Problem Solving Network

• Daniel D. Corkill
• Victor R. Lesser

AAAI Conference 1982 Conference Paper

Unifying Data-Directed and Goal-Directed Control: An Example and Experiments

• Daniel D. Corkill

Effective control in a multi-level cooperating knowledge source problem solver (such as Hearsay-II) requires the system to reason about the relationships among competing and cooperating knowledge source (KS) instantiations (both past and potential) that are working on different aspects and levels of the problem. Such reasoning is needed to assess the current state of problem solving and to develop plans for using the system’s limited processing resources to the best advantage. The relationships among KS instantations can be naturally represented when KS activity is viewed simultaneously from a data-directed and a goal-directed perspective. In this paper we show how data- and goal-directed control can be integrated into a single, uniform framework, and we present an example and experimental results of sophisticated focusing using this framework.