Integrating Object and Constraint Technologies for Stating and Solving Resource Scheduling Problems

Resource scheduling is a difficult and time consuming problem, especially when human resources are involved. The major reason of this difficulty emerges from the complexity and the dynamically changing nature of the regulations that restrict the solutions of problem. Constraint Logic Programming (CLP) is currently considered as the most powerful computational paradigm for stating and solving such problems. However, the statement of complex regulations is still extremely hard and results in a large semantic gap between the implied formalization and the problem regulations, as perceived by the end-user. Ideally, the end-user should be able to represent and modify directly the regulations, but since a high level formalism is not available, the interference of an expert programmer is always required. In this paper, we suggest an Object-Oriented meta-representation for the abstract, natural specification of regulations. With the target of obtaining an efficient scheduling system, we describe the relationship between our formalism and the CLP one, laying the foundation for automatic transformation between the two representations.