Abstract
A major problem in implementing distributed intelligent systems is to achieve a proper distribution of data and control. As a vehicle for the implementation of distributed knowledge based systems, the distributed logic programming language DLP is introduced. DLP combines logic programming with object oriented features including inheritance. Moreover, DLP supports parallelism and communication between concurrently active objects. The distinguishing feature of DLP is its support for backtracking over the results of a rendez-vous between objects, thus enabling search by distributed backtracking. As an example the implementation of a distributed medical (toy) expert system is described, illustrating the use of both the declarative and control features of the language. The object oriented paradigm provides a natural means of distributing the problem solving tasks, following the hierarchical representation of the knowledge.
Get full access to this article
View all access options for this article.