Run-time adaptation of knowledge-intensive processes through AI techniques: research challenges and some solutions
Andrea Marrella
, Department of Systems and Computer Science and Engineering at Sapienza – Universitá di Roma

Tuesday, February 14th, 12pm-1:30pm
Computer Science and Engineering Building – CSREB 3033

A knowledge-intensive process is one in which the people performing such process is involved in a fair degree of uncertainty. This is due to the high number of tasks to be represented and to their unpredictable nature, or to a difficulty to model the whole knowledge of the domain of interest at design time. Typically, a knowledge-intensive process can not be modeled sufficiently by classical, static process models and workflows, because as the knowledge-intensive process proceeds, the sequence of tasks depends so much upon the specifics of the context. An important aspect of knowledge-intensive processes is that they constantly evolve and often it is unpredictable the way in how they unfold. To deal with exceptions and
uncertainly introduced by such processes, the need for flexible and easy adaptable Process Management systems (aka, PMSs) has been recognized as one of the critical success factors for any PMS.

The main focus of this presentation is to discuss about how a modeling approach towards a declarative specification of process tasks, i.e., comprising the specification of input/output artifacts and task preconditions and effects, allows to layer planning techniques on top of traditional PMSs, in order to enable run-time process adaptation, which can be done without defining explicitly any recovery policy. A prototype implementation based on the above theory has been developed through the IndiGolog execution framework.