Skill-based Resource Allocation using Genetic Algorithms and Ontologies

Allocation of tasks to resources is a crucial issue for an organisation. This paper investigates the use of genetic algorithms, combined with an ontology, in producing optimal allocations of tutors to tutorials. The ontology is used to support partial matching; it helps determine which tutors’ skills (i.e. resources) are ontologically close to the optimal skill. These are then matched against the skill requirements of the task. The other novel feature of this GA implementation is the steady state duplication elimination algorithm, which (when coupled with 10% best chromosome crossover) produces more solutions with better fitness than a standard steady state GA.