Self-reconfiguration and optimisation of knowledge meshes with similar knowledge points

摘要

This paper presents a new approach to the self-reconfiguration and optimisation of knowledge meshes (KMs), based on the user's function requirements and similarity measure. A new similarity measure combined of the function similarity and layer similarity between similar knowledge points is defined, and the properties proved. A simple similarity algorithm for KMs is proposed. Aiming at user's maximum function-satisfaction, the method to the self-reconfiguration of KMs is presented. First, several appropriate KMs according to the enterprise requirements are selected and preprocessed by similarity measure. Second, a similarity threshold is given and the similarity is calculated for every pair of knowledge points in the selected KMs. The two knowledge points can be recognised as the same when the similarity between them exceeds the given threshold, and then the knowledge point with the lower function-satisfaction degree is replaced by the one with the higher function-satisfaction degree. Third, the KM multiple set operation expression is optimised by the hybrid genetic-tabu algorithm. Finally, a new KM obtained by operations on KM multiple sets can be mapped into an agent mesh (AM) for automatic reconfiguration of complex software systems. Based upon the above, the KM's optimisation are illustrated by an actual KM example which corresponds to the management information system (MIS) software used in a vehicle body plant at Nanjng in China, which shows the method to be very effective.