Abstract Establishment of indexing framework of business process case-base is important in creating a processes casebase, searching and matching the process cases, improving the quality of case-matching and providing scientific decision support for case-based reasoning (CBR). Based on the features of process cases,a threelevel indexing framework for business process casebase is proposed, which consists of industry level, business one and process informatization one, to realize process case classification storage, quick retrieval and matching. The process informatization level is the core of the three-level indexing framework, and the process case clustering is realized by aiNet, which is well known in the Artificial Immune System (AIS). Taking advantage of AIS in noise tolerance, self-organization, and memory and avoiding local optimization, the artificial immune antibodies construct a three-level indexing that forms the cluster centers of the process cases. The procedure of constructing the three-level index in business process case-base was researched empirically by using 105 enterprise samples.
|