Research On Trustworthy Software Ontology Model Based On Immune Algorithm

It is known that software system plays an important role in the information society, and it has been applied in all areas of national economic and social development. However, the current quality of software systems leaves much to be desired, software invalidation that frequently occurred deeply influenced our social lives, and more seriously, led to great losses. Thus, how to improve the creditability to better serve our society has become an important research focus among areas of software theory and technology application. Based upon ontology theory, this paper abstracts software systems from a high level, meanwhile, brings in the mechanism of biological immune system to establish software ontology system, which can offer software systems immune function similar to organisms, and greatly improve the reliability, usability and integrality of software systems.The paper begins with the concept of ontology, sets up an ontology semantic model based on the concept of network, which includes concept net ontology model (CNO), concept net evolution model and self-organization of ontology semantic model. Then, with reference to immune feedback mechanisms of biological immune system precisely recognizing, responding to and then eliminating the intruding antigens, ontology model is established based on immune arithmetic. This model is abstracted into four layers, which are semantic layer, evolving control layer, service layer and immune filtering layer. The semantic layer constructed by CNO determines and controls the system behavior. The evolving control layer is in charge of self-learning and evolving, and systems can get artificial immune vaccine in time via external learning interface so that semantic layer is able to adjust and evolve by itself. The service layer realizes all business functions for the system, providing kinds of service interfaces for the outside. And the immune filtering layer realizes system inner protection by timely filtering illegal service request and answering unknown service request from outside. At last, this paper take billing system for instance, its being put into instantiation by using UML has realized system self-evolvement in kinds of situations.The ontology model of credible software based on immune algorithm possesses abilities of self-study, self-adjustment and ongoing improvement. Even though user demands or knowledge change, it can adjust and improve itself to adapt quickly to these changes, thus ensure its high credibility and provide a methodology support for credible software development.