Research And Application On Ontology-Based Software Maintenance Konwledge Management System

Software maintenance is the last phase of software life cycle, and it is also the longest phase in software life cycle. With the cost of software maintenance become more and more expensive, the domain researchers start to concentrate on how to more greatly reduce the cost and improve the efficiency of the software maintenance.According to the actual situation of software maintenance, the present problems and difficulties are summarized and in addition, the author points out the necessity of making best of the existent knowledge about maintenance resources and puts forward a new concept called Ontology-based Software Maintenance Knowledge Management, which has two layers of innovation. The first innovation is from the view of how to mitigate the work of maintenance, and its emphasis on managing software maintenance through the angle of knowledge management, is very different from the former engineering-like methods. By extracting and sharing software maintenance knowledge, we can locate and resolve the maintenance requests quickly. The second innovation is its applying ontology theory to solve the semantic problems occurring in the process of knowledge sharing. It uses unified concepts and its structure to get a common understanding of some specific domain knowledge among people.On the basis of understanding the therories of ontology and knowledge, the author also puts forward software maintenance knowledge and builds up software maintenance domain ontology. Furthermore, the author designs and implements an Ontology-based Software Maintenance Knowledge Management System (OSMKMS) which enables to share and manage maintenance knowledge. The most marked character of this system is its three layers structure mode of B/S, and its ontology database and knowledge database in order to implement maintenance knowledge storage, search and update. Furthermore, the retrieval modes, such as navigation retrieval and Case-based Reasoning retrieval, can put the existent maintenance knowledge into the real maintenance work effectively. Finally, in the last part of this thesis, the author summarizes the work and presents the next work of maintenance knowledge management.