THE CHECK AND PROCESS OF KNOWLEDGE-BASE QUALITY OF MINING ENGINEERING EXPERT SYSTEM

The quality of knowledge in knowledge base is a kernel factor determining whether the mining engineering expert system is of high performances. In this paper, possible knowledge faults of knowledge base in mining engineering expert system and their reasons are analysed at first. Two check methods based on rule-relation tree and running history of knowledge base are proposed. These methods can rather completely check out possible knowledge faults in knowledge base. Some efficient measures are also suggested to process various possible faults occuring in different locations which are very useful to develop expert systems for mining engineering and to improve the performance and quality of their knowledge base.