SYSTEM FOR EXTRACTING KNOWLEDGE OF TYPICALITY AND EXCEPTIONALITY FROM A DATABASE OF CASE RECORDS.

摘要

Knowledge tree building system iteratively divides database of case records Cheng Shu conceptually significant clusters. Each cluster is automatically assigned unique conceptual sense according to its unique typicalness and exception in knowledge tree; It is required for not having existing domain-dependent knowledge. This system makes full use of quantitative and qualitative case record data available in all varieties. Artificial intelligence application such as pattern classification and nonmonotonic reasoning are particularly well adapted to by the building of systematic knowledge tree.