Derived data update in semantic databases.
I. A. Chen · University of Southern California Digital Library · 2014
The derived data update problem involves the transformation of modification of derived facts into corresponding changes to base facts or other derived facts. Using a database schema defined with a generic semantic database model which includes derived data specification, techniques and algorithms are provided for appropriately modifying base data when derived data are changed. Two approaches to derived data update, namely the schema-based approach and the rule-based approach, are presented here. The schema-based approach involves a structural algorithmic solution to derived data update, while the rule-based approach applies artificial intelligence rule inferencing techniques to determine derived data update propagation. Experimental prototype systems based upon the two approaches have been designed, developed, and tested. A comparative analysis of the two approaches and their corresponding prototypes, which emphasizes the organization of meta data, is also included. The study reveals that the schema-based approach is more efficient, but the rule-based approach provides more effective utilization of meta data. (Copies available exclusively from Micrographics Department, Doheny Library, USC, Los Angeles, CA 90089-0182.)