A Unified Approach to Data and Meta-data Modification for Data/Knowledge Bases

Abstract

A unified approach to the modification of the spectrum of information facts in a data/knowledge base is proposed. Derived data update, conceptual schema evolution, and exception handling are indeed different policies to resolve database inconsistency due to update. We propose a unified update approach that utilizes and balances the above three update policies to handle updates at both data and conceptual schema (meta-data) levels. An experimental prototype embodies the unified update algorithm; this prototype supports a two-level database model (a higher level object model and a lower level triplet representation), unified update algorithm, update heuristics, and an end-user interface. Several examples are described to demonstrate the prototype as well as the techniques underlying it.