PATTERN BASED DATABASE KNOWLEDGE EXTRACTION

Relational databases are still the most commonly used knowledge storage in enterprise applications. The process of application integration includes integration of those databases. Semantic technologies seems to have wide area of usage in this field. The first step of semantic integration is interpretation of relational data within chosen semantic formalism. Nevertheless, straightforward interpretations are not the most effective. The approach presented in this paper is based on interpretation of frequently occurring representational patterns. Described ideas has been implemented and successfully used for integration of ArcGis with CIM, and in construction of ABB manufacturing data warehouse, facilitating its ETL phase.