Data Mapping and Ontology Development for Credit Data

This document is a technical report detailing the challenges faced by Dun & Bradstreet in managing their complex data products related to company credit worthiness and contact information. It outlines a significant analytic project aimed at inventorying 160,000 unique data elements associated with their products. The report describes the difficulties in mapping these products, particularly due to a shortage of experts who understand the intricate systems involved. It explains the creation of an ontology to address the mapping issues, highlighting the problem of high cardinality in data elements related to addresses. The report emphasizes that many of the 90,000 elements associated with addresses represent different contexts rather than distinct attributes. Finally, it presents a solution involving the development of an interactive web-based system that classifies these elements for clearer, more efficient mapping to the source products.