David Jensen (1998). "Prospective Technology Assessment of Knowledge Discovery Systems." Presentation at Data Mining: The Search for Business Intelligence. Symposium presented by the Central Ohio Chapter of ACM. May 15. Columbus, Ohio.

Abstract

A poorly designed system for knowledge discovery and data mining (KDD) can be worse than no system at all. It can encourage complacency, misinform users, and impede the development of better systems. This talk presents a set of design patterns – high-level features of KDD techniques that address particular tasks. It also provides examples of implemented designs, partially based on a major federal study of KDD systems for financial fraud detection.

---

Feedback Fineprint