SQL: 1999, formerly known as SQL3
ACM SIGMOD Record
Machine Learning
Introducing Data Administration into a Business Organization
Proceedings of the Sixth International Conference on Entity-Relationship Approach
Proceedings of the 2003 ACM SIGMOD international conference on Management of data
The entity-relationship model: toward a unified view of data
VLDB '75 Proceedings of the 1st International Conference on Very Large Data Bases
Hi-index | 0.00 |
Data mining models are often implemented as program code and stored in an ad hoc fashion. In this paper we describe a methodology for developing model bases that can be implemented with only an extended relational database. This method stores models as what we call lightweight functions, which are straightforward textual representations of function values. Many applications use models that admit this approach. Business applications in particular, which can have a very large number of specialcase rules or business logic, are suitable for development with lightweight functions. Financial and forecasting applications give another example. We argue that the Lightweight Model Base offers some advantages for these applications. Introduction of lightweight stored functions in relational databases is a way to integrate the software-engineering methodology of tabledriven programming. This methodology advocates storing functions and data in tables. The computing process is just a mechanical evaluation of "joining" data and function relations. It would make stored business logic transparent for understanding and maintenance as relational data. Table-driven programming has much in common with statistics and data mining, and is a natural framework for combining data mining with databases.