Welcome!
To use the personalized features of this site, please log in or register.
If you have forgotten your username or password, we can help.
|
 |
An inductive logic programming query language for database mining
Extended abstract
| |
|
An inductive logic programming query language for database mining
Extended abstract
Luc De Raedt1 
| (1) |
Department of Computer Science, Katholieke Universiteit Leuven, Celestijnenlaan 200A, B-3001 Heverlee, Belgium |
Abstract
First, a short introduction to inductive logic programming and machine learning is presented and then an inductive database
mining query language RDM (Relational Database Mining language). RDM integrates concepts from inductive logic programming,
constraint logic programming, deductive databases and meta-programming into a flexible environment for relational knowledge
discovery in databases. The approach is motivated by the view of data mining as a querying process (see Imielinkski and Mannila,
CACM 96). Because the primitives of the presented query language can easily be combined with the Prolog programming language,
complex systems and behaviour can be specified declaratively. Integrating a database mining querying language with principles
of inductive logic programming has the added benefit that it becomes feasible to search for regularities involving multiple
relations in a database. The proposal for an inductive logic programming query language puts inductive logic programming into
a new perspective.
Keywords database mining query language - inductive logic programming - relational learning - inductive query language - data mining
Fulltext Preview (Small, Large)
 References secured to subscribers.
|
|
|
|
|
|