Data Mining and
Knowledge Discovery Special
issue on
Inductive Logic Programming and
Knowledge Discovery in Databases
guest editors: Saso Dzeroski and Nada Lavraťc
Jozef Stefan Institute, Ljubljana, Slovenia
Knowledge Discovery in Databases (KDD)
is concerned with identifying interesting patterns in data and describing them in a
concise and meaningful manner. In KDD, machine learning tools are often used for data
mining and are thus present in many KDD systems and applications. However, most of these
tools use a propositional representation of both the data analysed and the knowledge being
discovered, mining in effect a single relational table in a given database.
Inductive Logic Programming (ILP) can
be viewed as machine learningin a first-order language, where both the data analysed and
the patternsconsidered can involve several relations in a relational database. Using ILP
tools for data mining offers several advantages, including the expressiveness of
first-order logic as a representation language, the ability to use structured data as well
as various forms of background knowledge and the ability to use language bias provided by
the user to define the search space of patterns considered.
The special issue on Inductive Logic
Programming and Knowledge Discovery in Databases of the journal Data Mining and Knowledge
Discovery welcomes papers that focus on algorithms and applications that involve the
discovery of knowledge expressed in a relational or first-order formalism. An indicative,
but nonexaustive list of topics is given below.
* Declarative biases for KDD
* Extending the pattern representation language in
classification and clustering to include relations and first-order-logic
* Practical schemes for encoding prior knowledge for use in data mining and KDD
* Logic-based inductive query languages
* Combining probabilistic approaches with ILP
* Use of ILP to understand/visualize/explain complex models mined from data (i.e.
as postprocessor on a mining engine)
* Scalability of ILP to large database mining problems
* Pre- and post-processing steps for applying ILP to real-world problems
* Use of ILP in novel data mining settings
* Embedding ILP into the KDD process
* Innovative Knowledge Discovery applications of ILP
* ILP and Text Mining
* Mining the Web with ILP |