Multi-Relational Data Mining

Luc De Raedt Katholieke Universiteit Leuven Definition Multi-relational data mining is the subfield of knowledge discovery that is concerned with the mining of multiple tables or relations in a database. This allows it to cope with structured data in the form of complex data that cannot easily be represented using a single table, or an attribute-value representation as is common in machine learning. Relevant techniques of multi-relational data mining include those from relational learning, statistical relational learning and inductive logic programming.