Report on the International Workshop on Pattern Representation and Management (PaRMa'04)

The increasing ability to quickly collect and cheaply store large volumes of data, and the need for extracting concise information to be efficiently manipulated and intuitively analyzed, are posing new requirements for Database Management Systems (DBMS) in both industrial and scientific applications. A common approach to deal with huge data volumes is to reduce the available information to knowledge artifacts (i.e., clusters, rules, etc.), hereafter called patterns , through data processing methods (pattern recognition, data mining, knowledge extraction). Patterns reduce the number and size of the original information to manageable size while preserving as much as possible its hidden / interesting content. In order to efficiently and effectively deal with patterns, academic groups and industrial consortiums have recently devoted efforts towards modeling, storage, retrieval, analysis and manipulation of patterns with results mainly in the areas of Inductive Databases and Pattern Base Management Systems (PBMS).