Pattern Warehouse: Context Based Modeling and Quality Issues

摘要

abstract_textpThe conceptual and logical modelling of pattern warehouse is introduced by taking a term 'context' as virtual separator and a snowflake based schema is trying to accommodate for pattern arrangement. Moreover, various patterns are arranged in a hierarchy according to the 'context of knowledge' they carry, and it is also taken as the way of pattern's vertical normalization. The 'context of knowledge' has been incorporated in logical modeling so that query processing may be improved by revealing the sense of the underlying domain of source data. Consequently, four quality characteristics as a road map for better pattern warehouse design are investigated and suggested to incorporate during the beginning of conceptual modelling. A significant issue of pattern refreshment has been discussed in order to reflect the changes that have occurred in the sources, i.e. all the patterns extracted from pattern warehouse are guaranteed a consistent view of the same pattern extracted from the source data./p/abstract_text