Immune algorithm for bitmap join indexes

Bitmap join indexes are designed to prejoin the facts and dimension tables in data warehouses modeled by a star schema. They are defined on the fact table using attributes which belong to one or many dimension tables. The index selection process has become an important issue regarding the complexity of the search space to explore. Thus, the indexes can be defined on several attributes from several dimension tables (that may contain hundreds of attributes). However, only a few selection algorithms were proposed. In this article, we present a bitmap join indexes selection approach based on artificial immune algorithm. An experimental study was conducted on the dataset generated from APB-1 benchmark in order to compare the artificial immune algorithm with other algorithms.