Multi Relational Rules Mining in Data Warehouse

Multi relational rules mining is a new area of data mining research. There three kinds of algorithm currently. One is ILP approach, the second is AVL upgrading approach, and the last is mould a relational database into a single table to mine with traditionally single table technology. Due to the drawback of these methods, a new ideal to mine rules in data warehouse is proposed. The data schema constructing, primitive language and mining algorithm are analysis. The algorithm concerns the rules from one table to another only, but considers the single tablepsilas rules. The experiments show it has a good accuracy due to utilize the relational information in data warehouse.