Philosophies and Advances in Scaling Mining Algorithms to Large Databases

Data mining has become increasingly important as a key to analyzing, digesting and understanding the flood of digital data. Achieving this goal requires scaling mining algorithms to large databases. Many classic mining algorithms require multiple database scans and/or random access to database records. Work in this area focuses on overcoming limitations imposed when scanning a large database multiple times or accessing records at random is costly or impossible, as well as innovative algorithms and data structures to speed up computation. In this paper, we focus on illustrating scalability principles by highlighting some of the key innovations and techniques.