Comparison of View-Size Estimation Algorithms in OLAP

It must be quick, accurate, and reliable when the size of views are estimated in OLAP. Many me- thods to deal with view-size estimation apply specific statistical assumptions but their error may usually be large. In comparation, probabilistic techniques have slower speed, but the estimate has higher accuracy and reliability by using less memory. Several hashing-based view-size estimation methods were introduced and analyzed experimentally in this paper. The results showed that the Adaptive Counting provided accurate estimates regardless of the size of view, and its estimated speed remained constantly fast as the memory budget increased.