Obtaining Quick Results for Approximate Answers
Qaizar Ali Bamboat, Oliver Dunemann · 2001
In recent times the size of the databases has grown many folds and many of the decision support systems need a very fast and interactive response from these DBMS, but the queries to retrieve the results are at times very complex. In this work we have forwarded a new approach to this problem, by pre-computing the summary statistics in the form of histograms on the data and evaluating the random samples from them in order to answer the given queries quickly, but approximately. We believe that most of the decision support systems or OLAP applications can tolerate small errors in terms of getting the answers quickly. This work constitute a part of our larger work which is the development of an efcient Incremental data analysis engine.