Optimized Processing of a Batch of Aggregate Queries over Hidden Databases

A tremendous amount of data is concealed behind form-based interfaces that communicate any user query to their data store to deliver query answer. These interfaces limit the number of retrieved search results to the top-k matching tuples that are sorted using a proprietary ranking function; the database owner may also impose constraints on the type of the processed queries. These limitations make it difficult to process a batch of queries, in this paper; we tackle the problem of processing a batch of aggregate queries with the minimal number of queries sent to the hidden database to overcome the interface limitations. We are proposing a novel technique that makes use of the results of the fired queries to answer new aggregate queries without any additional cost. The proposed method is compared with the classical techniques of processing aggregate queries; it is evaluated through the estimation relative error and query cost. The results show that our method is more efficient than other methods in terms of query cost, so we can process a batch of queries with the minimal cost.