A Comparative Study of Data Structure for Efficient Query Processing

Abstract

Organisations are flooded with enormous amounts of data from many sources in the big data era. Effective query processing strategies are crucial for deriving significant insights and guiding well-informed judgements. MapReduce and other big data analytical tools have emerged as key concerns for numerous businesses and research teams. At the moment, similar tasks are submitted repeatedly using multi-queries that are converted into MapReduce jobs. Therefore, taking use of these related tasks may present opportunities to avoid performing MapReduce calculations repeatedly. Since then, other analyses have focused on the sharing possibility to improve multi-query processing. Therefore, the primary goal of this work is to thoroughly examine and distinguish between two predicate-based filter-based sharing opportunity strategies that are already in use: relaxed MRShare and MRShare. As far as predicate-based filters among multi-query questions, the relaxed MRShare technique performs noticeably better than the MRShare for shared data, according to a comparative study conducted using the TPC-H benchmark.