Date of Award


Publication Type

Master Thesis

Degree Name



Computer Science


Computer Science.


Morrissey, Joan,




Different from a centralized database system, distributed query processing involves data transmission among different sites and this communication cost is a dominant factor compared to local processing cost. So, the objective of distributed query optimization is to find strategies to minimize the amount of data transmitted over the network. Since optimal query processing in distributed database systems has been shown to be an NP-hard problem, heuristics are applied to find a near-optimal processing strategy. Previous research has mainly focused on the use of joins, semijoins, and hash semijoins (Bloom filters). The semijoin is a commonly recognized operator, which provides efficient query results. As a variation of semijoin, the composite semijoin is beneficial to do semijoins as one composite rather than as multiple single column semijoins. The Hash semijoin (which uses a Bloom filter) is used to minimize the cost of a semijoin operation. This thesis report provides a summary of each category of query processing techniques and optimization algorithms. Also in this thesis, we propose a new algorithm called Composite Semijoin Filter by combining the idea of composite semijoins, Bloom filters and PERF joins. One of the advantages of this algorithm is to avoid collisions. The algorithm is evaluated and compared with initial feasible solution (IFS) and another filter-based algorithm. It has been shown that the algorithm gives substantial reduction on relations and the total cost.Dept. of Computer Science. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis2004 .Z58. Source: Masters Abstracts International, Volume: 43-01, page: 0249. Adviser: Joan Morrissey. Thesis (M.Sc.)--University of Windsor (Canada), 2004.