Date of Award
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Distributed query processing is important for Distributed Database Systems. Through the past years, the research focus in distributed query processing has been on how to realize join operations with different operators such as semi-join and Bloom Filter. Experiments show that using bloom filters, the hash-semijoin, almost always does better than semi-join for the query processing. However as long as you use bloom filter, you cannot avoid collisions. So in order to get the cheaper processing, some of the past work uses two or more bloom filters to do the hash-semijoin. However several factors still affect the cost and optimization result. (1) How to decide the perfect number of the bloom filters, and what kind of bloom filter should be chosen. (2) There is no way to avoid collisions when utilizing bloom filters. (3) With bloom filter, we cannot keep the exact location information of the joining attributes (loss of join information). (4) With bloom filter, we never can combine the useful composite semi-join in the process. Taking the idea of PERF join into account, why not use the bloom filter (hash-semijoin) concept but come up with a new kind of filter "Complete Reducing Filter" (CRF), which can avoid the disadvantages of bloom filter, as well as inherit the advantages of it? We propose and implement a new algorithm called Complete Reducing Filter (CRF) based on PERF join, which can keep the join location information, as well as lower transmission cost (because it's still using the filter concept). At the same time, CRF can combine the composite semi join into the process, which overcome the impossibility if only using a bloom filter. With the variation of the bloom filter, we try to achieve better performance with lower cost. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis2003 .Z535. Source: Masters Abstracts International, Volume: 42-03, page: 0979. Adviser: Joan Morrissey. Thesis (M.Sc.)--University of Windsor (Canada), 2003.
Zhang, Yue (Amber)., "Variation of bloom filters applied in distributed query optimization." (2003). Electronic Theses and Dissertations. 4508.