Qi Yang, Clement T. Yu, Chengwen Liu, S. Dao, Gaoming Wang, Tracy Pham
{"title":"A hybrid transitive closure algorithm for sequential and parallel processing","authors":"Qi Yang, Clement T. Yu, Chengwen Liu, S. Dao, Gaoming Wang, Tracy Pham","doi":"10.1109/ICDE.1994.283073","DOIUrl":null,"url":null,"abstract":"A new hybrid algorithm is proposed for well-formed path problems including the transitive closure problem. The CPU time for computation is O(ne), and blocking technique is incorporated to reduce the disk I/O cost in disk-resident environment. The new features of the new algorithm are that only parents sets instead of descendant sets are loaded in from disk, and the computation can be parallelized efficiently. Simulation results show that our algorithm is superior to other existing algorithms in sequential computation, and that linear speedup is achieved in parallel computation.<<ETX>>","PeriodicalId":142465,"journal":{"name":"Proceedings of 1994 IEEE 10th International Conference on Data Engineering","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1994-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of 1994 IEEE 10th International Conference on Data Engineering","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICDE.1994.283073","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
A new hybrid algorithm is proposed for well-formed path problems including the transitive closure problem. The CPU time for computation is O(ne), and blocking technique is incorporated to reduce the disk I/O cost in disk-resident environment. The new features of the new algorithm are that only parents sets instead of descendant sets are loaded in from disk, and the computation can be parallelized efficiently. Simulation results show that our algorithm is superior to other existing algorithms in sequential computation, and that linear speedup is achieved in parallel computation.<>