{"title":"XJoin:一种新颖的XML数据整体小枝匹配方案","authors":"S. Sheu, N. Wu, Jyun-Bo Chen","doi":"10.1109/ITRE.2005.1503091","DOIUrl":null,"url":null,"abstract":"As the skeleton of a query tree, structural join operations, namely holistic twig joins, are central to XML query processing. There are techniques proposed recently to process the query tree as a whole using a chain of linked stacks at tree nodes to verify partial results. The idea is to exploit pre-built indices on participating elements to quickly skip mismatches among element instances. However, instances-based skipping requires multiple traversals from index tree root to leaf pages. The capability of expediting query execution is very limited. In this paper, we propose a novel index structure to allow skipping groups of unqualified elements in entirety. Every index page is accessed at most once regardless of join selectivities. In the sequel, our solution requires minimal I/O cost and thus significantly reduces overall execution time. The empirical evidences demonstrate over 50% performance improvement over extant competitors is achievable by our new approach.","PeriodicalId":338920,"journal":{"name":"ITRE 2005. 3rd International Conference on Information Technology: Research and Education, 2005.","volume":"09 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2005-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"XJoin: a novel holistic twig matching scheme for XML data\",\"authors\":\"S. Sheu, N. Wu, Jyun-Bo Chen\",\"doi\":\"10.1109/ITRE.2005.1503091\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"As the skeleton of a query tree, structural join operations, namely holistic twig joins, are central to XML query processing. There are techniques proposed recently to process the query tree as a whole using a chain of linked stacks at tree nodes to verify partial results. The idea is to exploit pre-built indices on participating elements to quickly skip mismatches among element instances. However, instances-based skipping requires multiple traversals from index tree root to leaf pages. The capability of expediting query execution is very limited. In this paper, we propose a novel index structure to allow skipping groups of unqualified elements in entirety. Every index page is accessed at most once regardless of join selectivities. In the sequel, our solution requires minimal I/O cost and thus significantly reduces overall execution time. The empirical evidences demonstrate over 50% performance improvement over extant competitors is achievable by our new approach.\",\"PeriodicalId\":338920,\"journal\":{\"name\":\"ITRE 2005. 3rd International Conference on Information Technology: Research and Education, 2005.\",\"volume\":\"09 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2005-06-27\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"ITRE 2005. 3rd International Conference on Information Technology: Research and Education, 2005.\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ITRE.2005.1503091\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"ITRE 2005. 3rd International Conference on Information Technology: Research and Education, 2005.","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ITRE.2005.1503091","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
XJoin: a novel holistic twig matching scheme for XML data
As the skeleton of a query tree, structural join operations, namely holistic twig joins, are central to XML query processing. There are techniques proposed recently to process the query tree as a whole using a chain of linked stacks at tree nodes to verify partial results. The idea is to exploit pre-built indices on participating elements to quickly skip mismatches among element instances. However, instances-based skipping requires multiple traversals from index tree root to leaf pages. The capability of expediting query execution is very limited. In this paper, we propose a novel index structure to allow skipping groups of unqualified elements in entirety. Every index page is accessed at most once regardless of join selectivities. In the sequel, our solution requires minimal I/O cost and thus significantly reduces overall execution time. The empirical evidences demonstrate over 50% performance improvement over extant competitors is achievable by our new approach.