XML with Incomplete Information: Representation, Querying, and Applications

点击翻译按钮获取中文摘要

Data on the Web - particularly XML data - is often incomplete andinconsistent, due to such factors as the lack of centralisation andcontrol over data quality. While the transfer and extension ofrelational tools to deal with XML data has been a central theme indata management research over the past decade, the standard databasetoolbox offers us little in terms of handling ofincompleteness. Indeed, it is one of the most notoriouslyunderdeveloped and most often criticised aspects of relationaldatabases. In addition, the flexibility of XML leads to many ways inwhich incompleteness of data can be accommodated, in addition to thestandard relational null values.There has not yet been any detailed study of incompleteness in XML.Our main goal is to conduct such a systematic study, and develop itsapplications in the area that underlies data management tasks on theWeb -- the use of data across multiple independent applications.We shall investigate models of XML with incomplete information andalgorithmic techniques for querying such data, paying particularattention to the correctness/complexity tradeoffs and to the practicalityof algorithmic tools. We shall investigate the fundamental role ofincompleteness in applications that involve the movement of XML data,such as integration of data from various sources or moving databetween peers according to mappings between their schemas. We shalldevelop a specification and algorithmic toolbox for dealing withincomplete information as it arises in such applications.