[Date Prev][Date Next][Date Index]

Talk Announcement: Four talks on Monday 24th of May



-------------------------------------------------------------------------------------------------------------------------------------------------

Dr. Reinhard Pichler, Siemens AG Österreich
Montag 24. Mai, 2004, 9 Uhr, Seminarraum 184-2

? Efficient Evaluation of Xpath Queries?

XPath has been proposed by the World Wide Web Consortium (W3C) as a practical
language for navigating in XML document trees and extracting information from these trees. The importance of XPath stems (1) from its potential application as an XML query language per se and (2) from the fact that it is at the core of several other XML-related technologies like XSLT, XPointer, XQuery, etc. Since XPath and related technologies will be tested in ever-growing deployment scenarios, its implementations need to scale well both with respect to the size of the XML data and the growing size and intricacy of the queries. Alas, this requirement is by no means met by existing XPath processors. In particular, they are only able to handle short and simple XPath
queries.
In this talk, the behaviour of existing XPath processors is sketched and the source of their rather poor performance is identified. This analysis is supported by test results with the currently most
popular XPath processors.
Only recently, a much more efficient XPath evaluation method has been proposed, which is based on the so-called ?Context-Value Table Principle?. In the second part of this talk, efficient XPath evaluation via the ``Context-Value Table Principle'' is explained by means of a running
example.
Moreover, some ideas for further improvements of this method will be shown. On the one hand, these ideas aim at a speed-up of the evaluation of arbitrary XPath queries. On the other hand, we identify two interesting fragments of XPath for which even further optimizations are
possible.

-------------------------------------------------------------------------------------------------------------------------------------------------

Dr. Sven Helmer, Universität Mannheim
Montag 24. Mai, 2004, 11 Uhr, Seminarraum 184-2

?Query Optimization in XQuery ?

The use of XML (eXtensible Markup Language), which was designed for representing semistructured data, spreads rapidly. In its wake XQuery became the quasi-standard for XML query languages. Due to the fact that XML is totally different from previous data models, we have to develop new methods for processing and optimizing queries. After briefly introducing XML and XQuery, we present an optimization technique for nested XQuery queries. This technique relies on
unnesting the queries.

-------------------------------------------------------------------------------------------------------------------------------------------------

Dipl.-Ing. Dr. Christoph Koch, TU Wien
Montag 24. Mai, 2004, 14 Uhr, Seminarraum 184-2

? A Visual Query Language for Complex-Value Databases?

In this talk, a visual language, VCP, for queries on complex values is proposed. The main strength of the new language is that it is purely visual: (i) It has no notion of variable, quantification, partiality, join, pattern matching, regular expression or any other construct proper to logical, functional or other dataase query languages and (ii) has a very natural, strong and intuitiv designmetaphor. The only operation is that of copying and pasting in either a schema or an example database tree. Such a tree only uses three kinds of nodes, namely set/bag-typed branches, tupletyped
branches and value leaves.
We show that despite its simplicity, VCP precisely captures the complex-value (set or bag) algebra without powerset, or equivalently, monad algebra (over sets or bags) with union and difference. Thus, its expressiv power is precisely that of the language that is usually considered to
play the role of relational algebra for complex-value databases.
This language also preserves data provenance through queries, a porperty not known to be shared by any other query language. We discuss the importance of this featuer in scientific -- in particular biological ­ databases. Moreover I present a vision for tracking data provenance on a
future, more robust semantic web.
This work is in part joint work with Peter Buneman.

-------------------------------------------------------------------------------------------------------------------------------------------------

Dipl.-Ing. Dr. Thomas Lukasiewicz, Università di Roma ?La Sapienza?
Montag 24. Mai, 2004, 16 Uhr, Seminarraum 184-2

? A Data Model and Algebra for Probabilistic Complex Values?

We present a probabilistic data model for complex values. More precisely, we introduce probabilistic complex values relations, which combine the concept of probabilistic relations with the idea of complex values in a uniform framework. We elaborate a model-theoretic definition of probabilistic combination strategies, which has a rigorous foundation on probability theory. We then defina an algebra for querying database instances, which comprises the operations of selecton, projection, renaming, join, Cartesian product, union, intersection and difference. We prove that our data model and algebra for probabilistic complex values generalizes the classical relational data model and algebra. Moreover, we show that under certain assumptions, all our algebraic operations are tractable. We finally show that most of the query equivalences of classical relational algebra carry over to our algebra on probabilistic complex value relations. Hence, query optimization techniques for classical relational allgebra can easily be applied to optimize queries on probabilistic
complex value reltions.