Faculty Candidate COLLOQUIUM

Relational Division - Algebra, Algorithms, and an Application

Ralf Rantzau

IBM Silicon Valley Labs,

Computer Science Faculty Candidate

Thursday, March 11, 2010
11:00 a.m. - 12:00 p.m.
Fuller Labs 320

Abstract:

Relational division is an operator of the relational algebra that realizes universal quantifications in queries against a relational database. Expressing a universal quantification problem in SQL is cumbersome. If the division operator would have a counterpart in a query language, a more intuitive formulation of universal quantification problems would be possible. Although division is a derived operator it can be expressed using other basic algebra operators the performance of queries involving a division problem can be significantly increased if it is implemented as a separate operator in a query processor.

In this talk, we take a look at the infrastructure needed to embed division-like operators into a database system: algorithms implementing the operator and algebraic "laws" to be able to optimize queries and parallelize query execution. An excellent application area for queries that benefit from the presence of a division operator is frequent item set discovery, an important data mining task. This talk will try to shed more light on "one of the least well understood operators of the relational algebra" (Chris Date).

______

Ralf Rantzau received his Ph.D. in Computer Science in 2004 from the University of Stuttgart,Germany under Prof. Bernhard Mitschang and Prof. Rudolf Bayer, where he studied how to bring data mining algorithms closer to database systems. In 2004, he joined IBM Almaden Center as a post-doc to work with Dr. Rakesh Agrawal. He was involved in projects such as context-sensitive ranking, auditing queries in databases with evolving schemas, discovery services to enable supply chain traceability of RFID-tagged items, as well as privacy enforcement for data streams. His inventions led to several patent applications.

In 2006, Ralf joined IBM Silicon Valley Labs to transfer privacy research technology into new software products. Recently, he worked on software solutions for business intelligence and data governance.

He received a Best Paper Award at ICDE 2006. His EDBT 2002 paper was selected among the best papers of the conference. Currently, he serves as a PC member at the conferences DEXA, SAC, and ICDE.

His research interests span problems in advanced query processing and optimization, data mining, data warehousing, data governance, and privacy.

Host: Prof. Michael Gennert

Refreshments will be served.

Maintained by webmaster@cs.wpi.edu
Last modified: 03/08/2010