Seminars at the Faculty of Informatics

The Faculty of Informatics is pleased to announce a seminar given by Dr. Dan Olteanu


TITLE: MayBMS - A Database Management System for Uncertain and Probabilistic Data

SPEAKER: Dan Olteanu, Oxford University Computing Laboratory

DATE: Thursday, February 28, 2008

PLACE: USI Lugano, room A22, Red Building (Via G. Buffi 13)

TIME: 09:30 - 10:30




Today, uncertainty is commonplace in data management scenarios dealing with data integration, sensor readings, information extraction from unstructured sources, or manually entered information. In these scenarios, uncertainty arises from the existence of alternatives for mapping schemas of different sources or non-identical record duplicates, different interpretations of sensor data, multiple extraction possibilities from unstructured data, and several possible readings of manually filled forms.


In this talk I will present my recent work on MayBMS, a database management system for efficiently managing large uncertain databases.

MayBMS has two key components: (1) U-relations, a succinct and purely relational representation system for uncertain data, and (2) world-set algebra, an expressive uncertainty-aware query language. I will discuss properties of U-relations and world-set algebra and show how they naturally overcome severe limitations of existing representation formalisms (like Washington's probabilistic databases) and of uncertainty-aware query languages (like Stanford's TriQL). For instance, U-relations are strictly stronger and can be exponentially more succinct than probabilistic databases, yet allow for the reuse of mature relational storage, indexing, and scalable query processing techniques. In contrast to TriQL, world-set algebra is generic and conservative over relational algebra, yet expressive enough to allow for what-if queries and the conditioning of databases using new evidence.



Dan Olteanu is University Lecturer at Oxford University Computing Laboratory, Academic Fellow of Oxford e-Research Centre, and Fellow of St Cross College. His research covers various aspects of data management, including data integration, processing streams of Web data, and uncertain and inconsistent databases.