Welcome to the Database group

News

Dutch broadcaster BNN tests the intuitive train planner developed at the Database Group. Their verdict: “ingenious”, and “approved for elderly”. Picture of Kien Tjin-Kam-Jet proudly in the back (in Dutch). See the You Tube movie at the right side of this page.

See the treinplanner in action at: http://treinplanner.info http://snipdex.org/ns/

The Database Group

The Database group mission is to provide data management to create added value on top of autonomous data sources.

Nowadays huge amounts of data are produced by both humans and devices connected to the Internet. This has led to information overload and a decrease in trust of data, as well as an increase of privacy threats. However, applications need high data quality providing privacy. Our research extends database systems with functionality to filter and compute relevant information, to reduce the unreliability of data, and to protect privacy.

·	Search on semi-structured data. Specifically, XML databases and distributed search

·	Enriching uncertain data. Specifically, data integration and streaming data

·	Security and Privacy. Specifically, search in encrypted data and secure handling of medical data.

Applications: web search, sensor network, e-Health, ambient intelligence.

Head of the Database group is Prof.dr. Peter M.G. Apers. Research of the Database group is an integral part of the Strategic Research Orientation ASSIST, Strategic Research Orientation NICE and the Strategic Research Orientation ISTRICE of CTIT.

The Database Group contributes amongst others to the following research systems and projects:

Pathfinder is our XQuery to relational compiler developed at the Database Group in cooperation with CWI Amsterdam and University of Tübingen. The Pathﬁnder project is an exploration of how far we can push the idea of using mature relational database management technology to design and build a full-ﬂedged XML database. Pathﬁnder requires only local extensions to the underlying DBMS’s kernel, such as the staircase join operator, a join recognition logic in our compiler, as well as a careful consideration of order properties of relational operators. The compiler is part of MonetDB/XQuery, which is claimed to be the world's most efficient XQuery database system. Website: http://www.pathfinder-xquery.org

PF/Tijah (Pathfinder/Tijah, pronounce as "Pee Ef Teeja") is a flexible open source text search system developed at the Database Group in cooperation with CWI Amsterdam and University of Tübingen. The system is integrated in Pathfinder and can be downloaded as part of the MonetDB/XQuery database system. PF/Tijah is used to aid research in information retrieval at the University of Twente, including the application of language models to search, entity retrieval, expert search and the efficient implementation of the W3C candidate recommendation XQuery Full-Text. Website: http://dbappl.cs.utwente.nl/pftijah/

StreetTivo is a project run in cooperation with the Human Media Interaction (HMI) Group and CWI Amsterdam that will bring video annotation techniques - such as a large-vocabulary speech recognition system, shot detection, low-level feature detectors for query-by-example, high-level feature detectors such as a face detector, etc. - to everybody's living rooms. StreetTivo connects hard disk recorders in a peer-to-peer network, enabling the distribution of workload of video annotation over many small machines. Website: http://streettivo.sf.net

InstantDB opens up a new alternative to protect personal data over time. It is based on the assumption that long lasting purposes of data can often be satisfied with a less accurate, and therefore less sensitive, version of the data. In our data degradation model, called Life Cycle Policy model, data is stored accurately for a short period, such that services can make full use of it, then degraded on time progressively decreasing the sensitivity of the data, until complete removal from the system. Website: http://eprints.eemcs.utwente.nl/11408/

SensorDataLab is our sensor lab environment. It is used to identify issues, develop ideas and validate solutions for sensor data management. The SensorDataLab addresses a localization scenario with four different sensor networks and about 60 deployed devices. In the context of this lab meta data management, streaming data, manual sampled data and provenance data are collected and managed. The goal is to come with storage solutions to support querying, annotating and processing of the different data as well as combinations of the different data types.

Website: http://www.sensordatalab.org

Presentation

In May 2011 Peter Apers, head of the Database Chair, has given a presentation about the research and teaching activities of the DB group. Just follow the link “presentation DB” to read it.