The Information Management Group
The Information Management Group (IMG) conducts research into the design, development and use of data and knowledge management systems. Such research activities are broad in nature as well as scope, including basic research on models and languages that underpins activities on algorithms, technologies and architectures. Challenging applications motivate and validate our research, in particular the Semantic Web and e-Science.
IMG Seminar 17th March 2010Title: Adaptive Join Processing in Pipelined PlansPresenter: Norman Paton Venue: 13.00 17th March 2010 in 1.3, Kilburn Building Abstrtact: In adaptive query processing, the way in which a query is evaluated is changed in the light of feedback obtained from the environment during query evaluation. Such feedback may, for example, establish that misleading selectivity estimates were used when the query was compiled, leading to the optimizer choosing an inappropriate join order or unsuitable join algorithms. This paper describes how joins can be reordered, and the join algorithms used replaced, while they are being evaluated in pipelined plans. Where joins are reordered and/or replaced during their evaluation, the approach avoids duplicating work that has already been carried out, by resuming from where the previous plan left off. The approach has been evaluated empirically, and shown to be effective for improving query performance in the light of misleading selectivity estimates. Labels: seminar |
IMG Seminar 10th March 2010Title: Feedback-Based Annotation, Selection and Refinement of Schema Mappings for DataspacesPresenter: Khalid Belhajjame Venue: 13.00 10th March 2010 in 1.3, Kilburn Building Abstrtact: The specification of schema mappings has proved to be time and resource consuming, and has been recognized as a critical bottleneck to the large scale deployment of data integration systems. In an attempt to address this issue, dataspaces have been proposed as a data management abstraction that aims to reduce the up-front cost required to set-up a data integration system by gradually specifying schema mappings through interaction with end users in a pay-as-you-go fashion. As a step in this direction, we explore an approach for incrementally annotating schema mappings using feedback obtained from end users. In doing so, we do not expect users to examine mapping specifications; rather, they comment on results to queries evaluated using the mappings. Using annotations computed on the basis of user feedback, we present a method for selecting from the set of candidate mappings, those to be used for query evaluation considering user requirements in terms of precision and recall. In doing so, we cast mapping selection as an optimization problem. Mapping annotations may reveal that the quality of schema mappings is poor. We also show how feedback can be used to support the derivation of better quality mappings from existing mappings through refinement. An evolutionary algorithm is used to efficiently and effectively explore the large space of mappings that can be obtained through refinement. The results of evaluation exercises show the effectiveness of our solution for annotating, selecting and refining schema mappings. Labels: seminar |
SNEE: Sensor Network Query Compiler/Optimizer :: Source Code ReleaseProject page: http://snee.cs.manchester.ac.We would like to announce the release of the SNEE sensor network query compiler/optimizer. SNEE (for Sensor NEtwork Engine) has been developed at the University of Manchester. It supports an expressive continuous, declarative query language over acquisitional streams, called SNEEql, using a software architecture that extends traditional distributed query processing techniques. SNEEql queries are compiled into query evaluation plans (QEPs) in the form of executable nesC/TinyOS code. SNEEql QEPs currently target the TOSSIM simulator and the Avrora Mica2/MicaZ instruction-level simulator (TinyOS 1.x only). We have also run SNEE-generated QEPs on Tmote Sky motes running TinyOS 2.x, and a further release in which the generated QEPs are fit for small-scale experiments on such mote-level hardware is planned for the next few months. The SNEE compilation/optimization architecture explicitly makes a broad range of query planning decisions that take into account the resource constrained nature of sensor networks, including:
SNEE compilation/optimization is also responsive to explicitly-stated QoS expectations, such as delivery time. A version of SNEE is being prepared that will enable multiple query evaluation as well as responsiveness to a broader range of QoS expectations than are currently supported by any comparable sensor network query processing software. SNEE was first developed in the DIAS-MC project funded by the UK EPSRC and is currently being significantly further developed and deployed in the SemSorGrid4Env project ( http://www.semsorgrid4env.eu/ ) funded by
the European Union. SemSorGrid4Env aims to provide enabling technology for the semantic discovery and integration of diverse sensor networks (and other data resources such as historical databases or satellite imagery) to support the development of on-the-fly data mashups involving streaming sensor data in the context of disaster response scenarios. One of the prototypes being developed as a demonstrator for this project is a fire-prevention application, and plans are underway to deploy SNEE in a sensor network for monitoring the risk of forest fires in Spain. The source code of SNEE has been released under the New BSD License, and is hosted at: http://code.google.com/p/snee/ At the above URL, you can download the source, with some simple examples. The quickest way to get up and running is to follow the 'getting started guide' link in the above page. The following associated publications are currently available: Overview paper: - Comprehensive Optimization of Declarative Sensor Network Queries. Ixent Galpin, Christian Y. A. Brenninkmeijer, Farhana Jabeen, Alvaro A. A. Fernandes, and Norman W. Paton. In SSDBM, pages 339-360, 2009. http://dblp.uni-trier.de/rec/ Available at: http://snee.cs.manchester.ac. Other papers about SNEE: - Validated Cost Models for Sensor Network Queries. Christian Y. A. Brenninkmeijer, Ixent Galpin, Alvaro A. A. Fernandes, and Norman W. Paton. In DMSN, (Article no. 8), 2009. http://dblp.uni-trier.de/rec/ Available at: http://snee.cs.manchester.ac. - A Semantics for a Query Language over Sensors, Streams and Relations. Christian Y. A. Brenninkmeijer, Ixent Galpin, Alvaro A. A. Fernandes, and Norman W. Paton. In BNCOD, pages 87-99, 2008. http://dblp.uni-trier.de/rec/ Available at: http://snee.cs.manchester.ac. - An Architecture for Query Optimization in Sensor Networks. Ixent Galpin, Christian Y. A. Brenninkmeijer, Farhana Jabeen, Alvaro A. A. Fernandes, and Norman W. Paton. In ICDE, pages 1439-1441, 2008. http://dblp.uni-trier.de/rec/ Available at: http://snee.cs.manchester.ac. More information is available at our project page: http://snee.cs.manchester.ac. We would be very grateful for any comments or questions, so do let us know your experiences, good or bad, if you try out the software. For this purpose, and as default for projects hosted by Google Code, the SNEE page ( http://code.google.com/p/snee ) comes with a Wiki as well It as Group Discussion fora. If you prefer to email us directly, please use {ixent,alvaro}(at)cs.man.ac.uk . External links to software/hardware platforms mentioned above: nesC/TinyOS http://docs.tinyos.net/index. TOSSIM http://www.eecs.berkeley.edu/~ http://docs.tinyos.net/index. Avrora http://compilers.cs.ucla.edu/ Mica2 http://www.xbow.com/Products/ MicaZ http://www.xbow.com/Products/ Tmote Sky http://www.sentilla.com/ |
Three papers from IMG members accepted at the EDBT 2010 conferenceEDBT ("Extending Database Technology") is one of the top conferences in the broad area of data management. We are happy to announce that this year's edition will feature three papers from IMG people, and they are:Khalid Belhajjame, Norman W. Paton, Suzanne Embury, Alvaro A. A. Fernandes and Cornelia Hedeler. Feedback-Based Annotation, Selection and Refinement of Schema Mappings for Dataspaces Paolo Missier, Norman Paton and Khalid Belhajjame. Fine-grained and efficient lineage querying of collection-based workflow provenance Kwanchai Eurviriyankul, Norman W. Paton, Alvaro A. A. Fernandes and Steven Lynden. Adaptive Join Processing in Pipelined Plans Congratulations especially to Norman, who is a co-author of all three! |
Web Accessibility Expert Interviewed By Spanish MediaA local Spanish media, NoticiasdeGipuzkoa.com, recently featured an interview with Dr. Simon Harper on Web accessibility and how it will give blind users the independence to purchase online. He discussed about the aspects of making a website accessible, the benefits of doing it, and the practicability of achieving it. Read the original full article at http://www.noticiasdegipuzkoa.com/2009/12/02/sociedad/euskadi/una-web-accesible-permite-a-un-usuario-ciego-comprar-online-y-asi-ganar-autonomia. For the translated English version of the article click here. |