Skip to content. | Skip to navigation

Informatik 5
Information Systems
Prof. Dr. M. Jarke
Sections
Personal tools
You are here: Home Teaching WS 09/10 Data Streams and Dataspaces

Data Streams and Dataspaces


Data streams are continuous streams of data items, which are usually relational tuples. Data items may arrive in multiple, rapid, time-varying, possibly unpredicated and unbounded streams, which poses new challenges for the data management of such data streams. Because of the amount, frequency and unboundedness of the incoming tuples, it is not possible to store the data in a traditional database system and process the queries on the stored data. A data stream management system has to offer support for continuous queries which might require approximation and adaptivity for query execution. Applications of data streams include financial applications, network monitoring, sensor networks, and web applications.

Dataspaces are a new concept in data management for diverse data integration applications in which the data sources are only loosely connected. Traditional data integration systems assume that schema and mappings of data sources and integrated system do not change very often, whereas in dataspaces, an incremental construction of the integration system is envisioned. In dataspaces, data integration at a detailed level is only done when it is really required; initially, a coarse grained integration can be done which only supports a limited set of functions. This is also called "pay-as-you-go" integration.

Registration

Registration is only possible from June 15 to June 28 through the central registration for seminars and lab courses of the Department of Computer Science.

Topics

There will be 12 topics available, which include

  • Probalistic Databases
  • Incomplete Databases
  • Query Processing in Data Streams
  • Data Stream Management Systems
  • Dataspace Systems
  • Schema Mappings

Links

Language

Presentations have to be done in English. The seminar paper can be written in German or English.

Schedule

June 15-28, 2009 Registration through the central registration for seminars and lab courses of the Department of Computer Science.
July 22, 2009, 3.00pm First meeting in the seminar room of Informatik 5

Introduction & Assignment of topics

October 20, 2009 First introduction for the seminar paper (1-2 pages)
November 10, 2009 First draft of the seminar paper (3-5 pages)
December 14, 2009 Draft version of the seminar paper (10-15 pages)
January 13, 2010 Final version of the seminar paper (15-20 pages)
January 20, 2009 Final version slides
January 27/28, 2010 Presentations (about 45mins each)

Prerequisites

Students must be enrolled in a computer science program (Bachelor-Informatik, Diplom-Informatik, SSE, or MI). Knowledge in database systems, knowledge representation and/or data modeling is required. Please provide the information about the attended courses while registering for the seminar. If you do not have the required knowledge (or you do not provide us the information that you have it), then it will be very unlikely that you are accepted for this seminar.

Document Actions