These are the old pages from the weblog as they were published at Cornell. Visit for up-to-date entries.

September 15, 2003

The Future of Database Research

Every few years a group of senior database researchers get together to do some self-assessment about the current state of database research and recommend new and future research areas. In the spring of this year there was such a meeting in Lowell, MA. It had been 5 years since the last meeting in Asilomar and from my point of view there had been significant achievements in the past years, so it would be interesting to see how the top of the database research field viewed these achievements and what they considered to be the next hot problems to be solved.

The meeting resulted in a report (pdf) that is certainly worth reading. There are 14 topics in the document that are considered to be worth of receiving extra attention from the research community.

  • Integration of Text, Data, Code and Streams - How to extend the traditional database into dealing non-structured data, with trigger & activity integration, and how to make data stream first class database citizens.
  • Information Fusion - How to perform information integration over different enterprises, often on an ad-hoc basis
  • Sensor Data and Sensor Networks - how to execute queries over large sets of sensors nodes and performance information integration over the results.
  • Multimedia Queries - 'find the scene where Spike kisses Buffy for the first time'.
  • Reasoning about Uncertain Data - Query processing at massive scale needs to move from a deterministic to a probabilistic model. How can we reason about the results of such queries.
  • Personalization - query results should depend on a 'personal profile'
  • Data Mining - Finding the really interesting stuff
  • Self Adaptation - self-manage the many tuning knobs of large database systems
  • Privacy - how to provide better data privacy integration into database systems
  • Trustworthy Systems - privacy is one aspect, but what about all the other security and integrity stuff?
  • New User Interfaces - Semantic Web?
  • One Hundred Year Storage - how do you deal with 100 year worth of information? How do you make it accessible?
  • Query Optimization - applied to semi-structured or stream processing.

Some of these points are not very surprising, they have been research topics for years, but I can see newer topics starting to dominate that are related to issue of scale in terms of data, customers and functionality. If these topics really have your interest it is fun to have a look at the transcript of the meeting by Mike Lesk and read how these top researchers, sometimes with egos-on-steroids, themselves deal with the lack of agreement on the hot research topics.

Posted by Werner Vogels at September 15, 2003 11:44 AM

New and Notable 10
Weblog: Sam Gentile's Blog
Tracked: December 31, 2003 12:00 AM