[Announcements | Schedule and Readings | Assignments and Quizzes | Syllabus/Policies ]

Reading Questions, Lecture 18 (11/15)

Papers: NiagaraCQ, Ariel

The first of these papers presents a "continuous query processing system" -- the idea is that users can pose queries over a sequence of XML documents that are arriving over the Internet (think stock quotes or news articles) and retrieve documents that match. Although the paper is written in terms of XML, most of the query processing techniques are quite similar to those seen in relational databases.

The second paper (which was written several years earlier) describes a "scalable trigger processor". This is an important paper because it inspired the work on NiagaraCQ, which in turn influenced a the recent trend in the database community to begin doing work on so-called "streaming databases" (about which we will read next week.)

As you read the papers, consider the following questions:

  1. Is there anything about NiagaraCQ that is specific to XML?
  2. How do the requirements of "continuous query processing" differ from traditonal query processing?
  3. How does NiagaraCQ efficiently evaluate queries?
  4. How do those techniques compare to the techniques proposed in the Ariel paper?

Samuel Madden (madden at csail dot mit dot edu)
Last modified: Fri Nov 12 21:02:34 EST 2004