[Berkeley] [CS Division] [Database Research]


 
CS 294-7

Post-Modern Database Systems:

Databases Meet the Web


In spring of 1999, Dr. Michael Carey of IBM Research visited Berkeley as the first annual Stonebraker Fellow. As part of his stay, he and Prof. Joe Hellerstein co-hosted this graduate seminar. This site captures the visitors' talks, and seminar discussion.

[Original Course Overview] [Speakers and Talks] [Reading List]
[Class Presentations] [Class Roster] [Instructors]

Speakers and Talks

Date Speaker Title (click for abstract) Slides Web Resources
Jan. 27, 1999 Dr. Michael J. Carey, IBM Almaden Research Center O-O, What Are They Doing to Relational Databases? (The Evolution of DB2 Universal Database) Powerpoint, HTML
Jan. 28, 1999 Prof. Michael Stonebraker, UC Berkeley/Informix/Cohera Preliminary Thoughts about Regres: A Next Generation DBMS Powerpoint Telegraph Website
Feb. 4, 1999 Prof. Eric Brewer, UC Berkeley/Inktomi The Rise of Infrastructure Services Ninja Website
Feb. 11, 1999 Prof. Alon Levy, University of Washington Web-site Management Based on Declarative Specifications Powerpoint Tiramisu Website
Feb. 18, 1999 Dr. Ashish Gupta, Amazon.com Some Data Integration and Database Issues in e-commerce
Feb. 25, 1999 Prof. Yannis Papakonstantinou, UC San Diego A TSIMMIS Retrospective Powerpoint TSIMMIS Website
Mar. 4, 1999 Prof. Jennifer Widom, Stanford University LORE: A Database Management System for XML Postscript LORE Website
Mar. 11, 1999 Adam Bosworth, Microsoft Corp. XML: A Common Framework for Moving Information Around the Web Powerpoint
Mar. 17, 1999 Prof. Michael J. Franklin, University of Maryland Adaptive Query Processing for Wide-Area Distributed Data Sources Powerpoint DIMSUM Website
Mar. 18, 1999 Dr. Laura Haas, IBM Almaden Research Center Cost-based Optimization of Queries Across Diverse Data Sources Freelance, HTML Garlic Website
Apr. 15, 1999 Prof. Pei Cao, University of Wisconsin Web Caching: Scalability, Functionality and Performance Issues Postscript WisWeb Website
Apr. 22, 1999 Anil Nori, Asera Inc. Databases for Internet Applications: Case Studies Powerpoint
Apr. 28, 1999 Dr. Udi Manber, Yahoo Behind the Scenes at a Web Portal

Overview

This course, offered for the first time in the Spring 1999 semester, will survey the brave new world of the web as viewed from the perspective of database systems. The web today is a terrific resource for global information and resource sharing – Your parents can look at your home page from across the globe, you can read today's news articles from your hometown newspaper while sitting in your apartment in Berkeley, you can buy books and CDs at a discount from web sites with impressively large inventories, and you can book travel arrangements without ever interacting with a human. But, to really succeed, you have to have a URL. Sure, you can use a web search engine – but much more powerful search tools, and probably more information structure, is going to be required in order to fully realize the potential of the web as a world-wide database system (as opposed to a world-wide heap of mostly unstructured documents). Moreover, today's database systems (which are largely based on 15-year-old system architectures) are not ideally suited to the web – they can handle most of the required data types, and new types can be defined for those that aren't covered, but they weren't designed to handle the imprecision, statelessness, or large volumes of requests that a popular web server must be prepared to cope with.

In a keynote speech at the ACM OOPSLA conference several years ago, Alan Kay said "HTML is what happens when you let physicists play with computers." We believe that (like all problems!) the web, in many ways, is a database problem. We also believe that there is much interesting database research to be done in order to make the world a better place for locating and utilizing data in a web context. In this course, we will survey the research literature in the vicinity of the web/database crossroads, examining a variety of web-related problems and some of their initial solutions. The goal of the course is two-fold: First, we (the instructors as well as the students) should all leave the course with a much better understanding of the web/database problem space and what's been done thus far. Second, we will hopefully develop a number of new and interesting ideas for relevant database research along the way.


Reading List

Adding Objects to Database Systems

The Web for Database Dummies

Database-Based Web Sites

Web Searching and Indexing

Heterogeneous Information Systems

Querying Semi-Structured Data

XML Promises and Reality

Heterogeneous Distributed Query Processing

Data Dissemination Approaches

Database Issues in E-Commerce

Caching Web Data

Wrap-Up


Class Presentations

Background:
Class Notes:


Instructors

Michael J. Carey, carey@almaden.ibm.com.

Joseph. M. Hellerstein, jmh@cs.berkeley.edu.