Agenda for DISW'96

The Distributed Indexing/Searching Workshop will span two days. The first day's goal is to identify areas for potential standardization through several directed discussion sessions. The second day's goal is to filter the list of issues to identify those most like to lead to useful standards.

The format for sessions during the first day is as follows:

  • Two plenary talks (lasting 15 minutes each) expressing opposing views on the session topic including 5 minutes of clarification questions. There will be a strict cutoff of questions based on time and focus (save the discussion for the breakout session).
  • A breakout session with three parallel tracks. The goal of the breakout sessions is to examine the potential for standardization efforts in three time periods: 1-3 months, 3-12 months, and 12+ months. A session chair will guide the discussion and prepare a short summary.
  • Session summaries from each chair and plenary discussion from workshop attendees.
  • Although the sessions have been designed to bring out controversy, it is not necessary to determine a winner; different approaches may be reasonable over different time frames or circumstances. The goal of the breakout sessions is to identify the issues and to suggest directions for standards efforts.

    Tuesday, May 28

    The first day of the workshop will expose possible directions for standardization efforts in the area of distributed indexing and searching. We have selected three areas based on the position papers submitted.

    The first session, distributed data collection, will address issues associated with the collection of data across the network. Is robots.txt adequate for future needs? What is the value of protocol- and programatic-based solutions?

    The topic for the second session is data transfer format. Early deployments may create a dominant standard such as the Virtual Software Library. Format negotiation enables interoperable access to multiple standards.

    The final session will examine the need for architectures that distribute search across several repositories. The most popular indexes today are constructed as centralized repositories in the mainframe model. More recently, meta-search engines have become more popular. Decentralized, topic-specific indexes take advantage of the restricted domain to add functionality. What is the role of repository access protocols like Z39.50 and other mesh-like models appearing on the Web? Is distributed searching a realistic paradigm for administratively decentralized resources?

    Schedule (with links to talks)

    Wednesday, May 29

    The second day will begin with a brief summary by the workshop chairs. This summary will contain a coherent overview of the efforts from the first day. The first session will discuss and enumerate potential standards directions. During the second session participants will break into groups to formulate written statements for incorporation into the workshop report.


  • Session V: Enumerate Standards Directions 9:00-12:00
    Synopsis of Tuesday Discussions(15 min.); Plenary Discussion (2 hours 45 min.)
  • Lunch 12:00-1:00
  • Session VI: Prepare Draft Reports 1:00-5:00

  • This page is part of the DISW 96 workshop.
    Last modified: Thu Jun 20 18:20:11 EST 1996.