Minutes for 25 April 2007 URW3 Meeting

Attendees:

Agenda:

  1. Welcome
  2. Work through any technical difficulties
  3. Discuss XG work plan
  4. Discuss and initiate process of identifying use cases
  5. Discuss policy for use of member-only and public mailing lists
  6. Other business

Scribe: Kathryn Laskey

Discussion summary:

  1. Welcome

    Participants introduced themselves

  2. Work through any technical difficulties

    We worked through the process of getting everyone logged in to IRC and Zakim agenda management.

  3. Charter and work plan

    Ken Laskey described charter and proposed work plan:

    The first five months we will:

    1. Develop use cases that show a compelling need for uncertainty reasoning and
    2. Compile a bibliography.

    During the sixth month we will identify a set of methodologies appropriate for which to identify information needs for each. The purpose is not to endorse particular methodologies or to develop applications, but to identify information needs that web services and web-based applications will need to support.

    During the next four months we will:

    1. Identify critical information needs for chosen methodologies
    2. Investigate and document commonalities in information needs across methodologies

    During the last month we will write a final report on our work.

    1. We should pass our results to other working groups as appropriate, and also keep them informed of our progress as appropriate
    2. Rules working group is the most likely to incorporate our work into Recommendation track.

    Comments on use cases, methodologies and semantics:

    - Mitch raised a concern that we are taking on an impossible task. People have been debating the semantics of uncertainty methodologies for decades and we cannot establish common semantics.

    - We agreed that we are not trying to resolve semantics, only to identify information needs for web applications and services for commonly applied formalisms. Different use cases for different formalisms can help us to identify these information needs.

    - Anne said we need to identify the questions we want to answer before getting bogged down in methodologies; Ken said that’s one major purpose of the use cases.

    - Mode of operation: people will propose use cases, coalesce in groups to discuss them. Identify relevant literature and appropriate methodologies. Connect the use cases to the methodologies. Through a feedback process, we categorize them and characterize the categories. Then we downselect to the critical information needs, decide what we can conclude, and write our final report.

  4. Discuss and initiate process of initiating use cases

    - We have a use case template

    - We will start by posting use cases on the wiki

    - There was support for finding use cases where there is data from a crawler or a user model; maybe some beginning experiments; benchmarks

    - Developing good use cases is difficult

    - Experiments are out of scope for us, but the use case can identify relevant experiments

  5. Discuss policy for use of member-only and public mailing lists

    - We are all subscribed to the member-only email list automatically, but each of us needs to subscribe to the public mailing list.

    - The discussion was continued by email and will be summarized at next

  6. Other business

    None

Action items

  1. Discuss preferences for public and member-only discussions on member-only list
  2. Ken to circulate name of public list and instructions for joining
  3. People send ideas for use cases to member-only list
  4. Ken will keep everyone informed of use cases
  5. Kathy will post minutes for this meeting