Information Model Review, 18 May 2005

Background

This meeting was proposed by Paul Watson and Carole Goble to

"[...] discuss the myGrid Information Repository (MIR), the Information Model that it supports and its relation to other myGrid components (e.g. the MIR browser and integration with Taverna). It will consist of two parts:

1. developers, users and potential users presenting their views and experiences on the current design and implementation
2. presentations and discussion of possible future options for the MIR, the Information Model and related components.

Topics to be covered include (but are not limited to):

• changes to the information model
• performance
• data visualisation
• security

Therefore, if you have experiences or views please prepare a short talk (up to 20mins) and e-mail <kmee@cs.man.ac.uk> so that we can schedule a slot for you. The meeting will start at 11 and end by 5pm."

See the InformationModel topic for more background. During development of the MIR, the model was refined and simplified for implementation: the version implmented is now held in a Rose XDE model in the MIR CVS module. The original information model included only persistent entities: see EScienceMediator for ideas on transient (event) entities.

Some of the presentations given on our UsersDay and DevelopersDay may be useful, too.

Time and place

Time: 1100 - 1700, Wednesday 18 May 2005

Place: Atlas Room 1, Kilburn Building, University of Manchester (directions).

The Atlas rooms are next door to the ESNW lab, the myGrid team's office.

Refreshments will be provided, with coffee from 1030.

Agenda

• 1100: Welcome and introduction - NickSharman
• 1130: Experience papers
  • Paul Watson: MIR - aims & issues
  • Jun Zhao: Experiences of using the information model for collecting provenance
  • Simon Miles: Provenance
  • Stef Egglestone: An assessment of the current myGrid storage system and suggestions for the future
  • DanieleTuri: KAVE and Taking LSIDs seriously
  • Pinar Alper: Modelling services: relating MIR & Feta
• 1300: Lunch
• 1330: Options papers
  • Tracy Craddock: Proposals for MIR (based on use case for MIR in relation to a workflow for identifying secretory proteins in bacteria)
  • Vic Tan: security architecture for myGrid
  • ...
• 1400: Breakout sessions, e.g.
  • Projects, programmes, people: navigating your stuff
  • Visualization
  • Provenance: storage, structure
  • Semantic annotation
  • Referencing externally-held data (e.g. SRB, database)
  • Security
• 1500: Reports back
• 1545: Discussion
• 1700: Close

Attendees

-- NickSharman - 12 May 2005

• InfoModel-Intro-NPS.ppt: Nick Sharman: Introduction

• InfoModel-Provenance-JZ.ppt: Jun Zhao: The information model & provenance

• InfoModel-Feta-PA.ppt: Pinar Alper: The information model & Feta

• InfoModel-KAVE-DT.ppt: Daniele Turi: KAVE and Taking LSIDs seriously

• InfoModel-securityoverview.ppt: Victor Tan: Overview of security issues in myGrid

• InfoModel-MIR-TC.ppt: Tracy Craddock: Proposals for MIR

• InfoModel-MIR-PW.ppt: Paul Watson: MIR Overview

Following the meeting, I've written the following to try and capture some of the ideas discussed at the meeting and since:

• Evolvingtheinformationmodel.doc: Evolving the Information Model: a brain dump

-- NickSharman - 10 Jun 2005