Use-Case: Browse a CIM Repository

Author: Phil Bentley
Version: 0.1
Date: 24 April 2008
Associated Work Packages: WP4, WP5, WP6

Purpose: Describes the generic scenario whereby an end-user performs a faceted search of the records in a CIM repository.

Actors: All actors.

Related Use-Cases:


This high-level use-case represents the general situation whereby an end-user wishes to perform a category-based browse through the set of metadata records contained in a CIM repository. By this we mean that the end-user does not provide specific search constraints, at least not in the first instance. Often, however, an initial browse approach will lead on to a more focussed query once the end-user has built up a broad picture of what is available. By selecting options/values from a range of supported categories the end-user is able to build up what we refer to as a "faceted search." It is envisaged that a faceted search should be capable of providing answers to the following types of queries:

  • Which datasets contain climate variable X ?
  • Which datasets are based upon (use) convection scheme X ?
  • Which datasets/experiments are based upon climate model M ?
  • Which datasets are suitable for acting as forcing data for climate model M ?
  • Which model (software) components can I couple to climate model M ?

Typical Course of Events

  1. The end-user navigates to the desired on-line CIM repository, possibly via the mechanism outlined in the Find a CIM Repository? use-case.
  2. The CIM repository home page displays a list of the categories (classifications) by which the records in the repository can be browsed. For example, by geographical region, originating data centre, earth system component (atmosphere, ocean, cryosphere, etc), physical parameter, and so on.
  3. The end-user selects a desired category and a particular item from that category, e.g. Earth System Component => "Atmosphere".
  4. The CIM web site displays all or, if there are many matches, a subset of the CIM records that match the chosen category. In the case of a subset of records, the web page includes familiar controls (e.g. << < > >>) for navigating through the full set.
  5. If applicable, the list of available categories (or category values) is modified (by the CIM web application) to reflect the valid choices supported by the current record selection.
  6. The end-user may then repeat step 3, as many times as required, to progressively refine the records that match the combination of search constraints.

Alternative Course of Events

  1. If the CIM repository is unavailable a warning message should be displayed. If this is because of scheduled downtime then some indication should be given as to the expected time of resumption of service.



Miscellaneous Notes:

To achieve consistency in the way that CIM repositories can be searched, the METAFOR project will need to identify and agree upon the core set of metadata descriptors that will be used to underpin the classification scheme and, by association, the browse interface.

Some of the examples of faceted searches given in the Summary section are contigent upon certain types of entities being described by the CIM repository. In addition to datasets (streams, runs, experiments, ensembles), it may be necessary to store separate descriptions for models, model components, job runs, etc.

Use-Case Diagram: