Rationale for Template

From XG Provenance Wiki
Revision as of 18:39, 3 November 2009 by Jcheney (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

The following documents were considered in creating the template:

  • SOCIAL: Use cases from the Social Web Incubator Group
  • EGOV: Template and [ use cases] from the eGovernment Interest Group
  • HCLS: A sample use case set from the Semantic Web Health Care and Life Sciences (HCLS) Interest Group
  • SECURE: Use cases from the Web Security Activity
  • PASOA: Informal and published use cases from the Provenance-Aware Service-Oriented Architecture project
  • TIPS: Ronald Reck's tips to this group on effective use case specifications

Most of the use cases followed a template, either explicitly or implicitly. On the other hand, the HCLS interest group did not, and it appears hard to analyse across the use cases or know where to start in documenting one. On this experience, we suggest that a template is desirable.

SOCIAL, SECURE, and PASOA consistently described use cases as short scenario stories involving a few actors. HCLS had a step-by-step scenario. EGOV use cases mixed scenario descriptions with discussion of standards and technologies, so were not strictly (just) use cases. We propose the Use Case Scenario, as a storyline, as the key part of the use case.

SOCIAL, SECURE, HLCS, and PASOA each identified some key point to the scenario: a goal in SOCIAL, a question being asked in SECURE and HLCS, and a piece of information being determined in PASOA. This is a useful way to distinguish the part of the scenario relevant to provenance/semantic web, so we propose both stating the use case Goal, and highlighting the most relevant steps in the scenario.

HLCS, EGOV, and PASOA describe use cases in particular domains, rather than commonly understood social interaction (as in SOCIAL and SECURE). We propose a Background section to explain the key aspects of the domain required to understand the scenario, if necessary. Also, HCLS illustrates an important case with regard to semantic web use cases: where current practice is to be replaced with something using SW technology. To allow this kind of use case, we propose a Current Practice section, to describe what happens pre-semantic web.

EGOV has a specific section on Problems and Limitations related to a use case. This seems important as a use case is often best understood by the cases in which it does not work well, or the reasons why it is non-trivial to achieve. Moreover, the template used by SOCIAL has a Preconditions field, which describes what limiting factors may exist to prevent the use case happening at all. We therefore propose to include a Problems and Limitations section.

TIPS suggests that the responsible Owner of a use case be clear, and so we follow this proposal. TIPS also points out that the list can grow long. It is not clear how detailed the use cases of this group will become, and some of those in the HLCS group are very detailed indeed. Therefore, we propose, at least to start with, to have each use case on a separate Wiki page, and develop an appropriate index (TIPS suggests a matrix by topic, as is done by SECURE) as the types of use case supplied become clear.

EGOV requires that use cases are easy to refer to, by name and ID, and SOCIAL also gives use cases short names. We propose to provide Short Names, and use the Wiki URL for each use case as a URI identifier.

It is not clear that the other fields of the more detailed templates used by SOCIAL and EGOV would be of benefit to the group at this level (as TIPS suggests).