From Library Linked Data
Revision as of 21:56, 14 June 2010 by Aisaac
Back to Use Cases & Case Studies page
Some guidelines for the curation of library linked data use cases:
- In the use case, it should be clear that linked data is a solution to a problem, not an end in itself. Therefore, try and describe use cases without mention of linked data, as far as possible, then bring up linked data as part of the solution.
- A useful use case is one that captures a whole set of similar scenarios. Thus, ideally the case should be described in general terms, while providing an example specific scenario to fix the ideas. In particular, you should ask whether technology or domain-specific elements that appear in the scenario are really essential to the scenario, or rather represent just one possible instance of it. If it is not essential, then it should be clearly demarcated as a specific example.
- Include enough details to highlight the technical challenges brought up by the use case
- bibliographic references: Prov-XG used mendeley to collect relevant literature: [Mendeley collection], maybe we want the same.
- none of the few curated use cases to date uses all of the sections in the template. Some discretion is needed when using a new section, if in doubt please consult with the curators to ensure uniform interpretation of the section's purposes
- the original intent of the use case may be domain- or technology-specific. That notwithstanding, this information should be preserved. As such it should find its place in the Background and Current Practice section, where it is clearly demarcated as motivation for the use case.
- the goal section should describe a user goal, as opposed to a "goal in the use of linked data". Once again, absence of any mention of linked data in this section is a good sign.
- the goal section should contain a short and crisp headline for the entire use case, which should then be expanded upon in the Use case scenario