W3C home > Mailing lists > Public > public-xg-lld@w3.org > August 2010

Re: Use Case template

From: Antoine Isaac <aisaac@few.vu.nl>
Date: Tue, 31 Aug 2010 22:45:08 +0200
Message-ID: <4C7D69D4.2020009@few.vu.nl>
To: Emmanuelle Bermes <emmanuelle.bermes@bnf.fr>
CC: "Haffner, Alexander" <A.Haffner@d-nb.de>, public-xg-lld@w3.org
Hi Alexander, Emmanuelle,

In fact I'd like to react on the "goal" section--even though I don't want to forget the others, they need attention as well :-)

I've been thinking on the discussion we had last week, on Alexander's mail and Kai's edition of Karen's case [1]. And the more I think, the more I'm convinced we should *not* try to enforce the case to present actors' (i.e., end users') goals. Or at least not in the template we circulate outside of this group.

I think many linked data cases will indeed not come with a clear-cut, user-oriented goal. This is just inherent to the linked data vision for which a vague motivation such as "We want to expose data as linked-data" (to re-use Kai's word) is plainly acceptable.

There are quite obvious cases, such as for data.gov.uk or other sites which have a "general service mission". They stop somehow half-way in the complete software development process. They point at useful apps as examples/success stories of course, but they don't try to enforce specific scenario(s). Note that this doesn't make their life easier, or less interesting: in fact their data must be generic/precise enough to accommodate a wide variety of relevant use cases (fully-fledged cases with real actors, this time ;-) )

Such cases will also happen in the library domain. This is I believe the case for id.loc.gov, for example. They point at some "benefits" [2] but those still aim at quite a variety of uses--and these remain *really* vague.

Would we ask use case contributors (especially these outside this group) to articulate such a statement, or even to do better by themselves? I think this could deter them, and that we should still allow both "goal approaches" to co-exist.

To me, it should be rather *our role* to try to make more explicit the goals at the level of actors/end users for the producer-oriented cases.
In fact it could be one great outcome of the group to come with a set of generic actor-level goals that can be re-used as motivation for the "We want to expose data as linked-data" cases to come. A kind of generic flyer for LLD.

Do not take me wrong, again I think Kai's work on Karen's case is excellent. And it's good that Karen is here to say whether she's happy or not with Kai's attempt. All this is really useful and can help our group to produce a really crucial bit of deliverable. But this should happen only for people who feel motivated enough for such an exercise :-) [3]

So, to sum up: we try to do it in the group in a later curation/analysis step, we do not request everyone else outside  to do it.
Do you think we could get consensus on that?



[1] http://www.w3.org/2005/Incubator/lld/wiki/index.php?title=Use_Case_Open_Library_Data&oldid=565
[2] http://id.loc.gov/authorities/about.html
[3] In fact given the time it took us to realize, I imagine that a template that tries to ensure appropriate actor-level goal may become very long and contribute to make the use case filling task (even more) tedious.

> Thank you Alex for clarifying the context.
> I'd like to bring this up again before our next call, because we need to
> reach an agreement on the use case template now.
> The discussion page at [1] is currently empty. So we will probably need
> to discuss the template again on next call.
> Maybe we can have a discussion on each part of the template and decide
> if we keep it / drop it / edit it.
> Following the discussions we already had, we could start with the
> following :
> Name -> no problem,  keep
> Owner -> no problem,  keep
> Background and Current Practice -> no problem,  keep
> Goal -> to be edited to make it clear that it's meant to be the goal of
> the actor in the scenario, not the goal of the use case
> Use Case Scenario -> no problem,  keep
> Target Audience  -> added by Joachim. The group finds it useful but it
> should be optional. To be edited : we need to add guidelines on how to
> fill it.
> Application of linked data for the given use case -> not discussed yet.
> Is that clear to everyone ?
> Existing Work -> to be edited to add prototypes
> Related Vocabularies  -> no problem,  keep
> Problems and Limitations -> not discussed yet. Is that clear to everyone ?
> Related Use Cases and Unanticipated Uses (optional) -> not discussed
> yet. Is that clear to everyone ?
> Library Linked Data Dimensions / Topics -> confusing. Drop it, or keep
> it only for curation ?
> References -> no problem,  keep
> Prototypes and Applications  -> added by Joachim. Drop it: content to be
> put under "existing work".
> Comments welcome on this proposal. I'm copying my mail in the discussion
> page of the template, so you're welcome to make your comments there.
> Emmanuelle
> [1] http://www.w3.org/2005/Incubator/lld/wiki/Talk:Use_Case_Template
> On Fri, Aug 27, 2010 at 12:19 PM, Haffner, Alexander <A.Haffner@d-nb.de
> <mailto:A.Haffner@d-nb.de>> wrote:
>  >
>  > Hi everyone,
>  >
>  >
>  >
>  > In addition to our telco from yesterday some comments to make sure
> the templates are applicable for upcoming UCs.
>  >
>  > First, I’d like to give some extra information regarding our template
> elaboration. Kai and I are both computer scientists so we are driven by
> experience of UC modeling (UML etc.) in the context of software
> development. As a consequence our chosen approach is similar to the one
> used in software engineering. It’s a user-centered approach. That means
> we try to identify user needs by analyzing the interaction of an actor
> (librarian, end user as data consumer, data provider – every imaginable
> user!) with a particular system (an already existing one or just an idea
> of a system).
>  >
>  > The conclusion of system requirements (in our case requirements for
> linked data in libraries) is in software engineering processes the next
> step and usually by use cases in this form pretty easy. However, this
> doesn’t mean this approach is best for LLD-XG needs.
>  > Nevertheless, we would first like to make sure that you got our
> thinking and then we can discuss the need to modify the UC gathering
> process to suit LLD XG requirements.
>  >
>  > From that on we should have a closer look to the single parts of the
> UC template and the descriptions therefore. The discussion yesterday
> showed the ambiguity of the goal-section. Karen stated the major goal,
> but actually we intended to highlight (1) the actor’s goal in this
> particular UC and (2) how linked data can support this specific actor’s
> goal. This is also pointed out by the comments of Kai in the Open
> Library UC [1].
>  >
>  > We have to make sure that the template is unambiguous regarding our
> common understanding of it’s purpose and intended use and after this our
> UC template should probably be ready to go…
>  >
>  > Cheers, Alexander
>  >
>  > [1] http://www.w3.org/2005/Incubator/lld/wiki/Use_Case_Open_Library_Data
>  >
>  >
>  >
>  > --
>  > Alexander Haffner
>  > Deutsche Nationalbibliothek
>  > Informationstechnik
>  > Adickesallee 1
>  > D-60322 Frankfurt am Main
>  > Telefon: +49-69-1525-1766
>  > Telefax: +49-69-1525-1799
>  > mailto:a.haffner@d-nb.de <mailto:a.haffner@d-nb.de>
>  > http://www.d-nb.de
> --
> =====
> Emmanuelle Bermès - http://www.bnf.fr
> Manue - http://www.figoblog.org
Received on Tuesday, 31 August 2010 20:45:46 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:35:55 UTC