Re: Business case for deploying a wiki

Hi Brian

Li Ding (cc'ed) would be a good contact person. We have used
(semantic) wikis heavily in maintaining our data, and in particular,
he created the data.gov wiki (http://data-gov.tw.rpi.edu/)

I will try to make some first inputs for the best of my knowledge.
More goes inline.

On Fri, Oct 23, 2009 at 4:37 PM, Brian Gryth <briangryth@gmail.com> wrote:
> Good afternoon all,
>
> I'd like to solicit some help from this group regarding the deployment of a
> wiki.
>
> My agency is exploring the possibility of deploying a wiki to provide a
> mechanism for the public to contribute in the development of content like
> FAQs, instructions, and a glossary for one of our filing applications.
>
>
>
> We feel that deploying a wiki will or hopefully will result in the
> following:
>
>
>
> 1.      Improve the content we provide to the public by ensuring that the
> most useful and easily-understood information is provided;
>
I believe wiki will help for this. For example, wiki allows you to use
templates to render data. It's often easier to update a template than
updating some hardwired UI code. It's also easy to visualize data if
you use a semantic wiki.

> 2.      Provide useful business intelligence that will improve public
> relations and improve future development;
>
> 3.      Build a sense of ownership of the information and the system the
> information supports; and
>
> 4.      Goodwill and trust for agency.
>
I'm positive for the rest 3 points as well, but not clear about their
precise meaning before I can make more detailed comments.
>
>
> We are working through concerns about giving up control over our content and
> about IT security.  I’m reaching for any assistance from the community,
> especially if you or your organization has deployed a wiki internally or
> externally.
We use wikis for: group portal; publication management; class
organization; working with collaborators; campus map; ontology
management, among many other things.

> We would like to know what questions the decision makers asked?  What concerns did people have?
a lot, from very high-level concerns to very geeky details..
* how it is related to current projects/funding?
* how our current data is added to wiki?
* how many data you can handle? how fast?
* how other people can know and use our data?
* how will you expose data in RDF? (for semantic web people)?
* I can't find information I want, can you improve the organization/layout?
* how to create a page?
* how to protect sensitive data? what's your policy?
* what if i forget password? what if we have many wikis? (one of the
most frequent email triggers)
* can i add HTML content to wiki?
* can you come to my office....

>  How did you address IT security concerns?
We do have a lot of sensitive data (e.g. email archives, internal
project journals etc.)

We use some access control extensions to protect data, by naming
conventions and groups, and other customizable policies. There are
some such extensions for Mediawiki or Semantic Mediawiki that work
generally well. HOWEVER, there are always some possible holes, and
none can be 100% safe - unless you have very restrictive user
registration control, or put the wiki under a fully protected path.

Thus, we have multiple wikis: one is open to everyone, one is partly
open with closed  user registration, and some are completely
protected.

The mess of maintaining multiple wiki is not easier. We tried to cope this with
* shared login using LDAP
* allow external data exchange (e.g. query) across wikis
* manual or automatic interlinks

> What would you do differently?  What worked?
Each of the "?" mentioned before would consume considerable time
before you find a real solution/workaround/alternatives/"it is
impossible".

In general, we can always find a way to build a workable wiki, even it
may not satisfy everybody all the time.

If I could rework some of my previous wiki projects, I would put more
documentation and user help pages on wiki, and have them easily
accessible (within 1 click from where it is needed). I would simplify
workflows so that it matches the current workflow (before using wiki)
as much as possible, and require as less training as possible.

I also wish we have hired a UI designer to create a good skin - a good
UI is as important as good content organization.

> Has the wiki been  beneficial?
That's for sure. I can't imagine how I, maybe for many of the rest of
our group, can work without our wikis.

> Anything else that you think would be of help to me and my  colleagues.
>
You may try semantic wikis (e.g. Semantic Mediawiki). If used
properly, it can save a lot of routine jobs and have your data
organized more nicely - you may check data-gov.tw.rpi.edu for an
example.

I hope this will help a bit. I'm happy to go on the discussion.

Jie

> I know this does not have a lot to do with publishing dataset, but I know
> this group is knowledgable and may be able to help me.  Please feel free to
> direct me to other resources as well.
>
> Thanks,
> Brian
>
> --
> Brian Peltola Gryth
> 715 Logan street
> Denver, CO 80203
> 303-748-5447
> twitter..com/briangryth
>

-----
Jie Bao
http://www.cs.rpi.edu/~baojie

Received on Saturday, 24 October 2009 04:36:07 UTC