W3C home > Mailing lists > Public > public-ddwg@w3.org > February 2007

Vocabulary process kick-start

From: Andrea Trasatti <andrea@trasatti.it>
Date: Mon, 26 Feb 2007 12:09:53 +0100
Message-Id: <8E219ADE-759F-4092-B6C3-6A922811AFA1@trasatti.it>
To: public-ddwg@w3.org

I would like to start the discussion about the process to define the  
Core Vocabulary entries for the DDR.

Let's start with a little background and explanation of why we need a  
process.
This group is going to produce a group Note with a vocabulary that  
should describe the core features and behaviours of mobile user- 
agents to allow a developer to produce a good presentation for the  
mobile.
The group has, in the past charter, ran a small survey among the  
members and identified a short list of device and user-agent features  
that are interesting to many site owners [1].

In this charter we have agreed to make 2 important steps, define a  
process to be used to create vocabulary entries and then collect the  
requests from internal and external groups and define the core  
vocabulary.

Step 1 is very important, in my opinion, because it also defines how  
the entries of the vocabulary will be collected, evaluated and  
eventually added (or not) to the vocabulary. In my mind, the process  
that this group will define, will also serve as a basis for the work  
that will be done by the maintainer of the vocabulary once the  
charter of this group will be over.
Step 2 is very important too, as is what will determine the success  
or not of this working group. It relies a lot on the quality of the  
process that we will define, but will also be the result of the work  
of all the people involved in this group and every external company  
or entity that is interested in participating in our activity. It is  
not hard to list all the things we would like to see in the  
vocabulary, but it will be very hard to actually restrict all that  
information to something that is considered "core" and that makes  
everyone happy with.

We can't start defining the entries of the vocabulary until we have  
defined how we will collect and evaluate the requests. First things  
first, then.

The process should be formal enough to make sure that requests are  
meaningful and that can be converted into vocabulary entries, but it  
should also be lightweight as this group is not a company with 10  
employees working on the vocabulary fulltime and because we do not  
want it to be too complicated to discourage requests.

For the above reasons, my initial proposal is as follows:
a) provide a common method for people to request a vocabulary entry.  
This should be as simple as a web form.
a.1) the request should provide at minimum the description of what  
entry is being requested, what is the entry needed for and maybe a  
suggested name.
a.2) the request may include the ontology definition (using Protégé)
b) the group will receive the request, evaluate it, discuss it
b.1) the group may contact the individual or company that requested  
the entry for more explanations
c) the group will consider if the entry should be created
d) the group will, if not provided, create a new entry in the  
vocabulary using a tool such as Protégé to edit the ontology OR edit  
or approve the one provided by the requesting entity
e) the group will add the entry to the Core Vocabulary

At bullet d what happens is simply the creation of the entry using a  
tool such as Protégé. This will provide the strong data typing that  
is required for this vocabulary and will also provide the ease of a  
visual tool to produce the human-unreadable, but machine-friendly RDF  
ontology definition of the entry.

In my mind the group will receive a lot of requests during this  
group's life. There should be a period, around the end of life of  
this group in which the group might review the entire vocabulary and  
drop entries that are not considered core, too similar to others or  
try to merge, group or review entries requested in the past.

Once all this is done, using Protégé anyone will be able to download  
the full vocabulary, browse it, check it and fit it into his database  
or application and use a DDR that will be defined in the other group  
Note and Recommendations.

I would like to remind that this is a first description of how the  
process might work. The group would like to hear comments and  
suggestions about the process.


Andrea Trasatti
Blog: http://trasatti.blogspot.com/
W3C invited expert

[1] http://www.w3.org/2005/MWI/DDWG/wiki/CoreVocabulary
Received on Monday, 26 February 2007 11:10:06 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 12:13:50 GMT