Gerard Meijssen wrote:
There is a difference between stronger citation standards and better citation technology. I am all for better citation technology. I am completely against raising the entry level of people to contribute to the Wikipedia project.
I agree we should avoid alienating new users as much as possible. So what do you think of introducing new citation/evidence mark-up which the renderer will then use to automatically flag "evidence holes" within an article (see mock-up: http://meta.wikimedia.org/wiki/Image:WikiTextrose_article_citations.png )? This would probably be an immediate milestone for the project as it is something we could implement without Wikidata.
I am well aware of what Wikidata is. Wikidata is the implementation of relational technology within the Mediawiki software. Off itself it provides you with no functionality. A database design is necessary to consider if it possible to create the functionality that you describe.
I'll put together a datamodel diagram once the functional requirements of the system are more well-defined. I've deferred doing so because I wanted to get feedback from the rest of the community first, including their reaction to the basic concept. I'll have a high-level entity-relationship diagram up soon, though.
Anthony DiPierro wrote:
Introducing detailed citation features would require
either abanoning
that simplicity or abandoning the concept of writing
the wikitext
directly. Both of these would be significantly
detrimental to the
Wikipedia project in the short term.
The mark-up would look something like:
[[cite:ISBN:123456789:p. 10|"cited text"|"paraphrase/article text"]]
To me that doesn't seem more complicated than most other basic wiki mark-up.
Ray Saintonge wrote:
This is all theoretically very interesting, and I
cannot oppose it.
Nevertheless until someone is ready to code this it
won't happen .
Meanwhile, many of us who concern ourselves with
content still have to
go on with life without waiting for you to do the
coding, which could
take a long time.
What would you suggest that we non-technical people
do in the meantime?
Erik Möller is working on Wikidata right now, which may be functional before the end of the year. Once this is done hopefully I can put together the beginnings of a prototype.
In the meantime people can help by updating the Meta project pages: [[m:Wikicite]] [[m:WikiTextrose]] . Fleshing out the functionality of the system is particularly important at this stage. Think of how you would use citation data. What would be the best way to visualize/present it? What kind of searches would you like to be able to do with it? Once you have some ideas create mock-ups or rough sketches of them and add those to the project page. People with expertise in library science could also help by creating lists for the sorts of data we should be capturing in the "card catalog" portion of the database. I've been using the "Functional Requirements for Bibliographic Records" as a starting point: http://www.oclc.org/research/projects/frbr/ (SJ- can you confirm whether this is an appropriate reference?). If someone could use it to begin to put together the list of attributes we need for the "card catalog" that would be very helpful.
Using the text relationship database, editors can
now
see at a glance what is authoritative within a particular literature. The article renderer now
takes
It is less cut-and-dried than this; one useful
comparative view would
be the authority-ranking of major essays/articles in a
field assuming
School-of-thought A is correct in its assumptions,
and the
authority-tree assuming some rival School B is correct in its
assumptions.
Certainly, though this depends on the field. In mathematics I'd bet there is relatively little contention; in the physical sciences probably more so, and much more so in the social sciences. Hopefully we can come up with formulas sophisticated enough to identify such patterns (maybe including schools of thought). I would like to see a visualization of citation data for someone like Edward Said, for example :)
virtuous circle begins- a citation based upon a
work
of popular history is exchanged for one relying
upon a
more specialized work, which is later exchanged
for a
scholarly monograph or journal article, which in
turn
encourages reference to primary sources, etc. By
this
process Wikipedia becomes not just accurate, but scholarly and state-of-the-knowledge.
By this process, the claims of the popular works are
also being
verified or disproven by Wikipedia authors over time;
hopefully that information
can be passed on to the book editors/publishers --
as they too enter
the digital age.
Right, and not necessarily just popular works if the Wikipedian is armed with strong enough data. I think we would all be proud the day such a feedback loop is introduced.
__________________________________________ Yahoo! DSL Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com
On 12/6/05, Jonathan Leybovich jleybov@yahoo.com wrote:
Anthony DiPierro wrote:
Introducing detailed citation features would require
either abanoning
that simplicity or abandoning the concept of writing
the wikitext
directly. Both of these would be significantly
detrimental to the
Wikipedia project in the short term.
The mark-up would look something like:
[[cite:ISBN:123456789:p. 10|"cited text"|"paraphrase/article text"]]
To me that doesn't seem more complicated than most other basic wiki mark-up.
Maybe complicated isn't what I'm looking for. But consider the following and whether or not you'd enjoy editing it by hand:
'''Roy [[cite:ISBN:123456789:p. 7|"Roy Orbison's middle name is Kelton"|"Kelton"]] Orbison''' ([[cite:ISBN:123456789:p.9|"He was born in Foo, Bar on April 23 of 1936"|"[[April 23]], [[1936]]"]] – [[cite:ISBN:123456789:p.11|"He died that same year, on the 6th of December"|"[[December 6]], [[1988]]"]]), [[cite:ISBN:123456789:p.13|"They called him "The Big O""|"nicknamed "The Big O""]], was ...
Maybe I'm misunderstanding how these cites would be used, because that was hell; it was even worse than I had thought before going through the actual exercise.
I don't think wiki markup is the proper solution for this. And that means significant redesign. Feel free to prove me wrong here, though, and show us a working model which is just as easy to edit as Wikipedia.
I think it's a great idea, I just think it's years ahead of its time (and that assumes it's designed independently of Wikimedia, cramming it through Wikimedia development processes would only hinder it).
Anthony
Anthony DiPierro wrote:
Maybe complicated isn't what I'm looking for. But consider the following and whether or not you'd enjoy editing it by hand:
'''Roy [[cite:ISBN:123456789:p. 7|"Roy Orbison's middle name is Kelton"|"Kelton"]] Orbison''' ([[cite:ISBN:123456789:p.9|"He was born in Foo, Bar on April 23 of 1936"|"[[April 23]], [[1936]]"]] – [[cite:ISBN:123456789:p.11|"He died that same year, on the 6th of December"|"[[December 6]], [[1988]]"]]), [[cite:ISBN:123456789:p.13|"They called him "The Big O""|"nicknamed "The Big O""]], was ...
Interesting example! Correct bibliography should not override reality checks. Saying that someone born in 1936 died in "that same year", which also happens to be 1988 leads me to the conclusion that 1936=1988. :-) This may not have been Anthony's intention, but if what we are trying to say becomes so obscured by citations this is an omen of our future problems.
Maybe I'm misunderstanding how these cites would be used, because that was hell; it was even worse than I had thought before going through the actual exercise.
I don't think wiki markup is the proper solution for this. And that means significant redesign. Feel free to prove me wrong here, though, and show us a working model which is just as easy to edit as Wikipedia.
I think it's a great idea, I just think it's years ahead of its time (and that assumes it's designed independently of Wikimedia, cramming it through Wikimedia development processes would only hinder it).
Perhaps. When I asked my own question about what the rest of us can do I was not interested in a lot of theoretical material about what library scientists put into card catalogues. I was considering the point of view of a normal Wikipedian (assuming such an animal exists) who is about to write an article and who already has adequate references that he is ready, willing and able to use. His problem is to find a practical way to enter the citations in a way that will scale with the proposed theoretical framework.
The ultimate database seems more and more like the computer geek's response the the unified field theory of physicists. There are still many agnostics among us who would like to see proof of 42's existence.
Ec
On 12/6/05, Ray Saintonge saintonge@telus.net wrote:
Anthony DiPierro wrote:
Maybe complicated isn't what I'm looking for. But consider the following and whether or not you'd enjoy editing it by hand:
'''Roy [[cite:ISBN:123456789:p. 7|"Roy Orbison's middle name is Kelton"|"Kelton"]] Orbison''' ([[cite:ISBN:123456789:p.9|"He was born in Foo, Bar on April 23 of 1936"|"[[April 23]], [[1936]]"]] – [[cite:ISBN:123456789:p.11|"He died that same year, on the 6th of December"|"[[December 6]], [[1988]]"]]), [[cite:ISBN:123456789:p.13|"They called him "The Big O""|"nicknamed "The Big O""]], was ...
Interesting example! Correct bibliography should not override reality checks. Saying that someone born in 1936 died in "that same year", which also happens to be 1988 leads me to the conclusion that 1936=1988. :-) This may not have been Anthony's intention, but if what we are trying to say becomes so obscured by citations this is an omen of our future problems.
Well, yeah, that was intentional. As I was performing the exercise I thought about how not all references are going to be as neat and clean as containing the exact statement in the original. What if the reference has a big long paragraph starting with "In 1988, Orbison began [blah blah whatever]" and ending with "He died that same year". Do you quote the whole paragraph, do you use ellipsis (in hindsight I guess that would be the best solution), do you just add [1988] after "that same year" (in which case why bother with the exact quote in the first place)? It's not so cut and dry.
I was thinking about this yesterday and I imagined some even cooler things that could be done, such as scanning in the actual page itself (it could be hosted on a separate site, with lots of access restrictions, under fair use). But now we're talking even more of a pipe dream than the original plan.
Maybe I'm misunderstanding how these cites would be used, because that was hell; it was even worse than I had thought before going through the actual exercise.
I don't think wiki markup is the proper solution for this. And that means significant redesign. Feel free to prove me wrong here, though, and show us a working model which is just as easy to edit as Wikipedia.
I think it's a great idea, I just think it's years ahead of its time (and that assumes it's designed independently of Wikimedia, cramming it through Wikimedia development processes would only hinder it).
Perhaps. When I asked my own question about what the rest of us can do I was not interested in a lot of theoretical material about what library scientists put into card catalogues. I was considering the point of view of a normal Wikipedian (assuming such an animal exists) who is about to write an article and who already has adequate references that he is ready, willing and able to use. His problem is to find a practical way to enter the citations in a way that will scale with the proposed theoretical framework.
Well, I think a good starting point would be to simply put the reference(s) in the comment field. Then someone or some software could later go through those references and apply them to the appropriate text (based on the diff). I go back and forth as to whether or not it'd be a good idea to have a separate field for this (and whether or not to require it to be filled out, at least for non-minor edits).
Putting in a reference with every single edit is probably a good idea (reverts of vandalism being at least one exception), but I can't even force myself to do it, so I guess it's best kept as some elusive target rather than a real requirement. It'd certainly slow down editing, especially if you got carried away with it (fixing a spelling error and referencing the page in the OED).
The ultimate database seems more and more like the computer geek's response the the unified field theory of physicists. There are still many agnostics among us who would like to see proof of 42's existence.
Ec
Well, yeah, to really do it well you almost need artificial intelligence, the holy grail of computer science.
Anthony
Anthony DiPierro wrote:
On 12/6/05, Ray Saintonge saintonge@telus.net wrote:
Anthony DiPierro wrote:
Maybe complicated isn't what I'm looking for. But consider the following and whether or not you'd enjoy editing it by hand:
'''Roy [[cite:ISBN:123456789:p. 7|"Roy Orbison's middle name is Kelton"|"Kelton"]] Orbison''' ([[cite:ISBN:123456789:p.9|"He was born in Foo, Bar on April 23 of 1936"|"[[April 23]], [[1936]]"]] – [[cite:ISBN:123456789:p.11|"He died that same year, on the 6th of December"|"[[December 6]], [[1988]]"]]), [[cite:ISBN:123456789:p.13|"They called him "The Big O""|"nicknamed "The Big O""]], was ...
Interesting example! Correct bibliography should not override reality checks. Saying that someone born in 1936 died in "that same year", which also happens to be 1988 leads me to the conclusion that 1936=1988. :-) This may not have been Anthony's intention, but if what we are trying to say becomes so obscured by citations this is an omen of our future problems.
Well, yeah, that was intentional.
Great! I suspected that possibility, but couldn't be sure.
As I was performing the exercise I thought about how not all references are going to be as neat and clean as containing the exact statement in the original. What if the reference has a big long paragraph starting with "In 1988, Orbison began [blah blah whatever]" and ending with "He died that same year". Do you quote the whole paragraph, do you use ellipsis (in hindsight I guess that would be the best solution), do you just add [1988] after "that same year" (in which case why bother with the exact quote in the first place)? It's not so cut and dry.
It needs to be adaptable to the different writing styles of different authors.
I was thinking about this yesterday and I imagined some even cooler things that could be done, such as scanning in the actual page itself (it could be hosted on a separate site, with lots of access restrictions, under fair use). But now we're talking even more of a pipe dream than the original plan.
It will be interesting to watch how Google gets throught the courts with its book search feature. If we're lucky the courts will give us some valuable guidance on points that matter to us.
Maybe I'm misunderstanding how these cites would be used, because that was hell; it was even worse than I had thought before going through the actual exercise.
I don't think wiki markup is the proper solution for this. And that means significant redesign. Feel free to prove me wrong here, though, and show us a working model which is just as easy to edit as Wikipedia.
I think it's a great idea, I just think it's years ahead of its time (and that assumes it's designed independently of Wikimedia, cramming it through Wikimedia development processes would only hinder it).
Perhaps. When I asked my own question about what the rest of us can do I was not interested in a lot of theoretical material about what library scientists put into card catalogues. I was considering the point of view of a normal Wikipedian (assuming such an animal exists) who is about to write an article and who already has adequate references that he is ready, willing and able to use. His problem is to find a practical way to enter the citations in a way that will scale with the proposed theoretical framework.
Well, I think a good starting point would be to simply put the reference(s) in the comment field. Then someone or some software could later go through those references and apply them to the appropriate text (based on the diff). I go back and forth as to whether or not it'd be a good idea to have a separate field for this (and whether or not to require it to be filled out, at least for non-minor edits).
Putting in a reference with every single edit is probably a good idea (reverts of vandalism being at least one exception), but I can't even force myself to do it, so I guess it's best kept as some elusive target rather than a real requirement. It'd certainly slow down editing, especially if you got carried away with it (fixing a spelling error and referencing the page in the OED).
When I was doing more with Wikisource I raised the possibility of synchronized side-by-side edit boxes. I had translations in mind at the time, but it could work equally well for annotations and references. For word references, I look forward to having Wiktionary fulfill that function, but even with 106,000 entries we are still far from being comprehensive enough to do that well.
Documenting is tedious business. With the meaning of words in particular I find that people add meanings from their own memories, and that can be far from accurate. That sphere also has an ongoing debate between descriptivism and prescriptivism. It's easy to spend an hour documenting a single word in simple cases. Common words can be far more problematical.
As much as we may want to see everything documented a reasonable balance needs to be found that will still encourage contributors to go as far as they want beyond minimum standards. We need to establish minimum standards that can even vary with the type of article. Biographical articles about living persons would certainly require a higher standard than biographies of Pokémon characters.
Ec
wikimedia-l@lists.wikimedia.org