[Commons-l] Yotophoto (free image search engine) developments

[bawolff]

Well my biggest comment (although this is not really related to the
subject at hand) is that yotophoto doesn't actually index things. If
I'm trying to find a free image, I generally search yotophoto and
mayflower. Mayflower finds stuff 10 times more often then yotophoto.
yotophoto theoretically should be a superset of mayflower, how come
they don't find anything one you search them (I've maybe found images
with them twice).
</rant>
-bawolff
> ---------- Forwarded message ----------
> Date: 10-Mar-2007 07:51
> Subject: Re: [Provide feedback] Wikimedia Commons photos
>
> Brianna,
>
> First, sorry for the delayed response and thanks for your comments. I
> especially appreciate feedback from people in your position as I want
> Yotophoto to be embraced by those contributing to  important projects
> like Wikimedia Commons.
>
> Right now, most of the links to go back to Wikipedia instead of the
> Commons for the main reason that awareness of the Wikipedia project is
> higher than that of Commons. It's just easier for people to get their
> head around it if they are unfamiliar with Commons.  (As an aside, I
> find it pretty tough  explaining how everything works with regards to
> 'free' licensing and how we index only third party sites and so on.
> You wouldn't believe how often people write us to ask if they can use
> one of "our" images or how they can upload their own images to us).
>
> One thing I also like about linking to the Wikipedia page over the
> Commons one is that the list of pages that link to the image seem to
> be more accurate on Wikipedia than Commons. Is this your experience as
> well? Any feedback on this would be helpful. See #3 below for a little
> more on why this is relevant.
>
> Having said that, I'm working a new version that will change a few
> things. Here's an overview:
>
> 1) Every site we index will have it's own "about" page. Here we will
> explain a bit about what each site is about, how to contribute, what
> kind of licenses the site allows and so on. In the case of the
> Wikipedia and Commons we intend to outline the relationship between
> the two and how Wikipedia will often hold a 'fake' image for images
> that come from Commons as you described.
>
> [I told him about Check Usage.]
>
> 2) Most images will have an intermediate preview (probably in a pop-up
> lightbox). From here we can provide a link to *both* the Commons and
> Wikipedia pages for an image. What do you think?
>
> 3) Yotophoto will eventually index images using captions from the
> various Wikipedia articles that include said image. So if an image
> doesn't have a complete description on it's own page - we can still
> index it based on all the captions used for that image throughout
> Wikipedia. Commons tends to have decent image descriptions but for
> images only on Wikipedia taken from third party sites (say a PD US-GOV
> image), the descriptions are often lacking.
>
> [I said:
> On Commons we have enough trouble forcing people to annotate their
> files with basic information like licensing and author, let alone a
> decent content description. Many people when they submit files, are
> submitting them with a specific purpose in mind and are focused on
> that, not on the possibility that the file might be re-used in the
> future for something else. Also when one is looking at an image, it is
> often quite self-explanatory, so the user doesn't feel the need to
> explain the purpose or background to the file beyond three words. Of
> course that doesn't stack up well for searching since machines can't
> "read" images well at all (or even, mostly, at all).
> ]
>
> 4) Yotophoto would like to enable one click republishing of images
> found so people can repost to their own blogs etc. To do this we will
> provide a copy and paste HTML code they can use. This code will link
> back to the image source (ie. Commons) as well as to the image
> license. Sites like Commons will get tons of back-links and exposure
> in this way I think. I like to think of this as the "Flickerization of
> Wikipedia" (and Commons).
>
> [I forwarded this part to wikitech-l to see what they say.]
>
> 5) The depth of the index will increase massively. We intend to one
> day have the entire Commons and Wikipeida image collections searchable
> (right now it is only a fraction) including not only photos but other
> types of images as well like maps, charts, flags and so on.
>
>
> I'd love to know what you think, and if you have any further
> suggestions. As well, feel free to invite anyone from the Commons to
> submit their ideas to me as well. I hadn't seen the Mayflower tool
> before but it looks quite nice -  I want to look into it more.
>
> Regards,
>
>
>
>
>
> On 2/26/07, brianna.laugher at gmail.com <brianna.laugher at gmail.com> wrote:
> > Brianna Laugher sent a message using the contact form at
> > http://yotophoto.com/feedback.
> >
> > Hi, I'm a volunteer on the Wikimedia Commons website. I just wanted to say
> > how impressed I am with Yotophoto at the moment. Displaying the license
> > and the image dimensions in the results is particularly awesome. A couple
> > of comments:
> >
> > Is it possible to do domain-specific searches? (like in Google,
> > site:commons.wikimedia.org) - doesn't seem like it, so far
> >
> > As you may or may not know, when projects like Wikipedia use images from
> > Wikimedia Commons, it creates 'fake' image pages like this one:
> > http://en.wikipedia.org/wiki/Image:Common_frog.jpg (note the red tab and
> > the tag 'This is a file from the Wikimedia Commons. The description on its
> > description page there is shown below.'). To better promote Wikimedia
> > Commons, it would be preferable if such images could be referred back to
> > their Wikimedia Commons page, rather than the 'fake' page at Wikipedia.
> > Would that be possible?
> >
> > Also, another Wikimedia editor has developed an image search engine
> > specifically for Wikimedia Commons called 'Mayflower', which you might be
> > interested in checking out:
> > http://tools.wikimedia.de/~tangotango/mayflower/
> >
> > As the default wiki search engine for media performs very poorly, we tend
> > to be very happy whenever we find out about nice products like these :)
> >
> > thanks!
> >
>
>
>
> ------------------------------
>
> _______________________________________________
> Commons-l mailing list
> Commons-l at lists.wikimedia.org
> http://lists.wikimedia.org/mailman/listinfo/commons-l
>
>
> End of Commons-l Digest, Vol 22, Issue 12
> *****************************************
>