Web collection tools

mc (apparently) - 07:49pm Jul 13, 2004 PST
via email

Matt Neuberg writes
>Sometimes a new idea is so simple, you can't believe no one's
>thought of it before.

<http://db.tidbits.com/getbits.acgi?tbart=07739>

He's referring to Webstractor, the browser that lets you collect and
edit web pages.

We agree that this is a really cool idea. But it's not that no one's
thought of it before. So, just by way of info, in 2001/2002 a
research prototype called Hunter Gatherer was published that lets
users create collections from web page components. (Indeed we have a
provisional patent on the collection-making approach).

Descriptions of the tool, related papers and a video can be found at
http://shaka.dgp.toronto.edu/hg/overview

With Hunter Gatherer, the focus was on letting users collect chunks
of web pages of interest into what we call "collections" - which can
be added to and subtracted from, revised and so on.

This is the main difference btwn the HG approach and Webstractor.
With HG, you indicate what part of the page you'd like to add to a
collection, hit a to add, and that's it: the component is in your
collection. With Webstractor you're picking up the entire page and
*then* you prune it.

Our empahasis was on reducing steps to make a collection - something
our research showed lots of people do, but do poorly because their
are too many steps involved to make a collection.

<http://www2002.org/CDROM/refereed/130/index.html>

In Webstractor, it seems that to make collections, you need to copy
and paste from all of the collected documents into a new one. We
automate the collection building; we also make it possible to share
collections by URL, so you have a copy and your pals can have a copy
- and your pals can edit their version without changing yours.

Hunter Gatherer was implemented as a proxy service. It would be
fairly straight ahead to translate this service to a plug-in but as
researchers, development wasn't the core mandate. The idea of
proxy/plugin was to let users use their own rather than new tools to
do information work. It was also to push standards like XHTML,
JavaScript and xPath. If anyone would like to collaborate with us on
converting HG to a plugin and continue the research, we'd love to
hear from you.

If you'd like to try hunter gatherer, we're still interested in a
variety of research questions about the collection-making approach.
send an email to
<mc-res at dgp.toronto.edu> and you'll get an automatic reply on how
to set up the tool. Please note it's not perfect; it's a concept.
but the way it supports the interaction for collection making is even
easier than Webstractor.

Thanks,
mc schraefel