[drupal-devel] An idea for Drupal in disaster relief efforts
scott at 4th.com
Thu Sep 1 02:23:29 UTC 2005
Good evening, Drupalites!
My brother, who is a mechanical engineer but not a computer geek, approached me
with an idea for a web application that could be of help to the victims of
Hurricane Katrina and other natural disasters in the future. As he was
describing the idea to me, I kept thinking, "Hmmmm.....Drupal can do that.
Hmmmm....Drupal can do *that*, too. And...yeah, it can do *that* as well.
And...well, it could do *that* also, with just a little code tweaking."
I don't have the server resources to host this, but it occurs to me that someone
on the Drupal list just might be able and willing to do so. Todd (my brother)
said it was fine to post the idea to the list, and if someone else wants to
implement it, he's okay with that. It's not about ego or credit, it's about
This isn't totally a "development" idea, because I think Drupal already has
most of the features that are needed, but I'm posting to the developer list
because it needs the help of people with that level of technical skill to
evaluate the idea and implement it if it's found to be feasible. Apologies
if this offends anyone. Since this is slightly off-topic for the list, I do
suggest that responses be sent off-list to me, scott at 4th.com. If there is
enough interest, I'll set up a dedicated email list on my own server.
Here's the idea...
In almost any major disaster, one of the worst losses for families is the kind
of irreplaceable property that represents memories. That would include things
like photo albums, school ribbons, certificates and placques, kids' art work
from school, and so on. Many times these items are destroyed, but not always.
Sometimes wind or water carries lightweight items for many kilometers and
deposits them somewhere. They end up getting thrown away because whoever finds
them has no way at all to locate the owner, since the items seldom have any
identifying names or addresses on them -- and that's a tragedy. It's bad enough
that families lose their homes and tangible property, but to have family
memorabilia survive the disaster itself but then be thrown away is a heartbreak
in its own right.
What Todd proposes is an Internet-based web site to unite people who find photos
or other memory-items after a disaster, with the people who lost them. In essence
what's needed is a large, keyword-indexed image database.
Ideally, one would use some kind of image-recognition software to automatically
catalog images -- especially faces -- based on visual characteristics and provide
a search facility akin to Google's image-search function. But given volunteer
workers, this could be done using the human brain's old-fashioned but very
sophisticated image recognition capabilities.
Imagine a web site where anyone could upload an image and assign taxonomy
keywords to describe it. For example, you could have vocabularies for:
* Item type
Photograph, digital picture of an object (and different types of that, etc.),
drawing, painting, color, monochrome, album or boxed collection of photos,
placque or certificate, ....
* Type of image (for photos only)
Individual portrait, group portrait, candid activity snapshot, scenic ...
* Description of main living subject(s), if applicable (multiple select)
Female, male, child, senior citizen, teen, adult, infant, animals,
African, Caucasian, Asian, Hispanic, couple, family with children,
* Description of setting, if applicable
Outdoors, indoors, religious setting, at home, at work, museum or exhibit,
school or college, daytime, nighttime, twilight, crowded scene, city, rural
* Non-living items prominent in photo, if applicable
Ocean, river or lake, mountains, trees, large building, house or apartment,
motor vehicles, aircraft
* Descriptive items for objects (if entry is a photo of a found object)
Use a free-tagging vocabulary for this
* Location where found
Use the "location" module to support this
* General keywords not covered above
Use a free-tagging vocabulary for this
* Description of item in the body of the node -- searchable in the usual
Drupal search index
Even if you had 100K photos in the system, this level of detail would narrow
most searches down to a few hundred items, which may be a bit tedious but
*is* feasible for family members searching for their lost memories.
How do you enable people to reclaim their items? To a certain degree, you would
have to rely on trust. Maybe you simply use the "contact" module to enable them
to send an email to the person who uploaded the image, and you leave it up to
those individuals to work out the trust issues and arrange return of the item.
Or you could use an e-commerce module with a micropayment that simply
establishes the identity record for the person who reclaims an item, without
actually "charging" them any substantive amount. I'm not sure how you would
approach this, but there are smarter people than me on this list!
To prevent inappropriate posting, you would need to have images go into a
moderation queue where a team of editors could verify that they aren't porn or
commercial ad banners or other icky stuff, then approve them for publication.
Obviously, this would need to be hosted on a fairly beefy server, because it
has the potential to get a lot of hits in a short period of time.
Among the spinoff benefits something like this could bring:
* Family members with an RSS reader could actually "syndicate" a search that
matches a particular item type in which they have a special interest, and
they would be notified when a new item of that type was posted by someone.
(You'd need an "and" rather than an "or" search, probably.)
* Eventually the system could be expanded to other found-belongings categories.
Maybe one would even want to have subdomains, on different hosts, for various
major categories, e.g., "photos.somedomain.org" and "objects.somedomain.org",
with "www.somedomain.org" simply being an overview page and links to the
other two servers.
* At worst case, at least with photos, if someone accidentally or intentionally
claims an original that isn't theirs, and the real owner later asks for it,
you still at least have the scanned image uploaded to Drupal. That's better
than nothing, which was what they would have had if the system were not in
Drupal has taxonomies that make the categorization very powerful. Drupal scales
well to a large number of nodes. Drupal has an embedded search engine. Drupal
intrinsically organizes images along taxonomy lines. In other words, Drupal
has a lot of the features that this system would need.
I'm not sure if this is feasible or not. My inclination is to think that it is,
given the proper server resources. It would certainly be a fantastic service
for the Open Source community to provide to the real world, and it would be a
project that could be re-used for future disasters that might occur anywhere
in the world.
This would have the incidental benefit of being *very* good public relations
for the Drupal project, if it were implemented by a coalition of Drupal
developers or consultants.
If anyone is interested in pursuing this, and has the time and resources to
devote, please feel free to take the idea and run with it. I'd appreciate an
email from anyone who decides to take this on, because I'd like to be involved
in helping to build or moderate the site, but there's no feeling of "ownership"
of the idea. For all practical purposes, the idea is released into the Open
Source community in a Creative Commons sort of way, in hopes that it can be
used by someone to benefit other human beings.
Kind regards, and thanks for reading.
Scott Courtney Drupal user name: "syscrusher" http://drupal.org/user/9184
scott at 4th dot com Drupal projects: http://drupal.org/project/user/9184
More information about the drupal-devel