Mobilizing Youth

Library Patrons as Data Workers?

Was just trolling the Web for references to Mobilizing Generation 2.0 and came across an interesting series of posts related to library blogging here.  Got me thinking about using game mechanics and a human algorithm to engage library patrons in the development of valuable data sets. My comment on David Lee King's blog copied below:
----------------------------------
Thanks for the references to my book (very flattering). Have been reading through your thread here - and love the idea of asking customers to take action.

Riffing on that idea... I starting thinking about some recent stories I've come across about game mechanics driving data collection (see this post: using-games-to.html).

What if you tied the library related action to the gathering of data that was, itself, actionable by the library... and then added a game mechanic (such as competition), so that young people would be more incentivized to participate?

In one of the examples mentioned in the above post, they've created a game where two people try to type the same word to describe a given photo. The two people play a game, but the organizer gets human-powered image recognition... and with lots of people playing the game, they get a valuable database for free!

So thinking (now very roughly) about libraries, I'm wondering a) what kind of data a library needs... and b) what kind of data library patrons might be able to develop.

I'm sure you can answer (a) better than I. For (b), I started wondering if there might be an opporunity to build out Semantic Web ontologies and data to support them. Although I'm sure there's a raging debate about the usefulness of the Semantic Web approach in the librarian community, there may be some interesting/useful connections here. At a library, people are already thinking about how one chunk of information relates to another... there might be some simple way to harvest the value in this thought process to build up a dataset that improves understanding of information on the WWW.

Rough thoughts... thanks for firing a spark... has been interesting to think through...

Game Mechanics for Volunteer Now!

I'm working on ideas for adding game mechanics to the Volunteer Now application. Please comment and let me know what you think.

Volunteer Now application summary:
Enable people to volunteer their expertise in spare time, via a mobile device. More details here.

Proposed game mechanics:

1. Every volunteer task is assigned a given number of points (by a volunteer captain).
   
2. Each time you complete a task, you win these points.
3. At any time, you can donate your points to a nonprofit of your choice - and a sponsoring organization will donate real dollars to match your points.
4. If you've reached a given number of points (say 100), you get a free dinner at a local restaurant, where you must dine with at least 2 other volunteers who also have 100+ points. And together, over dinner, you select a nonprofit to be the recipient of your points.
5. If you have 3-5 people, the sponsoring organization will match your points at 1x in dollars. 6-10 people the match will be at 2x. 10+ people, and there will be a 3x match.
6. You will also be required to upload a photo of the dinner event to the web site.

The idea is you get a double win for your volunteer effort... and also get to meet with people in your locale. What do you think? What improvements can I make to this process? Is it too complicated? Will it work? Please comment - I'm going to give away a free copy of my book to someone who comments!!

Here's a rough diagram describing the process

Using games to collect valuable data

A friend just pointed me to http://www.gwap.com/gwap/ which is a collection of games that use a folksengine (distributed human intelligence) to collect/build valuable data sets. The Nasa Clickworker program broke ground on this concept... which was inspired by the success of Wikipedia. Another recent entry in the field was the Free Rice game, where you solve a vocabulary puzzle and for each one you get right, a sponsor donates $n to a food aid organization. What's different about Gwap is that the results of your game directly contribute to the database, which is the same end-result as Clickworkers - but NASA didn't phrase their program as a game (although if they did, it may have been even more successful).

Gwap offers a series of games to accomplish the following:
* have people identify/tag photos
* identify/tag music
* build a database of common sense
* outline objects in photos
* rank perceived quality of photos

The creators of GWAP wrote an academic paper describing the concept in which they call it a "human algorithm." I heard rumor some time ago that someone(can't remember who) approached some large media/internet company (can't remember which one, google maybe) and said something like "i know how to tag your entire image library in under a year at almost no cost to you. if you hire me, i'll make it happen." and this person's big idea was to create this sort of game. The heresay is that this person was then hired.

It's of particular interest to me as we're working on the Volunteer Now application that intends to, similarly, harvest people's spare time and to direct it towards doing good. I wonder if we should consider a gaming angle? Can the concept of aggregating human knowledge via a distributed game work in the context of doing social good

Other games of this sort mentioned in the academic paper (haven't checked these out yet): Open Mind, Mind Pixel, Cyc

Added June 11:
This is brilliant! Just came across the guy who invented CAPTCHAs is now turning all of the thousands of hours used to solve captchas into productive time. He's scanning old books that need to be digitized (and that are too hard for standard OCR to handle) and using them as captchas - so that all of the human smarts gets put to good use. I love it!  He's also doing an image tagging game ESP game and a language translation game. Some interesting stats also cited:

9 billion hours are spent every year on solitare
7 million  spent to build empire state building
20 million spent to build panama canal

http://www.wired.com/techbiz/it/magazine/15-07/ff_humancomp
http://www.pbs.org/kcet/wiredscience/rss/media/wirs105_it02_hackglow_480.mp4