New service to monitor mapping deployments

[Rob Baker]

Hi all,

Somewhat in response to questions from the PhD students but also
wanting to provide a heads up as I think it'll be of interest to the
Ushahidi and larger participatory mapping community as well, I'm
currently working on a research project with Internews to analyze key
vectors within the entire Crowdmap data set (just shy of 13,000
deployments as of Oct 2011).

This project, CrowdGlobe, set to launch in January with a report of
our findings in March, will identify key points of information to find
similar patterns within deployments, the context of these deployments,
and the long tail of deployments that never made it off the ground.
We will also be including information from some key, non-hosted
deployments as well, such as those in Japan and Haiti.

I do have one question for the group here: as I'm currently querying
these databases for the information we need, I'm wondering what
programs and/or formats you all might use to store and analyze large
datasets. Google Fusion tables may not scale to this size; Excel
might but I feel like there must be better options out there. Any
thoughts?

Thanks very much. We're excited to get this information out soon and
will certainly be soliciting feedback before and after. We're also
certainly interested to hear if there is any specific information
about not just Ushahidi but this field in general as it's our hope to
extend this project beyond this platform and into other services.

Rob
rrb...@gmail.com (robb...@ushahidi.com)

[Heather Ford]

Great news, Rob. This is really necessary work so I'm really glad you're working on it.

I would personally love to see more information on verification tagging e.g. proportion of verified to unverified tags, rate at which information becomes verified through tagging.

Not sure about formats -- I'm guessing .csv and excel files will work but I'm not a quantitative researcher so I'll leave it up to those to answer that part :)

best,
Heather.

Heather Ford
Ethnographer: Ushahidi / SwiftRiver
http://ushahidi.com | http://swiftly.org
@hfordsa on Twitter
http://hblog.org

[Amanda]

My 2¢ - .csv is my favorite. It's simple, easy to parse, and doesn't
require any special software to open.

> > rrba...@gmail.com (robba...@ushahidi.com)
>
> Heather Ford
> Ethnographer: Ushahidi / SwiftRiverhttp://ushahidi.com|http://swiftly.org
> @hfordsa on Twitterhttp://hblog.org

[Matteo Beltramin]

great work. I'm very interested in this project.
I think it wiil be very useful for my thesis

[george chamales]

One issue that you'll possibly hit is that the default ushahidi csv export format puts all the categories in one cell separated by commas.  This makes it a bit difficult to parse by category.  One way to deal with that would be to have the categories as their own column, with a 1 in that column if a given report is in that category.

Also, I don't believe that the csv export has support for custom forms, however I don't know how many people on Crowdmap are using that feature.

george

George Chamales
Rogue Genius LLC

http://roguegenius.com
Phone:  718.288.7718
Skype:  notgeorge