[liberationtech] Trackography

[ɣęƈƞą]

On Fri, Jan 9, 2015 at 3:30 AM, Tim Libert <tlibert at asc.upenn.edu> wrote:


>  I have scant details of my project on this site, but the general idea of
> webxray is to do fairly large analyses fairly fast.  I’ll have some
> publications/research forthcoming which will detail it more.  if you would
> like to be in the alpha ping me off-list: http://webxray.org/
>
>
Thanks. I'll see you are focused in Health information, that's a very good
point (I was thinking also to target pornography.. funny correlation
happen, when the tracker are shared, look at this
https://www.eff.org/deeplinks/2014/07/white-house-website-includes-unique-non-cookie-tracker-despite-privacy-policy
:) )

After the CCC presentation, I got introduced to http://commoncrawl.org/ .
This is a large collection of crawled web page. I still have to understand
how can I integrate these data with Trackography, and will be the same
question you can pose for your project... because I feel that we have not
to collect all the same data and make our analysis, but:

1) be strong in a specific kind of analysis (or target) and describe, as
best, the data we have
2) public an openAPI, like
https://github.com/vecna/trackmap/blob/master/RESTfulAPI.md
3) lets other make apps/visualisation/analysis using our data
4) enhance our analysis with external sources

This kind of approach can scale.
We can make our research and our findings, and is very important otherwise
the whole collection is useless. but they don't scale.

"Others" can with different point of view and multiple data sources, make
more analysis and usage. I think that can be more effective in the long
term.

good luck with your project, I'll be in touch asap :)
C
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.stanford.edu/pipermail/liberationtech/attachments/20150109/74748d4c/attachment.html>