In the QS world I’d like to live in, our personal data would be easily available to us to learn from using many different methods and tools. Here are some conditions I think would make this easier:
Data can be exported from the various systems we use into a simple format for exploration.
We can store and backup our data using whatever method we want.
We can share our data with whomever we want.
We can rescind permission to look at our data.
We can flow our data into diverse visualization templates and analytical systems.
I’ve tried to express these needs briefly and simply, but any of them - and certainly all of them together - require changes in the systems we currently use, and these changes may be challenging for technical, business, social, and political reasons.
I know many people in our community have worked on parts of this problem, and I’m interested in your comments and ideas.
“We can store and backup our data using whatever method we want.” One tool I’m curious about here is Social Safe. I just downloaded a trial version. It looks like Social Safe both backs up your social data and allows you to export to a csv file. That’s what I’m talking about. This is not detailed QS biometric data, but it’s a piece of the puzzle, and possibly a model of one component of the “QS world I want to live in.”
Eri Gentry and I moderated a “work your own data” event last night where we experimented with sharing our experiences with long term data sets, and discussed the idea of an anonymized data repository managed by QS Labs that would work something like this (from the perspective of a user):
*I log in using a password I generate.
*I upload my data into a directory.
*This data is accessible to others, associated only with an anonymous user ID
*Anybody (?) can experiment with visualizing data in this directory, and the visualizations can be tagged with the user ID
*I can view the visualizations of my own data, but not others.
I know this is a very simple outline, and there are many other more complex scenarios, such as sharing, groups, etc. But I wanted to write down this simple case, and ask for comment!
What do you think?
Great session last night, with lots of ideas floating around.
A few comments:
allow ‘conversations’ between the visualizer/processor and the person who uploads the data. A visualizer could post a question to the anonymous uploader, and the uploader could respond. I realize this could be a bit cumbersome, without direct e-mail communication, but it could help with understanding the data.
Not sure if we have to limit who can see the visualizations of the data. I might like to see what somebody else’s data visualization looks like
Encourage people to upload some metadata that can help with understanding the data, or see patterns between datasets. Whatever the upload is comfortable sharing, for example:
- I walk to work everyday
- my walkscore is 78
- I traveled for 3 weeks at the beginning of November
- I go for long bikerides
I’d like to share more data (both with the public and with specific companies and organizations). The main reason I don’t isn’t that I can’t do so anonymously, but that the current (automated) sharing options are often all or nothing.
For example, I’d like to make my Foursquare check-ins public, but only restaurant check-ins, after one week, and without the exact time. I could publish my check-ins anonymously, but it wouldn’t be too hard to figure out whose data it is. I’d like to share my gym check-ins with a company that rewards me for working out, but have to grant them access to all my check-ins.
It’s nice to be able to export data in bulk from a service; this lets me clean up data before publishing it. But this approach only seems practical for one-off projects with a fixed duration.