Hi Helene, all,
I had an exchange with Ken who's behind Opinion Space, forwarding him your mail. I'm posting his response below.
Regarding how this tool could help the Commons, my view is as a starting point for the collective exploration of complex ideas. The starting subject (work) is deceptively simple, as it incorporates a range of related subjects e.g. economy, values and so on.
This tool can also help with evaluations. We need 20-40 more participants, ideally encouraged to take time to read and rate the textual responses of others, as the system needs at least 4-5 evaluations of each idea to generate a good estimate of its value.
In addition, and subsequent to this step, we can look at how the different suggestions may link up with one another. But more on this later stage, after we get a good sampling of answers and ratings.
On which subject, would it be possible to send this discussion to the wider ECC mailing list?
Cheers,
Tina
> This tool looks very interesting, though its subtleties (criteria for
> size, color, proximity etc) are not obvious at first sight, and it's hard
> to have an overview of what people have responded to select the ones you
> want to react to. I also had difficulties to post a second idea. Maybe I
> need to 'earn more points'? :)
Thank you for this helpful feedback! We will add more details about the
criteria and size etc to the About page. Regarding overview, once the
systems collects more input and ratings, we can analyze the data to
produce a graphical display...As for posting a second idea, just
register under a new userid to post additional ideas (we should explain
this also on the About page).
> Reading a bit more on how the algorithms work, I find interesting that the
> ideas are positioned by proximity of opinion based on a filter/profile
> question, and that the scoring model (though I didn't get how scores were
> expressed) took into account the proximity of the profiles of ideator and
> rater. It's also interesting to see that this has been around for a few
> years. So now there must be some data available on its effectiveness and
> possibilities of improvement.
We will present this paper at a conference next week:
http://goldberg.berkeley.edu/pubs/MIT-LINC-OS-Learning-S13.pdf
There are many areas for improvement and we'd be happy to talk further
with you! We are now beginning a new phase of the project to add support
for non-English participants...