On 2018-05-18 14:05:10 UTC, Deleted wrote:

imported from 2606

On 2018-05-23 12:41:30 UTC, Deleted wrote:

https://groups.google.com/forum/#!msg/tosdr/ldc1it-6taY/Y83sTYVMVHoJ Hi,

this is something that came up in the opennotice confcall just now -
if you look at for instance http://tos-dr.info/topics.html#track you
see the following data point texts:

You must use your real name (yahoo)
You must use your real name (facebook)
You must use your real name (xing)
You must provide your legal name (github)

These (at least the first three) are obviously the same thing (even
though we could weigh them differently depending on the type of
service).

Right now, we are not reflecting that, as data point texts are just
free texts. But we could start doing that, and that way we would build
up a valuable ontology of typical data points that we see recurring a
lot.

Once we have such a list, sites could pro-actively auto-review
themselves, and in particular, if sites use templated ToS texts like
the ones Docracy works on, then these could come pre-reviewed with
such standardized data point texts. The weighing would then still
depend on the specific service type, although even that could default
to a common suggested value.

Of course you could also then try to detect commonly used phrases in
ToS texts, trying to treat human-readable texts as if they were
machine-readable, but that's another extra step. For now just
standardizing the data point texts would already be a very valuable
thing, i think.

Thanks Veronica Par and others on the confcall for the suggestion!

Ciao,
Michiel

On 2018-05-30 12:50:25 UTC, Deleted wrote:

imported status as declined