On Sun, Oct 7, 2012 at 11:44 AM, Dave Sutter <[email protected]> wrote: > I like the idea of an automated quality checker for map edits. That can be > seen as an advancement to the simple rules based checking done when a commit > is done in JOSM. > > It might be easier to train the classifier if it concentrates on bad edits or > bad commits rather than bad mappers.
Hopefully. Perhaps my understanding of stats and analysis aren't up to this task. Most of the things I try to measure seem like tiny signals buried in tons of noise. And there really is a lot of variation in normal mapping. I know that I've committed single changesets that look like bad mapping in isolation. I'm sure that most of us have. I love what I'm seeing in some of the new tools. changemonger is giving human readable summaries of changesets. The interest in activity streams looks promising. There are really cool things on the horizon. _______________________________________________ talk mailing list [email protected] http://lists.openstreetmap.org/listinfo/talk
