Research: Community/People Dasboards

Mozillians
#1

Hello,

I’m doing a research about Community/People dashboards other organizations or companies are using to measure their human resources performance, status and health.

So far I’ve found a lot of technical dashboards to track non-human stuff, so if you happen to know an org that it’s using something like I’ve previously described, I would be really interested in taking a look at it, or even just know what does it look like (if it’s not public).

Thanks!

Update: Note that I’m not looking for HR tools to manage people, but more to visualize people.

#2

what attributes do you hope to track? Is this just work, or does it include things like reputation?

This was an interesting talk at OSCON:

http://www.slideshare.net/geekygirldawn/network-analysis-people-and-open-source-communities

#3

Still to be defined, but the idea is to track things outside the usual “commits to github”, “posts in mailing lists”…

We plan to run 1:1 with volunteers in these communities to know things that can’t be automated and have to be filled manually.

#4

This was one area that CBT did research on last year and as I recall it was decided to reject everything that existed and just build our own solution. As we do at Mozilla.

The solution Christie Koehler thought was a clear winner is http://bitergia.com/

It’s used by a number of prominent open source communities, integrates with nearly everything we use that isn’t custom built, and Mozilla Webmaker is listed as one of their users!

I think @adamlofting may know something more about this.

1 Like
#5

Yes, I can add some more info here :smile:

Bitergia are great. They are an extremely helpful and intelligent team who even volunteered on our pre-Portland-All-hands contribution analysis project here: https://wiki.mozilla.org/Contribute/analysis

The tools they built are great if you’re looking at participation in open source software from a development point of view (Code / Bugs / IRC / mailing lists) but this was only part of what we were looking at last year. For context, the Foundation data I was looking at included a couple of hundred people who contributed this way, out of the ten thousand or so total contributors. So we needed many systems.

I did use one of their tools (Bicho) in the infrastrucuture that I pieced together to help get data out of Bugzilla for Webmaker.

If I was tasked with building a dashboard for a specific engineering project today, I’d definitely start with their tools.

The Webmaker dashboard Bitergia built for us (very kindly) was a proof of concept we hoped would have wider adoption, but it wasn’t an org priority so it hasn’t moved any further since then. We were missing some data, like mailing list archive files and IRC logs, but these could be solved if needed. One other challenge we hit was that our Bugzilla history is so large it was going to take a very long time to get the data into Bitergia. That’s why the demo just looked at Webmaker.

This was part of the reason why the data we cared about was being joined up in our internal project Baloo, because our data team could copy / transform / manipulate the data directly rather than rely on ‘scraping’ processes. I should note that no one is actively working on Baloo right now.

–

While that’s some extra background on our work, that doesn’t really answer the question of visualization.

The general challenge you’ll face is joining up data from many different types of interactions in a single format (this is the data warehousing excercise that Baloo was focussed on) and then using that central data to build visualizations.

#6

Thank you all for the conversation here, this is really great.

We are actually going to talk to Bitergia tomorrow, and yes, they have fantastic solutions for open source projects focusing on development. I think this is a project that Mozilla will need to tackle at some point (I hope rather soon :wink: ).

What we’re looking for this quarter is something that we can build with data we can crunch more manually. We understand the difficulties of creating a dashboard and we first want to test the usefulness of the information and build user stories before we move into creating more complex dashboards.

Apart from that, what we need to understand is what is the information that will help us support communities better and understand their health, where there is potential, where Mozilla can help.

In a way is also about developing “indexes” that help us understand better the situation in a particular community. Those don’t necessarily need to have a lot of quantitative data behind.

So if you have any pointers in that direction that would be really great or if you want to get involved in general.

Let me know if there are more questions,

Thanks

  • Rosana