Sustainability of Stack Exchange Q&A communities | Hacker News Reader

Sustainability of Stack Exchange Q&A communities

arxiv.org

15 pointsatomashevic10 comments

Hi HN,

Here's a paper we've been working on for some time. Essentially, we looked at differences in failed vs successful SE communities on the same topic (e.g. failed vs launched Astronomy SE) and tried to identify what made them different in the early days (first 6 months).

I hope it may be useful/interesting to those interested in (q&a) community building.

Here's the abstract:

Knowledge-sharing communities are a fundamental element of any knowledge-based society. Understanding how they emerge, function, and disappear is thus of crucial importance. Many social and economic factors influence sustainable knowledge-sharing communities. Here we explore the role of the structure of social interactions and social trust in the emergence of these communities. Using tools from complex network theory, we analyze the early evolution of social structure in four pairs of StackExchange communities, each corresponding to one active and one closed community on the same topic. We adapt the dynamical reputation model to quantify the evolution of social trust in these communities. Our analysis shows that active communities have higher local cohesiveness and develop stable and more strongly connected cores. The average reputation is higher in sustainable communities. In these communities, the trust between core members develops early and remains high over time. Our results imply that efforts to create a stable and trustworthy core may be crucial for building a sustainable knowledge-sharing community.

What type of efforts was most useful in creating a sustainable community?

We didn't focus on individual contributions, only on community-level metrics. The reputation metric we used rewards frequency of contributions (questions, answer, comments) and that turned out to make a difference in building a strong core.

Also, core is not static, sustainable communities have the ability to "pull" new users into the core, so new, enthusiastic users are not discarded by the existing core.

If you are interested in individual efforts and user archetypes take a look at another work: https://arxiv.org/abs/1709.05111

After years of answers and over 90,000 reputation earned, I recently had my Stack Overflow account banned, with no explanation. I am one of only 70 people on the site with a Go silver badge:

https://stackoverflow.com/help/badges/1665/go?userid=1002260

So Stack Exchange doesn't seem to care about contributors anymore.

I feel GitHub issues is the new StackExchange in terms of solution quality for specific library or framework questions.

What did support tell you?

I'm sorry to hear that.

One of the reasons we focused on smaller niche SE communities is because Stack Overlflow became completely "another beast" where the processes of valuing contribution and building reputation work in a very different way (or don't work at all) on a large scale.

After my account was banned too, I looked around and eventually joined wotas.net. It's very small and you don't get my answers but... at least it's something. Sharing this in the hope that other SO exodus want to join as well (more people = better answers)

One small piece of anecdata which kind of supports the theme of cohesiveness being important.

I've been a part of a relatively small Stackexchange site (security.stackexchange.com) for 11.5 years and over that time I've made friends with other Sec.SE participants which have lasted to this day. Most of the early core users aren't that active on the site any more, but quite a few of us still hang out on-line and meet up when the opportunity presents.

Thanks for sharing! I made mistake in my comment when I called closed Stackexchange sites "failed" because the core users might have migrated to some other platform or formed another form of a meaningful community.

Reply on news.ycombinator.com