OpenIDEO is an open innovation platform. Join our global community to solve big challenges for social good. Sign Up / Login or Learn more

Reputation management systems in online communities

We can learn from research and practice on reputation management in online communities that address the problem of how to "bubble up" the best contributions or users in a dynamic community.

Photo of Joel Chan
2 11

Written by

Many online communities (e.g., user-driven map systems, user/customer review forums/communities) face an analogous problem of how to "bubble up" the best for recognition. This problem can be decomposed in a number of ways, but some major components include the input problem ["what information should go into determining a given entity's "reputation"], the computation problem ["how should we combine/weight the inputs to generate sensible reputations"], and the diffusion problem ["how should this reputation information be diffused in the community"]. I think there are potential pitfalls and "good tricks" to be gleaned from such avenues as ["how can I find the best products", "how can I find the most helpful reviews"] and Quora ["how can I find the best users to follow", "how can I find the most interesting/helpful/accurate answers to questions I'm interested in"]. Other examples I can think of include StackOverflow, 9GAG, Reddit, Pinterest, and of course, OpenIDEO! 

There is a well-developed field of academic research (in computer science and related fields) called reputation modeling, which studies how communities solve this problem in better or worse ways. Here is a link to a recent workshop on reputation modeling:

One key insight that I think can be gleaned and transferred is the idea of moving away from "top-down" moderation of reputation to really trusting the community to bubble up the best, and building a system around that. For a problem like identifying the best companies who are innovating and producing sustainable value, I believe the torrent of information can prove to be too much for even the most diligent curators, especially if we want the system to be dynamic and sensitive to change, free of bias, and friendly to the addition of "newcomers". Something like the "spotting network" inspiration that leverages the power of bottom-up-driven social media (combined with clever "reputation management" algorithms and architectures and combine all of the inputs in a sensible way) is, I think, a promising direction to pursue.


Join the conversation:

Photo of DeletedUser


Great direction for a Concept - another crowd-sourcing reference point: Yelp -

View all comments