Forum:A Universal/Common Biorep System

Question

[Obi Griffith]

Is anyone aware of a multi-site rep system?

We've seen how powerful "rep point" systems can be for encouraging community-driven altruistic contribution to open resources. Obvious examples are Biostar and various stack exchanges. Even Github in some way encourages committing code to public repos by providing nice summaries and visualizations of your contributions to various projects. That can also be a way to demonstrate rep. Of course, github is so intrinsically useful to the submitter that its not as much of an issue. Is there something comparable at wikipedia? Its not clear to me what the incentive for editing a wikipedia article is. But, I think there is a clear relationship between community size and amount of incentive needed to promote contribution. Wikipedia is used by the whole world and as such is more likely to benefit from the rare obsessives who help us all out by summarizing every star trek episode. Creating content on Biostar can sometimes be a high-cost, low-value enterprise. But it seems to have just enough momentum to keep growing. I do believe badges and rep points are one factor that keeps people coming back.

Now, suppose that I am launching a bioinformatics resource and hoping to benefit from some amount of crowd-sourcing. My resource is likely going to be too specialized to have an audience large enough where a rep point system would work. So, does it make sense to have a universal/common BioRep system that could be used across different resources? Validating/curating data in my specific resource would be just one way to get such points. Answering biostar/seqanswers questions, editing gene wikis, etc would be others. This would act as a common digital currency for academic contributions. Almost like bitcoins that accumulate by “intellectual mining” in various ways. They should not be totally universal. You might want to subdivide according to field/domain (e.g., Bio, Chem, etc). Given how many different web communities use some concept of rep I’m surprised this doesn’t exist yet.

Thoughts?

[Jeremy Leipzig]

I'm tired of rep. I would rather have a $5 Starbucks card than 10000 rep.

[Istvan Albert]

I going to upvote this

[Istvan Albert]

but actually that would be a neat thing to implement - I am going to look into that $5 starbuck card thing, seriously, if only there were a way to award this

[xb]

It requires careful thought and vision to do so, or we would run a risk of disturbing the peaceful and friendly community - even with a cent of financial motivation.

[Chris Miller]

Anyone who has a best answer accepted during the month of December gets entered into a drawing for a 10 dollar Starbucks card. Non-controversial, easy to implement. Metafilter does similar things with the community choosing fantastic posts and gifting them a little something, and it's always seemed like a nice thing to me.

[Chris Miller]

Hah - I'll take you up on that. Convince Istvan to transfer your rep to me and I'll mail you a Starbucks card. :) It might also work if we could convince him that we're really the same person

[Sean Davis]

The central problem here is both technical and cultural. Promotion in either academia or industry currently largely ignores the community involvement measured by such "rep" systems.

[Chris Miller]

Why not go all the way and implement Whuffie? https://en.wikipedia.org/wiki/Whuffie

[Istvan Albert]

we will play around with these concepts soon - we'll probably open a discussion on it once we are closer to implementing

[Chris Miller]

In case it wasn't clear, my comment was satirical. Having a universal rep system, or even a bioinformatics-wide rep system seems fraught with peril to me. Any implementation will be full of holes and assumptions, in the same way that various journal altmetrics are. I'm not saying it's intractable, just that it's hard.

Answer 1

[xb]

I think it is more of a question to determine the valuation of such properties within and between (sub-) communities. A nice topic, Obi!

[Obi Griffith]

Yes. I agree. This could be a key barrier to development of such a system. We already see it within one community. The question of how to value different kinds of contribution (question, answer, comment, news item vs detailed response). But, as you point out, this will be even more challenging when trying to combine across communities.

Answer 2

[Damian Kao]

A sensible way of measuring rep is probably what's stopping from these centralized communities from taking off. People are going to disagree on what is considered significant or not contributions. I think ResearchGate is attempting to do something like this with their RG points.

Answer 3

[Istvan Albert]

This is something that I often think about - I have come to believe that there is a fundamental flaw in the classic reputation system of the likes that the current Q&A sites (Biostar included) have. Better known people get a lot of points even for "fluff" and really valuable content garners less. Upvotes are often used to express agreement with a sentiment rather than necessarily value. So the reputation obtained via direct voting is biased.

I would like to see a more internal, intrinsic value - for example having cited posts or often viewed posts may be a better reflection of the value of one's contributions. Or give moderators the ability to rate posts on the content (for example rate it as high-information-content).

I will try to play around with these concepts in upcoming releases.

[xb]

I agree! Looking forward to such upgrade. Thanks, István!

[Biomonika (Noolean)]

What about a few hours delay before displaying name and photo next to the answers and display just reputation meanwhile? Then you wouldn't be tempted to vote for fluff from your favorite user :-) On the other hand, knowing who submitted what is extremely useful, so this approach assumes that people actually come back to re-read question and check for new answers.

[skymningen]

You would still recognise the well known, continuous posters from their writing style and the kind of topic they replied to. But I like the idea to at least add measures like Istvan suggested to the already available rep gaining (and loosing). Which is the main advantage of reputation over starbucks goodies: You can loose it. Which makes it a measure of trust, a biased one, but at least a hint.