We call our algorithm Peoplerank. It works similar to the original Pagerank.
Instead of ranking websites -- it ranks identities. Instead of tracking links -- it tracks attention.
It’s also a 2nd order metric. This means that it matters not only who pays attention to you, but also who pays attention to the people who pay attention to you. And so on.
We currently use only data from Twitter. However, we are already experimenting with incorporating data from other platforms as well as from from decentralized sources.
Improving accuracy of the results is our number one priority. The primary mean to do that is through inferring multiple streams of data.
We have already been doing this using data from Twitter. We are cross-referencing multiple metrics and this has led to massive improvements.
We believe that there are still massive improvements to be made when we start incorporating other sources of data.
Can anybody access this data?
Not yet, but our vision is to make all the data available in a permissionless manner.
In other words, anybody will be able to use it to build their own products, services or otherwise without asking us for permission.
Is this algorithm going to be decentralized?
Yes, we are committed to decentralizing this project down the line.
We cannot provide a specific timeline at this point.
Do you collect private data?
No. The algorithm is using only public data. We do not intend to collect nor use private data for this purpose.
Are you going to connect accounts from different platforms?
We intend to connect identities across various platforms, BUT it will be a self-sovereign system, where each identity is controlled by their owner.
We think that most people will voluntarily tie their identities from multiple platforms, because this will raise their status. Combining attention from multiple platform, should raise one’s influence. We also think there will be people who will not care about this and will prefer not to tie them together. And that’s fine.
Why reputation? Shouldn’t you be rather tracking technical ability?
There is a problem with the idea of tracking technical ability. The following thought experiment illustrates this well:
How would you track technical ability of a surgeon? We know that certificates and titles are not a good proxy. So perhaps we should derive a “performance” metric, e.g. the number of patients that die/survive on their table? One can assume that if such a metric would exist, surgeons would intentionally avoid taking on any difficult cases, because their incentive would be to pick as many “easy to solve” patients.
The problem is that for many cases “expertise” escapes quantification. We believe that in such cases, the best way to quantify expertise is through quantifying how a given expert is perceived by his or her peers.
How are the scores calculated?
Attention Score indicates one’s share in the group’s attention. It’s a 2nd order metric. This means that it matters not only how much attention someone pays you, but also how much attention this person receives themselves.
Your Attention Score is calculated in the following way:
[number of people who pay attention to you in a given cluster]
[average of their Attention Scores]
[% of their attention that is allocated to you]
Attention Scores are on a scale from 0 to 1000.