Quote from: tm2013 on April 22, 2016, 02:34:31 AM

Quote from: Melancholy on April 22, 2016, 02:10:21 AM

Quote from: JJ12880 on April 22, 2016, 01:32:41 AM

Friends,

I am searching adding a user trust system to the SmartChain, and was interested if anyone had any suggestions. There are many variables that can be used in such a system, such as age of the account, their verified status, and a trust rating provided by other users.

Assuming that a system is in place for one user to provide a rating of another user, what type of system would be easiest to use? A finite levels system is one option, (level one, level 3, level 3) etc, or a "points" system where a user accrues points based on ratings provided by others.

Another thought I had was that a users rating of another user could in turn be based on trust of the person doing the rating, such that old well known and trusted persons have greater weight.

Let me know what your thoughts are.

JJ

This is a cool idea! One option would be to basically have a +/- rating system where you can either trust a user (+) or avoid a user (-). It would essentially be a ratio or something like:
lamptoast (+53/-4) <-- ratio showing both rating totals
or
lamptoast (+49) <-- net rating (positives minus negatives)
or
lamptoast (93%) <-- ratio as a percent

We could also include the other factors you mentioned such as account age or even account activity (high, low, etc)
(just realized this is a lot like bitcointalk's system, heh)

IMO a "level" system might be gimmicky, depending on how its implemented.

I'd love to hear what others have to say!

One thing that must be considered is anti-fraud measures. It would not be good if SmartChain trust could be rigged and used for malicious purposes, as has been the case with things like Yelp and the Bitcointalk trust system. A cost in Radon could help, but it must not be so high that it excludes users from sharing their trust experience.

It would need to balance out, such that the following scenario could take place:
Alice gives a false negative trust rating to Bob.
In response, Alice receives a certain level of negative trust for this action from other users.
The weight of Alice's trust rating would go down to a certain level, neutralizing the effect of her false negative trust rating on Bob, and any other users she may have attacked.

Other thing that may take into consideration is the use of bots to gain trust, like someone receive positive trust from new participants and those same participants give trust between them.