Reputation V2
A blockchain based ratings system which closely resembles a real life ratings model with a more balanced power model between a merchant/service-provider and the consumer.
About
A blockchain based reputation system allows immense amount of analytics. For this, we need consistent and regular rating behaviour by users. Pain points of the existing system included 1. Platform manipulation - Merchants can buy reputation score or remove comments/ratings hurting business without improving. 2. Stale rating - Once a large reputation score is achieved, bad performance doesn’t reduce it due to an averaging. By creating a mechanism of incentivising users to provide ratings every time and for every experience, it attempts to bring promote consistent ratings and more real reputations based on activity. * It solves problems like Platform manipulation and stale reputation scores. * It can be applied to various categories - Merchant platforms, B2B, B2C * It introduces a decaying reputation and tokens and rewards activity and participation. * DAO controlled reputation calculation mechanisms(replaceable calculations, updatable thresholds) REP token - The users are rewarded a REP token for every review. - REP token can be burnt to boost reputation score(within limits). This helps upcoming businesses improve Reputation score. - Customers can redeem their tokens for discounts by transferring REP to businesses. - REP can’t be traded across accounts and hence can’t just be bought. - REP tokens and Reputation score both have a decay rule built into them. Being active is the only way to keep it up. [nb: I proposed this idea in a previous hackathon(no code implementation). However, this time I have added more detail and some code implementation.]
How it's made
- The mechanisms that require calculation(reputation and reconciliation) might require a lot of on-chain updates. Still focused on this on and evaluating rollup options to minimize transactions. - There are user safety aspects as the rating can indicate user location and identity. Evaluating a ZKP based approach with randomization and batching.