Who Should Represent Me? Similarity-Based Recommender Systems for Vote Delegation in DAOs

Delegation is a key governance mechanism in many Decentralized Autonomous Organizations (DAOs) that allows users to transfer their voting power to other users in a revocable manner. The aim is to increase participation in governance decisions to counter increasing voter apathy among DAOs. However, current delegation dashboards do not assist in finding an appropriate delegate in terms of voting alignment. We introduce a similarity-based delegate recommender that identifies suitable delegates using historical voting behavior. Utilizing five years of off-chain voting data from five exemplary DAOs, we evaluate ten similarity metrics, measuring how well the recommended delegates reproduce the original voters’ decisions. Furthermore, our approach results in a boost in terms of participation up to a factor of 52.