Crowdfunding Platform Recommendation Algorithm Based on Collaborative Filtering

Abstract

Crowdfunding platforms are a novel method of internet fundraising. The production of data from the crowdfunding platform has increased, but the benefits of this data have not. This has led to a state of "information overload." User-specific recommendation systems driven by data mining can play an effective role in solving this problem. This study introduces a collaborative filtering system that is driven by its end-users by fusing user input with the closest neighbor technique from machine learning. The results of this test demonstrate that the algorithm can provide useful project suggestions to those who utilize the crowdfunding site.