This article inspired me so much that I based my shitty undergrad senior thesis on it. My idea was to predict the trend of the ratings by using I think a traili

This article inspired me so much that I based my shitty undergrad senior thesis on it. My idea was to predict the trend of the ratings by using I think a trailing weighted average, weighted to the most recent window. It managed to generate more "predictive" ratings of the following 6 months based on the Amazon dataset I used, but I doubt it would have held up to much scrutiny. I learned a ton though! Edit: Link to paper, which looks like it actually attempts to use a linear prediction algorithm. https://github.com/rkuykendall/rkuykendall.com/blob/e65147f6ddc1185e04f60082a6089c16581c0123/assets/ranking-linear-prediction/Kuykendall-Ranking_Linear_Prediction.pdf https://github.com/rkuykendall/rkuykendall.com/blob/e65147f6...