Towards Beer Recommendaons for MovieLens Funing Xu Mentor: Max Harper Department of Computer Science and Engineering, University of Introducon Evaluaon Results Conclusion MovieLens is a website that helps users find beer We evaluated around 4000 users with at least 40 rangs who logged in In this study, results showed when we blend movies built by Grouplens Research Lab. It is mainly within six months. Under different blending strategies both in age a lile popularity and age, the number of using Item-Item collaborave filtering algorithm to condion and popularity condion, the top 20 movies are evaluated on movies in wish list will increase and RMSE will generate predicon score for movies. Currently, the selected metrics. decrease. The NDCG metric showed that orders of recommendaons are only based on movies’ blending of popularity and age only have a predicon score. The wish list metric measures how many movies in user’s top 20 minor negave impact on recommendaon recommendaons are also in this user’s wish list. The Root mean square qualies. In this study, we tried to blend movie age and movie deviaon (RMSE) metric is a typical metric which popularity to the movie predicon score with different measures the accuracy of recommendaons. The Normalized discounted The results indicates when we blend strategies. Furthermore, we evaluated different cumulave gain (NDCG) measures the quality of recommendaons popularity and age to original predicon blending strategies with metrics in the filed of varying from 0.0 to 1.0, with 1.0 represenng the idea value. The score, the recommendaons become less recommender system and informaon retrieval. following charts show the results of our study. personalized. Meanwhile, more recommended movies appear on user’s wish list, which shows that recognion of recommendaons is promoted.

Selected Reference

1. McNee, Sean M., John Riedl, and Joseph A. Konstan. "Being accurate is not enough: how accuracy metrics have hurt Figure 1: wish list metric with popularity Figure 2: wish list metric with age recommender systems." CHI'06 extended abstracts on Human factors in compung systems. ACM, 2006. Methods 2. Amatriain, Xavier, and Jusn Basilico. "Nelix recommendaons: beyond the 5 Two aributes of movies are chosen as blending factor, stars." Nelix Tech Blog 6 (2012). which are last year popularity and movie release date.

Popularity of a movie is the number of rangs it received.

Since the value of predicon score, popularity and movie Figure 3: RMSE metric with popularity Figure 4: RMSE metric with age release date has different scale, we first convert these value according to percenle rank. Therefore, all of these values are under the same scale of zero to one. Acknowledgement

In the end, we have two groups of new score which This work was founded by the University of decide the order of recommendaons as follows. Minnesota Undergraduate Research Opportunity Program(UROP). new!score1! = !A! × !prediction!percentile! + !B! × !popularity!percentile new!score2! = !A! × !prediction!percentile! + !B! × !movie!release!date!percentile A! + !B! = !1 Figure 5: NDCG metric with popularity Figure 6: NDCG metric with age