Web-based Movie Recommender System Using User Preference Models With Adjusted Euclidean Distance Similarity Measure

In our day to day life we come across questions like “which movie do we have to watch?” “which animated movies are best to watch?” etc.Recommendation is an act of saying something is good and deserve to be chosen.In this work,collaborative filtering technique is used based on user preference models in order to construct a system concerning movies that provides more precise recommendations and predictions with reduced computational complexity.In order to improve the performance of CF recommender system,we propose similarity measures and integrate the system into a web-based recommender application.One of the user preference models is focused on percentages of different movie genres a user has watched and the other is on the average ratings a user has given to different genres of movies.These models less depend on the rating data that users have given to the items,but more on item types to accurately predict user preference and has much less computing complexity for no need in finding the co-rated items.We use these user preference models with AED(Adjusted Euclidean Distance)similarity measure to calculate similarities among users.We evaluate our proposed similarity measure based on watched genre ratio preference model by experiments on Movie Lens 100 k and ml-latest-small datasets and compare the performance with Euclidean similarity measure.The results show that AED based on watched genre ratio preference model outperforms in all aspects than Euclidean.