A Movie Recommender System Using Modified Cuckoo Search

Recommender systems (RS) are data sieving method that recommends on the basis of liking of users and category of products from big data. In the same way, videos and movie recommender systems give some technique which helps the users to classify movie according to user's similar interest. This mechanism process makes recommender systems very important tool for Web site and various digital marketing applications. This paper focuses on the movie recommender system by using data clustering and nature-inspired algorithm. K-means algorithm is widely used algorithm for clustering due to its simple nature, handling large amount of data and running time is low. But it falls into local optima due to its randomly generated initial centroids. This algorithm can achieve global optimum solution if it is integrated with nature-inspired algorithm. This paper integrates k-means with nature-inspired algorithms (bat, firefly, cuckoo, modified cuckoo search) on data set (movielens). Outcomes are compared on the basis of objective function, less the objective function better the results. Our proposed system (k-mean modified cuckoo search) gives improved outcome than other algorithms.