Hybrid Personalized Recommender System Using Modified Fuzzy C-Means Clustering Algorithm

Recommender Systems apply machine learning and data mining techniques forfiltering unseen information and can predict whether a user would like a givenresource. This paper proposes a novel Modified Fuzzy C-means (MFCM) clustering algorithm which is used for Hybrid Personalized RecommenderSystem (MFCMHPRS). The proposed system works in two phases. In the firstphase, opinions from the users are collected in the form of user-item ratingmatrix. They are clustered offline using MFCM into predetermined numberclusters and stored in a database for future recommendation. In the secondphase, the recommendations are generated online for active users usingsimilarity measures by choosing the clusters with good quality rating. Wepropose coefficient parameter for similarity computation when weighting of theusers’ similarity. This helps to get further effectiveness and quality ofrecommendations for the active users. The experimental results using Irisdataset show that the proposed MFCM performs better than Fuzzy C-means (FCM) algorithm. The performance of MFCMHPRS is evaluated using Jesterdatabase available on website of California University, Berkeley and comparedwith fuzzy recommender system (FRS). The results obtained empiricallydemonstrate that the proposed MFCMHPRS performs superiorly