In the present field of E-business (Electronic business) environment, personalized web service recommender systems are used most commonly to produce recommendations in a large amount of information’s and suggestions on database. Most widely used technique for recommender system is the Collaborative filtering (CF). Whenever the rating database acquires a change of increase in the number of users and items then there is rapid increase in space and calculation complexities that results a shortage of scalability. This Scalability shortage is the most significant problem in the existing collaborative filtering based recommendation systems. Hence, for solving these scalability and data sparsity problems in this, item-memory based CF is proposed in this paper. The proposed CF provides more personalized web service recommendations to users with the help of item clustering prediction. The problem of Data sparsity in the proposed collaborative filtering can be handled by the Case Based Reasoning (CBR) together with the average filling and consequently scope to the item-memory based CF is reduced by performing item-memory clustering using the Genetic Algorithm (GA) in large datasets. Analysis on the proposed CF depicts an increased accuracy compared to traditional CF techniques and the experimental outcome of proposed method gives a better precision, recall and FI-scores.
Volume 12 | Issue 7
Pages: 714-720
DOI: 10.5373/JARDCS/V12I7/20202054