Recommendation System for E-commerce Using Alternating Least Squares (ALS) on Apache Spark

Recommendation system can predict the ratings of users to items by leveraging machine learning algorithms. The use recommendation systems is common in e-commerce websites now-a-days. Since enormous amounts data including users’ click streams, purchase history, demographics, social networking comments and user-item are stored databases, volume getting bigger at high speed, sparse. However, recommendations predictions must be made real time, enabling bring benefits human beings. Apache spark well suited for applications which require speed query data, transformation analytics results. Therefore, developed this research implemented on Spark. Also, matrix factorization using Alternating Least Squares (ALS) algorithm a type collaborative filtering used solve overfitting issues sparse increases prediction accuracy. problem arises as rating In alternating least squares method Spark MLlib developed. shows that RMSE value significantly reduced ALS 0.870. Consequently, it shown suitable training explicit feedback set where provide items.