Apply hybrid recommender system using genetic algorithms and singular value decomposition (SVD)‎

Abstract EN

Due to the increasing development in using modern technology for different purposes, recommendation systems have become essential in providing solutions to all users.

Recommendation systems can be used for various applications or on the internet to purchase products or to surf the web for scholarly articles, to watch videos and other activities.

Many statistical and mathematical suggestions have been made to make different systems capable of giving accurate recommendations that can be of benefit to users in various fields.

One of these suggestions is Singular Value Decomposition (SVD) which is the most popular method in previous research.

In addition to that, (SVD) has two important factors in its performance, which are, (? , ?) and choosing these two factors plays an important role in the accuracy of recommendation systems.

Choosing these two factors was a heavily discussed topic in previous academic research.

In this paper, the use of artificial intelligence method Genetic Algorithm (GA) to design a hybrid system with the Singular Value Decomposition (SVD) to specify the best values of (? , ?), and hence acquiring an accurate recommendation system with better speed.

After several experiments, the results indicated that the hybrid system better system according to time.

The hybrid system (GA) is also better than the system uses (SVD) in accuracy and time.