<![CDATA[Most of recommender systems are based on content can be helpful to find recommendation book suitable for reader but it only consider about a liked book by user without considered about the disliked one. For solving the problem, a recommendation based on content of liked and dislike book by user is done. In this research, we applied Binary Particle Swarm Optimization(BPSO) to select feature from book that the reader like and K-Nearest Neighbor(KNN) are use for classify book data which had the closest distance to the book that the reader liked and disliked. Testing the accuracy of the recommendations is done by comparing the results of recommendations in the book data that do not apply feature selection with book data that applies selective features to test the effect of application of feature selection on book recommendations. The result of book recommendation accuracy testing with feature selection give a better recommendation for user than recommendation without feature selection.]]>
Copyrights © 2019
