Recommender systems work by having accurate information about users designed forto offer the products to the customer., especially on a commercial website. The collaborative filtering method is one of the popular recommender systemssystem approaches that trytries to identify similar users or items, based on their previous transactions on them. The low accuracy of the suggestions is one of the major worries inconcerns about the collaborative filtering method, that becausedue to the sparseness of a rating matrix of users-items. To solution ofsolve this problem, many methods have been introduced to enhance the accuracy by discovering association rules byand using evolutionary algorithms such as particle swarm optimization. However, their runtime performance does not fulfill the needneeds of thea fast recommender system. In this article, an efficient method of producing cred association rules with higher performanceperformances based on a genetic algorithm is proposed. Experiments were performed for the data set of MovieLens. The experimental evaluation of a system based on our algorithm outperforms the performance of the multi-objective particle swarm optimization association rule mining algorithm by about 10 %.

The text above was approved for publishing by the original author.

Previous       Next

無料の試し

メッセージをご記入ください
修正依頼の言語をご選択ください

チェックアウト私たちの iPhone/iPad 校正 アプリ!

eAngel.me

eAngel.me is a human proofreading service that enables you to correct your texts by live professionals in minutes.