A New Recommender System based on Cooperative Co-evolution Algorithm

Expansion of global networks and storing theextensive amount of information in various websites,create a serious needs for filtering the irrelevantinformation in a personalized way. collaborativefiltering or recommender system is a filtering techniquethat allows incorporation of the profiles which can beimplicitly learned from previous activities [1]. we haveproposed the coco-cf1 as an effective method suitablefor collaborating filtering running in a jini-gridcomputing2 platform and operational in a distributedenvironment. the coco-cf generates representativerecords from stored preferences and seeks for theanswer with the best fitness in the recommender system.we have considered the user satisfaction rate, feasibilityof available results, user familiarity and averageresponse time as the evaluation factors. also we havefocused on mean absolute deviation, mean square errorand ranked evaluation as the performance evaluationparameters. the obtained results confirm that thecoco-cf is a successful method for collaborativefiltering.