Hybrid Approach for a Knowledge Recommender Service: A Combination of Item-Based and Tag-Based Recommendation

An exponentially increasing of knowledge in a knowledge management system is the main cause of the knowledge overload problem. A development of knowledge recommender service embedded in the knowledge management system becomes a challenging task. This paper proposed a hybrid approach by combining an item-based recommendation technique, also known as a collaborative filtering technique, with a tag-based recommendation technique, also known as a content-based filtering technique. To evaluate the performance of the proposed hybrid approach, a group of knowledge management system users were invited to be participants in this research study. As a criterion, the participants were asked to use the prototype of a knowledge management system embedded with the knowledge recommender service for 6 months. This would guarantee that each participant’s interaction with knowledge items could be recorded. A confusion matrix was then used to compute an accuracy of the proposed hybrid approach. The result of the experiment revealed that the proposed hybrid approach outperformed the item-based approach and the tag-based approach. Hence, the proposed hybrid approach was a promising technique for a knowledge recommender service in the knowledge management system.