Usage-based Object Similarity

摘要: Recommender systems are widely used online to support users in finding relevant information. They can be based on different techniques such as content-based and collaborative filtering. In this paper, we introduce a new way of similarity cal- culation for item-based Thereby focus the usage an object not object's claim hypothesis that indicates content similarity. To prove use learning objects accessible through MACE portal where students query several architectural repositories. For these objects, generate profiles their monitored within MACE. We further propose recommendation apply usage- calculation real systems.