As electronic commerce has penetrated into the publication business, personalized content recommendation has drawn much attention in recent years for automated informational service. In prior literature, several studies have used the concepts of content-based filtering or collaborative filtering to recommend books, articles, and news, among other media. However, either of these methods has limitations, and different content domains may have various needs in making recommendations. To address this gap, we design a hybrid system and apply it to the recommendation of research articles. Our method has the merits of both content-based and collaborative filtering. More importantly, such a hybrid solution is effective in addressing the problems of handling new users and new papers. These problems cannot be solved easily by conventional recommendation approaches, such as K-Nearest Neighbors and (KNN) and Frequent-Pattern Tree (FP-tree). The performance of our proposed system was evaluated in an experiment on published JECR papers to show superiority over benchmarks. Overall, this study makes contributions to information systems (IS) and electronic commerce literature and practice, and suggests that a hybrid solution as presented by our proposed system could better serve readers of academic journal to enhance service quality and user satisfaction.