Data, recommendation techniques, and view (DRV) model for online transaction

With the development of information technology, online transactions, including E-commerce, have been developed. Accordingly, recommendation systems were developed to facilitate customer preferences and increase business revenue. In this paper, our analysis shows that each of these systems was implemented to facilitate the recommendation process of a specific product or service category and applied to a dedicated context. The issue here is if the business provides more than one category of products and/or services it needs to utilize more than one approach to have an effective recommendation process. That would make it more complicated to implement and with a high cost. In addition, each of these systems was developed to overcome a specific problem. There is no guarantee that the system developed to address a dedicated problem could overcome the other problems. Examples of these problems include cold-start, data sparsity, accuracy, and diversity. In this paper, we develop Data, Recommendation Technique, and View (DRV) model. We consider this model to be a foundation for a generic framework to develop recommendation systems that overcome the issues mentioned.