An attribute-based scheme for service recommendation using association rules and ant colony algorithm

With the rapid development of m-commerce, predicting user's navigation and making the service recommendation become more and more important. Most researches focus on predicting user's navigation using context history and user preferences. But, the influence of the attributes of a service has been ignored. Simultaneously, some attributes are variable, so the recommendations are changeable. Therefore, the paper proposes an attribute-based scheme for service recommendation based on CASUP (context-aware system considering user preference). In proposed approach, the services are classified into several service clusters, and the service recommendations are carried out using Apriori algorithm and ant colony algorithm. Finally, the proposed model is validated by several simulation experiments, which demonstrate the effects of the service attributes in m-commerce.