Predicted Model based on Boltzmann Restricted Machine for Web Services Recommendation

Abstract

With the growing number of Web services, it becomes more difficult for users to choose the best ones that meet their needs and get the best quality of service (QoS). Many times, a user will only invoke a small number of services, which leaves many QoS values for those services blank. Therefore, users can't select the best services based on their QoS values. This problem can be solved by proposing a predicted model for recommending appropriate services. This model uses the Restricted Boltzmann Machine (RBM) to predict which of the services with missing QoS values we can recommend to users. We evaluate our model using the WSDREAM dataset. Experimental results indicate that the proposed model is well performed and gets better results compared to other models.