Web service recommendation based on event ontology

Abstract

Web service recommendation is an important means to help users quickly discover and select services. Service recommendation combined with situational information is one of the development trends of Web services in the future. This paper proposes a Web service recommendation method based on event ontology, which abstracts Web service requirements into service event classes, according to the historical service records of the target users, Web service recommendations combined with service event ontology and content similarity calculation are recommended, and event class elements and hierarchical models in service event ontology are used as reference, calculate the service similarity and generate the fusion recommendation result according to the weight; Then, based on the current situation status of the user, the user’s current situation is located to a node on the event ontology situation model, the predictive recommendation of services is based on the semantic relationship between event classes in the event scenario model. Finally, the service with the highest similarity among the two parts of service recommendation is recommended to users. The experimental results show that this method can better improve the accuracy and diversity of service recommendation.