Adaptive Model for Web Service Recommendation
I. Sultan, Torkey
E. Khedr, Ayman
Kamal Alsheref, Fahad
MetadataShow full item record
The Competition between different Web Service Providers to enhance their services and to increase the users' usage of their provided services raises the idea of our research. Our research is focusing on increasing the number of services that User or Developer will use. We proposed a web service recommendation model by applying the data mining techniques like Apriori algorithm to suggest another web service beside the one he got from the discovery process based on the user’s History. For implementing our model we used a curated source for web services and users which also contains a complete information about users and their web services usage. We found a BioCatalogue: A Curated Web Service Registry for the Life Science Community, and we tested our proposed model on it and 70 % of users chose services from services that recommended by our model besides the discovered ones by BioCatalogue.
Web service are one of the important inventions in our technology world, because it offered several properties such as Interoperability, Usability, Reusability and Deployability. Also it can be integrated into applications over networks through a structured programming interface. Software applications written in various programming languages and running on various platforms can use web services to exchange data over the Internet. Web services have become as a commodity in market while the users had become as a customer who search for a suitable service for his work. Thinking of web service suggestion as commodity and customers raise the idea of using idea of Recommender systems that used by E-commerce sites. These systems are used to suggest products to their customers. The Products recommendation process can be done through analysing several shared properties between customers like nationality, site, demographics, customer’s behaviour and buying history. Through analysing these properties the customer future buying behaviour can be predicted, and by the same way when user wants to search for a web service he writes his query then the service discovery agent find the suitable service for him, our model works after the discovery model finishing his process by finding the required web service and the user accepts this service then our model works as a service recommender based on the analysis of users history.  For example when user searches for a service of text mining then after finding it our proposed model analyses all records of users that used this service and gets list of other services that they used. Then the recommender model applies data mining methods like Association rule that using Apriori algorithm to find recommended services, then the user will decide to choose any of the recommended services. Applying this approach leads to expand the usage of web services that provided by the web services provider, and it gives the users more information about the most used services that used by other users who used the same service. The rest of this paper is organized as follows. Section 2 provides the background and motivation. Section 3 provides the related works that used in our research. Section 4 defines our proposed model and the modifications that had been done on the original one. Section 5 studies our methods in a case study. Finally, Section 6 contains the conclusion and future work.
- Articles