International Journal of Science and Research (IJSR) ISSN (Online): 2319-7064 Index Copernicus Value (2015): 78.96 | Impact Factor (2015): 6.391 Volume 6 Issue 9, September 2017 www.ijsr.net Licensed Under Creative Commons Attribution CC BY Web Service Recommendation System Incorporating Location and Personalized QoS Prediction Nitin V. Tawar 1 , Supriya A. Kinariwala 2 1 Dept. of Computer Science and Engineering, Marathwada Institute of Technology, Aurangabad, Maharashtra, India 2 Dept. of Computer Science and Engineering, Marathwada Institute of Technology, Aurangabad, Maharashtra, India Abstract: With the invent of new technologies and advancements in business, more and more organizations are working towards building automated systems which work over the internet throughout the world for communication purpose and the availability of huge number of web services for different requirements makes the task of business organizations. Due to presence of a wide variety of web services available for the user, it is a difficult task to choose a particular service satisfying the requirements. In this paper we have studied various previous works under recommending user with services which best suits the requirements so that he can select and use the best service available and we then propose a recommendation system, which presents the user with a list of recommended services as per his needs which are near to him physically as well as which satisfy security requirements. Keywords: Collaborative filtering, QoS Prediction, Personalized influence, Recommendation, Similarity Computation 1. Introduction Web services are interoperable components built for internet communication. This communication takes place using XML messaging. Web services are self-contained components which are made for application-to-application interaction over the internet using XML messaging. The web service architecture is composed of SOAP, WSDL and UDDI. Due to wide variety of web services available over the internet and with the increase in the world wide acceptance of the web services, business organizations prefer to use existing web services so as to focus on the business process rather than working on developing a web service which fits into their requirements. If user chooses to build a service, it consumes resources and incurs cost on user as well as provider side. Selecting a high quality web service among the available web services for a particular task is again a tedious job because user has to select a service as per his needs[1]. The user can not test for each and every web service by invoking and recording whether it is performing as needed or not, because invoking such a huge number of services is not possible. Even if some user tries to evaluate the web service, it may happen that the best services are not available at that time or network condition may affect web service response. It brings the necessity of a system which provides you with the recommendations of web service requested based upon parameters like response time or some other mentioned by the user. To increase the usability of a web service among the available set of web services, it is described by Quality of Service (QoS) parameters which are non functional characteristics of web services. QoS plays a vital role in the selection of high quality web services. These non-functional parameters include response time, throughput, usability, availability, reliability, etc[2]. It is a very difficult process to acquire all the QoS values of all the candidate web services due to network conditions at that instant and unavailability of that service at that time. Along with these issues services may give different QoS for different users. It necessitates the need of location parameter of the user to be considered in recommendation. If a user and web service lie in same location or region they may provide better QoS as compared to the ones lying far away. Web service recommendation system is one in which high quality web services are selected on the basis of QoS parameters and location and are presented to the user as per his or her requirements. Recommendation systems have become very popular in recent years and are utilized in a variety of areas. Recommendation system typically includes two kinds of results- through collaborative filtering and content based filtering. Collaborative filtering approaches make recommendations based on the past experiences of QoS parameters of the other users for a particular item requested. It finds similar items based on the similar QoS parameters. First missing values are predicted based on QoS similarity and user similarity as shown in Figure 1. Some of the web services are available for use only in a particular area. It is irrelevant to use a WS in India that is operable in USA. Existing approaches consider users‟ past experiences i.e. QoS parameters on web services for web service recommendation. Such methods might provide poor recommendation due to not taking in consideration the parameters like users‟ locations and hence the accuracy of the recommendation could be very low. To enhance the prediction accuracy and improve the quality of recommendations we need to take locations into consideration and implement location based and QoS based recommendation system [3]. Paper ID: ART20176986 1610