Qos prediction for web services based on Restricted Boltzmann Machines
- PDF / 1,284,306 Bytes
- 21 Pages / 595 x 791 pts Page_size
- 70 Downloads / 227 Views
Qos Prediction for Web Services Based on Restricted Boltzmann Machines
Le Van Thinh
Received: 16 September 2017 / Accepted: 14 November 2017 / Published: 31 December 2017 © The Society of Service Science and Springer 2017
ABSTRACT Recently, reliability prediction of Web services has become very important in related research communities. Especially, predicting the Quality of Service (QoS) for active users has been a hot issue of research and application. On the other hand, with the rapidly growing in number of service providers and users, resulting in a large number of data sets. It has a big impact to the QoS such as managing and monitoring for describing functional and nonfunctional characteristics of Web services. Therefore, we will certainly struggle at processing large data sets in the future, unless the issues are resolved quickly before it happens. In that context, QoS prediction on big data set is an urgent problem to be solved. In this paper, we present a new model for handling this problem based on a Restricted Boltzmann Machines, it is called Two Layers Model (TLM) and also proposed to add a new method for evaluating prediction of the field of web service quality. We use this model to cope with large data sets and the model has been used by the efficient learning and inference procedures for predicting QoS values of Web services. Our experiments were performed on two data sets in the WSDREAM data set and the experimental results have proved the effectiveness of the proposed model. KEYWORDS Web service, Two Layers Model, Restricted Boltzmann Machines. Le Van Thinh ( ) Faculty of Information Technology, Tuy Hoa Industrial College, Phu Yen Province, Viet Nam e-mail: [email protected]
198
Le Van Thinh
1. INTRODUCTION Web services are a system which is designed applications to allow the computer to communicate with each other via network. Service providers were established based on the needs of users, with increasing demands of users and a highly profitable business environment with considerable potential for growth. As a result, many Service providers tend to move in the same direction as business and marketing plans, which provides the direction on the potential market environment, leading to multiple providers may offer functionally identical services. On the other hand, QoS also depends on many other factors such as transmission on the network, implementation mechanism and hardware. Therefore, the user desires to know how to choose a QoS from service providers such as latency, availability and reliability. However, almost service providers nowadays focus on the profit and growth. Moreover, in e-commerce and other online environments, the QoS properties of each service are pre-determined by the service providers, but because of its promotion purpose, they often exaggerate to attract more users. Hence, QoS is one of the important factors to build service-toward applications (Serhani et al. 2005; Hinton 2012; Issa et al. 2009). The QoS property may have different values to different users (e.g., resp
Data Loading...
