The number of cloud service providers offering a wide range of services is increasing every day. This rate is exponentially proportional to the rate of increase of users and usage of smart devices as advent of technology has drastically improved when compared to previous decades. Preferences of users are never unique and cannot be constrained within a certain definition or documentations. This preference may be defined with a number of attributes which are categorized into functional and non-functional requirements of individual users. This also challenges the architecture as the volume of data is immeasurable and dynamic. The proposed system should be able to identify the functional and non functional requirements from the unstructured and structured databases. Since user preferences may vary to a large extreme in terms of quality and quantity, initial selection plays an important role in recommending the web services. These preferences may not be clear enough in some cases and handling them needs special attention to narrow down to appropriate decisions. Feedbacks are collected in all similar recommendation systems in this domain, which we have transformed to have a trust based strategy to form an association between previous user preferences. Interests that are said to conflict among one another has to be addressed to improve the efficiency of overall system. The system should be able to balance the recommendations based on trust parameter and likability of new users. These issues have to be answered when such a system is designed and implemented to return highly desirable response to users of cloud services. Existing systems deals with a tackling mechanism of one or two conditions stated but this proposal intends to deliver a procedural set of actions to integrate and respond to all requirements. The demanded response is estimated carefully by determining trust values and branching through associative ranks to display the list of web services to opt from. The efficiency of this strategy is evaluated through continuous experiments and investigations.
Volume 11 | 02-Special Issue
Pages: 271-279