Adaptive Quality Recommendation Mechanism for Software Service Provisioning

In this paper, we propose an adaptive quality recommendation mechanism to help software service providers understand the dynamism of quality demand from the majority of requesters accurately. The unique feature of our approach is that based on the intra-cluster proximity index (called the icp-index) that we propose, the granularity of service clustering can be adjusted dynamically to meet the wide variation of service providers who want to target their services to different groups of clients. Experiments show that our approach is more accurate and flexible to identify the need of service quality of requesters than existing solutions such as simple averaging, minimum-maximum-mean, or traditional interval-range data clustering.