QuARAM Recommender: Case-Based Reasoning for IaaS Service Selection

Abstract

Cloud computing enables elastic resource provisioning on demand and removes the boundaries of resources' physical locations. The number of cloud-based services is on the rise due to the growing interest from both providers and consumers. These services are characterized by a large number of features or properties, which makes the automatic service selection and deployment challenging. This paper proposes QuRAMRecommender, a cloud infrastructure service recommender framework based on case-based reasoning (CBR) that supports effective service selection. QuRAMRecommender supports decision making that accommodates the customer's preferences and feedback. We show the feasibility of our approach through a prototype implementation that elaborates on the main features of our system. The experimental results suggest that case-based reasoning is a viable option for recommending cloud services that best fit the customer's requirements.

A hybrid approach to automatic IaaS service selection

“…and knowledge about the transition from one service to another. The implementation and evaluation of the Case-Based recommender is presented in detail in Soltani et al [2,6]. Our experimental results show up to 90% precision of recommended services using case-based recommender.…”

“…In order to address the second challenge in providing automatic service selection (automatic evaluation, selection and integration of services) we primarily use case-based reasoning for evaluation and selection of services that can accommodate the application [2]. We incorporate the application's requirements, and the customer's priorities in our decision making using a set of similarity measures and weights in similarity calculation.…”

“…The QuARAM Service Recommender platform is a comprehensive extension to our previous work [2,6]. In that previous work, we described a recommendation service for an application using the case-based reasoning approach.…”

“…In this paper, we describe in detail the models and techniques used in QuARAM Service Recommender to address the aforementioned challenges in recommending suitable services for cloud application deployments and extend our previous work on a platform for service selection [6]. The platform is a part of the QuARAM (QoS-aware cloud application management) framework [7] (Fig.…”

“…It is anticipated that the fluidity of, and the competition within, the cloud computing market will grow as the technology matures. This will encourage providers to adopt a wider range of mechanisms, such as discounts and incentives, to attract potential consumers [2].…”