원문정보
초록
영어
The theory and technology of cloud computing have been widely adopted by many small and medium enterprises and individuals for their business systems or personal affairs. To facilitate the evaluation processes of cloud service creditability, the present paper proposed a new multi-attribute evaluation theory. In addition, a utility and collaborative filtering-based evaluation method was presented, which targeted actual problems existing in the processes, such as missing data or inconsistency of data dimensions. The new method utilizes Enhanced Lance and Williams Distance, which is based on Jaccard similarity coefficient, to measure the similarities between different cloud services; it also applies utility theory to data unification and integration. In the later part of this paper, simulation experiments were conducted to test the validity and rationality of the proposed method.
목차
1. Introduction
2. Processes of Cloud Service Evaluation
3. Collaborative Filtering-based Data Prediction
3.1. Calculating Similarity of Various Cloud Services
3.2. Choosing Similar Cloud Services
3.3. Missing Data Prediction
4. Utility-based Data Integration
4.1. Data Unification
4.2. Data Integration
5. Case Study
5.1. Results
5.2. Discussions
6. Conclusion
Acknowledgements
References