Cloud Harmony Search Algorithm For Optimization Of Composite Knowledge As A Service

With the continuous progress of parallel computing、distributed computing、 virtualization, cloud computing is gradually developed as a new business calculation model based on Internet. And it gets wide attention from business community and academia. Knowledge as a service (KaaS) is an integration of knowledge and service, it provides a new direction for knowledge management and knowledge innovation. Facing the continuous expansion of knowledge information and complex processing tasks, the demand of users for KaaS constantly tend to be more diversified and complicated. So that KaaS with relatively single function have been unable to meet the requirements of users. With the result that, it has become an inevitable trend to composition multiple KaaS together into a powerful and efficient service.The development of cloud computing offers a novel idea for the progress of intelligent optimization algorithm. Using the powerful computing and storage capacity of cloud computing system and paralleling the intelligent optimization algorithm, it can quickly and efficiently solve the problem of mass data processing. Harmony search algorithm is a heuristic global search algorithm and it has a simple structure、easy to implement、fast convergence and good robustness, etc. In recent years, harmony search algorithm has successful application in many composite optimization problems. But most of them are used in serial mode, which has certain limitation on the solving efficiency.In this dissertation, the key technology of cloud computing is used to parallel harmony search algorithm and brings forward cloud harmony search algorithm (CHS). For solving the problem of composite KaaS based on the quality of service (QoS), it is realized that the parallel solving about the relatively time-consuming part of harmony search algorithm in MapReduce. What’s more, an improved cloud harmony search algorithm (ICHS) is presented in this dissertation. Using the method of Skyline to initialize harmony memory, it could improve the algorithm efficiency. And using the method of TOPSIS to choose the final knowledge as a service, it could better solve this issue. The experimental results illustrate that the ICHS has high efficiency on solving composite KaaS.