There are so many Deep Webs in Internet, which contains a large amount of valuable data, This paper proposes a Deep Web data extraction and service system based on the principle of cloud technology. We adopt a kind of multi-node parallel computing system structure and design a task scheduling algorithm in the data extraction process, in above foundation, balance the task load of among nodes to accomplish data extraction rapidly; The experimental results show that cloud parallel computing and dispersed network resources are used to extract data in Deep Web system is valid and improves the data extraction efficiency of Deep Web and service quality.
Research on the Computer Graphics Rendering Technology based on GPU and Parallel Computing System