| Features like complexity, dynamic and weak connectivity as well as communication latency and limited bandwidth make the efficiency of data synchronization hard to be improved and high-efficient consistency of mobile data synchronization scheme based on the traditional centralized architecture difficult to be realized in mobile computing environment. To solve these problems, this thesis proposes the hybrid data synchronization mechanism(HDSM) for mobile computing, which makes full use of the advantages of centralized architecture and Ad hoc architecture, and builds a series of self-organization domains(SOD) which serve as management units of synchronization process so as to cut down the synchronization data traffic and lighten the synchronization server load through realizing localization of the data communication. In order to shorten the synchronization response time and speed up the data distribution in SOD, this thesis put forward the capacity-value-based data distribution strategy(CDDS) which establishes data distribution paths of SOD tree based on the comprehensive processing capacity of mobile nodes to implement the high-efficient data transmission for mobile nodes in SOD. This thesis also proposes the increment capture strategy based on track changes(ICSTC) which captures operation logs with triggers and obtains pure increment data by purification to cut down the traffic of synchronization data, reduce the cost of transmission and shorten the time of synchronization.To ensure the quality of the data users collected and provide to the users or decision-makers the most accurate and efficient data, the issue of finding truth has been paid extensive attention. Because of the issues of inaccuracy and complication of truth finder algorithm, this thesis proposes a truth finder algorithm based on entity attribute(TFAEA) which is on the basis of iterative computing system with source reliability and fact accuracy, considering the interactive degree among facts and dependent degree among sources and simplifying the truth finder algorithm. In order to improve the accuracy of the truth finder algorithm, this thesis makes combination of one-way text similarity degree and conflict degree among facts to calculate the interactive degree among facts. At the same time, this thesis also put forward source saturation to calculate the dependent degree among sources. At last, on the basis of HDSM and TFAEA, this thesis designs and constructs an e-commerce shopping system and elaborates the function of the system architecture as well as the implementation of each module. |
PDF Full Text Request