For the efficient shape analysis of massive, heterogeneous and complex 3D models, an optimization method for 3D model shape based on optimal minimum spanning tree was proposed. Firstly, a model description based on 3D model Minimum Spanning Tree (3D-MST) was constructed. Secondly, local optimization was realized by topology and geometry detection and combination of bilateral filtering and entropy weight distribution, obtaining optimized MST representation of the model. Finally, the shape analysis and similarity detection of the model were realized by optimized Laplacian spectral characteristics and Thin Plate Spline (TPS). The experimental results show that the proposed method not only effectively preserves shape features of the model, but also effectively realizes sparse optimization representation of the complex model, improving the efficiency and robustness of geometric processing and shape retrieval.

In order to apply the Internet of Things (IoT) gateway to various scenarios, a light-weight rules engine was proposed, through which users can define personalized rules on demand. However, the limited resource for computing and storage prevents the traditional rules engine, such as JRules, being applied on the IoT gateway directly. By the "related facts" attribute added to each rule and the mechanism of "Agent-Inference", both the running time and the response time of the rules engine were reduced. Adding the "related facts" attribute to each rule can reduce the number of the rules involved in matching operations, and the mechanism of "Agent-Inference" can reduce the waiting time for available rules. Based on these methods, a Faster Light-weight Rules Engine (FLRE) was implemented and applied to IoT gateways. The experiments on different-size data sets showed that the running efficiency was increased by 8%-30% with adding "related facts" attribute, and the response time was decreased by 7%-35% with using mechanism of "Agent-Inference". The evaluation shows the two methods are effective to apply the light-weight rules engine to the IoT gateway.

In order to achieve complementary advantages and information sharing of heterogeneous workflow systems among enterprises, concerning the workflow showing heterogeneous distribution and other characteristics, an interface design of heterogeneous workflow processes interconnection based on Web service was proposed. For the interconnection of heterogeneous processes, the solution of heterogeneous workflow interconnection was described from call interface，call mode and call return respectively. Taking example of SynchroFlow workflow process described by XPDL and ODE (Open Dynamic Engine) workflow process described by BPEL, the process calls between the workflows was achieved.

Content Delivery Network (CDN) adopts data replication or data caching methods to accelerate dynamic Web applications. A userbased data partition method especially for Websites that provide personal information publishing was proposed, such as forum, Web log service providers: partitions data into many parts according to users it belongs to, distributes to database systems near users. This method achieves better scalability because the write database queries are assigned to different databases.