The network and parallel computing domains play a key role in shaping future research and development activities in academia and industry. We selected three papers with high quality from NPC 2007 after no less than 30% revision. All submissions were reviewed by at least three reviewers on relevance and technical contents of the papers. We believe all of these papers and topics will not only provide novel ideas, new results, work in progress and state-of-the-art techniques in this field, but also stimulate future research activities in the area of network and parallel computing.
In the first paper, Cao et al. propose a new web segmentation method – Iterated Shrinking and Dividing on the basis of image processing technology and characteristics of web pages. First, the web page is saved as image, which is pre-processed by an edge detection algorithm such as Canny. Then dividing zones are detected and the web image is segmented repeatedly until all blocks are indivisible. This method can be used to analyse the web pages such as detecting similar visual layout. Experiments show that the algorithm is suitable for web page segmentation and does well in expansibility and performance.
In the second paper, Chen et al. integrate a field programmable gate array (FPGA) and an application specific integrated circuit (ASIC) hardware module into an encrypt-on-disk object store system to avoid re-encryption in the cryptographic storage system. In the system, almost all computationally expensive cryptographic operations are through a FPGA/ASIC hardware module. Once creator revokes some users, objects do not need re-encryption. Finally, an encrypt-on-disk object store prototype system is implemented using FPGA in the software solution with tested and effective performance.
In the last paper, Liu and Hu present a high performance storage system for RDF documents. The system employs optimised index structures for RDF data and efficient RDF query evaluation. The index scheme consists of three types of indices. Triple index manages basic RDF triples by dividing original RDF graphs into several sub-graphs. Path index manages frequent RDF path patterns for long path query performance enhancement. Context index is optional for context oriented RDF data and temporal RDF data. This paper describes the organisation of index structures, shows the process of evaluating queries based on the index structures, and provides a performance comparison with existing RDF databases through several benchmark experiments.
Finally, we would like to thank Professor Ivan Stojmenovic, the editor-in-chief of International Journal of Parallel, Emergent and Distributed Systems and the authors for their excellent contributions and patience in assisting us. The fundamental work of all reviewers on these papers is also very warmly acknowledged.