Abstract:Software distributed shared memory (DSM) system has constructed a virtual shared memory abstract on cluster, which combines the programmability of shared memory and fine scalability of cluster. So it is widely studied. Software DSM system is easy to fail because it is a distributed system, some kinds of fault tolerance are necessary for it to be more practical. A recoverable and portable software DSM system, JIACKPT (JIAjia with ChecKPoinTing), has been designed and implemented to tolerate the fault of system. JIACKPT, based on JIAJIA, has adopted the checkpointing technology. By maintaining the strict global consistent state and using some optimization techniques, JIACKPT has gotten high performance. The experimental results on an 8-node PC cluster show that the checkpoint overhead is less than 10% of the whole execution time when checkpoint is done once per minute. JIACKPT also has good portability and can run on several operating systems, such as Linux, Solaris, etc. JIACKPT is a practical recoverable software DSM system.

Abstract:The motivation of adopting the Generalized LR (GLR) parsing algorithm to construct parsers for programming languages is presented. A multi-level strategy for optimizing a GLR parser and the necessary runtime control mechanisms are introduced to the GLR parsers for flexible disambiguation and for invoking semantic actions attached to grammar rules while avoiding the “delayed semantic action” problem. The algorithm has been implemented in VPGE, a visual parser generation environment. Experimental results show that the speed of the generated GLR parsers is comparable to LALR(1) parsers generated by GNU’s Bison when parsing deterministic programming languages.

Abstract:Knowledge flow is the knowledge creation, distribution, and reuse among participants. In knowledge intensive organizations, business process control and knowledge assess management are closely related to each other. Workflow management is an important technology for business process control. Yet knowledge management mechanisms can not be represented by current workflow process definition meta models. An innovative extended workflow process definition meta model is first proposed for integrating the above two aspects. Based on that, modeling and control of knowledge flows are studied. A knowledge flow modeling approach is proposed by using five kinds of knowledge flow components to represent knowledge distribution and reuse, cooperation and communication among participants. To deal with dynamic elements in knowledge flows, an adaptive knowledge flow control approach with corresponding algorithms is proposed based on resource constraints, changes of knowledge requirements, and time constraints. This paper presents a beneficial approach for the effective integration of workflow and knowledge management technologies.

Abstract:Because of the rules of dynamic directive nesting and binding, some of the thread context in OpenMP programs can only be totally determined at runtime. However, by compiling time static analysis, nesting type can be partly determined and this information can be passed to other compiling phases to guide later translation and optimizations. Since the binding and nesting may span the procedure boundaries through calls, local and global analyses are not enough. It is the interprocedural analysis that provides the most required ability. By integrating information into traditional interprocedural analysis, the nesting type information of procedures is propagated along call graphs. And later translation and optimization phases can bind this global information with local information inside the procedure to determine the nesting types at compiling time. The results demonstrate that in typical science and engineering workload the nesting type is highly determinable at compiling time, and the application of this information may achieve less runtime overhead and the reduced code size.

Abstract:There usually are many attributes, called small-range attributes, with small number of different values in massive relations. The number of combination values of these attributes is also very few in massive relations so that there are a lot of repeated combination values of these attributes in massive relations. It is important to remove the repeated combination values to improve the efficiency of storing and querying massive relations. A compression method for removing the repeated combination values is proposed in this paper. To compress a massive relation, the method partitions the relation into two small relations: one consists of the small-range attributes and the other consists of the rest attributes. The key problem is to identify the small-range attributes. The NP-hardness of this problem is proved, and two approximate algorithms are proposed to solve this problem. The compression algorithms and the query processing based on the compressed method are also discussed. Experimental results show that the compression method has high compression ratio and enhances the query processing performance.

Abstract:During the process of mining maximal frequent item sets, when minimum support is little, superset checking is a kind of time-consuming and frequent operation in the mining algorithm. In this paper, a new algorithm FPMFI (frequent pattern tree for maximal frequent item sets) for mining maximal frequent item sets is proposed. It adopts a new superset checking method based on projection of the maximal frequent item sets, which efficiently reduces the cost of superset checking. In addition, FPMFI also compresses the conditional FP-Tree (frequent pattern tree) greatly by deleting the redundant information, which can reduce the cost of accessing the tree. It is proved by theoretical analysis that FPMFI has superiority and it is revealed by experimental comparison that the performance of FPMFI is superior to that of the similar algorithm based on FP-Tree more than one time.

Abstract:How to efficiently evaluate massive XPaths set over an XML stream is a fundamental problem in applications of the data stream. The current methods can not fully support the commonly used features of XPath, or can not meet the space and time requirement of the data stream applications. In this paper, a new tree automata based machine, XEBT, is proposed to solve the problem. Different from traditional ones, XEBT has the following features: First, it is based on tree automata with a powerful expressiveness, which can support Xpath {[]} without extra states or intermediate results; Second, XEBT supports many optimization strategies, including DTD based XPath tree automata construction, partial determination to reduce the concurrent states at running time with limited extra space costs, and the combination of bottom-up and top-down evaluation. Experimental results show that XEBT supports the complex Xpath and outperforms the former work in both efficiency and space cost.

Abstract:This paper proposes a probabilistic model incorporating long-term learning to estimate a dynamic user model. By using RF sequence as the user pattern, the approach can gradually update the prediction of current user based on matching the current user pattern with the user patterns in log. Compared with the invariant user model in PicHunter, the model is capable of dynamically adjusting when more user actions are observed, thus provides more accurate prediction for probability distribution. Experimental results on 11 000 images show that this approach can improve the retrieval accuracy apparently.

Abstract:Abstract: Although only more than one year since session initiation protocol (SIP) had been delivered, SIP has already become the research hotspot in telecommunication and network fields. More than 30 RFCs and drafts were delivered by SIP WG, including core protocol, QoS, security, message header and method extension, interconnection with PSTN, firewall and NAT traversal, application, multi-message body, instant messenger, etc. This paper introduces the SIP standard and the newest research trends in its related fields, and points out the future research direction and the promising foreground application.

Abstract:Content-Based publish/subscribe systems have recently received an increasing attention. Efficient routing algorithms and self-configuration are two key issues in the area of large-scale content-based publish/subscribe systems. Although many routing algorithms have been proposed, none of them fully exploits multicast to enhance system performance and save network bandwidth. In addition, the vast majority of currently available publish-subscribe middleware has ignored this self-configuration problem. This paper first proposes a hierarchical system model with multicast clustering. Then a hybrid routing algorithm is presented, which can fully exploit multicast in order to reduce the used network bandwidth. Moreover, a multicast clustering replication protocol and a content-based multicast tree protocol are presented for coping with the node or link failures and rebuilding the event dispatcher trees. Experimental results reveal that the system has better routing efficiency and lower cost, and guarantees the self-configuration characteristic.

Abstract:Researchers of law investigate the relevant law features and identification of computer evidences, while computer scientists investigate the technological features and acquisition methods of computer evidences. As an interdiscipline of law and computer science, computer forensics must be studied from the view angle of law, computer science and their combination. The separation of them may result in the confusion of identification in law and technology in computer evidences. In this paper, relevant issues of computer forensics to criminal evidence, law and technology are jointly investigated. Since the technological methods and tools of computer forensics are important, a typical case study is presented and analyzed. Finally based on the analysis of the deficiency for the current work on law enforcement and technology, future work on the improvement of computer forensics is discussed from the viewpoint of both law and computer technology.

Abstract:On the basis of the current single layer Markov chain anomaly detection model, this paper proposes a new two-layer model. Two distinctly different processes, the different requests and the system call sequence in the same request section, are classified as two layers and dealt with by different Markov chains respectively. The two-layer frame can depict the dynamic activity of the protected process more exactly than the single layer frame, so that the two-layer detection model can promote the detection rate and degrade the false alarm rate. Furthermore, the detected anomaly will be limited in the corresponding request sections where anomaly happens. The new detection model is suitable for privileged processes, especially for those based on request-response.

Abstract:Internet applications are developed rapidly. It is increasingly important for router itself to improve the ability to deal with networking congestion. Traditional Poisson model is unfit for Internet networks with burst flow. But self-similarity model suitable for Internet networks has not been used widely in practice because of its complex model and complicated calculation. By describing the practical buffer performance in routers, a new fuzzy congestion control model based on queues and a congestion control algorithm based on the model are presented. In the algorithm, all kinds of packets are firstly classified into queues according to their own priorities. Then the buffer state is divided into three phases, including normal, congestion avoidance, and congestion according to their buffer usage ratio. The three phases are crossover each other because of their fuzziness. Then by combining the whole congestion control, with the part congestion control, the fuzzy algorithm is carried out. Theoretical analysis and NS stimulation results show that the proposed algorithm has better networking performance in the fairness of all connections, compared with the traditional schemes, especially keeping from being affected by the connections with congestion. It really improves the routers’ ability to deal with network congestion.

Abstract:Blind digital watermarking, which can detect watermark without using the original image, is a key technique for practical intellectual property protecting systems and concealment correspondence systems. In this paper, a blind detection method for the digital image watermark is discussed. The theoretical research shows that the orthogonal projection sequence of a digital image is one-to-one correspondence with this digital image. By this conclusion, a kind of blind watermarking detector with good performance is designed and realized. To calculate the correlation value between the image and watermark, the pixel information of digital image is not adopted, but the orthogonal projection sequence of this image is adopted. Experimental results show that this watermark detector has not only the good robustness to Gaussian noise, but also the very strong resistant ability to translation and rotation attacks. Performance of this watermark detector is better than that of the general detector designed by using the pixel value information directly. The conclusions are useful for the research in the future.

Abstract:Natural interaction based on pen gesture is efficient for innovation in conceptual design. This paper presents a hierarchy concept model for pen gesture design based on the interactive informal constraint sketch and context-awareness. Furthermore, it discusses the constraint foundation and solution. Feature gesture as an instance of gesture applications in conceptual design is given. Compared with the traditional modeling mode, the feature gesture is convenient. Gesture-based interaction gives a natural way to record the brain streaming quickly. This paper provides an efficient method for innovative design to improve human-computer interaction.

Abstract:2-D polygon blending has been widely used in 2-D character animation, pattern matching, and geometric modeling. Previous algorithms tend to use the polygon’s geometric elements such as edge length, angle, area, skeleton to associate the regions on the two shapes that look alike. However they ignore the visual features of two polygonal shapes. This paper presents a new 2-D polygon blending method based on their visual features correspondence. By decomposing the two polygons into sub-polygon pairs with similar visual features compatibly, each sub-polygon in the source polygon can be smoothly transformed into the corresponding sub-polygon in the destination polygon. Since the feature-decomposition vertices are introduced, user can control the visual features correspondence flexibly and intuitively. Experimental results show that the two polygons can be blended as user prescribed whilst keeping the feature correspondence and feature preservation.

Abstract:This paper presents an effective nesting method for leather manufacturing, such as automobile interior decoration, etc. After the profiles of leather sheets and stencils are obtained, they are discretized to make the processing independent of the distinct geometry. The constraints of profiles are thoroughly considered. A heuristic bottom-left placement strategy is employed to sequentially place stencils on sheets. The optimal placement sequence and rotation are deterimined using a simulated annealing based genetic algorithm (SABGA). A natural concise encoding method is developed to satisfy the possible requirements of the leather nesting problem. Experimental results show that the proposed mehtod not only can be applied to the normal two-dimensional nesting problems, but also can be especially suited for the placement of multiple two-dimensionally irregular stencils on multiple two-dimensionally irregular sheets as well.

Abstract:Key Program is a kind of programs supported by National Natural Science Foundation of China (NSFC). It aims at improving the development of science and the continuable development of national economy and society. In this paper, the 2004 year’s application for funding and supporting Key Program by Computer Science Division of Information Science Department are summarized. While reviewing principles are provided, the problems are analyzed. Finally the suggestions are given.