Journal of Computer Applications

Review of QoS mechanisms for mobile Ad Hoc networks

ZHNAG Peng CUI Yong SUN Lei

2009, 29(3): 625-632.

Asbtract ( )

PDF (1980KB) ( )

Related Articles | Metrics

There are difficulties to guarantee QoS of mobile Ad Hoc networks due to their features of dynamically changing topology and wireless communication. Current QoS techniques for mobile Ad Hoc networks are divided into five categories: link layer solutions, network layer solutions, transport layer solutions, cross layers solutions and cross networks solutions. The main link layer solutions are QoS MAC protocols. The main idea in network layer solutions is QoS routing. Wireless TCP is the main scheme to guarantee QoS in transport layer. To choose route and balance load, cross layer solutions were researched to provide network architectures to support QoS. The main idea in cross network solutions is to propose QoS solutions between mobile Ad Hoc networks and fixed networks. These QoS solutions were introduced and analyzed.

Design and analysis of IBCbased TLS handshake protocol

Chang-Yan PENG Quan ZHANG Chao-Jing TANG

2009, 29(3): 633-637.

Asbtract ( )

PDF (1043KB) ( )

Related Articles | Metrics

To overcome the disadvantages of complex certificate management and long handshake latency which exist in current widelyused Transport Layer Security (TLS) protocol, two improved handshake protocols using identitybased cryptography were designed. One protocol adopted identity-based encryption scheme, and the other used identity-based authenticated key agreement scheme fit for mutual authentication. Results of security analysis and performance simulation show that IBC-based protocol has commensurate cryptographic computation overhead compared with certificatebased scheme with equivalent security, but it decreases communication overhead, and has prominent superiority in handshake latency; therefore, the scheme achieves higher efficiency.

Distributed group key management scheme based on cluster collaboration in WSN

Wei-Ni ZENG Ya-Ping LIN

2009, 29(3): 638-842.

Asbtract ( )

PDF (1029KB) ( )

Related Articles | Metrics

Sensor network is vulnerable to attacks when deployed in unattended fields, and no node can be trusted all the time thus can act as a key server. In this paper, a distributed group key management scheme was proposed. In the new scheme, the group rekeying process took full advantage of the characteristics of the inter-cluster and intracluster communication. When the number of the compromised nodes inside one cluster was smaller than t and a compromised node inside the cluster was detected, the group rekeying would be initiated by the cluster and the compromised nodes would be revoked by inter-cluster collaboration; otherwise, the group rekeying would be initiated by one neighbor cluster of the nodes, and the nodes will be revoked by intracluster collaboration. Compared with the existing schemes, the proposed scheme can rekey realtime, provide a high level of security and reduce communication overhead at the same time.

Research on EPA network without collecting switchers

2009, 29(3): 643-645.

Asbtract ( )

PDF (624KB) ( )

Related Articles | Metrics

Because EPA protocol realizes accessing media without collision, it supports nodes to share accessing media. Under the specifications of EPA, three EPA network structures were designed without collecting switchers supporting line topology. Methods followed to realize the network form and its real-time property was analyzed.

An efficient RP2P network based on hierarchical dividing

Yuan LI Shiping CHEN

2009, 29(3): 646-648.

Asbtract ( )

PDF (782KB) ( )

Related Articles | Metrics

RP2P algorithm combines arbitrary neighbor selection, typically used only in unstructured P2P networks, with a Distributed Hash Table (DHT) ring. It is the first of its kind to resolve requests in d hops with a chosen probability of 1－c. However, the capacities of the hosts participating in the network, such as bandwidth, memory, CPU, are very different, which will affect the efficiency of the whole system. On the other hand, the shock caused by some of the nodes in the network frequent joining in/departing from the system is also one of the factors affecting the performance. This paper analyzed the capacities of the nodes and proposed an efficient RP2P network based on hierarchical dividing. It improves the efficiency of the system, and solves the problem of system shocks.

Opportunistic coding based on source directed relay

Jian QIN BAIWEI Yang Ou LI

2009, 29(3): 649-651.

Asbtract ( )

PDF (530KB) ( )

Related Articles | Metrics

In this paper, the authors combined the idea of source directed relay and opportunistic coding, defined the butterfly and chain structure, proposed a new annulus structure of network coding further, and thus made the network coding much more easier to realize in wireless environment. The simulation results on platform of NS2 show that opportunistic coding can make the throughput and the energy efficiency higher obviously.

Bi-directional selection mechanism based on reputation

2009, 29(3): 652-654.

Asbtract ( )

PDF (579KB) ( )

Related Articles | Metrics

A bi-directional mechanism, BDS_R, was proposed to deal with the deficient fairness and low effectiveness in trust model. The BDS_R mechanism granted the selection freedom to the request peer and service peer for providing better services actively and increasing the transaction success rate. Experimental results show that: bidirectional selection mechanism based on trust has higher transaction success rate, and better performance in avoiding network resource waste, reducing network flow, and motivating peers.

DBCTIA algorithm based on binary object optimization used for load balancing of multiple links

Ben-Lai LIANG

2009, 29(3): 655-657.

Asbtract ( )

PDF (753KB) ( )

Related Articles | Metrics

The algorithm DBCTIA (Delay and residual Bandwidth’s Covariance Target Iteration Algorithm) based on binary object optimization for load balancing of multiple links was provided based on IP traffic engineering and Delay Target Iteration Algorith (DTIA) for delay balancing of multiple links. DBCTIA algorithm can balance the delay and residual bandwidth of multiple links by the second adjusting of IP routing tables size and amount according to the covariance of links delay and residual bandwidth. Experimental results show the balancing of multiple links' delay and residual bandwidth is remarkable by DBCTIA and the power value of multiple links after the balancing by DBCTIA is larger than that after the balancing by DTIA.

Adaptive algorithm for dynamic traffic grooming in WDM mesh networks

SUN Su-yun

2009, 29(3): 658-661.

Asbtract ( )

PDF (849KB) ( )

Related Articles | Metrics

Aiming at the problem of dynamic traffic grooming in WDM mesh network, the paper presented an adaptive algorithm for dynamic traffic grooming based on layered auxiliary graph network model. According to different traffic requests and grooming capabilities in the current network state, the algorithm applied corresponding strategy of traffic grooming for allocating network resource exactly, in order to reduce average network block probability. The simulation results show that the network performance can be improved significantly by the proposed algorithm.

P2P traffic identification method based on node status

2009, 29(3): 662-664.

Asbtract ( )

PDF (557KB) ( )

Related Articles | Metrics

Random port, blurred protocol features, as well as HTTP tunneling technology have become development trends of P2P application. A new method based on node status was proposed for P2P traffic identification. Node status obtained by multilevel identification module was hashed into a table, and then a trusted list was created by this hash table. Through the definitions of the active factor, incremental factor and link cache, P2P traffic generated by port jump and protocol encryption can be identified quickly and effectively.

Routing of P2P based on trust management

Hong-Lin WANG

2009, 29(3): 669-671.

Asbtract ( )

PDF (754KB) ( )

Related Articles | Metrics

Each node in P2P networks acts as a router; therefore, selecting appropriate note to store-and-forward packets in searching a certain file can improve the efficiency. The authors constructed a dynamic topology model based on trusting scheme, gave the approaches to assess node reliabilityrate and detected fault nodes with encrypted messages sending. The congestion of nodes caused by too many links of high-reliability-rate nodes could be avoided in terms of setting response threshold of CPU's response time. To improve the performance of the networks, the updating algorithm of nodes linking based on trusting idea was designed, and it was verified by emulation experiments.

Improved method of TCP-friendly congestion control protocol

Zhang Da-min Lin-Zhou-Ting CHEN

2009, 29(3): 672-674.

Asbtract ( )

PDF (612KB) ( )

Related Articles | Metrics

TCP Friendly Rate Control (TFRC) is a congestion control mechanism designed for non-TCP flows. Due to that not all losses are related to congestion in wireless network, to adopt initial TFRC will induce maloperations. In the paper, biased queue method was used to distinguish congestion losses from random losses. It would prevent unnecessary congestion control and would improve the quality of transmission in wireless networks. Through simulations and experiments, it is found that the method can greatly improve the stability of transport rate, finally accomplishes QoS of application layer.

Rate control algorithm based on standard H.264

Yu Yang HeXiao Hai

2009, 29(3): 675-677.

Asbtract ( )

PDF (744KB) ( )

Related Articles | Metrics

During the video communication process, an improved rate control algorithm was developed based on H.264 video coding standard in order to make the output data stream better adapted to different channels. Contrary to the MAD prediction of the basic unit only considering the relationship of the time in the coding standard, a new prediction method was introduced combining time prediction and space prediction. Compared with coding standard, the experimental results illustrate that the new algorithm can control rate more precisely and increase the peak signal noise ratio of the output video stream effectively. It proves that the new algorithm can make the output data stream of the encoder more stable.

Study and implementation of IPv4/ IPv6 transition technology based on multi-core

YANG Zhi-Yi YANG LI Xiao-Yang

2009, 29(3): 678-680.

Asbtract ( )

PDF (624KB) ( )

Related Articles | Metrics

According to the problem in the transitional period of IPv4/ IPv6 and the shortages of the mainly techniques during this period, this paper proposed the IPv6 tunnel technology based on multi-core, designed its model, used the multi-core flow technology, the lock-technique and the multi-core loadshare model to speed up transmitting, protected the critical resources and improved the whole performance. Finally, the IPv6 tunnel technology based on multi-core is proved to have better transition and service performance than that based on singlecore.

Research on Tableau algorithm for Horn-Extended DL

2009, 29(3): 681-685.

Asbtract ( )

PDF (828KB) ( )

Related Articles | Metrics

Description logic and logic program are two important kinds of knowledge representation, with interesting, yet orthogonal expressive power. In order to ensure the decidability of the combination of description logic and logic program, Motik restricted rules to DL-safe ones. Based on the work of Motik, Horn-extended DL, by extending description logic with Horn clauses, was proposed. Then the Tableau algorithm for Horn-Extended DL was given. Finally, the correctness and efficiency of the algorithm were verified by a numerical experiment.

Approach to pointcut definition based on conceptual model

2009, 29(3): 686-689.

Asbtract ( )

PDF (1038KB) ( )

Related Articles | Metrics

This paper proposed a visual approach to pointcut definition to tackle the fragile pointcut problem of aspectoriented programming. The pointcut definition was based on a conceptual model not directly on how the base program was structured, which had effectively decoupled the pointcuts from the base program's structure. Therefore, it improved the robustness of the pointcut. The conceptual model was, as an abstraction of the AspectJ pointcut descriptor, achieved through extending the UML meta-model elements. Therefore, it was called pointcut model. The Rational Rose was also extended, so as to support the pointcut model and automatically generate the aspect skeleton code from the model.

Software reliability modeling based on grey system theory

JIN Ang

2009, 29(3): 690-694.

Asbtract ( )

PDF (942KB) ( )

Related Articles | Metrics

Grey theory is applied to the study of software reliability modeling. Through modeling with history failure data, the nextstep failure time was predicted. To verify the accuracy of the Grey Model (GM), the authors applied it to a group of generally accepted data, and compared the results with that of traditional software reliability growth models by computing the SRE measure. The experimental results show that in most cases, the GM ranks first as far as the short-term predictability is concerned.

Web application test model based on action

He-Gang Fu Lu Yan-Jun Gang Zeng

2009, 29(3): 695-699.

Asbtract ( )

PDF (809KB) ( )

Related Articles | Metrics

The testing method based on state transition is an effective way to test the dynamic behavior of Web application. By improving the existing Web testing method, this paper proposed a new testing model based on action to test complex dynamic Web pages in Web application system. The methods consisted of two steps: 1) describing the system structure of Web application with WARD; 2) describing the inner page structure of complex dynamic Web page with actionbased SMD. At last, the results above were combined to model the whole Web application. Experiment on practical Web applications approves that the method is comprehensive and accurate.

Study and implementation of software automatic testing system based on STAF

LI Xia-An Zhi-Bo Chen

2009, 29(3): 699-701.

Asbtract ( )

PDF (743KB) ( )

Related Articles | Metrics

A software automatic testing system based on Software Testing Automation Framework (STAF) was designed and implemented to solve the problems in current automatic testing solutions. This system could be employed to manage the process of automatic testing, and support the execution and monitoring of automatic testing tasks in multiplatform network environment. The system was implemented to improve the degree of automation, enhance the efficiency of testing, shorten the testing period, and increase the trustworthiness of the testing results.

Research of connector model oriented component composition

2009, 29(3): 702-704.

Asbtract ( )

PDF (583KB) ( )

Related Articles | Metrics

The channel of dynamic component composition is connectororiented. Concerning the disadvantages in present connector models, a connector model with more granularity was presented. In contrast to the connectors in the existing connector models，this connector model initiated calls and managed their returns，and used connectors to encapsulate control in a component model. The evolvement of connector was given, thus the model of component composition was implemented based on connectors. In the end, an example was introduced.

An improved static program slicing algorithm

Long-Jie ZHANG

2009, 29(3): 705-707.

Asbtract ( )

PDF (726KB) ( )

Related Articles | Metrics

This article proposed an improved static program slicing algorithm, and applied it to software reverse engineering. To resolve the interprocedural call, the article converted it to intraprocedural call by establishing a suggestive table, which predigested the complexity of establishing the program dependence graph greatly. By analyzing the type of variables in the object program, the article offered several algorithms to do equivalent transformation to the linear operation codes of the program, which reduced the scale of the sliced program. In the end, the improved algorithm was proved to be effective by a slicing example.

Process modeling based on extended UML activity diagrams

2009, 29(3): 708-711.

Asbtract ( )

PDF (861KB) ( )

Related Articles | Metrics

Because of lacking effective method for describing the process of simulation, this paper presented a modeling way based on extended-UML activity diagrams, and constructed two individual models for the pre-code process, which concentrated on info-activity and physical activity. The simulation results reveal that this method can not only describe the particular work of modeling, but also be propitious to communication between the user and the developer.

UML-based approach to generate testing sequence of polymorphic software

CHEN Lian-ping Yi ZENG

2009, 29(3): 712-715.

Asbtract ( )

PDF (809KB) ( )

Related Articles | Metrics

Polymorphism is one of the three most important characteristics of objectoriented technology. It not only enhances the flexibility of programming and reuse of codes, but also brings new challenges to software testing. The cause of its appearance and its formalization expressing were analyzed. Under the guidance of controlflow graph, the information of class diagram was integrated with collaboration diagram, traditional Interprocedural Restricted Control-Flow Graph (IRCFG) was extended with polymorphic information and class diagram information. After analyzing the coverage criteria, an approach to generate testing sequence with polymorphic information was given.

Research on system of design reuse for parts based on ontology

2009, 29(3): 716-721.

Asbtract ( )

PDF (1158KB) ( )

Related Articles | Metrics

Design reuse of parts plays an important role in product design. The major difficulty of design reuse of parts is that the information of parts is difficult to express, manage and retrieve. The design reuse of parts based on ontology was studied. A framework of parts information expression and an approach of ontology modeling were proposed. The key technologies of retrieval and storage ontology information were given. Lastly, a reuse system prototype for parts was designed, and an example was given to verify the feasibility and effectiveness of the system.

New shape reconstruction method for little convex polyhedrons

Qing-Shuang Zeng

2009, 29(3): 722-724.

Asbtract ( )

PDF (1059KB) ( )

Related Articles | Metrics

A projective silhouette based method was proposed to reconstruct the shape of little convex polyhedron. First, the vertical section projective silhouette images of little convex polyhedrons taken in every parallel rotary angle were processed, and the corresponding set of silhouette sequences and the set of cross section projective length curves were acquired. Next, the index of all the potential surface planes of measured object was gotten from the curves set, then these planes were filtered by comparing with every silhouette sequence, and the parameters of all real planes were obtained by computing. Finally, the shape of little convex polyhedron was reconstructed. Experimental results show that the proposed method can achieve the shape reconstruction of little convex polyhedron accurately. Compared with other methods, it has such advantages as condensed device and fast process speed, and fits for the engineering applications aiming at the little convex polyhedron workpieces.

Corner detection using multi-scale representation of contour orientation rate based on B-spline

2009, 29(3): 725-728.

Asbtract ( )

PDF (1234KB) ( )

Related Articles | Metrics

Multi-scale representation for contour orientation rate was defined based on B-spline scale space, and a new corner detection algorithm was proposed by exploring multi-scale contour orientation rate and its multiplication. The local extrema of multiscale multiplication were directly defined as candidate corners. Multi-scale multiplication not only included the feature information of orientation rates at several scales, but also could enhance the contour orientation rates of corners while suppressing the rates of contour points corrupted by boundary or quantization noise. Thus, the good corner detection results could be obtained only through using a global threshold and that multi-scale detection of the candidate corners was truly performed. Finally, quantitative analysis of corner detection localization performance was performed, and the comparing experiments illustrate that the proposed algorithm possesses good detection and localization performance.

Edge collapse simplification based on similar curvature

2009, 29(3): 729-731.

Asbtract ( )

PDF (768KB) ( )

Related Articles | Metrics

To reserve more important shape features in the low-level model, an improved edgecollapse triangle mesh simplification algorithm was presented. On the basis of Garland's simplification algorithm, the concept of similar curvature was introduced and embedded into the original Garland's quadric error metric, so that the metric can not only measure distance error but also reflect geometric variations of local surface. The experimental results show that the improved algorithm effectively retains the detail characteristics of the model, and obtains a better simplified effect.

Shape and texture acquisition of objects in real scenes based on volumetric presentation

2009, 29(3): 732-735.

Asbtract ( )

PDF (1369KB) ( )

Related Articles | Metrics

In traditional threedimensional (3D) reconstruction methods based on images, the correlated feature matching between images is one of the factors affecting the fidelity and accuracy of modeling. A 3D reconstruction method of object in real scene based on volumetric presentation was proposed, which avoided the feature matching and simplified the complexity of modeling. Firstly, the silhouettes of object in every image were extracted. Secondly, the minimal bounding box of object in scene volume was traversed in a single-pass order, after that the volume element was projected back to input images and checked if it was consistent according to color invariants. Finally, the realistic volumetric model was rendered using calculated 3D information. Experimental results demonstrate that the reconstructed model with the method in this paper is more realistic.

Moving targets detection using local OGHM

2009, 29(3): 736-738.

Asbtract ( )

PDF (1107KB) ( )

Related Articles | Metrics

The efficiency of the 3D OGHM detection algorithm will be at a very low efficiency level if the resolution of the video is high. Concerning this problem, this paper proposed a new idea in the sense of the resolution of the image that participated in the computation. It is to limit the computation of OGHM in some special regions (for example, the neighbors of the potential targets), which is important for motion detection, while to prevent other unimportant pixels from the computation of detection. Following the idea, the authors proposed a new extended version of 3D OGHM named local OGHM (LOGHM) for motion detection, and discussed respectively the efficiencies of LOGHM and its preexistence. The experimental results show that LOGHM has higher efficiency and better antinoise performance compared with 3D OGHM algorithm.

Moving object detection algorithm based on multithreshold for PCNN

Ying-jie LIU

2009, 29(3): 739-741.

Asbtract ( )

PDF (782KB) ( )

Related Articles | Metrics

Motion detection has a wide range of applications in many artificial intelligence implementations. An improved motion detection algorithm was proposed. Gaussian mixture model of classical algorithm was used, and background and foreground were classified by using Pulse Couple Neural Network (PCNN). PCNN was modified, multi-threshold was adopted to detect object and simple maximum entropy rule was applied to end iteration. Also, a new learning rate was proposed in the model updating stage. Experimental results show that the proposed algorithm improves the ability of detection, noise restraining, stability and so on.

New algorithm for moving object extraction in MPEG compressed domain

Chun-mei HE

2009, 29(3): 742-746.

Asbtract ( )

PDF (1173KB) ( )

Related Articles | Metrics

The motion vector field directly obtained from the video stream is sparse and unreliable. It will lead to low accuracy of the extracted motion objects. This paper proposed a novel compressed-domain motion object extraction algorithm for MPEG video stream. The motion vector field directly obtained from the video stream was firstly processed by reliability test, densification and filtering. Blockbased region merging was used to obtain the rough region of every motion object. These motion regions were decoded to get the original pixel values, which were modeled with Gaussian mixture model by statistical method. The final accurate motion objects were extracted by threshold decision. Simulation results demonstrate that the proposed approach can get rid of the noisy motion vectors, then the accuracy of segmentation is greatly improved, and a high processing speed is achieved. For video stream in QCIF format, 50 frames per second can be achieved.

Improved Ribbon Snake algorithm for automatic road generation

HU Yang ZU Ke-Ju LI Guang-Yao

2009, 29(3): 747-749.

Asbtract ( )

PDF (1403KB) ( )

Related Articles | Metrics

In order to modify the incomplete road extraction caused by shadow, shelter and noise in high resolution remote sensing images, a Ribbon Snake model with width information was established based on the geometric characteristics of road. To overcome the great dependence of interior parameters and the easy being affected by the complex background of Ribbon Snake, a B-spline Ribbon Snake was constructed, where the smoothness of the Snake was implicit in the B-spline formulation and the flexibility of the Snake was adjusted by the number of control points. The road network segmentation results show that the improved B-spline Ribbon Snake can obtain a more accurate and smoother segmentation and is more robust to noise.

Carving surface modeling of images based on quadratic Bezier curved surface

Chen Yutuo Yang Weimin XIONG QiuBao HU Liang HAN Xuli

2009, 29(3): 750-754.

Asbtract ( )

PDF (1531KB) ( )

Related Articles | Metrics

Carving surface was directly modeled using the quadratic Bezier curve surface according to the grey or color pixels information of the image with clear texture and simple construction. This method segmented the image, analyzed and computed the pixel values and their position parameters for the sub-images, and used those values as feature parameters for constructing the surface patches of the model. The modeling was implemented by limited smooth connection of surface patches. The established surface model can describe the basic structure characteristics of the original objects in the image, and satisfy the requirements of digital carving processing. It has low complexity and strong practicability.

Assessment method of image interpolation based on edge primitive

Yu-Qing HE

2009, 29(3): 755-756.

Asbtract ( )

PDF (1079KB) ( )

Related Articles | Metrics

Image quality assessment methods are divided into subjective ones and objective ones. Subjective assessment methods are difficult to use in actual applications. As a statistical parameter, PSNR is commonly used in objective way about image fidelity; but sometimes it is not consistent with the subjective way. This paper proposed an assessment method of image interpolation based on edge primitive, which was a special reduced-reference image quality assessment. The experimental results show that this method is easier to implement and accords with HVS. Combined with other methods, this method can be used as a reference method in quality assessment of image interpolation.

Object tracking algorithm with adaptive color space based on CamShift

2009, 29(3): 757-760.

Asbtract ( )

PDF (1805KB) ( )

Related Articles | Metrics

Considering the poor performance that CamShift algorithm only applies to track targets with certain color, an improved algorithm named adaptive color space tracking algorithm was proposed. Using the new measurements, the current color space was selected dynamically according to the average distance between objects and backgrounds. With the introduction of the mechanism in similarity analysis, time cost was decreased. The experimental results show the new algorithm can track multicolor targets in complex backgrounds more accurately.

Application of Speech recognition technology in English pronunciation error correction

Bo ZHAO Xiao-hong TAN

2009, 29(3): 761-763.

Asbtract ( )

PDF (813KB) ( )

Related Articles | Metrics

Many applications of computer-aided language learning system lack of evaluation & feedback of learning speech. Auto speech recognition could provide a score based on likelihood. To describe an oral English training system. Besides normal speech evaluation, error detection and prosody modification based on the result of phoneme alignment and phoneme recognition are provided by this system. Recommendations for improvement queried from correction knowledge base offer learners a timely help. Current experiment shows most mistakes made unconsciously by high-level 2nd language speaker could be corrected by it.

Study on a novel method of speaker recognition

ZHANG Wei-Jie Wan－Chun FEI Liang-Jun XU Lu LIU

2009, 29(3): 764-767.

Asbtract ( )

PDF (784KB) ( )

Related Articles | Metrics

The speech signals were analyzed by means of shorttime Fourier transform, and the time series clusters that were composed of the same frequency's energy series of each speech were gained. Between every column vector serie and the serie consisting of every row vector's average value, the linear regression equations were established based on time series pretreatment and mathematical statistics. The deterministic parts and stochastic parts of the time series cluster were separated, and the feature parameters were extracted from speech signals of the speaker, then the speech signals were recognized. The experimental results show that the highest average recognition rate is 97.94%, with the distance parameter described in this paper and three frequencies on the speech set of 194 speech signals pronounced by eight speakers.

Musical score recognition of solo music signal based on WAV file

2009, 29(3): 768-770.

Asbtract ( )

PDF (719KB) ( )

Related Articles | Metrics

Most work about musical score recognition is based on MIDI format file; however, most musical data are stored in audio frequency format. Therefore, this paper discussed a musical score recognition method from WAV format file including solo music signal, adopted the Discrete Cosine Transform (DCT) combined with energy, and presented a principle of least derived notes. The music signal was divided into several subsections and the length of each section was equal to the duration of one semiquaver. The base frequency was obtained using one dimension DCT operated on each section. The extreme points of the musical notes were found by comparison with the adjacent sub-sections. And then, the musical scores were got using the principle of least derived notes. The experiments show that this method is very effective.

Combination method of Fisher theory and principle component for multi-sensor information fusion

Shu-Pping WAN

2009, 29(3): 771-773.

Asbtract ( )

PDF (502KB) ( )

Related Articles | Metrics

Aiming at the target recognition problem of multisensor with multiple characteristic indexes, a new fusion method was proposed combining Fisher discrimination and principle component. The method merged the number of discrimination functions to reduce the workload of recognition, and utilized the theory of Fisher discrimination to recognize targets. It is very suitable for the recognition of multitargets, very simple to compute, and can be performed on computer easily. The applied example proves that the method is effective.

Recognition algorithm of QR code for embedded systems

2009, 29(3): 774-776.

Asbtract ( )

PDF (613KB) ( )

Related Articles | Metrics

An algorithm to identify QR code for embedded systems was proposed, and verified on NIOS embedded system. The experimental results indicate that this algorithm can pick up the bar code mark information accurately and fast, and the effect is good.

Decision tree classification of remote sensing images based on vegetation indices

2009, 29(3): 777-780.

Asbtract ( )

PDF (1803KB) ( )

Related Articles | Metrics

In order to explore the applications of ETM+ remote sensing data to urban landscape pattern analysis, the decision tree classifier based on See5 was developed and its generation strategy was discussed in detail. Taking Xuzhou city as the study area, spectral features and ten vegetation indices, including Normalized Difference Vegetation Index (NDVI), Greenness Vegetation Index (GVI), Ratio Vegetation Index (RVI) and so on, were used and extracted for decision tree classification. By comparing the classification results with decision tree classifier based on spectral features only, vegetation indices used in the processing of remote sensing images classification could advance the classification accuracy. The result also shows that the decision tree classifier is effective to landscape pattern classification from remote sensing images based on various features.

Improved Mean Shift real-time multiple faces tracking algorithm

2009, 29(3): 781-784.

Asbtract ( )

PDF (1117KB) ( )

Related Articles | Metrics

A real-time multiple faces tracking algorithm based on Mean Shift algorithm was presented in this paper. An adaptive kernel window was adopted for each object, and a ranked sequential tacking strategy was proposed to solve the occlusion problem. Multi-accessory information was introduced to match the tracked objects and target windows. Furthermore, in order to improve the tracking speed and robustness, Kalman filter was introduced to predict the position of the object window. The experimental results show that the proposed multiple faces tracking algorithm can track multiple faces robustly and in real-time.

Color-based method for face detection and location

2009, 29(3): 785-788.

Asbtract ( )

PDF (1229KB) ( )

Related Articles | Metrics

Integrating with skin color and facial feature, a fast method was proposed for detecting and locating the face. In normalized RGB color space, skin pixels and lop pixels could be extracted by skin lotus. If a candidate face region did not contain lip pixels, it could be eliminated. The intensity of eye region is higher than that of others in the B component of normalized RGB color space because of sclera and pupil glisten. A region growing method was used to produce potential eyes, and then the geometry feature of eyes and lip were used to locate the face. The results of experiments indicate that the proposed method is robust and effective.

Recognition and description system of mathematical formulae for printed documents

CHEN De-Yu ZHU Xue-Fang SU Xiao-Cheng HANG Yue-Qing

2009, 29(3): 789-791.

Asbtract ( )

PDF (670KB) ( )

Related Articles | Metrics

The establishment of mathematical formula recognition system needs more investigation on mathematical formulae themselves, recognition methods of printed characters and mathematical symbols and their descriptions. On the basis of the experimental system of mathematical formulae recognition and descriptions, the above issues were discussed in the paper. The proposed system could recognize some formulae structures, characters and mathematical symbols. The formulae in image form could be converted to the corresponding text form with the effective descriptions of the results using MathML.

Online visual inspection method for O-ring number on the pipe connectors

2009, 29(3): 792-794.

Asbtract ( )

PDF (1170KB) ( )

Related Articles | Metrics

In the visual inspection application for the O-ring number on the chrome-plated pipe connectors, the orientations of the pipe connectors are unfixed, and there are often shadows and hotspots in the image. To prove the stability of the image processing results, the basic image processing procedures of thresholding, thinning and size filtering were adopted. Moreover, the threshold value was calculated by using P-tile method, the lines with branches or with large curvatures in the thinned image were broken by using neighborhood operation, and the broken lines in the sizefiltered image were linked based on the geometric information of these lines. The processing tests on real images show that the accurate number of O-rings can be fast acquired with the image processing algorithms presented in this paper.

Flatness defect pattern recognition with data mining technology

Xiao-yan Zhao

2009, 29(3): 795-797.

Asbtract ( )

PDF (590KB) ( )

Related Articles | Metrics

The flatness defect pattern recognition based on data mining technology was proposed. In order to solve low accuracy of normal BP (Back Propagation) network, a novel data mining algorithm based on hierarchical BP model was presented. The new model with binary tree structure reduced prediction range of each network and adopted several networks for degree elevation. Compared with the normal BP model, the new system precision was improved remarkably. The experimental results show this method can meet the requirements of the producing process.

Research and implementation of RBFNN classification algorithm for digital signal modulation recognition

2009, 29(3): 798-800.

Asbtract ( )

PDF (507KB) ( )

Related Articles | Metrics

A new Radial Basis Function Neural Network (RBFNN) classification algorithm for the digital signal modulation recognition was proposed. Subtraction clustering algorithm and Least Mean Square (LMS) algorithm were used to adaptively train the number and location of the hidden layer center and the weight coefficient of output layer considering all the characteristic parameters, and the best interface in the multidimensional space could be found. Meanwhile, the blindness and randomness of the number and position of the hidden layer center were avoided. The simulation results show that the algorithm can improve correct recognition rate with the same characteristic parameters.

Zero digital image watermarking method against rotation attack based on Contourlet transform

Jie ZHAO

2009, 29(3): 801-804.

Asbtract ( )

PDF (1414KB) ( )

Related Articles | Metrics

In order to improve the ability against rotation attack of zero digital image watermarking method, this paper proposed a method based on Contourlet transform against rotation caused by distortion, LogPolar Mapping (LPM) was used to transform rotation problem in Castesian coordinate to translation property in LPM coordinate. After translation, an exhaustive search was adopted to detect test whether image involves watermarking or not. The experimental results demonstrate that the proposed method can obtain better visual effect; meanwhile, it is robust enough to some image degradation processes such as adding noise, cropping, filtering and JPEG compression.

Improved negative selection algorithm for network anomaly detection on high-dimensional data

Wen-Zhong GUO Guo-Long CHEN Qing-Liang CHEN

2009, 29(3): 805-807.

Asbtract ( )

PDF (799KB) ( )

Related Articles | Metrics

Negative Selection (NS) algorithm of artificial immunology has been successfully applied to anomaly detection on some lowdimension data, but the performance becomes unfavorable on highdimension data. Realvalued negative selection algorithm with variablesized detectors (VRNS) was applied to network intrusion detection and a variation of it (IVRNS) was proposed to improve the performance on highdimension data. In the improved algorithm, the detectors were used to control the coverage of them according to the overlap among the detectors. Experimental results prove the effectiveness of this novel algorithm on high-dimension data and a high detection rate with a lower false alarm rate in network intrusion.

Method of Intrusion Alert Correlation Based on Uncertain Knowledge Discovery

2009, 29(3): 808-812.

Asbtract ( )

PDF (1250KB) ( )

Related Articles | Metrics

To solve the problems of huge number, trivialness and dispersedness of intrusion alerts raised by intrusion detection systems, an intrusion alert correlation method based on uncertain knowledge discovery is proposed. A new uncertain sequence patterns discovery algorithm, named CWINEPI, is used to discover the sequence patterns of intrusion alerts in the knowledge discovery part. After filtered, the sequence patterns of intrusion alerts are translated into intrusion alerts condensed rules. The attack patters are obtained with correlating sequence patters, and converted into intrusion scenarios building rules. In the intrusion alerts correlation part, the alert correlation engine is constructed to correlate alerts produced by multi intrusion detection systems using patter matching with the rules obtained from knowledge discovery part. The alerts from Defense Advanced Research Projects Agency 2000 intrusion evaluation data (DARPA2000) validate the good performance of the knowledge discovery part. The correlation results of the alerts from real testing environment show the proposed method is efficient and feasible.

Research of data acquisition technology in program automatic unpacking

2009, 29(3): 813-816.

Asbtract ( )

PDF (747KB) ( )

Related Articles | Metrics

The hidden code and data extraction of packed program is an important issue of malware detection technology. The running characteristics of packed program were analyzed, the memory monitoring algorithm and dynamic identification algorithm of basic block were given, and the data acquisition system of program automatic unpacking based on QEMU emulator was described. Testing results show that the code and data of original program can be extracted, and executable behavior of program unpacking can be recorded.

Software protection technology based on selfsealing code block

Li-Guo Zhou

2009, 29(3): 817-822.

Asbtract ( )

PDF (941KB) ( )

Related Articles | Metrics

With regard to the shortage of trash instruction inserting using software protection of conventional technology, a new software protection technology was proposed based on the concept of selfsealing code block. With the help of the new technology, the software could be protected from static analysis and dynamic debugging. The generation method of random instruction sequence and relevant reverse instruction sequence for constructing selfsealing code blocks were emphasized, including random original instruction sequence generation based on instruction encode table and reverse instruction sequence generation based on reverse engineering. The corresponding algorithms and sample analysis were provided.

Method of conflict detection and elimination for security policy

2009, 29(3): 823-825.

Asbtract ( )

PDF (746KB) ( )

Related Articles | Metrics

Security policy is the base of system security management. It is difficult to eliminate policy conflict in the distribution system because of its complexity. Therefore, conflict detection and elimination of policy is the key to realize policybased system defense. This paper proposed a general formal description of policy, defined the logic relationship among the policy's elements, and classified the conflicts of policy, then discussed some conflict detection and elimination methods corresponding to the conflicts.

Research on probing characteristics of proxy server

Jie YANG Hui SHU

2009, 29(3): 826-829.

Asbtract ( )

PDF (797KB) ( )

Related Articles | Metrics

The communication characteristics of proxy server were investigated from the angle of remote host computer. The functions and classifications of proxy server were summarized, and its working model was proposed, together with analysis of its logic modules and factors that influenced communication. After the principle of probing characteristics of proxy server was analyzed, a probing algorithm for HTTP proxy was designed, and a probing system prototype was implemented. The system accomplished penetrating through specific authentication mechanism in Windows domain environment. At last, the system was tested and its performance was analyzed. It is concluded that it is applicable to a certain extent.

Clustering algorithm based on grid and fractal dimension

min-jun liang

2009, 29(3): 830-832.

Asbtract ( )

PDF (612KB) ( )

Related Articles | Metrics

Combining the approaches based on grid and fractal, a new kind of clustering algorithm called grid and fractal dimension based clustering algorithm (GFDC) was presented. It overcame the shortcoming of low clustering quality in traditional grid-based clustering method and solved the time-consuming problem of fractal clustering method. In the initial stage of the new algorithm, some initial clusters were got according to the grid density. Then in the expansion stage, the unclassified grids were categorized using the idea of fractal. Experimental results confirm that GFDC is able to capture arbitrary shapes and non-neighboring clustering, and can be applied to the massive and high-dimension dataset.

Multi-feature fusion method based on support vector machine and k-nearest neighbor classifier

2009, 29(3): 833-835.

Asbtract ( )

PDF (593KB) ( )

Related Articles | Metrics

The traditional classification methods only use one single classifier, which may lead to onesidedness, low accuracy, and that the samples nearby the Support Vector Machine (SVM) hyperplanes are more easily misclassified. To solve these problems, the multifeature fusion method based on SVM and KNearest Neighbor (KNN) classifiers was presented in this paper. Firstly, the features were divided into L groups and the SVM hyperplanes were constructed for each feature of training set. Secondly, the testing set was tested by SVMKNN method, and the decision profile matrixes were obtained. Finally, these decision profile matrixes were implemented by multi-feature fusion method. The experimental results on Iris data show that the forecast accuracy of the multi-feature fusion method based on SVMKNN classifiers increases by 28.7% and 1.9% than those of SVM and SVM-KNN methods respectively.

Method of constructing phylogenetic tree based on fuzzy clustering

2009, 29(3): 836-838.

Asbtract ( )

PDF (730KB) ( )

Related Articles | Metrics

The evolutionary history of various species can be disscussed by constructing the phylogeny trees. Therefore, many scientists focus on the research of the phylogeny trees. In this paper, the authors used the similarity matrix computed by 4D representation of DNA sequence and regarded it as a fuzzy matrix, then used the maximal tree method to construct the phylogeny tree. This method need not sequence alignment and the computation is simple. The experiments demonstrate its validity.

Splitting attribute selection method based on cost performance

刘星毅 LIU Xing-Yi

2009, 29(3): 839-842.

Asbtract ( )

PDF (746KB) ( )

Related Articles | Metrics

Cost-sensitive decision trees usually concern the discussion of the test cost and misclassification cost. During the classification process, splitting attribute selection is the most important. The paper analyzed the disadvantages and the advantages of the existing methods and proposed a novel method that combined the information ratio in information theory with the cost including the test cost and the misclassification cost to select the split attributes. The experimental results show that this method outperforms significantly the existing methods.

Method for Chinese short text classification based on feature extension

2009, 29(3): 843-845.

Asbtract ( )

PDF (614KB) ( )

Related Articles | Metrics

In this paper, based on the characteristics that short texts describe weak signals，a method based on feature extension (STCFE) was introduced to classify Chinese short texts. In this method, the correlation rules between feature items of training set and testing set were mined by FPGrowth algorithm, and then those rules were applied to extend the features of the testing set. Meanwhile, to classify Chinese short texts effectively, semantic information was introduced and the DEF term formula of words was improved in HowNet. Experimental results show that the proposed method performs well, and its Micro-F1 and Macro-F1 are higher than those of conventional approaches.

Automatic acquisition of customer reviews

Xing WU

2009, 29(3): 846-848.

Asbtract ( )

PDF (774KB) ( )

Related Articles | Metrics

In this article, the authors labeled the consumer reviews got from the Web, and extracted the features of customer reviews content, selected the features of different reviews with CHI approach. Support Vector Machine (SVM) was used to get the classification, then to classify the updated consumer reviews and mine the useful reviews. The experiment shows that this way has high recall and precision.

Query expansion based on context of document and search result

Jiang Hui

2009, 29(3): 852-853.

Asbtract ( )

PDF (419KB) ( )

Related Articles | Metrics

When editing a word processing document, we may search the Web by using a term in the document as an initial query and then modifying the query by adding keywords extracted from the text surrounding the search term. There are query expansion methods which use the text surrounding the search term in the initial result to weight candidate keywords in the source document to modify the query. However, this approach may lead to topic drift. To solve the problem, the initial results were filtered first and only the results containing similar contexts as that in the source document were selected to help choosing additional keywords. Experiments show that this method can get more appropriate additional keywords than other methods.

Extended algorithm for XML query reformulation

Li Rui Kui-Gui Wu

2009, 29(3): 854-857.

Asbtract ( )

PDF (738KB) ( )

Related Articles | Metrics

Query reformulation is a key issue of the data integration, and it automatically rewrites the users query request to the data source request directly. Recently the University of Michigan and the IBM AImaden Research Center presented a novel algorithm for constraintbased XML query reformulation, but it did not consider the problem of complex schema matching and limited the algorithm application. Based on the original query reformulation algorithm, an extended algorithm for XML query reformulation was presented, expanding the application scope of the original algorithm. The correctness and time complexity of the extended algorithm were also given.

Research on RBFNNaided adaptive UKF algorithm

Wen-Qiang GUO Zhi-Guang QIN

2009, 29(3): 858-861.

Asbtract ( )

PDF (687KB) ( )

Related Articles | Metrics

It is well known that the Kalman filter can be adopted to make unbiased estimation for system state with the measurement of the noise interference. However, neither the EKF algorithm nor the UKF algorithm can avoid filtering divergence. In terms of the adaptability of RBF neural network (RBFNN), this paper proposed a RBFNN-based algorithm to correct the output of the Kalman filter and further to avoid the output divergence. Both the simulation and the application results show that the output divergence can be effectively avoided by the presented filtering algorithm.

Study on samples regression based on priority ordered neural network

2009, 29(3): 862-864.

Asbtract ( )

PDF (557KB) ( )

Related Articles | Metrics

Samples regression is an important issue in the field of Dynamic Constructive Neural Network (DCNN). From the view of the mapping values of neural network, input side and output side training samples were separated into different sub-sets according to the algorithm based on Priority Ordered Neural Network (PONN), which was a more similar model of humanbeing and widely used in different fields because of its priority neurons in neural network. Then, subsets were constructed into PONN using a parallel algorithm which accelerated the whole training process. Lastly, the validity of this proposed method was tested on the different types of dataset. The results indicate the proposed method can effectively reduce the regression error.

Ant colony algorithm based on pheromone intensity

2009, 29(3): 865-867.

Asbtract ( )

PDF (583KB) ( )

Related Articles | Metrics

When selecting the path, all the existing ant colony algorithms take account of the information of pheromone and distance. In order to simulate the real ant better, this paper proposed a new ant colony algorithm based on the intensity of pheromone, which only depended on the intensity of pheromone when selecting the path. The path length (or distance) was used when initializing and updating the pheromone. A strategy of dynamic pheromone update was given. The experiments demonstrate that the proposed algorithm can obtain good searching results, and the speed of operation is more than five times faster than the existing ant colony algorithms.

Ensemble learning application in improvement of particle swarm optimization algorithm

Shi Yan 施彦

2009, 29(3): 868-870.

Asbtract ( )

PDF (678KB) ( )

Related Articles | Metrics

Ensemble learning is a methodology that combines multiple individual learners, and can get better results. In this paper, ensemble learning was applied to improve the performance of Particle Swarm Optimization (PSO) algorithm in three aspects: on particle level or dimension level; in the whole iteration process or periodical iteration process; on single population or multiple subpopulations. The experimental results of three typical function optimization problems show that PSO algorithm based on ensemble learning can improve performance compared with the standard PSO and some improved PSO algorithms.

CMAC application using triangulation in reinforcement learning

Fang-Yi SUN

2009, 29(3): 871-873.

Asbtract ( )

PDF (623KB) ( )

Related Articles | Metrics

A reinforcement learning controller based on CMAC neural networks using triangulation was studied and applied to the learning control of intercepting a ball in the RoboCup. By utilizing Kuhn triangulation based on simplex interpolation in the continuous state space of Markov Decision Processes (MDPs), the value functions of MDPs were approximated with linear triangulation so that the generalization ability of the CMACbased reinforcement learning controller could be improved. Simulation results on the learning control of intercepting a ball show that the CMACbased learning controller using triangulation is much more efficient than the learning controller based on CMAC uniform coding.

Research of subjective trust model based on credibility theory

Xiao-Bin WANG

2009, 29(3): 874-876.

Asbtract ( )

PDF (620KB) ( )

Related Articles | Metrics

Through discussing the evaluation methods of subjective trust in open network, the fuzzy quality of subjective trust was investigated by introducing credibility theory, and the quantitative evaluation approach of subjective trust was presented based on the criteria of maximizing expected value. Then under the situation that the key attributes of subjects and the weight of evaluation were characterized as fuzzy variables, a simulating method was given to calculate the expected value of subjective trust. Finally, the presented approach was proved to be feasible and effective by an example.

Generalized least squares support-vector-machine algorithm and its application

2009, 29(3): 877-879.

Asbtract ( )

PDF (475KB) ( )

Related Articles | Metrics

Least Squares Support-Vector-Machines (LS-SVM) algorithm is an efficient project about pattern classification on unclassifiable sample set condition. While dealing with many factual pattern classification problems, this algorithm reflects certain limitation. A generalized LS-SVM algorithm was introduced to further improve the applicability of LS-SVM. This new method was applied to radar range profile's recognition. The experimental results show that this new method can achieve better recognition effect.

Parallel algorithm for matrix multiplication based on De Bruijn network

Zhao-Quan CAI

2009, 29(3): 880-883.

Asbtract ( )

PDF (1249KB) ( )

Related Articles | Metrics

It is necessary to develop parallel matrix multiplication because of serial matrix multiplication costing much memory and time. Many different parallel matrix multiplications have been developed according to different network topologies. De Bruijn network is one of the most attractive network models, which has good parallel computing performance. It is simple to execute parallel matrix multiplication on De Bruijn network and easy to implement. In this paper, De Bruijn network was introduced, and then a parallel algorithm for matrix multiplication based on De Bruijn network was proposed. Its speedup ratio, efficiency and scalability were analyzed. It is proved by experiments and comparison that this algorithm is equal to Cannons in terms of time complexity.

Research and realization of casualty evacuation model in emergency management system

Fei-hu HU Hui-min CHEN Lin-yan SUN

2009, 29(3): 884-887.

Asbtract ( )

PDF (1019KB) ( )

Related Articles | Metrics

According to the urgent characteristics of casualty evacuation, this paper established a casualty evacuation mathematical model of multidisaster places, multihospital and multipatients with the goal of the shortest time. Considering the hospital resource competition of patients, Operation Table Method (OTM) was used to solve the model and realize the goal of the shortest time. The method has been realized on the provincial level information demonstration platform of emergency management, and can deal with different injury types and the patient priorities. With this platform and Geographic Information System (GIS) technology, the method is simulated and the effect is good.

View-dependent greenhouse scene modeling and interactive walkthrough

WEI wei

2009, 29(3): 888-391.

Asbtract ( )

PDF (1394KB) ( )

Related Articles | Metrics

A method for realtime and realistic rendering virtual agricultural scenes that include group of solar greenhouse was proposed in this paper. This method used viewdependent continuous LOD models to reduce the amount of polygons needed to be rendered actually, and a collision detection algorithm was implemented based on bounding box. The rendering speed of the scene was improved by using visibility culling techniques. In the constructed virtual agricultural scenes, large-scale plants stands were created rapidly by using geometric transformation, while shadow volume algorithm was implemented to render shadow in the scene. These significantly improved the realism of virtual agricultural scene. The simulation results demonstrate the proposed method can reduce the number of rendered elements effectively, and improves the rendering speed with high realism. This can meet the needs for realtime and interactive walkthrough in large scale virtual scenes.

Semantic Web-based context-aware service selection in task-computing

HUANG Run-Cai

2009, 29(3): 892-895.

Asbtract ( )

PDF (823KB) ( )

Related Articles | Metrics

The hierarchy of task-computing and the decomposition strategy of complex task were analyzed according to the characteristics of taskcomputing in pervasive environment. The architecture relied on the semantic representation of service and content based on shared ontology. It could not only improve the accuracy and recall, but also realize the functions such as knowledge sharing, capabilitybased search, autonomous reasoning and semantic matchmaking. Thus, a dynamic context-aware service selection mechanism based on semantic was proposed to filter and rank the matching services according to their dynamic context in order to optimize the discovery process and save the time and energy of users. In addition, the selection mechanism could be used to discover the useful static and dynamic context information to provide users with the most suitable and relevant services.

DSS for E-government project approval based on ontology and CBR

Liu Xiao-Wen Ke-Jin Hu

2009, 29(3): 896-899.

Asbtract ( )

PDF (933KB) ( )

Related Articles | Metrics

The advantages of integrating ontology and CaseBased Reasoning (CBR) for developing Egovernment project Decision Support System (DSS) were analyzed. The framework and decision support process of Egovernment project approval system were presented by focusing on representation method of Egovernment project cases and how to measure the similarity of those cases after the ontology theory imported into CBR DSS. The experimental results show that the method, to extract and express the knowledge of project cases by ontology and to organize and index the cases by these knowledge, effectively improves the precision ratio and recall ratio of cases search.

Evolutionary design of combinational circuit based on game genetic algorithm

徐海芹 Xu Hai-Qin

2009, 29(3): 900-902.

Asbtract ( )

PDF (532KB) ( )

Related Articles | Metrics

To improve the speed and efficiency of combinational logic circuit design, this paper presented a Game Genetic Algorithm (GGA). In GGA, each output of the circuit was regarded as a player. The connection and combination of logic gates of each output were regarded as strategies. GGA transformed the circuit design problem to an optimal game decisionmaking one. The choice of policy was realized by genetic algorithm, and then the game model was built. Simulation results show that the method is effective.

Anti-spam model based on semi-Naive Bayesian classification model

Bei Hui

2009, 29(3): 903-904.

Asbtract ( )

PDF (539KB) ( )

Related Articles | Metrics

Because Naive Bayes (NB) classification model is simple and effective, good efficiency can be achieved in antispam applications. On the other hand, the assumption of its attribute independence makes it unable to express its semantic dependence. This paper proposed a new antispam classification model based on semi-NB classification model, averaged on N one-dependence classification model. It relaxed the assumption of condition independence of each attribute. It was assumed that all attributes were dependent on one attribute (1-dependence). The average on N 1-dependence was regarded as the probability of each class label. This method is simple and efficient and decreases the classification error ratio.

Interbank cash concentration model based on absolute minimum fork tree

JI Mo-Li

2009, 29(3): 905-907.

Asbtract ( )

PDF (541KB) ( )

Related Articles | Metrics

Cash concentration is the cash flow that goes from the subordinate accounts in different banks to the head office account. Corporations adopt interbank cash concentration to improve the capital management efficiency as a whole. But the handling charges throughout the interbank transferring abruptly increase because of the frequent cash collections. That results in corporate profit deduction. This paper introduced an algorithm named absolute minimum fork tree and designed an interbank cash concentration model. In conclusion, due to models availability study the cost could be reduced efficiently.

Interaction model of community mining and topic detection and tracking

Xiao peng Tao

2009, 29(3): 908-911.

Asbtract ( )

PDF (755KB) ( )

Related Articles | Metrics

Community mining is an important application in the field of Web information mining. Topic detection and tracking is an important application in the field of text information study. Currently these two technologies are studied separately. To better apply these two technologies to complicated social networks formed by Internet, this paper combined them for research, discovered the relationships of community and topic, created static and dynamic interaction models for community mining and topic detection and tracking, and designed algorithms to mine communities, detect topics and track communities.

Coherence parallel algorithm of seismic data processing based on CUDA

LI KENLI

2009, 29(3): 912-914.

Asbtract ( )

PDF (913KB) ( )

Related Articles | Metrics

In seismic exploration interpretation, the application of coherent technology can clearly identify faults and stratigraphy, but the traditional calculation method cannot meet the need of the coherence body calculation from 3D seismic data. Based on CUDA (Compute Unified Device Architecture) platform, a coherence parallel algorithm was proposed. It could accelerate the speed of matrix multiplication with the performance of GPU cluster. Extensive experiments have been conducted in a PC with Intel Core2Due CPU and NVIDIA GeForce 8800 GT graphic card, and the results prove the efficiency of the proposed algorithm. Even though the actual speedups in production codes will vary with the particular problem, the results obtained here indicate that GPU can potentially be a very useful platform for processing large-scale seismic data.

Table of Content