����Ŀ¼

15 October 2019, Volume 5 Issue 5
Comprehensive Review
Summary of text classification methods
You YU, Yu FU, Xiaoping WU
2019, 5(5):  1-8.  doi:10.11959/j.issn.2096-109x.2019045
Asbtract ( 576 )   HTML ( 62)   PDF (687KB) ( 965 )
Figures and Tables | References | Related Articles | Metrics

How to effectively classify text has become a hot topic.Firstly,the concept of text classification,word segmentation,feature extraction and text classification methods were introduced,and the research actuality was summarized.And then the challenges of text classification related technologies were analyzed.Finally,the development trend of text classification was summarized.

Special Column: Routing Technology in Complex Network Environment
Spreading dynamics based key nodes identification in inter-domain routing system
Huihu ZHU, Han QIU, Junhu ZHU, Ziyi ZENG
2019, 5(5):  9-20.  doi:10.11959/j.issn.2096-109x.2019046
Asbtract ( 307 )   HTML ( 10)   PDF (1978KB) ( 147 )
Figures and Tables | References | Related Articles | Metrics

The inter-domain routing system is a critical infrastructure of the Internet and it is of great significance to protect the key nodes of inter-domain routing system.The key nodes identified by the existing methods can not reflect the importance of the nodes on the cascading failure process.The method of key nodes identification is proposed basing on spreading dynamics.A node importance evaluation model based on DDF-CFM model is proposed.,which could takes the failure effect caused by load redistribution and UPDATE messages propagation into account after node fails.The experiments turn out that the accuracy of this method is at least 7.3% higher than that of exist ing methods.And the experimental results show that in the scale of 10000 nodes,the failure of only 5 key nodes will lead to large-scale cascade failure of inter-domain routing systems.

QoS routing algorithm based on multiple domain architecture of SDN
Wei HUANG, Ran LU, Cuncai LIU, Sibo QI
2019, 5(5):  21-31.  doi:10.11959/j.issn.2096-109x.2019047
Asbtract ( 289 )   HTML ( 8)   PDF (1634KB) ( 207 )
Figures and Tables | References | Related Articles | Metrics

Traditional distributed network architecture constraints the innovation of routing algorithm.Software-defined network (SDN) provides a new solution for the optimization of routing algorithm.Previous researches show that the quality of service (QoS) routing issues are based on heuristic algorithm mostly,but these methods cannot be applied in large networks due to their high computing complexity.However,other algorithms have a lot of problems,which are high complexity or poor QoS performance,such as shortest path algorithm.This paper proposes A new QoS routing algorithm:LC-LD routing algorithm was proposed.LC-LD was based on SDN west-east interface and binds both delay constraint and cost constraint.keeping a good balance between computational complexity and algorithm performance.Finally,the simulation results show that LC-LD can possess both low computational complexity and high QoS routing performance.

Fermat point based connectivity restoration strategy in networks
Zhaobin ZHOU, Hongyan ZHANG, Xiaoding WANG
2019, 5(5):  32-38.  doi:10.11959/j.issn.2096-109x.2019048
Asbtract ( 275 )   HTML ( 3)   PDF (1049KB) ( 87 )
Figures and Tables | References | Related Articles | Metrics

The connectivity restoration ensures the availability and reliability of a network.Both of geometrical features and topological structures should be taken into consideration at the same time,without which previous works can hardly restore the connectivity with the least number of relay nodes.The Fermat point,the triangulation and the minimum spanning tree are integrated with the design of an efficient restoration strategy.The theoretical analysis indicate that the approximation ratio of the proposed strategy is $\frac{3\sqrt{3}}{4-\sqrt{3}}$and the complexity of which is O(n log n).Simulation results show that the proposed strategy outperforms other strategies in the number of relay nodes required.

New method for link prediction in directed networks based on triad patterns
Sheng CHANG, Hong MA, Shuxin LIU
2019, 5(5):  39-47.  doi:10.11959/j.issn.2096-109x.2019049
Asbtract ( 337 )   HTML ( 4)   PDF (1074KB) ( 134 )
Figures and Tables | References | Related Articles | Metrics

Almost all current studies on link prediction problem focus on undirected networks.Unfortunately,many complex networks in the real world are directed.Ignoring the direction of a link will overlook some important information or even make the prediction meaningless.Directly applying the methods for undirected networks to directed networks will reduce the accuracy of prediction.A new method for link prediction in directed networks based on triad patterns was proposed.The proposed metric compare the difference of triad structures between undirected and directed networks and use potential theory to filter the triad patterns.By statistics of triad closeness in various networks,new method calculate the similarity between nodes using the triad closeness index of a network as the weight for different triad patterns.Experiments on nine real networks show that accuracy of proposed method is 4.3% better than benchmark methods.

Frequent path discovery algorithm for financial network
Fang LYU, Fenghe TANG, Junheng HUANG, Bailing WANG
2019, 5(5):  48-55.  doi:10.11959/j.issn.2096-109x.2019050
Asbtract ( 1141 )   HTML ( 30)   PDF (890KB) ( 630 )
Figures and Tables | References | Related Articles | Metrics

With the proliferation of various illegal financial activities,more and more attention is paid to the research of finding criminal cues in financial network by scholars.The characteristics of the transaction data generated by bank accounts are analyzed in detail,and a general model of bank account transaction network is established.On this basis,a two-direction active edge searching method is proposed to solve the problem of evaluating the relationship strength between financial entities.And then,a breadth-first frequent path discovery algorithm with depth controlled is presented,with which the way how the financial flows is restored.Experiment results on the real bank data show that the above two methods are effective in solving the problem of peer prediction and financial tracking respectively.

Papers
Software patch comparison technology through semantic analysis on function
Yan CAO, Long LIU, Yu WANG, Qingxian WANG
2019, 5(5):  56-63.  doi:10.11959/j.issn.2096-109x.2019051
Asbtract ( 276 )   HTML ( 6)   PDF (1282KB) ( 217 )
Figures and Tables | References | Related Articles | Metrics

Patch comparison provides support for software vulnerability,and structural comparison has been developed.Based on summarizing binary files comparison and anti-comparison methods,comparison technology was proposed by semantic analysis on function to address the issue that structural comparison cannot carry on semantic analysis.Through traditional structural comparison,syntax differences in function-level were analyzed to find the maximum common subgraph.Then,the path cluster was built between the input and output of the function depend on program dependency analysis.Function output characteristics was established based on symbolic execution.Semantic differences of functions were compared by functional summaries.Based on the maximum isomorphic subgraph,the matched functions which there are possible semantic changes between was further analyzed.Ultimately,the experimental results showed the feasibility and advantages of the proposed method.

Video tampering detection algorithm based on spatial constraint and gradient structure information
Han PU, Tianqiang HUANG, Bin WENG, Hui XIAO, Wei HUANG
2019, 5(5):  64-79.  doi:10.11959/j.issn.2096-109x.2019052
Asbtract ( 229 )   HTML ( 7)   PDF (10915KB) ( 87 )
Figures and Tables | References | Related Articles | Metrics

The traditional video passive forensics method using only the principle of similarity between adjacent frames will cause a lot of false detection for the video with severe motion.Aiming at this problem,a video tamper detection method combining spatial constraints and gradient structure information was proposed.Firstly,the low motion region and the high texture region were extracted by using spatial constraint criteria.The two regions were merged to obtain the robust quantitative correlation rich regions for extracting video optimal similarity features.Then improving the extraction and description methods of the original features,and using the similarity of the gradient structure in accordance with the characteristics of the human visual system to calculate the spatial constraint correlation value.Finally,the tampering points were located by the Chebyshev inequality.Experiments show that the proposed algorithm has lower false detection rate and higher accuracy.

Reversible data hiding algorithm based on asymmetric histogram shifting
Yufen HE, Zhaoxia YIN, Jin TANG, Lei LIU, Shilei HUANG
2019, 5(5):  80-89.  doi:10.11959/j.issn.2096-109x.2019053
Asbtract ( 228 )   HTML ( 9)   PDF (2471KB) ( 129 )
Figures and Tables | References | Related Articles | Metrics

The shifting of two asymmetric histograms in opposite directions in data embedding respectively had produced the pixel compensation and restore effect,a better reversible data hiding algorithm based on pixel prediction was proposed,two asymmetric histograms of prediction error were generated on the more right and the more left side of zero value,when they were shifed in the second data embedding stage,more pixels would be restored to the original image pixel value to reduce image distortion and improve the image quality.Compared with the traditional algorithm,it reduces the amount of pixels involved in the histogram shifting and protects the quality of secret image.

Trusted platform for third-party cloud computing online evaluation and analysis technology
Bo ZHAO, Xiang LI, Fei YAN, Liqiang ZHANG, Huanguo ZHANG
2019, 5(5):  90-104.  doi:10.11959/j.issn.2096-109x.2019054
Asbtract ( 402 )   HTML ( 10)   PDF (4317KB) ( 187 )
Figures and Tables | References | Related Articles | Metrics

The service model of cloud computing creates a serious threat to user data security,the trustworthiness of cloud computing through third parties,must be evaluated then the healthy development of the cloud computing industry can be guaranteed.The research of the National High Technology Research and Development Program of China,“trusted platform for third-party cloud computing online evaluation and analysis technology” were summarized.The trusted evaluation model and architecture were described from the cloud platform,the dynamic evaluation technology,the credible evidence collection and the trusted test mechanism,the credibility of the quantitative evaluation method,and the prototype system integrated by tools ,to evaluate the credibility of cloud platform.

Research on system log anomaly detection based on deep learning
Yidong WANG, Peishun LIU, Gbin WAN
2019, 5(5):  105-118.  doi:10.11959/j.issn.2096-109x.2019055
Asbtract ( 680 )   HTML ( 44)   PDF (2296KB) ( 404 )
Figures and Tables | References | Related Articles | Metrics

The system log reflects the running status of the system and records the activity information of specific events in the system.Therefore,the rapid and accurate detection of the system abnormal log is important to the security and stability of the system.A log anomaly detection algorithm based on GRU neural network is proposed.Log parsing is implemented based on log key technology.Log anomaly detection is realized by using anomaly detection model of execution path and anomaly detection model of parameter value.The system has the advantages of less parameters and faster training.It improves the running speed while achieving higher detection accuracy,and is suitable for log analysis of large information systems.

Bimonthly, started in 2015
Authorized by:Ministry of Industry and Information Technology of the People's Republic of China
Co-sponsored by:Xidian University, Beihang University, Huazhong University of Science and Technology, Zhejiang University
Edited by:Editorial Board of Chinese Journal of Network and Information Security
Editor-in-Chief:FANG Bin-xing
Executive Editor-in-Chief:LI Feng-hua
Director:YI Dong-shan