Search

Copyright of Journal of Jilin University (Science Edition) / Jilin Daxue Xuebao (Lixue Ban) is the property of Zhongguo Xue shu qi Kan (Guang Pan Ban) Dian zi Za zhi She and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Aiming at the problem that existing image text cross-modal retrieval is difficult to fully exploit the local consistency in the mode in the common subspace, a cross-modal retrieval method based on improved graph convolution is proposed. In order to improve the local consistency within each mode, the modal diagram is constructed with a single sample as a node, fully mining the interactive information between features. In order to solve the problem that graph convolution network can only do shallow learning, the method of adding initial residual link and weight identity map in each layer of graph convolution is adopted to alleviate this phenomenon. In order to jointly update the central node features through higher-order and lower-order neighbor information, an improvement is proposed to reduce neighbor nodes and increase the number of layers in graph convolution network. In order to learn highly locally consistent and semantically consistent public representation, it shares the weights of common representation learning layer, and jointly optimizes the semantic constraints within the modes and the modal invariant constraints between modes in the common subspace. The experimental results show that on the two cross-modal data sets of Wikipedia and Pascal sentence, the average mAP values of different retrieval tasks are 2.2%~42.1% and 3.0%~54.0% higher than the 11 existing methods. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Jilin University (Science Edition) / Jilin Daxue Xuebao (Lixue Ban) is the property of Zhongguo Xue shu qi Kan (Guang Pan Ban) Dian zi Za zhi She and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Currently, there are two main limitations in single-label text classification tasks based on hierarchical graph attention networks. First, it cannot effectively extract text features. Second, there are few studies that highlight text features through the connection between text and labels. To address these two issues, a hierarchical graph attention network text classification model that integrates label information is proposed. The model constructs an adjacency matrix based on the relevance between sentence keywords and topics, and then uses word-level graph attention network to obtain vector representations of sentences. This method is based on randomly initialized target vectors and utilizes maximum pooling to extract specific target vectors for sentences, making the obtained sentence vectors have more obvious category features. After the word-level graph attention layer, a sentence-level graph attention network is used to obtain new text representations with word weight information, and pooling layers are used to obtain feature information for the text. On the other hand, GloVe pre-trained word vectors are used to initialize vector representations for all text labels, which are then interacted and fused with the feature information of the text to reduce the loss of original features, obtaining feature representations that are distinct from different texts. Experimental results on five public datasets (R52, R8, 20NG, Ohsumed, and MR) show that the classification accuracy of the model significantly exceeds other mainstream baseline models. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In order to improve the operational capability of automatic train supervision (ATS), reduce the software and hardware cost, a dispatch path search method is proposed to linked devices such as platforms and signals. Based on the station map data and topology theory, the proposed method acquires the connection relationship of the platform in a semi-automatic or fully automatic manner, and the connection data of the platform is stored in the data structure of the adjacent matrix. The path search rule is designed, in which both depth-first search (DFS) and breadth-first search(BFS) are supported to obtain the path group. Automatic or manual methods to meet the project requirements are also proposed. Take the platform as an application case, the method is adapted to train dispatcher and route trigger function related to paths of linked devices. The path group can be reloaded by ATS subsystem, which can improve the ATS performance effectively. The method has been applied to Changsha urban rail transit Line 3 and Line 4, Wuxi urban rail transit Line 4, to support plan editing and train scheduling, and achieves good results, such as the generation time of station map can reduce over 70%,incidental jamming of the route trigger is eliminated when there are more trains,and route files can be generated automatically within minutes (original manual generation took several days).

Copyright of Journal of South China University of Technology (Natural Science Edition) is the property of South China University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In order to solve the problem that the traditional sensitive information early warning algorithm is difficult to cluster information effectively, a new sensitive information early warning algorithm is proposed. Firstly, the sensitive information of high word frequency in the network was collected. Secondly, the accurate mining of high sensitive information was completed by setting a reasonable threshold. Thirdly, the high sensitive information was effectively clustered by deep neural network, and the adjacency matrix was constructed. Finally, the connection average of network parameter nodes was calculated. Combined with the above clustering results, the monitoring and early warning of high sensitive information were realized. The experimental results showed that the data sent by the network nodes using the traditional algorithm was not regular, and the warning response time was more than 2 s, while the network data using the deep neural network algorithm presents a fully connected state, and the sensitive information could show a high degree of consistency, and the warning response of the algorithm to the sensitive information could be completed in 1 s. The research shows that the early warning algorithm based on deep neural network can be applied in the field of network security.

Copyright of Journal of Donghua University (Natural Science Edition) is the property of Journal of Donghua University (Natural Science) Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Central China Normal University is the property of Huazhong Normal University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In order to discuss the minimum eigenvalue of adjacency matrix in the class of complementary graphs of the tricyclic graph with a given order of n and n－4 pendent vertexes, the unique graph whose minimum eigenvalue reaches the minimum is characterized. Based on the simple undirected connected graph,the minimum eigenvalue of the graph is studied from the structure of the complement graph, and the minimum eigenvalue of the adjacency matrix in the complement graph class of the tricyclic graph with a given order of n and n－4 pendent vertexes reaches the minimum unique graph when the value is λ( G ( 85;P](n-4)/2,(n-4)/2[XC符号2.eps;%90%90;P])C)[WTBZ]. The result shows that the associative graph adjacency matrix is a matrix which represents the adjacency between vertices, and its minimum eigenvalue is the minimum eigenvalue of graph, which can describe the essential properties of graph well. The conclusion from this research shows that the minimum eigenvalue of the complement graph of the tricyclic graph with a given order of n and n－4 pendent vertexes reaches the minimum eigenvalue, which provides certain reference for further study of the minimum eigenvalue of the adjacency matrix in the complement graph class.

There are many types of planetary gear trains， and each one corresponds to a transmission ratio calculation formula， which makes great difficulties in the comprehensive research of planetary gear trains and the development of general design software. According to the structural diagram of the planetary gear train， the corresponding adjacency matrix is written. Then， the basic loop of the train is searched from the adjacency matrix to construct the basic loop matrix. Finally， the basic loop matrix is transformed into the coefficient matrix of the equation of motion. And use MATLAB to automate this algorithm and solve the equation of motion to obtain the gear ratio of the planetary gear train.

In order to solve the problem that the traditional sensitive information early warning algorithm is difficult to cluster information effectively. a new sensitive information early warning algorithm is proposed. Firstly the sensitive information of high word frequency in the network was collected. Secondly, the accurate mining of high sensitive information was completed by set ting a reasonable threshold. Thirdly, the high sensitive information was effectively clustered by deep neural network and the adjacency matrix was constructed. Finally, the connection average of network parameter nodes was calculated. Combined with the above clustering on the monitoring and early warning of high sensitive information were realized. The experimental results showed that the data sent by the network nodes using the traditional algorithm was not regular. and the warning response time was more than 2 s. while the network dat using the deep neural network algorithm presents fully connected state. and the sensitive information could show a high degree of consistency and the warning response of the algorithm to the sensitive information could be completed in The research show that the early warning algorithm based on deep neural network can be applied in the field of network security. [ABSTRACT FROM AUTHOR]

Using appropriate compression techniques to compactly and accurately represent and store graph data with hundreds of millions of nodes and edges is a prerequisite for the analysis and operation of large-scale graph data. Compact graph data representation not only reduces the storage space of graph data, but also supports efficient operation on graph data. This paper summarizes the research progress of graph data compression technologies in graph data management from the storage point of graph data, and focuses on the following three compression technologies: compression technology based on adjacency matrix, compression technology based on adjacency list, and compression technology based on formal method. Their related representative algorithms, application scopes, advantages and disadvantages are discussed. Finally, the current situation and problems of graph data compression technologies are summarized, and the development trend of future graph data compression technologies is given. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Hebei University of Science & Technology is the property of Hebei University of Science & Technology, Journal of Hebei University of Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Control Theory & Applications / Kongzhi Lilun Yu Yinyong is the property of Editorial Department of Control Theory & Applications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The inertia of a graph G is defined to be the triple In(G) = {i+ (G), i- (G)，i0 (G)}，where i+ (G)，i-(G),i0 (G) are the numbers of positive,negative and zero eigenvalues of the adjacency matrix A(G) including multiplicities, respectively. Some bounds for the positive index of inertia of a graph under some graph operations including adding a vertex or an edge, subdivision of an edge, contracting of two vertices and the join of graphs, are obtained.

In order to make it convenient for computer automatic synthesize,the automatic type synthesis of planar kinematic chains with modified Assur-group adjacency matrix is proposed. First of all,aiming at the known amount of the components,the active ones and the Assur-group,the diagonal elements with the different combination modes is solved according to the Assur-group theory. Then the synthesis method and concrete process is summarized and the selected Assur-group adjacency matrix is defined. The kinematic chains are subdivided in terms of 3 conditions of pole number,active pole and Assur-group type,meanwhile the isomorphism is estimated. As a conclusion,the method has better regularity,high efficiency and is beneficial to eliminate the rigid chains and reduce the calculation workload of isomorphism identification. Especially,the method can be applied to the automatic type synthesis for the planar mechanism kinematic chains with multiple links and multiple degrees of freedom. In addition,the success is achieved in the application on the type synthesis for the planar closed kinematic chains with 12 links and 12-below-links. Finally,the type synthesis method is demonstrated with the example of kinematic chains with 6 links and single degree of freedom.

The isomorphism identification of structural type is a vital problem now. For dealing with the isomorphism generated in synthesis of planar configuration based on the synthesis method of Assur- group adjacency matrix,a methodology of Assur- group adjacency matrix isomorphism identification is proposed. First of all,according to the diagonal elements is derived by the amount of Assur- group and active number,the diagonal isomorphism identification is obtained. Then,the condition is classified into three kinds,first is sequence isomorphism when Assur- group is equipped,second is symmetrical configuration isomorphism after adding Assur- group and the last one is discovered in matrix synthesis,which would be recognized and deleted by this method. Finally,the methodology proposed is able to be achieved by computer,easy to control and takes high efficiency in isomorphism identification.

This paper proposes a searchable encryption model for graph data to address the problem that most of the current searchable encryption schemes cannot deal with complex data structures. The algorithms and security models of the scheme are described. Taking the encryption index of matrix structure into account， the detailed description of the searchable encryption scheme for graph data that supports adjacency queries is presented. The scheme utilizes pseudo-random functions and pseudo-random permutations to protect the privacy of the graph data and the encrypted index. Real models and ideal models test the security of the scheme. Compared with traditional searchable encryption schemes， the proposed scheme has more flexible searching ability and higher efficiency， which promotes broader application under big data environment. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Harbin Institute of Technology. Social Sciences Edition / Haerbin Gongye Daxue Xuebao. Shehui Kexue Ban is the property of Harbin Institute of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Due to the rapid development of modern network data, the existing algorithms for generating association rules can bring in a number of redundant rules which are far greater than the actual value of the rules. The redundant rules not only affect user analysis, but also reduce the utilization of the association rules. In order to eliminate the redundant rules, we propose a method for removing redundant association rules of business data based on the first order predicate formula, which uses the first order predicate formula to represent the association rules through the conversion of the equivalent formula. And the predicate formula is converted to the adjacency matrix by using the algorithm and the matrix equivalence, and the redundant association rules are deleted. Experimental raw data is from the UCI data set, the association rules are generated by Weka, and then the redundant rules are removed by Java and MATLAB. [ABSTRACT FROM AUTHOR]

Copyright of China Sciencepaper is the property of China Sciencepaper and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Aiming at the disadvantages in dynamics and algorithm time complexity in the existing web service composition methods, we propose an adjacency matrix-based web service composition approach. Sequential and parallel relationship among web services is represented by adjacency matrix. Abstract services are created by clustering，and the composition relationship among abstract services is established by domain experts. The Warshall algorithm is utilized to calculate the transitive closure of the adjacency matrix so that we can judge whether the the service request is met or not. Meanwhile the dynamic service composition flow is constructed. The proposed method applies well in service composition due to its simple operation and the O(n³) time complexity of the Warshall algorithm. [ABSTRACT FROM AUTHOR]

A total-coloring is called vertex-distinguishing (VDTC for short) if all the color sets are different. The minimum number of used colors is called the vertex distinguishing total chromatic number. This paper proposed an algorithm that looks for the conflict set, then used some rules to exchange step by step until the value of the objective function meets the requirements, when the coloring was accomplished. The experimental results show that the algorithm can get the vertex distinguishing total chromatic number of random graphs whose vertices number is given effectively, and the time complexity is less than 0 (n³). [ABSTRACT FROM AUTHOR]

The market basket analysis based on association rules is a powerful tool for physical and online retailers to stimulate consumption and increase profits via improving strategies on category management, store layout, associated promotions and recommendations. According to this demand, we propose AC algorism to get association rules and M-AC algorism which is used to obtain frequent k-itemset based on adjacency matrix and cut matrix technology. This method possesses such merits as simple and efficient operation, low cost, and memory property (meaning that there is no need to operate again when data updates). The effectiveness of this new method is proved by its successful application in a retail chains. [ABSTRACT FROM AUTHOR]