检索结果-内蒙古大学图书馆

Research on Trust Prediction from a Sociological Perspective

Journal of Computer Science & Technology 2015年第4期30卷 843-858页

作者：王英王鑫左万利 College of Computer Science and Technology Jilin University Changchun 130012 China Key Laboratory of Symbolic Computation and Knowledge Engineering Ministry of Education Changchun 130012 China School of Computer Technology and Engineering Changehun Institute of Technology Changchun 130012 China

Trust, as a major part of human interactions, plays an important role in helping users collect reliable infor-mation and make decisions. However, in reality, user-specified trust relations are often very sparse and follow a power law distribution; hence inferring unknown trust relations attracts increasing attention in recent years. Social theories are frameworks of empirical evidence used to study and interpret social phenomena from a sociological perspective, while social networks reflect the correlations of users in real world; hence, making the principle, rules, ideas and methods of social theories into the analysis of social networks brings new opportunities for trust prediction. In this paper, we investigate how to exploit homophily and social status in trust prediction by modeling social theories. We first give several methods to compute homophily coe？cient and status coe？cient, then provide a principled way to model trust prediction mathe-matically, and propose a novel framework, hsTrust, which incorporates homophily theory and status theory. Experimental results on real-world datasets demonstrate the effectiveness of the proposed framework. Further experiments are conducted to understand the importance of homophily theory and status theory in trust prediction.

关键词： trust prediction homophily coefficient status coefficient social theory matrix factorization

来源：评论

学校读者我要写书评

暂无评论

Detecting communities from signed network based on local search

Detecting communities from signed network based on local sea...

引用

2016 2nd International Conference on Electronics,Network and Computer engineering(ICENCE 2016)

作者： Xueyan Liu Bo Yang Xuehua Zhao Yi Yang School of Computer Science and Technology Jilin University Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University School of Digital Media Shenzhen Institute of Information Technology China UnionPay

Many researchers have begun to study signed networks which are widely existed in real world. In the signed network, the links are labeled the positive or negative sign to represent the active or passive relation between individuals, such as trusted or distrusted relation in social networks. Communities mining is still a great challenge to the domain of signed networks because of negative links. Unlike communities of unsigned networks, positive links mainly occur in the communities and negative links tend to occur between the communities in the signed networks. Nowadays, many methods which are based on global search for signed network community have been raised, and most of these methods require the global information at each iteration. Besides, determining the number of communities is an important problem for current algorithm for the lack of priori knowledge. To address above problems, a novel community detection method based on local information, is proposed for signed networks in this paper. The proposed method mainly includes two steps. In the first step, the number of communities is determined in terms of the centrality of nodes. In the second step, the local objective function is optimized by the local information of nodes, so the global objective function can also be optimized indirectly. Finally, the communities in signed networks are efficiently found. To validate the proposed method, the comparisons are made with other methods in the synthetic and real signed networks. The experimental results indicate that communities in signed networks can be efficiently found by the proposed method.

关键词： signed networks community detection local search

来源：评论

学校读者我要写书评

暂无评论

Video face recognition based on the convex hull model of kernel subspace sample selection

引用

Journal of computational and Theoretical Nanoscience 2016年第2期13卷 1436-1441页

作者： Wang, Yu Shen, Xuan-Jing Chen, Hai-Peng College of Computer Science and Technology Jilin University Changchun130012 China Applied Technology College Jilin University Changchun130012 China Key Laboratory of Symbolic Computation and Knowledge Engineering Ministry of Education Jilin University Changchun13001 China

This paper proposes a novel video face recognition method based on the convex hull model of kernel subspace sample selection. This method treats each video as an image set. Each image is represented as a point in the feature vector space, then each image set is characterized by a convex hull spanning its feature points. The similarity between image sets can be measured using the distance of closest approach between their convex hulls. Before using the convex hull model, the kernel subspace sample selection was used to discard points that were far from the fitted model. The sample selection on the kernel subspace reduces the influence of confounding factors such as motion and gestures. Through experiments using the YouTube Face Database, the algorithm achieved high recognition accuracy, which verified the validity of the algorithm. © 2016 American Scientific Publishers All rights reserved.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Link prediction in social network based on local information and attributes of nodes

引用

Journal of Physics: Conference Series 2017年第1期887卷

作者： Yingying Liang Lan Huang Zhe Wang College of Computer Science and Technology Jilin University 130012 Changchun China Key Laboratory of Symbolic Computation and Knowledge Engineering (Jilin University) Ministry of Education 130012 Changchun China Zhuhai College of Jilin University 519041 Zhuhai China

Link prediction is essential to both research areas and practical applications. In order to make full use of information of the network, we proposed a new method to predict links in the social network. Firstly, we extracted topological information and attributes of nodes in the social network. Secondly, we integrated them into feature vectors. Finally, we used XGB classifier to predict links using feature vectors. Through expanding information source, experiments on a co-authorship network suggest that our method can improve the accuracy of link prediction significantly.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Otsu multi-thresholding based on DCG3A for medical image

引用

ICIC Express Letters, Part B: Applications 2016年第8期7卷 1651-1656页

作者： Fang, Zheng Shen, Xuanjing Chen, Haipeng Feng, Yuncong College of Computer Science and Technology Jilin University No. 2699 Qianjin Ave. Changchun130012 China Key Laboratory of Symbolic Computation and Knowledge Engineering Ministry of Education Jilin University No. 2699 Qianjin Ave. Changchun130012 China

In order to solve the problem that multi-thresholding segmentation spends too much time finding the optimal solution in medical image segmentation, Otsu multi-thresholding based on dynamic combination of genetic algorithm and ant algorithm (abbreviated as DCG3A-OM) was proposed in this paper. Firstly, the algorithm uses genetic algorithm to generate preliminary partition results, and then converts them into the initial information of ant colony. Finally ant algorithm will approach to optimal values efficiently. While running genetic algorithm, the conditions of combination will be analyzed dynamically to avoid stopping genetic algorithm early or late. Besides, it increases the probability of variation and decreases the numbers of individuals which are calculated for the evolution rate of colony. The experimental results demonstrate that DCG3A-OM gets the optimal solution more quickly and efficiently than the others, and it can always get the optimal values mostly. © 2016 ICIC International.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Topic modeling for large-scale text data

引用

Frontiers of Information Technology & Electronic engineering 2015年第6期16卷 457-465页

作者： Xi-ming LI College of Computer Science and Technology Jilin University Changchun 130012 China MOE Key Laboratory of Symbolic Computation and Knowledge Engineering Jilin University Changchun 130012 China

This paper develops a novel online algorithm, namely moving average stochastic variational inference （MASVI）, which applies the results obtained by previous iterations to smooth out noisy natural gradients. We analyze the convergence property of the proposed algorithm and conduct a set of experiments on two large-scale collections that contain millions of documents. Experimental results indicate that in contrast to algorithms named ＇stochastic variational inference＇ and ＇SGRLD＇, our algorithm achieves a faster convergence rate and better performance.

关键词： Latent Dirichlet allocation （LDA） Topic modeling Online learning Moving average

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous data mining for planning active surveillance of malaria 15

Heterogeneous data mining for planning active surveillance o...

引用

ASE BigData and SocialInformatics, ASE BD and SI 2015

作者： Gu, Xiao Chen, Hechang Yang, Bo College of Computer Science and Technology Jilin University China Key Laboratory of Symbolic Computation and Knowledge Engineering Ministry of Education China

ISBN: (纸本)9781450337359

Malaria is one of the most serious diseases in the world, which is densely distributed in poverty and remote areas. In the prevention and control of malaria, active surveillance is more efficient than passive surveillance to discover the incidences timely and accurately. However, it is always faced with the challenge of how to allocate the limited sources, such as medical staff and medicine, appropriately so as to achieve a maximum infect. In this paper, we propose a novel method to characterize the spatiotemporal patterns of infection risk for active surveillance planning. Specifically, we propose a temporal heterogeneous diffusion network model to discover high risk areas timely, and a mixture optimization method to find high risk areas accurately. The validation on existing real-world data shows that our method outperforms the existing state-of-the-art both in terms of infection risk prediction and planning of active surveillance under different thresholds. © 2015 ACM.

关键词： Mixtures

来源：评论

学校读者我要写书评

暂无评论

A novel method to measure graph similarity 12

A novel method to measure graph similarity

引用

12th IEEE International Conference on E-Business engineering, ICEBE 2015

作者： Wang, Xu Ouyang, Jihong College of Computer Science and Technology Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Changchun China

ISBN: (纸本)9781467380027

A novel method to measure the graph similarity is proposed, where the labels, in-degrees, and out-degrees of the vertices in the graph are comprehensively considered in order to conquer the high complexity and information loss in the measurement process. In this proposal, the graph is decomposed into a generalized tree and its similarity can be measured by computing the cost of the node sequence triples in the generalized tree. The proposed method decreases the computational complexity to O(n2) in comparison with the other available graph similarity methods. An example is provided to illustrate that the method is feasible and effective. © 2015 IEEE.

关键词： Forestry

来源：评论

学校读者我要写书评

暂无评论

A bio-inspired method for locating the diffusion source with limited observers

A bio-inspired method for locating the diffusion source with...

引用

Congress on Evolutionary computation

作者： Yuxin Liu Chao Gao Xinyan She Zili Zhang College of Computer and Information Science & College of Software Southwest University Chongqing China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Changchun China School of Information Technology Deakin University Geelong VIC Australia

ISBN: (纸本)9781509006243

Locating the source of diffusion is a challenging problem in complex networks and has great practical significance for restraining rumors propagation and controlling epidemics spreading. An efficient locating method should have a higher locating accuracy with the minimum required information. Although existing locating methods based on observers consider the time delays of edges, they compute the time delays based on the shortest path, which may differ from the actual diffusion process. Moreover, the higher locating accuracy of traditional method with observers has a great dependence on the assumption that the propagation delays along edges follow a definite distribution such as the Gaussian distribution. In order to solve these shortcomings, this paper proposes a Physarum-inspired method to locate the diffusion source that is independence of the distribution of propagation delays. Our method quantifies the nutrient transportation process in the adaptive network evolved by Physarum, which is used to simulate the information or epidemic diffusion routes in a social network. Simulation results on various benchmark networks show that our method has a better performance in terms of error distance than that of Gaussian method without assuming the definite distribution of time delays. Together with the advantage that our method does not require the sender information of observers compared with existing methods, our method allows for a wider range of applications in the real-world networks.

关键词： Electron tubes Observers Diffusion processes Adaptive systems Propagation delay Conductivity

来源：评论

学校读者我要写书评

暂无评论

Pathological Brain Detection Based on Online Sequential Extreme Learning Machine

Pathological Brain Detection Based on Online Sequential Extr...

引用

2016 IEEE International Conference on Progress in Informatics and Computing

作者： Siyuan Lu Hainan Wang Xueyan Wu Shuihua Wang School of Computer Science and Technology Nanjing Normal University State Key Lab of CAD & CG Zhejiang University Key laboratory of symbolic computation and knowledge engineering of ministry of education Jilin University Key Laboratory of Statistical information technology and data mining State Statistics Bureau Department of Electrical Engineering The City College of New YorkCUNY

Magnetic resonance imaging(MRI) is a kind of imaging modality, which offers clearer images of soft tissues than computed tomography(CT). It is especially suitable for brain disease detection. It is beneficial to detect diseases automatically and accurately. We proposed a pathological brain detection method based on brain MR images and online sequential extreme learning machine. First, seven wavelet entropies(WE) were extracted from each brain MR image to form the feature vector. Then, an online sequential extreme learning machine(OS-ELM) was trained to differentiate pathological brains from the healthy *** experiment results over 132 brain MRIs showed that the proposed approach achieved a sensitivity of 93.51%, a specificity of 92.22%, and an overall accuracy of 93.33%,which suggested that our method is effective.

关键词： magnetic resonance imaging wavelet entropy online sequential extreme learning machine Classification pattern recognition

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：