A knowledge graph embeddings based approach for author name disambiguation using literals

Scientometrics - Scholarly data is growing continuously containing information about the articles from a plethora of venues including conferences, journals, etc. Many initiatives have been taken to...     

Hybrid clustering of multi-view data via Tucker-2 model and its application

With the modern technology fast developing, most of entities can be observed by different perspectives. These multiple view information allows us to find a better pattern as long as we integrate them in an appropriate way. So clustering by integrating multi-view representations that describe the same class of entities has become a crucial issue for knowledge discovering. We integrate multi-view data by a tensor model and present a hybrid clustering method based on Tucker-2 model, which can be regarded as an extension of spectral clustering. We apply our hybrid clustering method to scientific publication analysis by integrating citation-link and lexical content. Clustering experiments are conducted on a large-scale journal set retrieved from the Web of Science (WoS) database. Several relevant hybrid clustering methods are cross compared with our method. The analysis of clustering results demonstrate the effectiveness of the proposed algorithm. Furthermore, we provide a cognitive analysis of the clustering results as well as the visualization as a mapping of the journal set.     

基于随机重构图结构的图神经网络分类算法

目的:为了解决在特征聚合过程中的确定性传播所导致的节点相似性破坏和节点对邻域依赖性高的问题,构建基于随机重构图结构的图神经网络分类算法。方法:首先,随机特征变换根据学习的权重值对随机保留的部分节点特征进行增强,生成随机特征。然后,利用生成的特征计算融合系数,对原始图和k近邻图进行自适应融合,重构出随机图结构。最后,将提取的多支浅层特征加入到重构图结构的卷积层,使模型随着层数的加深可得到浅层信息的补充。此外,对联合优化分类损失和自监督学习损失,保持节点相似性和平滑性。结果:与其他节点分类方法在Cora、Citeseer和Pubmed数据集上进行半监督实验和全监督实验结果对比,本文的算法精度提高了0.9%～2.3%。结论:基于随机重构图结构的网络分类算法在节点分类任务中取得较好的性能。     

Indicators in a research institute: A multi-level classification of scientific journals

Indicators in a research Institute ought to be readable at several decision levels, and particularly with different break-downs of the publication set chosen as reference. Citation transactions between journals have been widely used to structure scientific subfields in ISI databases. We tried a seed-free structuration of SCI/CMCI journals (a) to test convergence of pure citation-built specialties (roughly 150) on SCI/CMCI journals with existing classifications at the subfield level (b) to explore the interest and the limits of this approach for upper levels of aggregation (roughly 30 fields). A few limits of journal-level classification are addressed. At the subfield level, the convergence is large with some discrepancies worth noticing. At the subdiscipline level, the method is not sufficient to achieve a satisfactory 30-level delineation, but gives a good basis for informed expert validation.     

A novel classification method for paper-reviewer recommendation

Reviewer recommendation problem in the research field usually refers to invite experts to comment on the quality of papers, proposals, etc. How to effectively and accurately recommend reviewers for the submitted papers and proposals is a meaningful and still tough task. At present, many unsupervised recommendation methods have been researched to solve this task. In this paper, a novel classification method named Word Mover’s Distance–Constructive Covering Algorithm (WMD–CCA, for short) is proposed to solve the reviewer recommendation problem as a classification issue. A submission or a reviewer is described by some tags, such as keywords, research interests, and so on. First, the submission or the reviewer is represented as some vectors by a word embedding method. That is to say, each tag describing a submission or a reviewer is represented as a vector. Second, the Word Mover’s Distance (WMD, for short) method is used to measure the minimum distances between submissions and reviewers. Actually, the papers usually have research field information, and utilizing them well might improve the reviewer recommendation accuracy. So finally, the reviewer recommendation task is transformed into a classification problem which is solved by a supervised learning method- Constructive Covering Algorithm (CCA, for short). Comparative experiments are conducted with 4 public datasets and a synthetic dataset from Baidu Scholar, which show that the proposed method WMD–CCA effectively solves the reviewer recommendation task as a classification issue and improves the recommendation accuracy.     

Validation of scientific topic models using graph analysis and corpus metadata

Probabilistic topic modeling algorithms like Latent Dirichlet Allocation (LDA) have become powerful tools for the analysis of large collections of documents (such as papers, projects, or funding applications) in science, technology an innovation (STI) policy design and monitoring. However, selecting an appropriate and stable topic model for a specific application (by adjusting the hyperparameters of the algorithm) is not a trivial problem. Common validation metrics like coherence or perplexity, which are focused on the quality of topics, are not a good fit in applications where the quality of the document similarity relations inferred from the topic model is especially relevant. Relying on graph analysis techniques, the aim of our work is to state a new methodology for the selection of hyperparameters which is specifically oriented to optimize the similarity metrics emanating from the topic model. In order to do this, we propose two graph metrics: the first measures the variability of the similarity graphs that result from different runs of the algorithm for a fixed value of the hyperparameters, while the second metric measures the alignment between the graph derived from the LDA model and another obtained using metadata available for the corresponding corpus. Through experiments on various corpora related to STI, it is shown that the proposed metrics provide relevant indicators to select the number of topics and build persistent topic models that are consistent with the metadata. Their use, which can be extended to other topic models beyond LDA, could facilitate the systematic adoption of this kind of techniques in STI policy analysis and design.

     

基于图着色理论的异构车联网时隙分配方案

针对异构车联网系统车辆较多导致时隙分配困难的问题,提出了一种基于图着色理论的时隙分配方案。该方案以考虑两跳内节点的图为模型,通过图着色的方法分配时隙,有效降低了隐藏终端带来的丢包;另外给出了一种高效实用的时隙重用分配算法。该算法根据度定义权值以确定车辆分配时隙数目,保证了公平性,提高了时隙重用,进而提高了消息发送的可靠性,同时也适用于网络拓扑多变的车联网场景。仿真结果表明,在车辆数为200、时隙数为100时,与传统时隙分配方法相比,该方案的车辆平均收包率获得大幅提升。此外,随着时隙重用的增加,车辆间干扰增强,从而导致平均收包率降低。研究还发现,增加车辆发射功率时,由于接收端信干噪比先增加后趋于不变,所以平均收包率也先增加后趋于不变。     

基于灰度的亚像素插值视觉测量方法

针对现有视觉测量中的检测代价高,精度低和速度慢问题,该文提出一种基于计算机机器视觉的紧密内插值亚像素测量方法。该方法基于线性插值算法的原理,结合常规边缘检测方法和图像的灰度曲线图,利用阈值分割和标准长度进行亚像素自适应阈值选择。为验证该方法的有效性,对标准量块长度进行测量实验,并分析测量系统的误差影响因素。相比于传统的Canny算子方法检测结果,该方法的平均测量准确度提升46.2%。实验结果表明该算法的测量精度较高,可以快速、精确地测量出物体的几何尺寸。     

A pilot study on the connection between scientific fields and patent classification systems

Methods to link academic research achievements with innovative industries have gained considerable awareness worldwide in recent years. Subsequently, responding to industries’ demand to reinforce the linkage between scientific research and industries is an issue awaiting urgent resolution for the government. Previous scientific pertaining to the linkage between scientific fields and (academic papers) technological fields (technology patents) primarily focus on non-patent research or university–industry collaboration. However, these studies failed to highlight the type of linkages between science and technological fields. Therefore, we conducted a pilot study to identify the core scientific fields in different technological fields. In addition to the proposed network maps linking scientific and technological fields, this study also identified the core scientific fields for patent development, including materials science, multidisciplinary; engineering, chemical; physics, applied; nanoscience and nanotechnology; and chemistry, physical. Due to the scarcity of research pertaining to the linkage of scientific fields and technological fields, the government, research and development units, and universities lack a framework for linking fundamental scientific research with the development of industry technologies. Therefore, in this study, we used an author–inventor network to analyze this research topic, expecting that the results can serve as a reference for further research.     

A new multiscale computational method for elasto-plastic analysis of heterogeneous materials

A new multiscale computational method is developed for the elasto-plastic analysis of heterogeneous continuum materials with both periodic and random microstructures. In the method, the multiscale base functions which can efficiently capture the small-scale features of elements are constructed numerically and employed to establish the relationship between the macroscopic and microscopic variables. Thus, the detailed microscopic stress fields within the elements can be obtained easily. For the construction of the numerical base functions, several different kinds of boundary conditions are introduced and their influences are investigated. In this context, a two-scale computational modeling with successive iteration scheme is proposed. The new method could be implemented conveniently and adopted to the general problems without scale separation and periodicity assumptions. Extensive numerical experiments are carried out and the results are compared with the direct FEM. It is shown that the method developed provides excellent precision of the nonlinear response for the heterogeneous materials. Moreover, the computational cost is reduced dramatically.     

一种基于矢量图的区域变化检测方法

提出了一种基于矢量图进行遥感影像的区域变化检测的方法,该方法通过将历史矢量图与当前影像进行匹配后叠加、设定全局变化系数与局部变化系数等步骤,对各矢量图斑区域内的影像DN值做统计分析,达到检测区域变化的目的.试算分析表明,全局变化系数比局部变化系数表现得更加敏感,更易对错判数目造成影响.利用历史矢量图做变化检测,不仅可以避免复杂的影像辐射校正工作,还可以充分利用历史形成的矢量库资料,实现数据复用,同时也便于矢量库的历史继承与发展.     

A novel pattern extraction method for time series classification

Multivariate time series classification is of significance in machine learning area. In this paper, we present a novel time series classification algorithm, which adopts triangle distance function as similarity measure, extracts some meaningful patterns from original data and uses traditional machine learning algorithm to create classifier based on the extracted patterns. During the stage of pattern extraction, Gini function is used to determine the starting position in the original data and the length of each pattern. In order to improve computing efficiency, we also apply sampling method to reduce the searching space of patterns. The common datasets are used to check our algorithm and compare with the naive algorithms. Experimental results are shown to reveal that much improvement can be gained in terms of interpretability, simplicity and accuracy.     

显影液QA,QC的科学方法

本文为自动洗片机用高温快速显影液的科学管理与质量保证提供了简便、科学的检测方法,力医(?)影像质量的提高提供了可靠的保证。具有科学性、实用性、普及性。     

一种计算代价敏感算法分类精度的方法

代价敏感普遍应用于解决分类不平衡问题,但代价敏感算法一直没有一个客观的评价标准.本文提出一种针对代价敏感算法的分类精度计算方法,以平衡精度替换总体精度来有效地评定代价敏感算法的分类性能.相比于传统的总体精度,该平衡精度不会忽略小类样本的贡献.通过代价敏感超限学习机对基因表达数据进行分类对比实验,结果表明,平衡精度可以更为客观、合理地表示代价敏感算法的分类性能.     

A self-consistent method for the description of the generalized conductivity of heterogeneous systems

Journal of Engineering Physics and Thermophysics - A generalization is presented of the self-consistent field method for the determination of the effective conductivity of heterogeneous materials...     

The Annus Mirabilis paper: years of peak productivity in scientific careers

Scientometrics - This paper defines the ‘miraculous year’ as the most productive year in academics’ scientific careers. It tests the hypothesis that annual productivity is...     

A numerical upscaling method for an elliptic equation with heterogeneous tensorial coefficients

The aim of this paper is to develop a new numerical method to determine the effective permeability on a coarse scale level of problems with strongly heterogeneous tensorial coefficients defined on a fine scale. This method stems from a primal–dual mixed formulation and the upscaled permeability is obtained by imposing the continuity of the numerical flux across the lines of discontinuities. Moreover, a finite volume scheme is built to numerically solve elliptic equations with this kind of coefficients. Finally, the method is illustrated by several numerical experiments. Copyright © 2002 John Wiley & Sons, Ltd.     

A new local meshless method for steady-state heat conduction in heterogeneous materials

In this paper a truly meshless method based on the integral form of energy equation is presented to study the steady-state heat conduction in the anisotropic and heterogeneous materials. The presented meshless method is based on the satisfaction of the integral form of energy balance equation for each sub-particle (sub-domain) inside the material. Moving least square (MLS) approximation is used for approximation of the field variable over the randomly located nodes inside the domain. In the absence of heat generation, the domain integration is eliminated from the formulation of presented method and the computational efforts are reduced substantially with respect to the conventional MLPG method. A direct method is presented for treatment of material discontinuity at the heterogeneous material in the presented meshless method. As a practical problem the heat conduction in fibrous composite material is studied and the steady-state heat conduction in unidirectional fiber–matrix composites is investigated. The solution domain includes a small area of the composite system called representative volume element (RVE). Comparison of numerical results shows that the presented meshless method is simple, effective, accurate and less costly method for micromechanical analysis of heat conduction in heterogeneous materials.     

A hybrid finite element method for heterogeneous materials with randomly dispersed rigid inclusions

A hybrid finite element approach is proposed for the mechanical response of two-dimensional heterogeneous materials with linearly elastic matrix and randomly dispersed rigid circular inclusions of arbitrary sizes. In conventional finite element methods, many elements must be used to represent one inclusion. In this work, each inclusion is embedded inside a polygonal element and only one element is required to represent one inclusion. In numerically approximating stress and displacement distributions around the inclusion, classical elasticity solutions for a multiply-connected region are employed. A modified hybrid functional is used as the basis of the element formulation where the displacement boundary conditions of the element are automatically considered in a variational sense. The accuracy and efficiency of the proposed method are demonstrated by two boundary value problems. In one example, the results based on the proposed method with only 64 hybrid elements (450 degrees of freedom) are shown to be almost identical to those based on the traditional method with 2928 conventional elements (5526 degrees of freedom).