数字科研时代的引文分析-基于被引频次分析的实证研究

从期刊被引频次的角度出发,采取实证研究的方法,选择国际权威的引文数据库Web of Science和著名的搜索引擎GoogleScholar,以《美国信息科学和技术学会杂志》为文献源进行相关分析,得出在数字科研时代引文分析有必要采取多个引文分析工具,使得引文分析能跟上时代发展步伐的结论。     

数字人文研究领域的知识图谱构建与分析——基于WoS文献关键词和引文上下文的实证

[目的/意义]引文是施引文献与被引成果的纽带，反映了后继者的借鉴和肯定。本研究在传统题录关键词网络的基础上，创新地将引文上下文关键词作为研究材料，所构建知识图谱不仅能揭示文献主题的深层次信息，也能够反映受众主观筛选和利用文献的知识过程。[方法/过程]选取数字人文为研究领域，获取3个文献集和两个引文文本集，构建两个无向的关键词共现网络和两个有向的基于文献引证的关键词网络。通过共现网络，观察数字人文领域知识的吸收与扩散；通过引证关键词网络，观察数字人文的形成与转化。[结果/结论]研究揭示数字人文的研究重点、核心领域与核心技术，从受众的角度为数字人文领域未来研究提供借鉴和参考。     

Exploring all-author tripartite citation networks: A case study of gene editing

Constructing academic networks to explore intellectual structure realize academic community detection, which can promote scientific research innovation and discipline progress, constitutes an important research topic. In this study, tripartite citation is fused with co-citation and coupling relations as a way of weighting the strength of direct citations, and all-author tripartite citation networks were constructed due to the contributions of all authors to the resulting publications. For purpose of exploring the potential of the all-author exclusive and inclusive tripartite citation networks, gene editing is taken as a case study. The extensive experimental comparisons are conducted with the traditional author single-citation networks and first-author tripartite citation network in terms of network structure characteristics, identifying core scholars, and exploring intellectual structures. The following conclusions can be drawn as follows: our all-author tripartite citation networks are able to help identify the most influential scholars in the field of gene editing, and the intellectual structures from exclusive tripartite citation networks are optimal.     

农林类高校图书馆引用认同探讨

将引用认同方法应用于农林类高校图书馆研究特点和研究网络分析。结合《中国引文数据库》,计算"211"农林类高校图书馆2009-2011年引用认同指标,构建其互引认同网络;计算其标准化 (h,g,R)指数,并对各引用认同指标与其相关性进行分析。研究表明:引用认同可以反映出这些图书馆研究特点和研究偏好及对国外成果吸收借鉴的情况,国外机构引文/引文比、平均引用半衰期和平均即年引用指数与标准化(h,g,R)指数呈正相关,平均即年引用指数与标准化(h,g,R)指数有较强的相关性。     

中文网络灰色文献分布探析——基于Google Scholar的样本分析

鉴于传统引文索引数据库的局限性,借鉴相关研究的做法,尝试利用Google Scholar抽样调查中文网络灰色文献的分布特征.验证这一工具可用性的同时,发现样本中灰色文献的整体开发利用水平不高,体现在:①研究和管理机构偏少;②灰色文献所占比例(或者说可见性)偏低;③样本灰色文献的半衰期与常规文献差别不大;④灰色文献来源机构单一.样本所得结论的适用范围还有待进一步研究验证.     

基于无监督图神经网络的学术文献表示学习研究

学术文献特征表示,是学术文献搜索、分类组织、个性化推荐等学术大数据服务的关键步骤。研究表明,图神经网络能够有效学习文献的特征表示,然而当前研究主要集中在有监督学习方法上,不仅对数据集的大小和质量的要求较高,且学习到的文献特征表示与具体任务高度耦合。基于此,本文将四种无监督图神经网络方法引入学术文献表示学习,从Cora、CiteSeer和DBLP (database systems and logic programming)数据集的引文网络、共被引网络和文献耦合网络中学习文献的表示向量,并应用于文献分类和论文推荐两大下游任务。研究结果表明,(1)深度互信息图神经网络适合于文献分类任务,对抗正则化变分图自编码器则在论文推荐任务上性能更佳;(2)Cora数据集上的结果表明,相较于共被引和文献耦合网络,引文网络更适合于学习通用的文献表示向量。     

Collaboration prediction based on multilayer all-author tripartite citation networks: A case study of gene editing

Academic collaboration prediction is considered to be an important way to help scholars expand their research horizons and explore a vast and suitable range of partners. However, existing studies mainly rely on historical collaborations for future predictions, which has limitations in digging into credible collaboration possibilities in a wide range of cross-disciplinary contexts. In view of this, this study tries to combine three typical citation relationships (including direct citation, co-citation, and coupling) to predict prospective collaborations based on citation information that reflects the characteristics of scholars’ knowledge structure and research habits, which is supposed to provide supplement and extension for traditional implementation. To this end, we construct all-author tripartite citation networks based on the bibliographic data in the field of gene editing, and apply the Node2vec and Multi-node2vec algorithms to predict collaborations between authors in both single and multiple layers. According to compare with that of link prediction indicators (including CN, AA, PA and RA, etc.) commonly used for traditional collaboration networks, it is found that the prediction results in the multilayer all-author tripartite citation network should be relatively more accurate. The results will be helpful for scholars in the field of gene editing to explore potential collaborators with an implicit research connection.     

专利引证分析工具的设计与实现

专利引证分析在专利分析中的作用越来越大。文章介绍了一个专利引证分析工具的设计与实现,包括数据的获取、清洗、转换和生成图形,并进行了简单的应用分析。同时,介绍了一个开源的可视化工具包Prefuse,希望对想通过用可视化手段进行文献分析的研究人员提供有益的启示。     

From conventional governance to e-democracy: Tracing the evolution of e-governance research trends using network analysis tools

The adoption of e-governing practices has revolutionised the administrative machinery of governments worldwide by improving efficiency, transparency, and accountability. Researchers and administrators often aim to identify emerging research fronts and the timeline of the evolution to forecast and implement technology. In this work, we systematically investigate the trajectory of the global evolution and emerging research fronts as well as the prospects for e-governance using citation network analysis. The growth curve fitted to the number of articles published per year shows that the research activities are still in the ascendant phase. We visualise the global main path of the citation network and investigate the patterns to trace the knowledge diffusion path, major milestones, and emerging research fronts. The cluster analysis identifies the major topics of research as administration and information system management, e-governance framework design, efficiency or quality evaluation, and the application of social networks and open data leading to e-democracy. The adoption of open data and social networking for user interactions with government that leads to participatory governance are the emerging research trends. We also identify research that can have a future impact based on network parameters. The results contribute to the literature by setting the focus of future research, and assisting administrators in selecting suitable models and methodologies, and manufacturers with the development of required technical devices suitable for the upcoming phase of symbiosis.     

PageRank variants in the evaluation of citation networks

This paper explores a possible approach to a research evaluation, by calculating the renown of authors of scientific papers. The evaluation is based on the citation analysis and its results should be close to a human viewpoint. The PageRank algorithm and its modifications were used for the evaluation of various types of citation networks. Our main research question was whether better evaluation results were based directly on an author network or on a publication network. Other issues concerned, for example, the determination of weights in the author network and the distribution of publication scores among their authors. The citation networks were extracted from the computer science domain in the ISI Web of Science database. The influence of self-citations was also explored. To find the best network for a research evaluation, the outputs of PageRank were compared with lists of prestigious awards in computer science such as the Turing and Codd award, ISI Highly Cited and ACM Fellows. Our experiments proved that the best ranking of authors was obtained by using a publication citation network from which self-citations were eliminated, and by distributing the same proportional parts of the publications’ values to their authors. The ranking can be used as a criterion for the financial support of research teams, for identifying leaders of such teams, etc.     

Testing the fairness of citation indicators for comparison across scientific domains: The case of fractional citation counts

Citation numbers are extensively used for assessing the quality of scientific research. The use of raw citation counts is generally misleading, especially when applied to cross-disciplinary comparisons, since the average number of citations received is strongly dependent on the scientific discipline of reference of the paper. Measuring and eliminating biases in citation patterns is crucial for a fair use of citation numbers. Several numerical indicators have been introduced with this aim, but so far a specific statistical test for estimating the fairness of these numerical indicators has not been developed. Here we present a statistical method aimed at estimating the effectiveness of numerical indicators in the suppression of citation biases. The method is simple to implement and can be easily generalized to various scenarios. As a practical example we test, in a controlled case, the fairness of fractional citation count, which has been recently proposed as a tool for cross-discipline comparison. We show that this indicator is not able to remove biases in citation patterns and performs much worse than the rescaling of citation counts with average values.     

科学知识扩散研究框架

通过对科学知识扩散相关文献的梳理,构建科学知识扩散的研究框架,并对研究对象、扩散关系表示、衡量指标、扩散模型等方面进行详细评述。扩散的对象包括期刊、学科、科研人员等,扩散过程主要以文献引证和作者合著关系表示。在实证中,基于文献引证关系的引文及引文网络分析是科学知识扩散的主流研究方法。衡量指标可以按照测度粒度分为文章、期刊、学科3个层次。现有科学知识扩散的模型研究以定性研究为主,定量化分析较少,常见思路为跨学科借鉴成熟模型。     

学科交叉研究综述

[目的/意义] 学科交叉研究意义重大,通过对学科交叉理论及实践的综述,分析讨论学科交叉研究存在的问题,并提出未来的研究内容和研究思路.[方法/过程] 通过对学科交叉理论及实践相关文献的系统调研、归纳和分析,结合对学科交叉类型和交叉动力学研究的综合分析,重点对当前已有的测度学科交叉程度的引文分析指标和学科交叉度计量的实践研究进行分析.[结果/结论] 当前的学科交叉研究分为宏观态势和微观内容两个层面,研究主体内容分为学科交叉类型研究、学科交叉动力学研究以及学科交叉性测度指标研究.学科交叉的测度属性分为学科多样性和学科聚合性,其测度指标分为3类:学科多样性测度指标、学科聚合性测度指标和综合性测度指标.学科交叉度计量的实践研究可以分为3类:基于交叉度的统计指标的计量、基于社会网络指标的计量以及利用多种指标的综合性计量.     

基于全代引证网络的高影响力专利知识扩散特征研究

［目的/意义］ 基于专利的全代引证网络对专利进行分类,对高影响力专利的知识扩散特征进行分析,为专利影响力的认识和评估提供重要参考。［方法/过程］ 以生物学家悉尼·布伦纳的专利为例,研究其专利和前向引证专利生成的专利全代引证网络,根据专利的直接引证数量和引证长度两个对专利扩散发挥重要作用的因素将专利分为四类,将具有高被引数量长引证路径的专利定义为高影响力专利,对这种专利的知识扩散特征进行分析。［结果/结论］ 研究发现,在专利的全代引证网络中"关键专利""重要专利"和"隐藏的高影响力专利"对专利的扩散影响巨大,全代引证网络中专利的领域变化也体现了知识的流动现象,知识扩散速度可以通过数字直接刻画专利时序网络特点。结合研究结果,对高影响力专利的特点有了更具体的认识,并为高影响力专利的评价提出新的思路。     

中文文献引文情感语料库构建

基于内容的引文情感分析克服了传统基于引用频次的引用同一化问题,是引文内容分析领域一个重要的研究热点。然而引文情感分析依赖于带标注的数据集,目前大规模高质量的引文情感语料资源匮乏,严重制约了该领域的研究。因此,本文在分析引文情感表达方式的基础上提出了一套适用于引文情感表示的标注体系,并详细阐述了语料库建设的技术和方法。采用人机结合的标注策略,借助完善的引文标注系统,构建了规模较大的中文文献的引文情感语料库。统计结果显示,在中文信息处理和科技管理领域情感褒义和贬义总的引用的占比分别为22%和6%,引文情感标注kappa值达到0.852,表明该语料库能够客观地反映作者的情感倾向性,可为论文评价、引文网络分析和情感分析等相关领域的研究提供数据支撑。     

用于科学结构分析的混合网络社团划分方法述评

[目的/意义]复杂网络的社团结构研究已逐渐成为科学家借助文献数据开展科学结构研究的有力工具，社团划分效果的不同对科学结构的解读有着举足轻重的影响。本文对混合网络社团划分方法进行梳理，以期对该领域的相关研究提供借鉴参考。[方法/过程]通过文献调研，阐明混合网络的概念与类型，从网络构建或算法革新角度对各类型混合网络的社团划分研究进行概述，也对支撑混合网络社团划分的经典算法进行简介。[结果/结论]通过系统地梳理总结不同类型混合网络的社团划分工作，为后续的网络分析研究提供研究的视角和方法，同时揭示其在科学结构研究中所面临的挑战与所具有的现实意义，展望今后可能进一步拓展的相关研究方向。     

Reward or persuasion? The battle to define the meaning of a citation

The history and development of the Science Citation Index (SCI) is an example of the power of users in defining and influencing the development of a new technology. The SCI was developed as a tool for the historian of science for the purpose of tracing the history of ideas, but it was appropriated by users for purposes for which it was unintended – as a tool for evaluating the literature, individuals, institutions, and countries. The development of a citation tool gave rise to a debate over what is actually measured by citations. The citation‐as‐reward camp views citations as indicators of quality and impact, whereas the citation‐as‐persuasion camp views citations as no more than rhetorical devices. While neither view can fully explain how authors use citations, citation‐as‐reward prevails as the dominant interpretation.     

Ranking scientific articles based on bibliometric networks with a weighting scheme

As the volume of scientific articles has grown rapidly over the last decades, evaluating their impact becomes critical for tracing valuable and significant research output. Many studies have proposed various ranking methods to estimate the prestige of academic papers using bibliometric methods. However, the weight of the links in bibliometric networks has been rarely considered for article ranking in existing literature. Such incomplete investigation in bibliometric methods could lead to biased ranking results. Therefore, a novel scientific article ranking algorithm, W-Rank, is introduced in this study proposing a weighting scheme. The scheme assigns weight to the links of citation network and authorship network by measuring citation relevance and author contribution. Combining the weighted bibliometric networks and a propagation algorithm, W-Rank is able to obtain article ranking results that are more reasonable than existing PageRank-based methods. Experiments are conducted on both arXiv hep-th and Microsoft Academic Graph datasets to verify the W-Rank and compare it with three renowned article ranking algorithms. Experimental results prove that the proposed weighting scheme assists the W-Rank in obtaining ranking results of higher accuracy and, in certain perspectives, outperforming the other algorithms.     

人文社科类期刊长效文献研究

[目的/意义] 长效文献是指出版多年后仍然被连续多年引用,或者引用频次相对较高的文献,研究这部分老化较慢的长效文献的影响力,有助于完整评价文献整个生命周期的价值。[方法/过程] 选取CSSCI数据库来源期刊（2016-2017）中九个学科在2012年-2016年间出版期刊文献的引文年代数据,通过共时观察法进行科学文献老化情况和长效文献分析的研究。[结果/结论] 期刊中引文年代较为久远的引用量来自于大量的低频次被引用的文献集合。长效文献数量较少,可分为3种类型：成长型、成熟型和衰退型,3种类型的长效文献会随着时间推移产生一定转化。具有长效文献较多的期刊未必影响因子高,除心理学学科外,其他每个学科都有且仅有一本具有显著长效影响力的期刊。长效文献的关键词与当前学科研究热点相似,但是长效文献的追溯年代更久远,在反映当前学科热点来源的同时,对未来学科热点预测将发挥作用。     

基于隐喻思维的引文分析研究与创新

[目的/意义]对引文分析领域的研究与创新进行回顾和展望。[方法/过程]运用隐喻思维方法,以"科学睡美人"和第二代期刊评价指标（EF和SJR）的创设为例,对引文分析领域取得的主要研究进展进行说明;同时,通过对生态学相关研究文献的调研分析,提出将生态学原理引入到未来引文分析研究中来,形成一个新的研究方向——引文生态学,并就若干重要研究议题进行简要说明。[结果/结论]隐喻思维方法在学术研究及方法论创新中具有重要作用,引文生态学的提出对引文分析领域的理论完善和应用创新具有重要价值。