| 专利引用关系形成的解释框架:一个指数随机图模型视角 |
| |
| 引用本文： | 杨冠灿,陈亮,张静,李纲.专利引用关系形成的解释框架:一个指数随机图模型视角[J].图书情报工作,2019,63(5):100-109. |
| |
| 作者姓名： | 杨冠灿 陈亮 张静 李纲 |
| |
| 作者单位： | 1.中国人民大学信息资源管理学院 北京 100872;2.中国科学技术信息研究所 北京 100038;3.武汉大学信息管理学院 武汉 430072 |
| |
| 基金项目： | 本文系国家自然科学基金项目"基于指数随机图模型的专利引用关系形成影响因素及机理研究"（项目编号：71403256）和国家自然科学基金项目"面向专利文本中实体关系抽取的远程监督方法研究"（项目编号：71704169）研究成果之一 |
| |
| 摘 要： | 目的/意义]近年来，围绕着专利引文网络结构特征的研究出现大量的研究成果，这些成果都从某种程度上折射出专利引文关系的形成受到来自属性特征之外关系特征的影响，而现有的以回归方法为基础的统计推断方法难以将这些因素纳入到分析框架中，因此，急需探索新的方法。方法/过程]从关系形成视角，专利引用关系形成可表示三种广义的关系形成过程：自组织影响过程、自身属性影响过程、网络协变量影响过程，并建立关系形成过程与网络配置间的映射关系，最终，形成一整套可用于理解复杂专利引用关系形成问题的解释框架。结果/结论]提出一整套可用于理解复杂专利引用关系形成问题的解释框架，该框架是未来进一步构建网络统计模型的理论基础，另外，解释框架包含丰富的网络配置项，预示着未来指数随机图模型在文献计量、科学网络分析上广阔的应用前景。
|
| 关 键 词： | 专利引用关系形成 解释框架 统计网络模型 |
| 收稿时间： | 2018-07-23 |
Framework for Explanations of Patent Citation Formation: An Exponential Random Graph Model Perspective |
| |
| Yang Guancan,Cheng Liang,Zhang Jing,Li Gang.Framework for Explanations of Patent Citation Formation: An Exponential Random Graph Model Perspective[J].Library and Information Service,2019,63(5):100-109. |
| |
| Authors: | Yang Guancan Cheng Liang Zhang Jing Li Gang |
| |
| Institution: | 1.School of Information Resource Management of Renmin University of China, Beijing 100872;2.Institute of Science and Technical Information of China, Beijing 100038;3.School of Information Management of Wuhan University, Hubei 430072 |
| |
| Abstract: | Purpose/significance] Although there have been efforts of scholars to answer the question, what’s determinants of patent citation formation are not solved satisfactorily. Scholars find formation of patent citation is influenced by the structure characteristics of patent citation network. However, the current framework of statistical inference methods based on logistical regression is failing to incorporate the above factors, so an innovative method need to be introduced. Method/process] From a tie formation perspective, patent citation formation represents three broad category of tie formation processes: attribute-based processes, self-organizing network processes and covariates processes. Furthermore, based on these processes, the paper establishes a mapping relationship between those processes with particular types of configurations. Finally, a framework is proposed for understanding the complexity of patent citation formation. Result/conclusion] The paper introduces a framework for understanding patent citation formation, which lays the groundwork for statistical network modeling in the future. In addition, broadly network configuration selection from the framework offers significant opportunities to extend existing bibliometrics and open new pathways in complexity of scientific network analysis. |
| |
| Keywords: | patent citations formation framework for explanations statistical network analysis |
|
| 点击此处可从《图书情报工作》浏览原始摘要信息 |
| 点击此处可从《图书情报工作》下载免费的PDF全文 |
