分类:Concept Mapping of Scientometrics and its applications in scientometrical studies

来自Big Physics
Jinshanw讨论 | 贡献2020年9月17日 (四) 21:56的版本 →‎Research Tasks and Motivations
(差异) ←上一版本 | 最后版本 (差异) | 下一版本→ (差异)


This is a short summary of its Chinese version 科学计量学多层网络构建和应用

Research Tasks and Motivations[1]

  1. To construct a multilayer network of Scientometrics, where researchers, research papers and concepts are the actors and reseachers-metoring-researchers, researcher-writing-papers, repapers-citing-papers, papers-working-on-concepts, and logic among concepts are the relations.
  2. To develop algorithms on the network for estimating influence (in various senses) of various actors, and various relations
  3. To share this network with members of this research community, potentially can be used in teaching courses or writing books of Scientometrics, reviewing research papers
  4. More generally, if the network construction and analysis can be done successfully for Scientometrics, the same can be done to many other fields/disciplines or even to the whole human knowledge.

The Key Ideas

  1. Network of concepts encodes informatons on properties of concepts/research topics such as research fronts, fundamental/core questions, hot (currently or ever heavily studied) topics
  2. With the network, and algorihtms revealing and making use of those properties, one may define most valubale (in various sense, such as creative, hotness-tracing, fundamental) actors
  3. Algorithms based on propagations over the whole network, might reveal something new, for example, concepts studied by highly cited papers might be more valuable than those studied by less cited papers, papers working on more fundamental concepts should be weighted more even with less recieved citations, and so on

Approaches

Data

  1. Research papers from journals of Scientometrics
  2. human annotated data on researchers and their academic trees
  3. human annotated concepts of Scientometrics and their network
  4. Textbooks of Scientometrics

Method

  1. Natural Language Processing[2][3][4][5][6], named entity extration technique and relation extraction technique in perticular
  2. Machine Learning from human annotated data and corpus of research papers
  3. Propogaton algorithms, such as PageRank-like ones and the General Input-Output Analysis[7]
  4. Expert examinations of the network and of the results of analysis

Validation

  1. of the multilayer network
  2. of the results of analysis

References

  1. Jinshan Wu. "Infrastructure of Scientometrics:The Big and Network Picture." Journal of Data and Information Science, vol.4, no.4, 2019, pp.1-12. DOI: 10.2478/jdis-2019-0017
  2. John Giorgi and Xindi Wang and Nicola Sahar and Won Young Shin and Gary D. Bader and Bo Wang, End-to-end Named Entity Recognition and Relation Extraction using Pre-trained Language Models, arXiv cs.CL 1912.13415(2019).
  3. A. Bordes, N. Usunier, A. Garcıa-Duran, J. Weston, and O. Yakhnenko, “Translating embeddings for modeling multi-relational data,” in Proc. Adv. Neural Inf. Process. Syst., 2013, 2787-2795.
  4. J. Weston, A. Bordes, O. Yakhnenko, and N. Usunier, “Connecting language and knowledge bases with embedding models for relation extraction,” in Proc. Conf. Empirical Methods Natural Language Process., 2013, 1366-1371.
  5. S. Riedel, L. Yao, A. Mccallum, and B. M. Marlin, “Relation extraction with matrix factorization and universal schemas,” in Proc. Conf. North Amer. Chapter Assoc. Comput. Linguistics: Human Language Technol., 2013, 74-84.
  6. Maximilian Nickel, Kevin Murphy, Volker Tresp, Evgeniy Gabrilovich. A Review of Relational Machine Learning for Knowledge Graphs. Proceedings of the IEEE 2016.
  7. Zhesi Shen, Liying Yang, Jiansuo Pei, Menghui Li, Chensheng Wu, Jianzhang Bao, Tian Wei, Zengru Di, Ronald Rousseau, Jinshan Wu, Interrelations among scientific fields and their relative influences revealed by an input–output analysis, Journal of Informetrics 10, 82-97(2016). doi:10.1016/j.joi.2015.11.002

本分类目前不含有任何页面或媒体文件。