A New Metric for the Analysis of Scientific Citation Network in Statistics

Toy example of citation network
The figure shows a toy example of citation network. In this network, each node represents an article and each link between nodes represents the citation. Node b cites node a directly and node e cites node a indirectly. The citation path from node e to node a is e→b→a, so the length of this path is 2.

Nowadays, the research funds are very limited, and all research institutes and universities are competitive against one another for these resources. Therefore, it is very important to have a fair method to judge the research performance, so that the resources are distributed in a reasonable and efficient way.

In this study, we develop a new method "Article network influence" (ANI) for evaluating the research performance of scientific articles. The database we used to evaluate the research performance is Web of Science database. It contains big volume of data and those citation data are used to construct the scientific citation network. Instead of only considering the direct citation, indirect citations are also considered since it might be possible that people using the latest article as reference and ignore the original article. Those citations with different citation path length will be weighted and be used as the final score of the research performance, ANI. The 36 years statistical article citation network is used as the demonstration of this method. The network is divided into several sub-networks, and ANI is applied to each sub-network. The top 20 influential articles in each sub-network are listed and the history trend can be observed from it. For the future research plan, ANI can be applied to other article citation network to have an insight of other fields.

Bibliographic information:
  • Article title: A New Metric for the Analysis of the Scientific Article Citation Network
  • Journal: IEEE Access
  • Issue Date: DECEMBER 2019
  • Volume: 7, Issue:1
  • On Page(s): 132027-132032
  • Print ISSN: 2169-3536
  • Online ISSN: 2169-3536
  • Digital Object Identifier: 10.1109/ACCESS.2019.2937220

Chang Lin-Hsuan, Department of Statistical Science