Visual analysis method of research fronts based on literature co-citation clustering

Abstract

The present invention relates to a research front visual analysis method based on literature co-citation clustering. The method comprises: establishing a literature database in the research field, andencoding the downloaded paper data; extracting a citing reference with high citations in the literature database as a core high-cited citing reference and an external reference with high citations asa high-cited external reference, and adding the high-cited external referenced to the core high-cited citing documents as the core high-cited citing references together; generating a direct-citationmatrix and a co-citation matrix of the core high-cited citing references; assigning a value of zero in the co-citation matrix to a value through numerical transformation of a corresponding position inthe direct-citation matrix; generating a co-citation network based on the co-citation matrix, and finding out dense subgraphs from the network; performing term extraction on frontier topics; and performing visual analysis and design on clustering results of the co-citation network.

Description

technical field [0001] The invention relates to the technical fields of bibliometrics and visual analysis, and uses a method combining co-citation and direct citation of documents to perform visual mining and evolution analysis on research frontiers in the field of artificial intelligence. Background technique [0002] Research Fronts represent the most dynamic areas of science and technology and attract the interest of many researchers. The detection of research fronts is very important for researchers to understand the latest field, view the state of research topics holistically, and seize important interdisciplinary opportunities. The development and connection of cutting-edge topics also provide researchers with a perspective to view research topics from time and field dimensions, and help inspire new discoveries. Typically, a frontier topic consists of a group of highly cited papers that represent the structural features of the current literature, especially those that...

AI Technical Summary

Problems solved by technology

There are many ways to visualize the citation relationship. In 1965, Derek J. de Solla Price pointed out the network relationship between scientific papers in his article "Scientific Paper Network" [2]. Applied Visual Analysis Techniques to the Detection and Time Evolution of Frontier Topics

Images