Yet another method for author co-citation analysis: A new approach based on paragraph similarity
Journal
Proceedings of the Association for Information Science and Technology
Date Issued
2017-01-01
Author(s)
Hsiao, T.-M.
Abstract
Copyright 2017 Tsung-Ming Hsiao and Kuang-hua Chen Co-citation analysis has been widely adopted to represent the intellectual structure of a discipline. In general, all co-cited author pairs are regarded equal. With the development of computer technology and easy accessibility of machine readable full-text articles, new weighting schemes for measuring co-citation strength (CCS) have been proposed. However, in previous studies, only distance and sentence similarity are used to adjust CCS when applying co-citation analysis. In this study, we propose a new approach to measuring CCS based on paragraph similarity. Investigation was carried out to compare our approach and traditional author co-citation analysis (TACA), as well as other different parametric ACAs. Preliminary results show that TACA and distance-based ACA (DACA) share many commonalities. In contrast, similarity-based ACAs reveal the different structure from that of TACA and DACA. However, differences in resulting network structure were still found between paragraph-similarity-based ACA (PACA) and sentence-similarity-based ACA (SACA). Compared to SACA, PACA produces less number of factors and clusters and moderate size of clusters.
Type
conference paper