Abstract
The keyword network analysis has been used for summarizing research trends, and network clustering algorithms play important roles in identifying major research themes. In this paper, we performed a comparative analysis of network clustering algorithms to find out their performances, effectiveness, and impact on cluster themes. The AAG (American Association for Geographers) conference datasets were used in this research. We evaluated seven algorithms with modularity, processing time, and cluster members. The Louvain algorithm showed the best performance in terms of modularity and processing time, followed by the Fast Greedy algorithm. Examining cluster members also showed very coherent connections among cluster members. This study may help researchers to choose a suitable network clustering algorithm and understand geography research trends and topical fields.
Recommended Citation
Lee, Youngho; Lee, Yubin; Seong, Jeong; Stanescu, Ana; and Hwang, Chul Sue
(2020)
"A Comparison of Network Clustering Algorithms in Keyword Network Analysis: A Case Study with Geography Conference Presentations,"
International Journal of Geospatial and Environmental Research: Vol. 7:
No.
3, Article 1.
Available at:
https://dc.uwm.edu/ijger/vol7/iss3/1