This study performs a webometric analysis to explore the communication characteristics of scientific knowledge in a national
scholarly Web space comprising top ranking universities and government supported research institutions in South Korea. We
found significant differences in scholarly communication activity as well as linking behavior among different subspaces in
addition to institutional differences. We also found the usefulness of the ADM approach in analyzing the metric data containing
extreme outliers and discovered the directory model as the most appropriate. Page counts were found significantly correlated
with inlinks as well as with outlinks at the directory level in the whole scholarly Web space.