- The clustering techniques produced thematically coherent but potentially granular subgroups within the broader mathematics, statistics, and computer science themes, as indicated by domain-specific terms.
- While the 250 sample case had balanced cluster distributions, the 500 sample case showed an asymmetry, suggesting more detailed subgroups within the larger statistical methods/theory cluster.
- Statistics remained a consistently relevant theme, with shifts between theory and applications. Computer science exhibited more transformative evolution, driven by technological advances, interdisciplinary nature, and practical applications.