Comparing hierarchical mathematical document clustering against the mathematics subject classification tree

Abstract

Mathematical publications are often labelled with Mathematical Subject Classification codes. These codes are grouped in a tree-like hierarchy created by experts. In this paper we posit that this hierarchy is highly correlated with content of publications. Following this assumption we try to reconstruct the MSC tree basing on our publications corpora. Results are compared to the original hierarchy and conclusions are drawn. © 2013 Springer-Verlag Berlin Heidelberg.