Skip to main content
Figure 2 | EPJ Data Science

Figure 2

From: Ground truth? Concept-based communities versus the external classification of physics manuscripts

Figure 2

Co-occurrence matrix of arXiv categories during year 2013. Built on the complementary dataset to arxivPhys2013, this matrix reflects the relationships between arXiv categories and allows to justify the meaningfulness of some remarkable discrepancies, like the merger of hep-th and gr-qc articles. Each non-diagonal element reflects the fraction of articles in which two specified categories have co-occurred. The diagonal cells represent the fractions of articles that have been assigned a single category, i.e. they concerns the articles of the arxivPhys2013 dataset. A normalization procedure has been performed such that each row of the matrix sums to 1. Thus, the aforementioned fractions correspond to the fractions of manuscripts that have been labeled with a given category.

Back to article page