Skip to main content

Table 3 Co-mention similarity measures summarization

From: Automated assessment of biological database assertions using the scientific literature

Dice

\(D(o_{1},o_{2})=2\times \frac {\mid Sentences(o_{1})\cap Sentences(o_{2})\mid }{\mid Sentences(o_{1})\mid +\mid Sentences(o_{2})\mid }\)

Jaccard

\(J(o_{1},o_{2})=\frac {\mid Sentences(o_{1})\cap Sentences(o_{2})\mid }{\mid Sentences(o_{1})\cup Sentences(o_{2})\mid }\)

Overlap

\(O(o_{1},o_{2})=\frac {\mid Sentences(o_{1})\cap Sentences(o_{2})\mid }{min(Sentences(o_{1}),Sentences(o_{2}))}\)

Cosine

\(Cos(o_{1},o_{2})=\frac {\mid Sentences(o_{1})\cap Sentences(o_{2})\mid }{\sqrt {\mid Sentences(o_{1})\mid \times \mid Sentences(o_{2})\mid }}\)

  1. The function Sentences(o) returns from a set of documents those sentences where the object o occurs