Exploiting Semantic Annotations and -Learning for Constructing an Efficient Hierarchy/Graph Texts Organization

Table 4

The results of the comparison of the correlation coefficient between human judgment with some relatedness measures and the proposed semantic relatedness measure.

Measure	Relevance correlation with M&C
Distance-based measures
Rada	0.688%
Wu and Palmer	0.765%
Information-based measures
Resnik	0.77%
Jiang and Conrath	0.848%
Lin	0.853%
Hybrid measures
T. Hong and D. smith	0.879%
Zili Zhou	0.882%
Information/feature-base measures
The proposed MFsSR	0.937%