where n is the number of fvs in the ontology o and sim(fvi, fvj) is a similarity between the two vectors. A
score of zero indicate that all FVs are unique. In general, we would like this score to be as low as possible
in order to discriminate the FVs, but this depends a lot on the quality of the ontology.
The Average FV Neighbourhood Similarity (AFvNS) score indicates the degree of overlap with
neighbouring entities and is defined as follows.