TABLE 3.

Evaluation results on our three datasets. Evaluation results on our three datasets. The top scores achieved for each dataset are highlighted in bold.

Ranking metric	MAP	MRR	RR at 5	RR at 10
Manually curated data
Poincaré	0.215	0.282	0.394	0.437
PPMI	0.281	0.343	0.431	0.583
Co-occurrence	0.192	0.282	0.333	0.375
spaCy	0.274	0.371	0.500	0.653
Word2Vec	0.297	0.385	0.514	0.625
DIISH	0.402	0.482	0.625	0.764
The Cook’s Thesaurus data
Poincaré	0.180	0.225	0.292	0.391
PPMI	0.178	0.225	0.291	0.377
Co-occurrence	0.175	0.252	0.331	0.414
spaCy	0.271	0.349	0.511	0.625
Word2Vec	0.266	0.358	0.503	0.617
DIISH	0.381	0.448	0.622	0.725
Food.com user review data
Poincaré	0.125	0.196	0.290	0.374
PPMI	0.179	0.274	0.347	0.456
Co-occurrence	0.117	0.203	0.267	0.335
spaCy	0.167	0.269	0.405	0.524
Word2Vec	0.197	0.326	0.460	0.560
DIISH	0.256	0.373	0.519	0.633