Hybrid semantic recommender system for chemical compounds in large-scale datasets

Table 4 Results of ONTO-RESNIK for the user 33142

Training	Relevant	Score	Top@20	Score
60561	*59484*	2.18	134258	7.59
62642	77367	6.82	61755	7.59
62664	77380	6.74	84082	7.59
62996	*77629*	6.60	84078	7.59
62997	84078	7.59	59949	7.48
62998	84082	7.59	90930	7.29
77314	*134230*	4.33	66139	7.29
77374			60381	6.87
77378			77367	6.82
77381			90775	6.82
77382			77380	6.74
77384			62471	6.65
77385			61847	6.65
77598			87452	6.65
77613			87799	6.65
77625			61713	6.65
77626			61329	6.65
77627			61334	6.65
77628			62534	6.65
84081			67164	6.65
84084

The table presents the training items for this user, the relevant items in the testing set, the scores of these items calculated using the ONTO-RESNIK algorithm and the top@20 recommendations, and respective scores. Underlined are the relevant items which were recommended (77367, 77380, 84078, 84082) and in Bolditalic the relevant items which were not recommended in the top@20 (59484, 77629, 134230)

ISSN: 1758-2946