TAFM: A Recommendation Algorithm Based on Text-Attention Factorization Mechanism

Table 3

Performance of different models in the test set.

Algorithm model	Model accuracy	AUC	Log_loss
AFM	0.82291	0.63090	0.4506
CCPM	0.82296	0.63170	0.4504
DCN	0.82564	0.69840	0.4245
DeepFM	0.82543	0.69830	0.4247
FNN	0.82554	0.69740	0.425
MLR	0.82336	0.65860	0.4412
NFM	0.82528	0.69670	0.4254
PNN	0.82585	0.69970	0.4239
xDeepFM	0.82352	0.63430	0.4491
wDeepFM	0.82560	0.69790	0.4248
TAFM	0.82694	0.73018	0.4144