Table 1 Performance evaluation of recommendation approaches constructed with the Fed-FR-MVD framework compared to other methods.

	Precision@10	Recall@10	HR@10	NDCG@10	F1@10	MRR@10	Coverage@10	AUC
T-SV (V₁)	0.46	0.375	0.56	0.61	0.42	0.46	0.75	0.8
T-SV (V₂)	0.45	0.365	0.55	0.6	0.41	0.45	0.74	0.79
Fed-SV (V₁)	0.43	0.34	0.535	0.585	0.385	0.43	0.73	0.78
Fed-SV (V₂)	0.42	0.33	0.525	0.575	0.375	0.42	0.72	0.77
FED-MVMF	0.44	0.35	0.545	0.595	0.395	0.44	0.74	0.79
FL-MV-DSSM	0.45	0.355	0.55	0.6	0.405	0.45	0.75	0.8
SEMI-FL-MV-DSSM	0.445	0.35	0.545	0.595	0.4	0.445	0.745	0.795
Fed-FR-MVD	0.47	0.37	0.565	0.615	0.425	0.47	0.765	0.815
FedCT	0.455	0.36	0.555	0.605	0.41	0.455	0.755	0.805
FedCDR	0.465	0.365	0.56	0.61	0.42	0.465	0.76	0.81

This table provides a comprehensive overview of the models’ performance across various evaluation metrics, focusing on accuracy, recall capability, ranking quality, and coverage. By emphasizing Top-10 evaluations—particularly the local and top-ranked results—this table aids in identifying the strengths and weaknesses of each approach, enabling tailored assessments based on specific application needs. The inclusion of AUC as a distinct metric further enriches the analysis, offering insights into model effectiveness beyond the top-ranked recommendations.
Significant are in value [bold].

Quick links

Search