A social trust and preference segmentation-based matrix factorization recommendation algorithm

EURASIP Journal on Wireless Communications and Networking

Table 4 RMSE comparison with baselines

Datasets	Dimension	Indicator	PMF	RSTE	SocialMF	TrustSVD	SPMF	Accuracy improvement
Epinions	K = 5	MAE	0.979	0.950	0.825	0.804	0.794	1.24%
	K = 5	RMSE	1.290	1.196	1.070	1.043	0.989	5.18%
	K = 10	MAE	0.909	0.958	0.826	0.805	0.762	5.34%
	K = 10	RMSE	1.197	1.278	1.082	1.044	0.974	6.70%
Ciao	K = 5	MAE	0.920	0.767	0.749	0.723	0.571	21.02%
	K = 5	RMSE	1.260	1.020	0.981	0.955	0.759	20.52%
	K = 10	MAE	0.822	0.763	0.749	0.723	0.565	21.85%
	K = 10	RMSE	1.078	1.013	0.976	0.956	0.756	20.92%

The italic values reflected inside Table 4 is the accuracy of the SPMF algorithm