Novel Recommendation System for Tourist Spots Based on Hierarchical Sampling Statistics and SVD++

Table 4

F1 values of different algorithms (the best value of each row is shown as 25.30, the suboptimal value of each row is shown as 27.44, and unit is %).

N’	SVD	SVD++	NMF	Slope One	KNN	KNN + Zscore	KNN + means	KNN Baseline	Co-Clustering	BPR	Our_HS	Our_Mix
1	21.77	25.13	23.78	21.83	24.93	24.32	24.09	23.30	25.30	15.03	0.62	25.19
2	24.37	27.44	24.51	24.36	25.43	22.75	23.16	23.55	25.93	16.11	1.46	27.90
3	21.93	27.69	20.73	22.55	24.86	22.09	23.26	24.21	27.08	15.84	1.96	27.80
4	23.72	26.09	22.94	22.64	25.38	24.64	23.54	24.74	26.10	16.51	2.98	28.08
5	23.91	28.11	24.74	23.77	25.46	24.68	23.81	24.56	26.57	16.79	3.37	29.11
6	25.56	23.49	24.86	24.99	25.63	26.06	23.99	25.52	26.92	15.89	3.69	27.86
7	27.52	28.83	25.33	25.74	26.07	26.21	25.98	25.90	28.60	18.08	4.67	29.83
8	26.74	28.58	25.65	25.82	26.24	26.24	26.31	26.04	25.39	18.69	4.97	30.42
9	26.98	28.74	26.22	26.72	26.33	26.35	25.62	26.26	27.05	19.83	6.15	31.05
10	28.11	28.23	25.56	27.00	27.11	26.24	26.48	26.59	27.27	19.96	6.90	32.86
	25.06	27.23	24.43	24.54	25.74	24.96	24.62	25.07	26.62	17.30	3.68	29.01