Hyperparameter ranges for all considered baseline models

Algorithms	topK	shrink	Similarity	Normalize	α	β	ℓ₂ norm
kNN algorithms	[5 to 1000]	[0 to 1000]	[cosine, jaccard, asymmetric, dice, tversky]	[True, False]	—	—	—
Graph based algorithms	[5 to 1000]	—	—	—	[0 to 2], strategy: uniform distribution	[0 to 2], strategy: uniform distribution	—
EASE^R	—	—	—	—	—	—	[1 to 1e⁷], strategy: log-uniform distribution

Optimal hyperparameter values

Algorithms	topK	shrink	Similarity	Normalize	α	β	ℓ₂ norm
ItemkNN	144	1000	cosine	True	—	—	—
UserkNN	146	1000	cosine	True	—	—	—
P₃α	496	—	—	False	0.7681732734954694	—	—
RP₃β	496	—	—	True	0.44477903655656115	0.5968193614337285	—
EASE^R	—	—	—	—	—	—	1e³

Statistics of the provided train-test splits

Users	Items	Interactions	Density	Data leakage	Interactions per user_min	Interactions per item_min	Interactions per user_max	Interactions per item_max	Testing users
78,578	77,801	2,240,156	0.00062	No	16	10	13313	1907	52,639

Full result table

Methods	R@20	R@40	N@20	N@40	T-Time (s)	P-Time (s)	Average P-Time Per User (ms)
Random	0.0004	0.0006	0.0002	0.0003	0	281.290	3.579
TopPopular	0.009	0.016	0.007	0.009	0.052	239.364	3.046
UserKNN	0.126	0.172	0.104	0.119	629.351	240.319	3.058
ItemKNN	0.152	0.205	0.127	0.144	627.185	211.171	2.687
P₃α	0.133	0.188	0.107	0.124	182.373	232.467	2.958
RP₃β	0.141	0.190	0.117	0.133	200.428	238.372	3.033
EASE	—	—	—	—	—	—	—
DCCF (Code)	0.092	0.138	0.071	0.086	5065.123	57.026	0.738
DCCF (Paper)	0.088	0.134	0.068	0.082

Optimal hyperparameter values

Algorithms	topK	Shrink	Similarity	Normalize	α	β	ℓ₂ norm
ItemkNN	144	1000	cosine	True	—	—	—
UserkNN	146	1000	cosine	True	—	—	—
P₃α	496	—	—	False	0.7681732734954694	—	—
RP₃β	496	—	—	True	0.44477903655656115	0.5968193614337285	—
EASE^R	—	—	—	—	—	—	1e³

Statistics of the provided train-test splits

Users	Items	Interactions	Density	Data leakage	Interactions per user_min	Interactions per item_min	Interactions per user_max	Interactions per item_max	Testing users
50,821	41,390	1,172,425	0.00040	No	5	1	1037	2513	50,821

Full result table

Methods	R@20	R@40	N@20	N@40	T-Time (s)	P-Time (s)	Average P-Time Per User (ms)
Random	0.0004	0.0007	0.0001	0.0003	0	117.449	2.311
TopPopular	0.025	0.037	0.013	0.017	0.028	91.554	1.801
UserKNN	0.213	0.295	0.132	0.153	186.934	88.685	1.745
ItemKNN	0.235	0.321	0.146	0.168	285.577	105.796	2.081
P₃α	0.231	0.317	0.141	0.164	133.445	98.886	1.945
RP₃β	0.245	0.337	0.149	0.173	135.346	99.059	1.949
EASE	0.218	0.303	0.132	0.154	3846.424	162.580	3.199
DCCF (Code)	0.189	0.266	0.112	0.132	4396.699	27.119	0.709
DCCF (Paper)	0.187	0.264	0.112	0.132

Optimal hyperparameter values

Algorithms	topK	shrink	Similarity	Normalize	α	β	ℓ₂ norm
ItemkNN	144	1000	cosine	True	—	—	—
UserkNN	146	1000	cosine	True	—	—	—
P₃α	496	—	—	False	0.7681732734954694	—	—
RP₃β	496	—	—	True	0.44477903655656115	0.5968193614337285	—
EASE^R	—	—	—	—	—	—	1e³

Statistics of the provided train-test splits

Users	Items	Interactions	Density	Data leakage	Interactions per user_min	Interactions per item_min	Interactions per user_max	Interactions per item_max	Testing users
47,939	41,390	2,357,450	0.00119	No	20	3	251	2408	47,528

Full result table

Methods	R@20	R@40	N@20	N@40	T-Time (s)	P-Time (s)	Average P-Time Per User (ms)
Random	0.0005	0.0010	0.0003	0.0005	0	126.208	2.632
TopPopular	0.010	0.017	0.007	0.009	0.064	107.504	2.242
UserKNN	0.087	0.131	0.063	0.079	358.630	143.295	2.989
ItemKNN	0.110	0.163	0.080	0.099	357.441	150.163	3.132
P₃α	0.109	0.162	0.080	0.099	40.439	104.470	2.179
RP₃β	0.116	0.173	0.084	0.104	76.275	131.423	2.741
EASE	0.093	0.143	0.067	0.084	1455.104	229.898	4.795
DCCF (Code)	0.064	0.101	0.045	0.058	2064.829	35.941	0.756
DCCF (Paper)	0.066	0.104	0.046	0.059