Datasets descriptions

	A	B	C	D	E	F	G	H	I	J	K	L	M
1	Name	Problem	Number of lables	Total size	Train size	Test size	Number of words	Modalities	sklearn linearsvm result	Best published result	Links	Additional information	Dataset variants
2	R8	Text categorization	8	7674	5485	2189	16560	text, label	macro-F1: 93.99, accuracy: 97.48	macro-F1: 97.9	https://paperswithcode.com/sota/text-classification-on-r8		orig wo punct, lemm, lemm wo stopwords
3	R52	Text categorization	52	9100	6532	2568	18436	text, label	macro-F1: 77.62, accuracy: 95.09	macro-F1: 95	https://paperswithcode.com/sota/text-classification-on-r52	very uneven lebel distribution	orig wo punct, lemm, lemm wo stopwords
4	AG's news	Text categorization	4	127600	120000	7600	59371	title, text, label	macro-F1: 91.93, accuracy: 91.94	Accuracy: 95.51	https://paperswithcode.com/sota/text-classification-on-ag-news		orig, lemm, lemm wo stopwords
5	Ohsumed	Text categorization	23	23166	10433	12733	30088	text, label	macro-F1: 40.30, accuracy: 45.40	Accuracy: 68.5	https://paperswithcode.com/sota/text-classification-on-ohsumed		orig, lemm, lemm wo stopwords
6	20NG	Text categorization	20	18821	11293	7528	62707	text, label	macro-F1: 83.54, accuracy: 84.29	Accuracy: 88.5	https://paperswithcode.com/sota/text-classification-on-20news		orig, lemm, lemm wo stopwords
7	DBPedia	Text categorization	14	630000	560000	70000	650142	title, text, label	macro-F1: 97.84, accuracy: 97.84	Accuracy: 99.38	https://paperswithcode.com/sota/text-classification-on-dbpedia		orig, lemm, lemm wo stopwords
8	IMDb	Sentiment analysis	2	50000	25000	25000	81012	text, score, label	macro-F1: 88.54, accuracy: 88.54	Accuracy: 96.21	https://paperswithcode.com/sota/text-classification-on-imdb	This dataset contains 50000 unsupervised documents
9	Amazon 2	Sentiment analysis	2	4000000	3600000	400000	1666110	text, label	macro-F1: 86.95, accuracy: 86.95	Accuracy: 97.6	https://paperswithcode.com/sota/text-classification-on-amazon-2
10	Yelp 5	Sentiment analysis	5	700000	650000	50000	319389	text, label	macro-F1: 55.31, accuracy: 56.00	Accuracy: 73.20	https://paperswithcode.com/sota/text-classification-on-yelp-5
11	Sogou News	Text categorization	5	510000	450000	60000	147540	title, text, label	macro-F1: 93.77, accuracy: 93.77	Accuracy: 97.25	https://paperswithcode.com/sota/text-classification-on-sogou-news http://academictorrents.com/details/b2b847b5e1946b0479baa838a0b0547178e5ebe8	Chinese dataset\
12	Eurlex	Text categorization	4108	51000	45000	6000	46804	header, main_body, title, type, attachments	macro-F1: 21, accuracy: 85			Multilabel	lemmatized
13	Atis	Text categorization	8	5634	4834	800	837	text, label	macro-F1: 90.78, accuracy: 88.41			intent	orig
14	Atis+Twitter	Text categorization	2	22941	17842	5099	13524	text, label	macro-F1: 100., accuracy: 100.			SPAM	orig
15	SMS	Text categorization	2	5016	3901	1115	7234	text, label	macro-F1: 96.73, accuracy: 98.22
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100