About MIND

MIcrosoft News Dataset (MIND) is a large-scale dataset for news recommendation research. It was collected from anonymized behavior logs of Microsoft News website. The mission of MIND is to serve as a benchmark dataset for news recommendation and facilitate the research in news recommendation and recommender systems area.

MIND contains about 160k English news articles and more than 15 million impression logs generated by 1 million users. Every news article contains rich textual content including title, abstract, body, category and entities. Each impression log contains the click events, non-clicked events and historical news click behaviors of this user before this impression. To protect user privacy, each user was de-linked from the production system when securely hashed into an anonymized ID. For more detailed information about the MIND dataset, you can refer to the following paper:

If you are interested in using this dataset in your research work, welcome to cite this paper:

Fangzhao Wu, Ying Qiao, Jiun-Hung Chen, Chuhan Wu, Tao Qi, Jianxun Lian, Danyang Liu, Xing Xie, Jianfeng Gao, Winnie Wu and Ming Zhou. MIND: A Large-scale Dataset for News Recommendation. ACL 2020.


Download

The MIND dataset is free to download for research purposes under Microsoft Research License Terms. Before you download the dataset, please read these terms and click below button to confirm that you agree to them.


This dataset can support many researches on news recommendation, and can be downloaded at:

We also provide the link to Microsoft Azure Open Datasets for eaiser access to our large dataset on cloud platform:

For more details about the data formats, please refer to this document:


In addition, to help the researchers get familiar with our data and run quick experiments, we release a small version of the MIND dataset by randomly sampling 50,000 users and their behavior logs from the MIND dataset. We name this dataset MIND-small. The training and validation sets of MIND-small can be downloaded at:


The implementation of several existing news recommendation methods and general recommendation methods can be found at Microsoft Recommenders.

Following is a step-by-step tutorial for generating predictions on the dataset:

Leaderboard

Rank Team AUC MRR nDCG@5 nDCG@10

1

Oct. 23, 2020
doubleQ 0.7129 0.3604 0.3956 0.4518

2

Oct. 23, 2020
dtsnu 0.7114 0.3568 0.3916 0.4485

3

Oct. 23, 2020
started2givingup 0.7113 0.3585 0.3933 0.4498

4

Oct. 23, 2020
fuxictr 0.7111 0.3578 0.3923 0.4489

5

Oct. 23, 2020
oahciy 0.7096 0.3540 0.3883 0.4454

6

Oct. 23, 2020
WONNIU 0.7095 0.3569 0.3916 0.4481

7

Oct. 23, 2020
NoahRS 0.7061 0.3552 0.3895 0.4459

8

Oct. 23, 2020
Ravox 0.7048 0.3505 0.3845 0.4416

9

Oct. 23, 2020
Qinne 0.7032 0.3496 0.3830 0.4397

10

Oct. 23, 2020
xyzhang 0.6980 0.3476 0.3814 0.4383

11

Oct. 23, 2020
YangZhenghong 0.6979 0.3480 0.3809 0.4377

12

Oct. 23, 2020
gcc_microsoft 0.6979 0.3479 0.3806 0.4373

13

Oct. 23, 2020
huailei 0.6968 0.3492 0.3812 0.4386

14

Oct. 23, 2020
hanyan 0.6964 0.3441 0.3764 0.4336

15

Oct. 23, 2020
Tmail 0.6952 0.3433 0.3755 0.4327

16

Oct. 23, 2020
lixmcm 0.6952 0.3438 0.3765 0.4335

17

Oct. 23, 2020
duchunning 0.6951 0.3428 0.3753 0.4324

18

Oct. 23, 2020
fanxiaoxing 0.6949 0.3448 0.3770 0.4336

19

Oct. 23, 2020
overlord 0.6942 0.3445 0.3767 0.4332

20

Oct. 23, 2020
veason 0.6941 0.3435 0.3751 0.4321

21

Oct. 23, 2020
hvv 0.6940 0.3456 0.3782 0.4349

22

Oct. 23, 2020
id500 0.6935 0.3451 0.3773 0.4344

23

Oct. 23, 2020
Ironball 0.6923 0.3440 0.3778 0.4351

24

Oct. 23, 2020
fit4you 0.6910 0.3413 0.3732 0.4307

25

Oct. 23, 2020
AndreaChao 0.6903 0.3413 0.3727 0.4299

26

Oct. 23, 2020
Ivy 0.6901 0.3414 0.3726 0.4299

27

Oct. 23, 2020
changebin 0.6888 0.3396 0.3706 0.4273

28

Sept. 01, 2020
yupei 0.6867 0.3404 0.3721 0.4288

29

Nov. 14, 2020
kkq10466 0.6862 0.3347 0.3644 0.4218

30

Oct. 23, 2020
Tsar 0.6862 0.3374 0.3676 0.4253

31

Sept. 01, 2020
majinc 0.6862 0.3397 0.3705 0.4274

32

Nov. 20, 2020
ConnollyLeon 0.6860 0.3363 0.3668 0.4237

33

Nov. 21, 2020
saltyfish 0.6856 0.3335 0.3633 0.4207

34

Nov. 15, 2020
qxe34517 0.6854 0.3345 0.3655 0.4229

35

Nov. 15, 2020
fzy37920 0.6854 0.3345 0.3655 0.4229

36

Nov. 29, 2020
Ruixinhua 0.6846 0.3335 0.3635 0.4209

37

Oct. 23, 2020
heroddaji 0.6845 0.3376 0.3677 0.4244

38

Oct. 25, 2020
faysir 0.6836 0.3343 0.3644 0.4219

39

Aug. 27, 2020
CyberFish 0.6834 0.3373 0.3683 0.4258

40

Oct. 23, 2020
again_and_again 0.6828 0.3365 0.3665 0.4240

41

Oct. 23, 2020
xinghuaz 0.6827 0.3385 0.3691 0.4259

42

Aug. 28, 2020
amgis3 0.6827 0.3320 0.3615 0.4196

43

Nov. 30, 2020
msj 0.6826 0.3335 0.3629 0.4202

44

Oct. 23, 2020
dengjia284 0.6814 0.3325 0.3614 0.4192

45

Oct. 23, 2020
ustc_jingang 0.6809 0.3337 0.3633 0.4208

46

Oct. 23, 2020
qqyysg 0.6799 0.3336 0.3644 0.4213

47

Nov. 13, 2020
YHu 0.6790 0.3338 0.3624 0.4192

48

Nov. 30, 2020
finegrained 0.6787 0.3346 0.3653 0.4221

49

Aug. 29, 2020
Reto-hitsz 0.6783 0.3288 0.3575 0.4153

50

Oct. 23, 2020
simple2better 0.6766 0.3311 0.3609 0.4184

51

Oct. 23, 2020
shahnawaz 0.6766 0.3282 0.3576 0.4149

52

Oct. 23, 2020
zidane4ever21 0.6710 0.3236 0.3523 0.4093

53

Oct. 23, 2020
dust 0.6691 0.3267 0.3539 0.4109

54

Nov. 28, 2020
TJ_Arthur 0.6646 0.3185 0.3460 0.4046

55

Oct. 23, 2020
aqweteddy 0.6566 0.3173 0.3425 0.3989

56

Oct. 23, 2020
xcw 0.6554 0.3172 0.3434 0.4000

57

Oct. 23, 2020
HJS_TJU 0.6512 0.3122 0.3357 0.3934

58

Oct. 23, 2020
AND-OR 0.6506 0.3079 0.3320 0.3898

59

Oct. 23, 2020
drew 0.6486 0.3087 0.3333 0.3905

60

Aug. 27, 2020
gaojx 0.6342 0.3112 0.3358 0.3919

61

Oct. 23, 2020
tcchiang 0.5950 0.2794 0.2974 0.3551

62

Aug. 26, 2020
learner 0.5847 0.2648 0.2786 0.3361

63

Oct. 23, 2020
only2233 0.5352 0.2408 0.2507 0.3075

64

Aug. 24, 2020
cs880141 0.5180 0.2327 0.2395 0.2953

65

Aug. 26, 2020
cdj0311 0.5059 0.2290 0.2350 0.2900

66

Aug. 26, 2020
zhoubang1 0.5009 0.2256 0.2316 0.2869

67

Aug. 25, 2020
bird 0.5002 0.2248 0.2304 0.2862