What do people use to access Machine Learning for Kids?

I use Cabin for analytics on Machine Learning for Kids. (If you’re not familiar with them, their blog post on how to do analytics in a way that prioritizes user privacy is worth a read – the approach is simple but elegant. And you can see a demo of what a Cabin dashboard looks like.).

I thought it might be interesting to share what Cabin tells me about who has used Machine Learning for Kids over the last seven days.

What Operating Systems are people using?

Operating System Uniques
Windows 404,873
iOS 132,971
macOS 67,848
Android 55,176
Mac OS 35,743
Chrome OS 23,536
Linux 21,852
Ubuntu 10,780
Chromium OS 8,484
HarmonyOS 408
Raspbian 31
OpenHarmony 17
PlayStation 13
Tizen 10
android 3

At work, I’m mostly surrounded by MacBooks and don’t often see a Windows computer. It’s easy to assume that is normal, so this is a reminder that I’m in a bit of a bubble. Windows is still dominant.

Interesting to see “macOS” and “Mac OS” separate (I was tempted to combine them, but I decided to leave the data I get from Cabin as-is.)

My favourite part of looking at this is wondering who are the thirteen people who visited my site from a PlayStation???

What Browsers are people using?

Browser Uniques
Chrome 371,578
Edge 113,331
Mobile Safari 89,409
Safari 80,781
Mobile Chrome 32,831
Firefox 30,271
GSA 24,375
Opera 4,944
Samsung Internet 3,787
Samsung Browser 1,986
Yandex 1,973
Snapchat 1,039
Chrome WebView 784
Huawei Browser 778
Instagram 749
WebKit 683
Android Browser 663
NAVER 454
Avast Secure Browser 337
WeChat 305
Mobile Firefox 249
MIUI Browser 223
Facebook 208
Whale 174
Electron 170
QQBrowser 126
AVG Secure Browser 111
DuckDuckGo 103
Vivo Browser 93
Baidu 89
Quark 89
HeyTap 80
Opera Touch 50
Chromium 49
Sogou Explorer 48
KAKAOTALK 36
Zalo 33
Silk 29
Ecosia 21
Smart Lenovo Browser 21
Line 20
LinkedIn 19
TikTok 18
MetaSr 12
Opera GX 11
UCBrowser 10
PaleMoon 5
baiduboxapp 4
360 3
Vivaldi 3
IE 3
QQ 3
Twitter 2
Oculus Browser 2
Brave 1
Bing 1
Sleipnir 1
Comodo Dragon 1

This is a loooong list – and I haven’t heard of some of them before!

Unsurprising that Chrome is so high up the list. More surprising that two users used a virtual reality headset! (Oculus Browser)

What Screen Size do people use?

Screen size Percentage
1024px – 2400px 57.8
2400px or more 28.8
1024px or less 13.4

This one surprised me. Nearly thirty percent of users have a huge monitor!

It’s a shame that the 58% category is so broad – there is a huge difference between 1024px and 2400px.

What Device Types do people use?

Device Percentage
Desktop 75.2
Mobile 20.9
Tablet 3.9

The nature of the projects that students make on the site (working for up to an hour in Scratch) means that I expect it to skew towards desktops and laptops. This one was roughly what I expected.

What Language do users use?

Language Uniques
English 330,229
Arabic 205,287
Spanish 71,151
Chinese 33,385
German 30,202
Korean 17,271
Italian 12,002
French 7,575
Turkish 7,573
Russian 7,514
Catalan 6,177
Greek 5,658
Vietnamese 4,610
Portuguese 4,098
Polish 3,090
Indonesian 2,746
Czech 2,542
Romanian 1,649
Lithuanian 1,340
Japanese 1,250
Ukrainian 975
Dutch 752
Thai 600
Hebrew 499
Danish 448
Swedish 443
Croatian 438
Estonian 434
Hungarian 261
Slovak 219
Persian 203
Serbian (Cyrillic) 139
Slovenian 119
Galician 105
Uzbek 69
Bulgarian 69
Finnish 68
Mongolian 62
Norwegian Bokmal 39
Kazakh 34
Basque 30
Welsh 29
Hindi 28
Latvian 27
Malay 22
Marathi 13
Tamil 13
Albanian 12
Burmese 12
Azerbaijani 11
Gujarati 7
Afrikaans 7
Bengali 7
Icelandic 7
Armenian 7
Amharic 6
Georgian 6
Yoruba 6
Quechua 5
Khmer 5
Telugu 5
Swiss German 4
Bosnian 4
Najdi Arabic 4
ml 4
ne 3
Norwegian 3
rwk 3
Oromo 3
Maori 2
chm 2
iw 2
Scottish Gaelic 2
Urdu 2
Xhosa 2
Luxembourgish 2
ksh 2
chrome 2
Occitan 1
Serbo-Croatian 1
Komi 1
Belarusian 1
Kyrgyz 1
Corsican 1
Dhivehi 1
ewo 1
scn 1
Sorani (Kurdish) 1
Sundanese 1
Oriya 1
Sindhi 1
Filipino 1
Punjabi 1
Aymara 1
Kannada 1
Pashto 1
Igbo 1

This was, disappointingly, more skewed towards English than I expected. There is a very long tail, with a few languages that I hadn’t heard of before!

What Countries are users from?

Country Uniques
Saudi Arabia 254,992
United Kingdom 75,632
Spain 54,847
India 45,255
United Arab Emirates 32,781
United States 29,434
Germany 23,836
Hong Kong 20,789
Mexico 19,449
Korea (South) 18,798
China 14,248
Indonesia 13,208
Italy 13,117
Viet Nam 8,817
Greece 8,466
Turkey 8,423
Singapore 6,641
Austria 6,292
Pakistan 4,611
France 4,600
Russian Federation 4,123
Canada 3,988
Romania 3,878
Egypt 3,811
Malaysia 3,622
New Zealand (Aotearoa) 3,571
Poland 3,431
Australia 3,328
Brazil 3,185
Czech Republic 2,900
Colombia 2,809
Thailand 2,713
Japan 2,405
Portugal 2,189
Switzerland 2,106
Kenya 2,062
Lithuania 1,891
Macau 1,875
Iran 1,856
Nigeria 1,537
Jordan 1,422
Ukraine 1,421
Taiwan 1,313
Morocco 1,312
Argentina 1,295
Netherlands 1,279
Kazakhstan 1,276
Qatar 1,257
Lebanon 1,256
Algeria 1,184
Ireland 1,170
Peru 1,160
Sri Lanka 1,099
Moldova 1,045
Chile 1,019
Mongolia 984
Bangladesh 833
Ecuador 807
Bermuda 785
Israel 781
Sweden 740
South Africa 710
Tunisia 695
Belgium 694
Bulgaria 650
Ghana 575
Estonia 565
Uzbekistan 564
Denmark 528
Armenia 524
Myanmar 477
Croatia (Hrvatska) 467
Philippines 443
Uruguay 440
Luxembourg 433
Senegal 352
Kuwait 349
Oman 348
Serbia 335
GG 333
Guatemala 329
Bahrain 321
Aruba 314
Kyrgyzstan 312
Slovak Republic 289
Albania 271
Venezuela 264
Cyprus 255
Nepal 254
Ethiopia 252
Hungary 251
Georgia 246
Maldives 241
Malta 241
Cambodia 239
Uganda 238
Yemen 231
Finland 222
Tanzania 210
Syria 208
XK 198
Iraq 195
Honduras 173
Bosnia and Herzegovina 170
Belarus 169
Slovenia 166
Puerto Rico 154
Guyana 134
El Salvador 134
Norway 132
Latvia 119
PS 118
Panama 114
Bolivia 112
Azerbaijan 111
Dominican Republic 98
Nicaragua 96
New Caledonia 90
Mauritius 86
Unknown 74
Guinea 73
Zambia 64
Paraguay 57
Macedonia 50
Botswana 46
Haiti 44
Iceland 44
Trinidad and Tobago 41
Jamaica 41
Costa Rica 39
Cameroon 37
ME 36
Liberia 31
Libya 29
Rwanda 28
Laos 25
Reunion 24
Guadeloupe 24
Mauritania 23
JE 19
Cote D’Ivoire (Ivory Coast) 17
Sudan 17
Zimbabwe 15
CD 14
Benin 12
Congo 12
Barbados 12
Namibia 10
Djibouti 10
Seychelles 9
Mozambique 7
Angola 7
Tajikistan 7
Somalia 7
IM 7
Madagascar 7
Mali 7
Papua New Guinea 6
Fiji 6
Brunei Darussalam 6
SB 6
Suriname 6
Gambia 5
Andorra 5
Togo 5
Montserrat 4
Anguilla 4
Sierra Leone 4
Afghanistan 4
Malawi 4
Martinique 4
Gabon 4
Virgin Islands (U.S.) 4
Saint Kitts and Nevis 3
Burkina Faso 3
Saint Lucia 2
Cuba 2
Turks and Caicos Islands 2
Cayman Islands 2
Bahamas 2
Saint Vincent and the Grenadines 2
Grenada 2
Antigua and Barbuda 2
Comoros 2
Chad 2
Tonga 2
Swaziland 2
Monaco 1
Lesotho 1
TL 1
French Guiana 1
Liechtenstein 1
Burundi 1
Virgin Islands (British) 1
Gibraltar 1
Belize 1
French Polynesia 1
Guam 1
Niger 1
Bhutan 1
CW 1
AX 1
BL 1
SS 1
Cape Verde 1

Cabin estimate user’s location from IP address. As they only doing it at a country-level, I assume it’s reasonably accurate. The sorts of things that lead to IP addresses pointing to a different country (such as the corporate network in my office exiting to the Internet in mainland Europe) don’t seem likely for the schools and code clubs that are my typical users.

The number of users from Saudi Arabia was unusual – I think there might’ve been an event in the last week.

How many cloud projects do users create each day?

Date Number of cloud projects created
Wed 22 Apr 20,805
Tue 21 Apr 22,055
Mon 20 Apr 20,079
Sun 19 Apr 13,751
Sat 18 Apr 11,121
Fri 17 Apr 12,260
Thu 16 Apr 18,009
Wed 15 Apr 19,359
Tue 14 Apr 17,630
Mon 13 Apr 15,198
Sun 12 Apr 11,179
Sat 11 Apr 6,191
Fri 10 Apr 11,660
Thu 9 Apr 18,533

Cabin only tells me unique visitors to the site. Not all visitors are students. Not all visitors create ML projects.

It’s difficult for me to know how many users actually use the site to create machine learning projects.

I don’t require users to log in unless they want to save a project and work on it across multiple days. I don’t have any data on what proportion of my users take advantage of the site’s support for using it anonymously. Anecdotally, I hear it’s more common than users who log in. (Primarily because it’s so much simpler to run a class with young children if they don’t need to enter a username and password!) But either way, any logins or user accounts I counted would only give a partial view.

If users store their projects on their own computer then I don’t need to enforce limits on how many projects they store or how big they get. But that means I don’t have any way to count how many such projects are created.

I can only count the projects that users store in the cloud, on my servers. That is how I get the numbers in the table above.

It can only be a partial view, but it’s still interesting. For example, weekends are noticeably quieter. That’s perhaps to be expected for a site used in a lot of schools.

I enjoy looking at data

I don’t have much of a conclusion here. This sort of data is useful as a guide for things like translation effort, and having reasonable expectations for browser capabilities as I consider moving more of the site’s computation out of the cloud and on-device.

Plus, I find stats interesting!

Tags: ,

Leave a Reply