Slide 1
Slide 1 text
TryEnglishLT (2019/6/27)
K-means clustering of users
using Slack reactions
@zawawahoge
beBit
Slide 2
Slide 2 text
•@zawawahoge
•beBit (joined in this April)
•Software engineer
• Infra engineer
• Data analysis
Love for Python!
Slide 3
Slide 3 text
No content
Slide 4
Slide 4 text
Do you use slack?
Slide 5
Slide 5 text
Many emojis are available!
Slide 6
Slide 6 text
Custom emoji can be added!
Slide 7
Slide 7 text
Emoji is actually text!
:sushi:
Slide 8
Slide 8 text
What is “Slack reaction”?
Slack reactions make communication easier!
Slide 9
Slide 9 text
Motivation:
Who uses emojis as like you use?
Slide 10
Slide 10 text
Slack API
Is Useful When you want
to...
• Get a List of
• emojis
• reactions
• posted messages
• profiles of all members
• Post messages
• Upload files
From Slack Web API
Slide 11
Slide 11 text
Slack API which I use TODAY
Ø /api/users.list
Ø To list team members
Ø /api/reactions.list
Ø To list reactions for a member
Ø /api/emoji.list
Ø To list all emojis
Slide 12
Slide 12 text
All reactions of all users are obtained
• 90,000 reactions
• 150 users
• 2,000 types of
emojis
Slide 13
Slide 13 text
No content
Slide 14
Slide 14 text
K-means clustering
1. Data shaping
Remove top3 emojis (all users use them)
2. Determine the number of clusters
4 clusters to get simple results
Then...
K-means clustering!
Slide 15
Slide 15 text
Results
Use default emojis
Use custom emojis
Use emoji to praise
酒(sake)
(60)
(38)
(11)
(46)
Slide 16
Slide 16 text
Good English skill will
Change your life