Slide 1

Slide 1 text

LINE Taiwan Data Engineering Team

Slide 2

Slide 2 text

Who we are: Mission AI-enable Applications Business Intelligence

Slide 3

Slide 3 text

Who we are: Mission AI-enable Applications Business Intelligence Marketing Tools LINE Family Services LINE TODAY LINE SHOPPING LINE SPOT LINE MUSIC LINE Sticker Timeline LINE POINTS Official Account Fact Checker LINE HELP TW LINE Travel Ads

Slide 4

Slide 4 text

What We Do: ML/DS Marketing Tools LINE Family Services LINE TODAY LINE SHOPPING LINE SPOT LINE MUSIC LINE Sticker Timeline LINE POINTS Fact Checker LINE HELP TW LINE Travel NLP Knowledge Graph Uplift Modeling Official Account Ads NER Classifier Duplication Detector Auto completion Keyword Extraction Related Search Text Generation User Tagging Data Analytics Recom- mendation

Slide 5

Slide 5 text

POI Recommendation Recommend personalized POIs - Increase user engagement - Consider cold start users #Recommendation #LINE SPOT

Slide 6

Slide 6 text

Article Auto-tagging Automatic extract keywords from articles - Increase page views - Find popular topics #NLP #Keyword Extraction #LINE TODAY #LINE SHOPPING #LINE Travel

Slide 7

Slide 7 text

Advertisement User targeting #Knowledge Graph #LINE POINTS Find new users by Knowledge Graph - Increase users

Slide 8

Slide 8 text

Advertisement User targeting #Knowledge Graph #LINE POINTS Find new users by Knowledge Graph - Increase users User A New User User B belong to belong to click click click OA 1 click? follow follow belong to click Category 2 Category 1 Campaign B Campaign A Campaign C

Slide 9

Slide 9 text

Official Account User targeting #Uplift Modeling #LINE Sticker Find persuadable users by Uplift Modeling - Reduce marketing cost - Reduce mute rate

Slide 10

Slide 10 text

Official Account User targeting #Uplift Modeling #LINE Sticker Find persuadable users by Uplift Modeling - Reduce marketing cost - Reduce mute rate No Yes No Yes Buy if not treated Buy if treated Uplift Model

Slide 11

Slide 11 text

What We Do: MLOps NLP Knowledge Graph Uplift Modeling NER Classifier Duplication Detector Auto completion Keyword Extraction Related Search Text Generation User Tagging Data Analytics Recom- mendation ML Ops Dev Notebook Feature Eng. Model version Monitoring GPU Automation

Slide 12

Slide 12 text

Data Engineer ML Engineer Data Scientist Team Structure Define the service metrics Data-driven decision Research & apply ML algorithm Massive dataset processing ML Pipeline (MLOps)

Slide 13

Slide 13 text

We Are Hiring! Data Engineer Data Scientist Data Pipeline architecture ML Pipeline (MLOps) Massive dataset processing Data-driven decision Define the service metrics Visualize and simplify the data 3+ years of experience Qualifications Responsibilities Responsibilities Advanced SQL skills A product mindset Experience building distributed service Qualifications Experience with big data tools (Hadoop, Spark,…) Experience applying ML to real world problems

Slide 14

Slide 14 text

Thank You