Slide 1
Slide 1 text
No content
Slide 2
Slide 2 text
> Jim Horng / LINE TODAY | Data TF
How Machine Learning helps
LINE Fact Checker
Slide 3
Slide 3 text
Agenda
> The Need for Fact Checker
> Overview of ML Components
> Overview of ML System
> Challenges and Future Work
Slide 4
Slide 4 text
※Source from Digital Society Project
Severity of Fake Messages
Slide 5
Slide 5 text
Fact Checker OA, Dashboard
OA - Query OA - Report Dashboard
Slide 6
Slide 6 text
Agenda
> The Need for Fact Checker
> Overview of ML Components
> Overview of ML System
> Challenges and Future Work
Slide 7
Slide 7 text
AI Knows Everything?
Slide 8
Slide 8 text
How ML Helps?
Slide 9
Slide 9 text
How ML Helps?
Similar
Messages Total Messages
Verified
Messages
ML Near-Duplication
ML Classification
Slide 10
Slide 10 text
Near-Duplication - Use Cases
Verified Fake Message:
"footage on Captain's Instagram Stories showed them wearing wedding rings on their both hands,
which proves Captain America and Captain Marvel get married in Las Vegas"
Query Result Type
Captain America and Captain Marvel get married in Las Vegas True Partial
The wedding in Las Vegas is hosted by Captain America
and Captain Marvel couple
True Semantically
Ironman and Black Widow get married in Los Angeles False Syntactically
Slide 11
Slide 11 text
Near-Duplication - Flow
> Long Text ➔ Full Match
• performs faster and trustworthy
> Short Text ➔ Partial Match + Fuzzy Tolerance
• 20% user query are partial texts of original
messages
( E.g. sentence, topic of an article )
Slide 12
Slide 12 text
Full Match - BERT Based
※Source from
https://towardsdatascience.com/bert-explained-state-of-the-art-language-model-for-nlp-
f8b21a9b6270
> Has Chinese pre-trained model
> Can extract sentence vector from
Upstream
> Can capture semantics based on
Context
Slide 13
Slide 13 text
Topic probability:
traffic: 3.7%, life: 38%, art: 49%, health: 2.54%, others: 1.6%,
sport: 4.3%, education: 0.7%, law: 0.16%
"They were in Vegas for the BillBoard Music Awards but, a few
hours later, footage on Captain's Instagram Stories showed
them wearing wedding rings on their both hands,
which proves Captain America and Captain Marvel get married
in Las Vegas "
Message Classification
Slide 14
Slide 14 text
Message Classification
> BERT + NN Layer
> Fine-Tuned (Accuracy) vs.
Pre-Trained (Speed)
Slide 15
Slide 15 text
Agenda
> The Need for Fact Checker
> Overview of ML Components
> Overview of ML System
> Challenges and Future Work
Slide 16
Slide 16 text
ML System Under the Hood
Serving
Training Data
Ingestion
Async
Execution
Model
Management
Model
Deployment
Index-based
Vector Search
(ANN)
Scheduling,
Orchestration
Slide 17
Slide 17 text
Agenda
> The Need for Fact Checker
> Overview of ML Components
> Overview of ML System
> Challenges and Future Work
Slide 18
Slide 18 text
Similarity 65% 46%
82%
Identify Message With Image
> VGG16 (ConvNet Configuration D)
> Convolution network extracts image features, Able to capture
objects and shape.
Slide 19
Slide 19 text
Duplicated Reported Messages
Training
Store as New
If New, by Near-
Duplication
Report
Message
Check By
Cache / Search
Engine
Store As
Cache / Search
Engine
House Keeping:
Find Duplicates
By Near-
Duplication and
Merge
Training
Store as New
If New, by Near-
Duplication
Report
Message
Slide 20
Slide 20 text
Knowledge Graph Model
> Knowledge Graph Triples
> Inference
> Fusion with other models
Slide 21
Slide 21 text
Thank You