Presented By
O’Reilly + Intel AI
Put AI to Work
April 15-18, 2019
New York, NY
Discover opportunities for applied AI
Organizations that successfully apply AI innovate and compete more effectively. How is AI transforming your business?
Be a part of the program—apply to speak by October 16.

AutoML in Chatbot Builder Framework

Jaewon Lee (LINE Corp.), Sihyeung Han (NAVER & LINE Corp)
4:05pm4:45pm Thursday, April 18, 2019
Implementing AI
Location: Rendezvous
Secondary topics:  Automation in machine learning and AI, Media, Marketing, Advertising, Models and Methods, Text, Language, and Speech

Who is this presentation for?

Deep Learning Model engineers who seek to implement better chatbot model and who always have any doubts with hyper-parameters for their own deep learning models, Data Coordinators who want to well-manage large data corpus by finding better embedding vector space for each query, Business Planners who want to apply Chatbot to their own business



Prerequisite knowledge

Attendees with high interest in NLP especially Chatbot are more than welcome to participate this session. This session does not require any deep scientific knowledge to understand complicated deep learning algorithms. However, attendees with some insights to use chatbot to their business model and know the difference between rule-based Chatbot and AI Chatbot might help them learn more.

What you'll learn

- Understand a detailed pipeline/structure of Chatbot Builder Framework - Understand each NLP engines, especially Dialogue Model - Learn how to adopt AutoML in Chatbot Builder Framework


Have you ever built Chatbot? There are lots of steps to build Chatbot and each step requires tremendous work to do. Our Chatbot Builder Framework only requires raw data corpus to create a high-performance chatbot on your own business domain. Just click ‘Build’ button on our framework. Then, our framework will manage data corpus by clustering into scenario, pre-process data, optimize hyper-parameters for deep learning models, propose the best ensemble-model for your domain and finally serve to multiple messenger platforms such as LINE, Facebook messenger. This entire pipeline will suggest the most optimized Chatbot model in the end and serve to users so that they can apply to their own business. Don’t hesitate to build Chatbot. We will easily provide simple, but excellent Chatbot for you.

On our Chatbot Builder framework, clustering all queries into similar clusters will help to easily manage large text/log data corpus. Also, auto hyper-parameter tuning will allow deep learning engineers to focus more on current model architecture or developing other cutting-edge models. In the end, each model will validate each status by computing self-evaluation scores from each model.

I would like to share the value of AutoML which allows us to provide better model and how AutoML can be applied in different NLP area, not just in Chatbot.

Here are some contents what I will going to cover on this session
— Contents
1. Necessity, but Difficulty of AutoML in NLP/Chatbot
2. Apply AutoML in Chatbot Builder Framework
3. Data Clustering and Hyper-parameter Tuning
4. Auto-Evaluation by models
5. Share current business cases in LINE by using AutoML in Chatbot

Photo of Jaewon Lee

Jaewon Lee

LINE Corp.

Jaewon Lee is a Data Scientist in NLP at NAVER & LINE Corp in South Korea. After his team, Company.AI, was acquired by NAVER & LINE Corp in July 2017, his team focuses on developing Clova Chatbot Builder Framework by applying NLP engines to lead customers to easily build and serve chatbot to their own business. Also his team focuses on NLP research to improve performance of their dialogue model. Previously Jaewon worked for Hana Financial Investment in South Korea as a Quantitative Data Analyst using machine learning algorithms to predict financial markets. He received B.S. degree in Applied Math and Statistics with Computer Science from Johns Hopkins University in 2015.

Sihyeung Han


Kevin Han is a business consultant and service planner of NAVER & LINE Corp, a Korean company known for the biggest domestic web portal(NAVER), mobile messenger(LINE), and AI-related solutions(CLOVA). A goal for his team is to provide ‘end-to-end’ AI service for clients, from improving dialogue model to consulting clients to create maximum value out of our Chatbot service. Previously, Kevin worked for IBM as a cognitive consultant. He holds a bachelor’s degree in Psychology and Business from New York University.

Leave a Comment or Question

Help us make this conference the best it can be for you. Have questions you'd like this speaker to address? Suggestions for issues that deserve extra attention? Feedback that you'd like to share with the speaker and other attendees?

Join the conversation here (requires login)