# 2021 NAVER Search Colloquium 발표 영상 공개

> Clean Markdown view of GeekNews topic #4496. Use the original source for factual precision when an external source URL is present.

## Metadata

- GeekNews HTML: [https://news.hada.io/topic?id=4496](https://news.hada.io/topic?id=4496)
- GeekNews Markdown: [https://news.hada.io/topic/4496.md](https://news.hada.io/topic/4496.md)
- Type: news
- Author: [xguru](https://news.hada.io/@xguru)
- Published: 2021-06-24T09:52:07+09:00
- Updated: 2021-06-24T09:52:07+09:00
- Original source: [naversearchconf.naver.com](http://naversearchconf.naver.com/)
- Points: 6
- Comments: 1

## Topic Body

총 6가지 세션으로 구성된 총 22개의 강연 동영상(한국어)

1. Learning to Rank

ㅤ- The Faimess-Utility Trade-off Problem

ㅤ- Neural Topic Models for Web Search

ㅤ- Optimizing Training Data in Learning to Rank

2. User Modeling

ㅤ- Fairness in Search & Recommender Systems

ㅤ- Classifying Web Documents by User Intent

ㅤ- Understanding POI: from Context to Embedding

3. eCommerce

ㅤ- AI-driven Shopping Search and Price Comparison for Billion-scale Products

ㅤ- Item2Item: Wider and Deeper Product Recommendation

ㅤ- Deep Local Search for Finding Hidden Meaning

ㅤ- Online Serving AD Relevance Model Using BERT

4. Platform

ㅤ- Search Engine and SRE Platform for Global Web Search System

ㅤ- Front-end Engineering for Building an Immersive Experience in Search

ㅤ- NAVER federated learning: Challenges and opportunities

ㅤ- NAVER Knowledge Graph Introduction

5. Vision

ㅤ- How to Improve Generalization in Visual Search

ㅤ- Creative Ways for Visual Search

ㅤ- StyleMapGAN: Exploiting Spatial Dimensions of Latent in GAN for Real-time Image Editing

ㅤ- Computer Vision for Autonomous Driving

6. Language AI

ㅤ- Knowledge Snippet: Direct Answer to a Search Query

ㅤ- Conversational AI with Big Generative LM

ㅤ- Meta-Learning for the low-resource domain

ㅤ- Search Applications with Big Generative LM

## Comments



### Comment 5488

- Author: xguru
- Created: 2021-06-24T09:52:13+09:00
- Points: 1

제목은 영어지만 1-1 세션 빼고는 모두 한국어 발표입니다.