Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Learning Dual Retrieval Module for Semi-supervi...
Search
Ryusuke_Tanaka
October 01, 2019
Technology
0
86
Learning Dual Retrieval Module for Semi-supervised Relation Extractionの紹介
Learning Dual Retrieval Module for Semi-supervised Relation Extractionの紹介です。
Ryusuke_Tanaka
October 01, 2019
Tweet
Share
More Decks by Ryusuke_Tanaka
See All by Ryusuke_Tanaka
医師向けQAサイトのための推薦システム開発
ryusuketa
1
1.7k
An Effective Approach to Unsupervised Machine Translationの紹介
ryusuketa
0
120
Universal Decompositional Semantics on Universal Dependencies
ryusuketa
0
88
動画視聴を整数倍(最大値)で_効率化するchrome extension作った
ryusuketa
0
83
双曲空間への単語埋め込みと QAサービスでの自然言語処理を 用いた推薦システムについて
ryusuketa
0
600
Other Decks in Technology
See All in Technology
エンジニアとして長く走るために気づいた2つのこと_大賀愛一郎
nanaism
0
210
Agentic Coding 実践ワークショップ
watany
24
19k
Databricks Free Edition講座 データエンジニアリング編
taka_aki
0
2.7k
Models vs Bounded Contexts for Domain Modularizati...
ewolff
0
210
BiDiってなんだ?
tomorrowkey
1
290
AI Agent Standards and Protocols: a Walkthrough of MCP, A2A, and more...
glaforge
0
450
なぜCREを8年間続けているのか / cre-camp-4-2026-01-21
missasan
0
980
ソフトとハード両方いけるデータ人材の育て方
waiwai2111
1
540
Kaggleコンペティション「MABe Challenge - Social Action Recognition in Mice」振り返り
yu4u
1
670
Kusakabe_面白いダッシュボードの表現方法
ykka
0
360
かわいい身体と声を持つ そういうものに私はなりたい
yoshimura_datam
0
250
形式手法特論:コンパイラの「正しさ」は証明できるか? #burikaigi / BuriKaigi 2026
ytaka23
17
6.4k
Featured
See All Featured
Visual Storytelling: How to be a Superhuman Communicator
reverentgeek
2
420
RailsConf & Balkan Ruby 2019: The Past, Present, and Future of Rails at GitHub
eileencodes
141
34k
Build The Right Thing And Hit Your Dates
maggiecrowley
38
3k
How To Stay Up To Date on Web Technology
chriscoyier
791
250k
Designing Experiences People Love
moore
143
24k
How Fast Is Fast Enough? [PerfNow 2025]
tammyeverts
3
430
Jamie Indigo - Trashchat’s Guide to Black Boxes: Technical SEO Tactics for LLMs
techseoconnect
PRO
0
47
Navigating the moral maze — ethical principles for Al-driven product design
skipperchong
2
230
Responsive Adventures: Dirty Tricks From The Dark Corners of Front-End
smashingmag
254
22k
Deep Space Network (abreviated)
tonyrice
0
35
SEO for Brand Visibility & Recognition
aleyda
0
4.2k
Conquering PDFs: document understanding beyond plain text
inesmontani
PRO
4
2.2k
Transcript
Learning Dual Retrieval Module for Semi- supervised Relation Extraction
h$>Y 4X?f7]Oc=l ,)8#d 4,),)l9N#e 4,)kl9NiI!,)8aKTD,)`m#G p]Oc=l9N.5, C6f7:j !+*%#ogH _L
! nB" bB" c=:j nB" ,)# q 4l9S<4PES<,+(U\ 4l9#6,) FBkeyword matching^A;^VZ@ 4l9103(&2#-'+/J R ! rlearning-to-rank#d WQZM[N
Introduction 5$-%/" entity 5$+-% 63#'47 ,1 .*)2&.*)2distant supervision0( !
self-training self-training: %" ,!-$unlabeled (*#. )
, proposed: '+& ,self-training-
relation relation
: relation prediction relation retrieval :unlabelled
Prediction !
! learning-to-rank" Pointwise Approach:
Pairwise Approach :
Unlabeled& ". unlabeled self-training & %, * →+(%,
→& ".)# /". !'$ )- EM
TACRED:
None