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
(Reading )Does BERT Know that the IS-A Relation...
Search
ando
January 23, 2023
Science
120
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
(Reading )Does BERT Know that the IS-A Relation Is Transitive?
Komachi lab
ando
January 23, 2023
More Decks by ando
See All by ando
博士論文公聴会資料
ando55
0
470
Is In-hospital Meta-information Useful for Abstractive Discharge Summary Generation?
ando55
0
270
(読み会)Evaluating Factuality in Text Simplification
ando55
1
210
(Reading) Relational Multi-Task Learning Modeling Relations between Data and Tasks
ando55
0
200
(Reading )Agreeing to Disagree: Annotating Offensive Language Datasets with Annotators’ Disagreement
ando55
0
180
(Reading) Preregistering NLP research
ando55
0
73
(Reading) Predictive Adversarial Learning from Positive and Unlabeled Data
ando55
0
160
Argument Invention from First Principles
ando55
2
350
Other Decks in Science
See All in Science
O(log n)-Approximation Algorithms for Bipartiteness Ratio
tasusu
0
130
チュートリアル:世界モデル
hf149
0
2.1k
プロジェクト「Azayaka」のSARの数式とジオメトリ
syuchimu
0
430
AIを用いた PID制御で部屋 の温度制御をしてみた
nearme_tech
PRO
0
200
生成AIが科学とRAにもたらしていること:メタサイエンスの視点から
rmaruy
0
120
Tensor Factorization Meets Deformed Information Geometry: Convex Relaxation under Deformed Algebra
gkazunii
0
130
明治薬科大学講義_ビッグデータ解析を支えるデータベース技術とクラウドコンピューティング
ktatsuya
1
150
共生概念の整理と AIアライメントの構想
hiroakihamada
0
240
因果探索の発展と展望
sshimizu2006
2
850
データベース02: データベースの概念
trycycle
PRO
2
1.4k
Kritische evaluatie van GenAI-output voor literatuuronderzoek
voginip
0
210
知能とはなにか -ヒトとAIのあいだ-
tagtag
PRO
0
110
Featured
See All Featured
The AI Revolution Will Not Be Monopolized: How open-source beats economies of scale, even for LLMs
inesmontani
PRO
3
3.7k
Agile Leadership in an Agile Organization
kimpetersen
PRO
0
200
Reflections from 52 weeks, 52 projects
jeffersonlam
356
21k
Conquering PDFs: document understanding beyond plain text
inesmontani
PRO
4
3k
Ecommerce SEO: The Keys for Success Now & Beyond - #SERPConf2024
aleyda
1
2.1k
How To Speak Unicorn (iThemes Webinar)
marktimemedia
1
520
How to Think Like a Performance Engineer
csswizardry
28
2.7k
Testing 201, or: Great Expectations
jmmastey
46
8.2k
Data-driven link building: lessons from a $708K investment (BrightonSEO talk)
szymonslowik
1
1.2k
Keith and Marios Guide to Fast Websites
keithpitt
413
23k
AI Search: Implications for SEO and How to Move Forward - #ShenzhenSEOConference
aleyda
1
1.3k
Collaborative Software Design: How to facilitate domain modelling decisions
baasie
1
280
Transcript
ACL2022 Short paper
- #&35͕*4"ؔͷ5SBOTJUJWFੑ࣭ͱͲͷఔҰக͢Δ͔ΛఆྔԽɻ - #&35ͷ༧ଌ*4"ؔͷਪҠੑʹશʹैΘͳ͍͜ͱ͕໌Β͔ʹͳͬͨ 5SBOTJUJWF A is-a B & →
A is-a C︖ B is-a C - #&35*4"ؔΛͷ"DDVSBDZͰ༧ଌͰ͖Δ - #&35͕"JTB#ͱ#JTB$Λਖ਼͘͠༧ଌͨ͠߹ɺ"JTB$ ˠͦͷ··ࣝϕʔεͱͯ͠ѻ͑ͳ͍ 2 概要
- ໊ࢺΛ༻͍Δ 3 Task definition 語義のペア Is-aであれば1, でなければ0
-\TcT㱨4 BOEIZQP T 㱵 ^ 8PSE/FUͷશͯͷϦʔϑηϯεɻ 48PSE/FUͷશͯͷޠٛͷू߹ɺIZQP T ҙຯTͷԼҐޠٛͷू߹Λද͢ɻ -
*4"ΛऔΓग़͢ • -͔ΒMFBGͷޠٛ/ΛϥϯμϜʹநग़͠ɺ/ΛSPPU3ʹଓͯ͠ɺύεQ /ˠ3 Λಘ Δ Q͔Β"ɺ#ɺ$ͷͭͷޠٛΛϥϯμϜʹநग़͠ɺ"ɺ#ɺ$ͷޠٛͷྫจΛಘΔ ͜ͷ݁Ռɺλϓϧʢ"ɺ#ɺ$ʣͱͭͷਖ਼ྫʢ"ɺ#ʣɺʢ#ɺ$ʣɺʢ"ɺ$ʣ͕ಘ ΒΕΔɻ - /PU*4"ΛऔΓग़͢ • -͔Βೋͭͷ/ /ΛϥϯμϜʹநग़͠ɺೋͭͷύεQ /ˠ3 ͱQ /ˠ3 Λ ಘΔQɺQ͔ΒͦΕͧΕ"⒌ɺ#⒌ΛϥϯμϜʹαϯϓϦϯά͠ɺޠٛ"⒌ɺ#⒌ͷྫจΛಘ Δ "⒌͕#⒌ˠ3ͷύε্ʹͳ͍߹ɺ·ͨͦͷٯͷ߹ɺෛྫʢ"⒌ #⒌ʣΛಘΔ - σʔληοτͰɺݸͷλϓϧ͕͋Γɺݸͷਖ਼ྫͱݸͷෛྫ͕͋Δɻ 4 Positive, Negative example抽出
- #&35ͷ࠷ऴͷϕΫτϧΛ༻ʢTVCXPSEBWFSBHFʣ - #&35ϑΝΠϯνϡʔχϯά͠ͳ͍ɺ࠷ۙʢOOʣྨثΛ༻ ˠγϯϓϧͳͷΛ͍͍ͨ - ϢʔΫϦουڑΛڑࢦඪͱͯ͠༻͍Δɻ 5 分類 語義2
語義1
- ςετηοτͰɺ " # ͱ # $ ͷฏۉϗοϓͦΕͧΕ ͱ Ͱ͋ͬͨɻ
- ͜ͷࠩɺ8PSE/FUͷྫจΛ࣋ͭޠ͕ٛۉʹ͍ͯ͠ͳ͍͜ͱʹىҼ͍ͯ͠Δɻ ʢ্ҐޠྫจΛ͍͍࣋ͬͯ͢ʣ 6 Hop数
- ϗοϓ͕গͳ͍߹ʢʣɺϗοϓ͕ଟ͍߹ʢʣΑΓߴ͍ਫ਼Ͱ*4"ؔ Λ༧ଌͰ͖Δ - ϗοϓ͕૿Ճͯ͠༧ଌਫ਼େ͖͘ԼͤͣɺϗοϓͰਫ਼ʹେ͖ͳΒ͖ͭ ͳ͍ɻ 7 実験結果 ?
- #&35͕ʢ"ɺ#ʣɺʢ#ɺ$ʣΛਖ਼͘͠༧ଌͨ͠߹ʹɺ*4"ؔʢ"ɺ$ʣΛਖ਼͘͠༧ଌ ͢ΔසΛଌఆ - Q "# ͷ " # ͷ͏ͪɺਖ਼͘͠༧ଌ͞Εͨ
" # ͷׂ߹ - Q "$c"# #$ ɺ " # ͱ # $ ͕ਖ਼͘͠༧ଌ͞ΕΔ߹ɺ " $ ͕ਖ਼͘͠༧ଌ͞ΕΔׂ߹ - "JTB$͕ৗʹਖ਼͘͠༧ଌ͞Εͳ͍͜ͱɺ#&35͕ཧతʹҰ؏ͨ͠༧ଌΛ͢ΔೳྗΛ ͍͍ܽͯΔ͜ͱΛ͍ࣔࠦͯ͠Δɻ 8 Prediction Ability for Transitivity