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
A Fast and Accurate Dependency Parser using Neural Networks
Search
youichiro
December 03, 2018
Technology
0
110
A Fast and Accurate Dependency Parser using Neural Networks
長岡技術科学大学
自然言語処理研究室
文献紹介(2018-12-04)
youichiro
December 03, 2018
Tweet
Share
More Decks by youichiro
See All by youichiro
日本語文法誤り訂正における誤り傾向を考慮した擬似誤り生成
youichiro
0
1.3k
分類モデルを用いた日本語学習者の格助詞誤り訂正
youichiro
0
61
Multi-Agent Dual Learning
youichiro
1
120
Automated Essay Scoring with Discourse-Aware Neural Models
youichiro
0
91
Context is Key- Grammatical Error Detection with Contextual Word Representations
youichiro
1
110
勉強勉強会
youichiro
0
63
Confusionset-guided Pointer Networks for Chinese Spelling Check
youichiro
0
150
A Neural Grammatical Error Correction System Built On Better Pre-training and Sequential Transfer Learning
youichiro
0
110
An Empirical Study of Incorporating Pseudo Data into Grammatical Error Correction
youichiro
0
150
Other Decks in Technology
See All in Technology
Databricksを活用してDELISH KITCHENのレシピレコメンドを開発した話
furu8
0
250
開発生産性向上サービスを作るFindyが自分たちで開発生産性を爆上げした組織づくりの歩み / Findy's path to boosting its own development productivity 2024-04-17
ma3tk
1
270
プロトタイピングによる不確実性の低減 / Reducing Uncertainty through Prototyping
ohbarye
3
190
マルチアカウント環境への発見的統制の導入
ch1aki
1
1.3k
Janus
bkuhlmann
0
480
普段有償でサポート業務をしているCSAが技術知見を無料で公開する理由
07jp27
1
630
Terraformあれやこれ/terraform-this-and-that
emiki
2
100
入社後初めてのタスクでk8sアップグレードした話.pdf
kkato1
0
380
長期間TiDBを使ってきた話 @ 私たちはなぜNewSQLを使うのかTiDB選定5社が語る選定理由と活用LT / Experiences with TiDB Over Time
chibiegg
2
580
転移学習とドメイン適応の基礎
kmatsui
2
570
Data and AI Governance: Existing Challenges and Emerging Trends
scotthsieh825
0
140
アプリがつくるNOT A HOTELブランド
hokuts
0
450
Featured
See All Featured
RailsConf & Balkan Ruby 2019: The Past, Present, and Future of Rails at GitHub
eileencodes
124
32k
Agile that works and the tools we love
rasmusluckow
323
20k
Bootstrapping a Software Product
garrettdimon
PRO
301
110k
Designing for humans not robots
tammielis
247
25k
個人開発の失敗を避けるイケてる考え方 / tips for indie hackers
panda_program
60
14k
How to name files
jennybc
64
92k
Mobile First: as difficult as doing things right
swwweet
216
8.6k
Done Done
chrislema
178
15k
Faster Mobile Websites
deanohume
296
30k
Side Projects
sachag
451
41k
CoffeeScript is Beautiful & I Never Want to Write Plain JavaScript Again
sstephenson
153
14k
Web Components: a chance to create the future
zenorocha
304
41k
Transcript
A Fast and Accurate Dependency Parser using Neural Networks Danqi
Chen and Christopher D. Manning Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) 2018-12-04 0
ü ü 1
• D • D P D D 2
3A B C B ) E AB : • ic
R p o L d f • ThgP GH F SL G n 2- 3 2- D ATek 2C Tb S a n 3 ( 2C Tek nI R HS n ,-3 ( 2C Tek I R HS A BC A T T C 3
- - He has good control . ROOT ))(- 4
( ( has good control . ROOT He ) -
5
( -) ) good control . ROOT He has (,
- 6
( ) ( ( good control . ROOT has ()-
,, He 7
( ) ( ( control . ROOT has good ()-
,, He 8
( ) ( ( . ROOT has good control ()-
,, He 9
( ) ( ( . ROOT has control ()- ,,
He good 10
) ) ) . ROOT has ( ) --, (
( ( ( He good control 11
) ) ) ROOT has . ( ) --, (
( ( ( ( He good control 12
) ) ) ROOT has ( ) --, ( (
( ( ( He good control . 13
) ) ) ROOT ( ) --, ( ( (
( ( He good control . has 14
• • 15
) TS Ø P Ø I ( - 16
• 17
18
19
Cube activation function: ! " = "$ 20
- 21
- - 22
ü C ü C c ü 23
24
ge l -2 15 l k H KKK H H
8G D / GEL A 6407 . D l d 6G8DH ED 8H 8GH D H KKK H H 8G D DEML G8DH ED 8H 8GH D l hb fTaOcm i H KKK H H 8G D DDEDE DE HH l N P U D D:L 58GH D D: ED :E N P U D D:L 8GH D 25