知識ベースを活用した自然言語処理の手法とその応用

知識ベースを活用した自然言語処理の手法とその応用

1. 1. 

2. 2. STUDIO OUSIA ‣ ‣ ‣ ‣ ‣ ‣ 2
3. 3. STUDIO OUSIA 最近のニュース ‣ 
 ‣ ✦ ✦ 3 Human-Computer Question Answering Match @ NIPS 2017
4. 4. STUDIO OUSIA 4
5. 5. STUDIO OUSIA ‣ ‣ ‣ 5
6. 6. STUDIO OUSIA ‣ 
 
 ‣ 
 6 

7. 7. STUDIO OUSIA
8. 8. Entity Linking
9. 9. STUDIO OUSIA ‣ ‣ ‣ ‣ 9 /wiki/Frozen_(2013_film) New Frozen Boutique to Open at Disney's Hollywood Studios /wiki/
 The_Walt_Disney_Company /wiki/
 Disney’s_Hollywood_Studios
10. 10. STUDIO OUSIA ‣ 
 ‣ 10
11. 11. STUDIO OUSIA ‣ 
 
 
 ‣ 
 11
12. 12. STUDIO OUSIA ‣ ‣ ‣ ‣ ‣ Wikipedia link graph Neighboring words of words and anchors Aristotle was a philosopher + Logic Science Europe Socrates Renaissance Metaphysics Philosopher Philosophy AvicennaAristotle Plato Yamada et al., Joint Learning of the Embedding of Words and Entities for Named Entity Disambiguation. CoNLL 2016. Yamada et al., Learning Distributed Representations of Texts and Entities from Knowledge Base. TACL 2017.
13. 13. STUDIO OUSIA 13 [CoNLL 2016] [TACL 2017]
14. 14. STUDIO OUSIA ‣ ‣ 14 

15. 15. STUDIO OUSIA ‣ ‣ ‣ 15
16. 16. STUDIO OUSIA 16 /wiki/Frozen_(2013_film) New Frozen Boutique to Open at Disney's Hollywood Studios /wiki/
 The_Walt_Disney_Company /wiki/
 Disney’s_Hollywood_Studios ‣ ‣ ‣
17. 17. STUDIO OUSIA 17 /wiki/Frozen_(2013_film) New Frozen Boutique to Open at Disney's Hollywood Studios /wiki/
 The_Walt_Disney_Company /wiki/
 Disney’s_Hollywood_Studios ‣ ✦ ✦
18. 18. STUDIO OUSIA ‣ ‣ ✦ ✦ ✦ 
 19 apple Apple Inc. Apple (fruit)
19. 19. STUDIO OUSIA ‣ ‣ 20 Mention Candidate Generation Mention Detection and Disambiguation
20. 20. STUDIO OUSIA ‣ 
 ‣ ‣ ✦ ✦ ✦ 21 Mention Candidate Generation Mention Detection and Disambiguation
21. 21. STUDIO OUSIA ‣ ‣ 22
22. 22. Noisy User-generated Text Shared Task
23. 23. STUDIO OUSIA ‣ ✦ ✦ ‣ ✦ ✦ ✦ ✦ 24
24. 24. STUDIO OUSIA 25 Performances of the proposed systems at segmenting entities Performances of the proposed systems at both segmentation and classification tasks NLANGP nrc multimedialab USFD
25. 25. STUDIO OUSIA 
 26
26. 26. STUDIO OUSIA ‣ 
 ‣ ‣ ‣ ‣ 
 
 28 With%the%assistence%of%his%chief%minister,%the%Duc%de%Sully,%he%lowered% taxes%on%peasantry,%promoted%economic%recovery,%and%ins:tuted%a%tax%on% the%Paule<e.%Victor%at%Ivry%and%Arquet,%he%was%excluded%from%succession% by%the%Treaty%of%Nemours,%but%won%a%great%victory%at%Coutras. Henry%IV%of%France
27. 27. STUDIO OUSIA ‣ ‣ ‣ 29 Neural Quiz Solver Neural Type Predictor Information Retrieval Models Answer Scorer (GBRT) Neural network models
28. 28. STUDIO OUSIA Words Entities Sum Average The protagonist of … Protagonist Novel Author … … Answers Franz Kafka Tokyo Calcium Dot Softmax ‣ ‣ ‣ ‣ 30 The protagonist of a novel by this author is evicted from the Bridge Inn and is talked into becoming a school janitor…
29. 29. STUDIO OUSIA ‣ ‣ ‣ ‣ Wikipedia link graph Neighboring words of words and anchors Aristotle was a philosopher + Logic Science Europe Socrates Renaissance Metaphysics Philosopher Philosophy AvicennaAristotle Plato Yamada et al., Joint Learning of the Embedding of Words and Entities for Named Entity Disambiguation. CoNLL 2016. Yamada et al., Learning Distributed Representations of Texts and Entities from Knowledge Base. TACL 2017.
30. 30. STUDIO OUSIA Words Entities Sum Average The protagonist of … Protagonist Novel Author … … Answers Franz Kafka Tokyo Calcium Dot Softmax ‣ ‣ ‣ ‣ 32 The protagonist of a novel by this author is evicted from the Bridge Inn and is talked into becoming a school janitor…
31. 31. STUDIO OUSIA ‣ ‣ ‣ 33 The, protagonist, of, a, novel… Neural Type Predictor (CNN) Words Person, Author
32. 32. STUDIO OUSIA ‣ ‣ ‣ 34
33. 33. STUDIO OUSIA ‣ 
 ‣ ✦ ✦ ‣ 35
34. 34. STUDIO OUSIA ‣ ‣ ‣ 36 Neural Quiz Solver Neural Type Predictor Information Retrieval Models Answer Scorer (GBRT) Neural network models
35. 35. STUDIO OUSIA ‣ ‣ ✦ ✦ 37 AI間の対戦でのシステムのスコア クイズエクスパートとの対戦の様子
36. 36. STUDIO OUSIA ‣ ‣ ‣ 39
37. 37. STUDIO OUSIA ‣ ✦ ✦ ✦ ‣ ✦ ✦ 
 40