Journal of KIISE (정보과학회 논문지)

Korean Institute of Information Scientists and Engineers (한국정보과학회)
권호 표지
DOI QR코드

DOI QR Code

Korean Semantic Role Labeling Using Semantic Frames and Synonym Clusters

의미 프레임과 유의어 클러스터를 이용한 한국어 의미역 인식

  • 임수종 (한국전자통신연구원 지식마이닝연구실) ;
  • 임준호 (한국전자통신연구원 지식마이닝연구실) ;
  • 이충희 (한국전자통신연구원 지식마이닝연구실) ;
  • 김현기 (한국전자통신연구원 지식마이닝연구실)
  • Received : 2016.01.05
  • Accepted : 2016.04.26
  • Published : 2016.07.15
⟨ Previous Next ⟩

Abstract

Semantic information and features are very important for Semantic Role Labeling(SRL) though many SRL systems based on machine learning mainly adopt lexical and syntactic features. Previous SRL research based on semantic information is very few because using semantic information is very restricted. We proposed the SRL system which adopts semantic information, such as named entity, word sense disambiguation, filtering adjunct role based on sense, synonym cluster, frame extension based on synonym dictionary and joint rule of syntactic-semantic information, and modified verb-specific numbered roles, etc. According to our experimentations, the proposed present method outperforms those of lexical-syntactic based research works by about 3.77 (Korean Propbank) to 8.05 (Exobrain Corpus) F1-scores.

기계학습 기반의 의미역 인식에서 어휘, 구문 정보가 자질로 주로 쓰이지만, 의미 정보를 분석하는 의미역 인식은 의미 정보 또한 매우 유용한 정보이다. 그러나, 기존 연구에서는 의미 정보를 활용할 수 있는 방법이 제한되어 있기 때문에, 소수의 연구만 진행되었다. 본 논문에서는 의미 정보를 활용하는 방안으로 동형이의어 수준의 의미 애매성 해소 기술, 고유 명사에 대한 개체명 인식 기술, 의미 정보에 기반한 필터링, 유의어 사전을 이용한 클러스터 및 기존 의미 프레임 정보 확장, 구문-의미 정보 연동 규칙, 필수 의미역 오류 보정 등을 제안한다. 제안하는 방법은 기존 연구 대비 뉴스 도메인인 Korean Propbank는 3.77, 위키피디아 문서 기반의 Exobrain GS 3.0 평가셋에서는 8.05의 성능 향상을 보였다.

Keywords

Acknowledgement

Grant : (엑소브레인-1세부) 휴먼 지식증강 서비스를 위한 지능진화형 WiseQA 플랫폼 기술 개발

Supported by : 정보통신기술진흥센터

References

  1. R.Johanssonand P.Nugues, "Dependency-basedSemantic Role Labeling of PropBank," Proc. ofthe EMNLP-2008, 2008.
  2. S.Lim et al., "Korean Semantic Role Labeling Using Domain Adaptation Technique," Journal of KIISE, Vol. 42, No. 4, pp. 475-482, Apr. 2015. (in Korean) https://doi.org/10.5626/JOK.2015.42.4.475
  3. M.Palmer et al., "Korean Propbank," Linguistic Data Consortium, Philadelphia, 2006.
  4. B.Kim, Y.Lee and J.Lee, "Unsupervised Semantic Role Labeling for Korean Adverbial Case," Journal of KIISE, Vol. 34, No. 2, pp. 112-122, Feb. 2007. (in Korean)
  5. A.Giuglea and A.Moschitti, "Semantic Role Labeling via FrameNet, VerbNet and Propbank," Proc. of the ACL-2006, 2006.
  6. M. Surdeanu et al., "The CoNLL-2008 Shared Task on Joint Parsing of Syntactic and Semantic Dependencies," Proc. of the CoNLL-2008, pp. 159-177, 2008.
  7. S.Lim, C.Lee and D.Ra, "Dependency-based Semantic Role Labeling Using Sequence Labeling with a Structural SVM," Pattern Recognition Letters, Vol. 34, pp. 696-702, 2013. https://doi.org/10.1016/j.patrec.2013.01.022
  8. S.Lim and H.Kim, "A Study of Various Machine Learning Features for Korean Semantic Role Labeling," Proc. of 9th KIISE and KBS Joint Symposium, pp. 125-128, Jun. 2015. (in Korean)
  9. M.Lesk, "Automatic Sense Disambiguation Using Mahine Readable Dictionaries: How to Tell a Pine Cone from an Ice Cream cone," Proc. of the 5th Annual International Conference on Systems Documentation, pp. 24-26, 1986.
  10. K.Kim, "Synonym and Antonym dictionary for Factoid Question Answering system," ETRI Exobrain subproject Final Reports, NATMAL Corp, 2014. (in Korean)
  11. Y.Bae and C. Ock, "Introduction to the Korean Word Map(UWordMap) and API," Proc. of 26th Annual Conference on Human and Cognitive Language Technology, pp. 27-31, 2014. (in Korean)
  12. [Online]. Available: http://exobrain.re.kr/onedintro
  13. S.Lim et al., "Korean Proposition Bank Guidelines for Exobrain," Proc. of 27th Annual Conference on Human and Cognitive Language Technology, pp. 250-254, 2015. (in Korean)
  14. J.Lim et al., "Korean Dependency Parsing Model based on Transition System using Head Final Constraint," Proc. of 26th Annual Conference on Human and Cognitive Language Technology, pp. 81-86, 2014. (in Korean)
  15. C.Lee, S.Lim and H.Kim, "Korean Semantic Role Labeling Using Structured SVM," Journal of KIISE, Vol. 42, No. 2, pp. 220-226, Feb. 2015. (in Korean) https://doi.org/10.5626/JOK.2015.42.2.220