Hedge Detection Using the RelHunter Approach

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Standard

Hedge Detection Using the RelHunter Approach. / Fernandes, Eraldo R.; Crestana, Carlos E. M.; Milidiú, Ruy L.
Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task. ed. / Richard Farkas; Veronika Vincze; György Szarvas; György Mora; Janos Csirik. USA: Association for Computational Linguistics (ACL), 2010. p. 64–69 (CoNLL '10: Shared Task).

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Harvard

Fernandes, ER, Crestana, CEM & Milidiú, RL 2010, Hedge Detection Using the RelHunter Approach. in R Farkas, V Vincze, G Szarvas, G Mora & J Csirik (eds), Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task. CoNLL '10: Shared Task, Association for Computational Linguistics (ACL), USA, pp. 64–69, 14th Conference on Computational Natural Language Learning - CoNLL 2010, Uppsala, Sweden, 15.07.10. <https://dl.acm.org/doi/pdf/10.5555/1870535.1870544>

APA

Fernandes, E. R., Crestana, C. E. M., & Milidiú, R. L. (2010). Hedge Detection Using the RelHunter Approach. In R. Farkas, V. Vincze, G. Szarvas, G. Mora, & J. Csirik (Eds.), Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task (pp. 64–69). (CoNLL '10: Shared Task). Association for Computational Linguistics (ACL). https://dl.acm.org/doi/pdf/10.5555/1870535.1870544

Vancouver

Fernandes ER, Crestana CEM, Milidiú RL. Hedge Detection Using the RelHunter Approach. In Farkas R, Vincze V, Szarvas G, Mora G, Csirik J, editors, Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task. USA: Association for Computational Linguistics (ACL). 2010. p. 64–69. (CoNLL '10: Shared Task).

Bibtex

@inbook{146b5884042d420797371b35d9a026a9,
title = "Hedge Detection Using the RelHunter Approach",
abstract = "RelHunter is a Machine Learning based method for the extraction of structured information from text. Here, we apply RelHunter to the Hedge Detection task, proposed as the CoNLL-2010 Shared Task. RelHunter's key design idea is to model the target structures as a relation over entities. The method decomposes the original task into three subtasks: (i) Entity Identification; (ii) Candidate Relation Generation; and (iii) Relation Recognition. In the Hedge Detection task, we define three types of entities: cue chunk, start scope token and end scope token. Hence, the Entity Identification subtask is further decomposed into three token classification subtasks, one for each entity type. In the Candidate Relation Generation sub-task, we apply a simple procedure to generate a ternary candidate relation. Each instance in this relation represents a hedge candidate composed by a cue chunk, a start scope token and an end scope token. For the Relation Recognition subtask, we use a binary classifier to discriminate between true and false candidates. The four classifiers are trained with the Entropy Guided Transformation Learning algorithm. When compared to the other hedge detection systems of the CoNLL shared task, our scheme shows a competitive performance. The F-score of our system is 54.05 on the evaluation corpus.",
keywords = "Informatics, Business informatics",
author = "Fernandes, {Eraldo R.} and Crestana, {Carlos E. M.} and Milidi{\'u}, {Ruy L.}",
year = "2010",
language = "English",
isbn = "978-1-932432-84-8",
series = "CoNLL '10: Shared Task",
publisher = "Association for Computational Linguistics (ACL)",
pages = "64–69",
editor = "Richard Farkas and Veronika Vincze and Gy{\"o}rgy Szarvas and Gy{\"o}rgy Mora and Janos Csirik",
booktitle = "Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task",
address = "United States",
note = "14th Conference on Computational Natural Language Learning - CoNLL 2010 : Shared Task, CoNLL 2010 ; Conference date: 15-07-2010 Through 17-07-2010",
url = "http://toc.proceedings.com/08986webtoc.pdf",

}

RIS

TY - CHAP

T1 - Hedge Detection Using the RelHunter Approach

AU - Fernandes, Eraldo R.

AU - Crestana, Carlos E. M.

AU - Milidiú, Ruy L.

N1 - Conference code: 14

PY - 2010

Y1 - 2010

N2 - RelHunter is a Machine Learning based method for the extraction of structured information from text. Here, we apply RelHunter to the Hedge Detection task, proposed as the CoNLL-2010 Shared Task. RelHunter's key design idea is to model the target structures as a relation over entities. The method decomposes the original task into three subtasks: (i) Entity Identification; (ii) Candidate Relation Generation; and (iii) Relation Recognition. In the Hedge Detection task, we define three types of entities: cue chunk, start scope token and end scope token. Hence, the Entity Identification subtask is further decomposed into three token classification subtasks, one for each entity type. In the Candidate Relation Generation sub-task, we apply a simple procedure to generate a ternary candidate relation. Each instance in this relation represents a hedge candidate composed by a cue chunk, a start scope token and an end scope token. For the Relation Recognition subtask, we use a binary classifier to discriminate between true and false candidates. The four classifiers are trained with the Entropy Guided Transformation Learning algorithm. When compared to the other hedge detection systems of the CoNLL shared task, our scheme shows a competitive performance. The F-score of our system is 54.05 on the evaluation corpus.

AB - RelHunter is a Machine Learning based method for the extraction of structured information from text. Here, we apply RelHunter to the Hedge Detection task, proposed as the CoNLL-2010 Shared Task. RelHunter's key design idea is to model the target structures as a relation over entities. The method decomposes the original task into three subtasks: (i) Entity Identification; (ii) Candidate Relation Generation; and (iii) Relation Recognition. In the Hedge Detection task, we define three types of entities: cue chunk, start scope token and end scope token. Hence, the Entity Identification subtask is further decomposed into three token classification subtasks, one for each entity type. In the Candidate Relation Generation sub-task, we apply a simple procedure to generate a ternary candidate relation. Each instance in this relation represents a hedge candidate composed by a cue chunk, a start scope token and an end scope token. For the Relation Recognition subtask, we use a binary classifier to discriminate between true and false candidates. The four classifiers are trained with the Entropy Guided Transformation Learning algorithm. When compared to the other hedge detection systems of the CoNLL shared task, our scheme shows a competitive performance. The F-score of our system is 54.05 on the evaluation corpus.

KW - Informatics

KW - Business informatics

UR - http://www.scopus.com/inward/record.url?scp=85123205923&partnerID=8YFLogxK

M3 - Article in conference proceedings

SN - 978-1-932432-84-8

T3 - CoNLL '10: Shared Task

SP - 64

EP - 69

BT - Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task

A2 - Farkas, Richard

A2 - Vincze, Veronika

A2 - Szarvas, György

A2 - Mora, György

A2 - Csirik, Janos

PB - Association for Computational Linguistics (ACL)

CY - USA

T2 - 14th Conference on Computational Natural Language Learning - CoNLL 2010

Y2 - 15 July 2010 through 17 July 2010

ER -

Recently viewed

Researchers

  1. Marco Waage

Publications

  1. How development leads to democracy
  2. An Adaptive Resonance Regulator Design for Motion Control of Intake Valves in Camless Engine Systems
  3. Performance of the DSM-5-based criteria for Internet addiction
  4. Inequality in the Transition from Primary to Secondary School
  5. Introduction: The representative turn in EU Studies
  6. Extraction of information from invoices - challenges in the extraction pipeline
  7. Magnesium-based metal matrix nanocomposites—processing and properties
  8. I Am Not A Hacker
  9. Pragmatics broadly viewed
  10. Dimensions, dialectic, discourse
  11. Non-acceptances in context
  12. What Role for Public Participation in Implementing the EU Floods Directive? A comparison with the Water Framework Directive, early evidence from Germany, and a research agenda
  13. Contrasting requests in Inner Circle Englishes
  14. Learning Analytics an Hochschulen
  15. From teacher-centered instruction to peer tutoring in the heterogeneous international classroom
  16. Nitrogen uptake by grassland communities
  17. Kommentar zu Ute Tellmann
  18. Operationalizing Network Theory for Ecosystem Service Assessments
  19. Thanking and responding to thanks in American English: Language patterning and contextual appropriateness
  20. Competition in fragmented markets
  21. Mapping the Order of New Migration
  22. Is There a Way Back or Can the Internet Remember its Own History?
  23. Case study: The development of a multi-material heat sink by Additive Manufacturing using Aerosint technology
  24. Synthesis and future research directions linking tree diversity to growth, survival, and damage in a global network of tree diversity experiments
  25. Predicting the future performance of soccer players
  26. Testing for a break in the persistence in yield spreads of EMU government bonds