a cascade model for proposi1on extrac1on in argumenta1on
play

A Cascade Model for Proposi1on Extrac1on in Argumenta1on Yohan Jo 1 - PowerPoint PPT Presentation

Feb 28, 2023 •437 likes •782 views

A Cascade Model for Proposi1on Extrac1on in Argumenta1on Yohan Jo 1 , Jacky Visser 2 , Chris Reed 2 , Eduard Hovy 1 1 Language Technologies Institute, Carnegie Mellon University 2 Centre for Argument Technology, University of Dundee 6th Workshop

  1. A Cascade Model for Proposi1on Extrac1on in Argumenta1on Yohan Jo 1 , Jacky Visser 2 , Chris Reed 2 , Eduard Hovy 1 1 Language Technologies Institute, Carnegie Mellon University 2 Centre for Argument Technology, University of Dundee 6th Workshop on Argument Mining August 1, 2019 � 1

  2. Argumenta*on Mining Argumentative Proposition Argument Structure Dialogue/Monologue Extraction Identification A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 2

  3. Proposi*on Extrac*on Argumenta*ve Discourse Unit (ADU) Segmenta*on • Original text is segmented into ADUs • Argumentative structure as relations between the ADUs Prior Work • Definitions of ADU boundaries 
 Stab and Gurevych, 2014; Stede et al., 2016; Peldszus and Stede, 2015; Al-Khatib et al., 2016 • Methods for auto-segmentation 
 Eger et al., 2017; Ajjour et al., 2017; Persing and Ng, 2016 A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 3

  4. Limita*ons of Segmenta*on-based ADUs • ADUs may lack important semantic information ▶ Referents of anaphors (A) She (Alice) complained to me (Bob) . (B) Bob is upset. ▶ Subject of a phrase (A) Alice knows Bob well but (B) kept the secret. (C) Bob should appreciate that. A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 4

  5. Limita*ons of Segmenta*on-based ADUs • ADUs may completely miss implicit propositions ▶ Reported speech (A) The doctor said we need more magnesium. ▶ Questions (A) Why would you spend your valuable money on tax? (B) Tax is a waste for nothing. ▶ Imperatives (A) Don't spend your valuable money on tax. (B) Tax is a waste for nothing. A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 5

  6. • What NLP challenges are there to obtain complete propositions ? • Can we get semantically improved propositions using standard NLP techniques? A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 6

  7. Data • 2016 U.S. presidential debates (Visser et al., 2019) ▶ Inference Anchoring Theory (Reed and Budzynska, 2011 ) ▶ 8,008 locutions (278 reported speech, 565 questions) ▶ Cohen's kappa: 0.610 (1) Segment utterance (3) Annotate propositions into locutions (ADUs) (2) Identify illocutionary acts instantiated by the locutions A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 7

  8. A Cascade Model for Proposi*on Extrac*on Alice: Bob stopped by my o ffi ce and complained, ``Why is the company not launching the new Utterance of a speaker service?'' I think I have explained to him already. AnaphoraResolution Anaphora- resolved utterance LocutionExtraction Locution2 Locution1 Y Speech IsReportedSpeech? SpeechIdentification N Y IsQuestion? Declarative QuestionTransform N Y Declarative IsImperative? ImperativeTransform N SubjectReconstruction Revision Propositions A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 8

  9. A Cascade Model for Proposi*on Extrac*on Alice: Bob stopped by my o ffi ce and complained, ``Why is the company not launching the new Utterance of a speaker service?'' I think I have explained to him already. Bob stopped by Alice's o ffi ce and complained, ``Why is the company not AnaphoraResolution launching the new service?'' Alice think Alice have explained to Bob already. Anaphora- resolved utterance LocutionExtraction Locution2 Locution1 Y Speech IsReportedSpeech? SpeechIdentification N Y IsQuestion? Declarative QuestionTransform N Y Declarative IsImperative? ImperativeTransform N SubjectReconstruction Revision Propositions A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 9

  10. A Cascade Model for Proposi*on Extrac*on Alice: Bob stopped by my o ffi ce and complained, ``Why is the company not launching the new Utterance of a speaker service?'' I think I have explained to him already. Bob stopped by Alice's o ffi ce and complained, ``Why is the company not AnaphoraResolution launching the new service?'' Alice think Alice have explained to Bob already. Anaphora- resolved utterance [L1] Bob stopped by Alice's o ffi ce and [L2] complained, ``Why is the LocutionExtraction company not launching the new service?'' Alice think [L3] Alice have explained to Bob already. Locution2 Locution1 Y Speech IsReportedSpeech? SpeechIdentification N Y IsQuestion? Declarative QuestionTransform N Y Declarative IsImperative? ImperativeTransform N SubjectReconstruction Revision Propositions A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 10

  11. A Cascade Model for Proposi*on Extrac*on Alice: Bob stopped by my o ffi ce and complained, ``Why is the company not launching the new Utterance of a speaker service?'' I think I have explained to him already. Bob stopped by Alice's o ffi ce and complained, ``Why is the company not AnaphoraResolution launching the new service?'' Alice think Alice have explained to Bob already. Anaphora- resolved utterance [L1] Bob stopped by Alice's o ffi ce and [L2] complained, ``Why is the LocutionExtraction company not launching the new service?'' Alice think [L3] Alice have explained to Bob already. Locution2 Locution1 Y [L2] complained, `` [L2'] Why is the company not launching the new service? '' Speech IsReportedSpeech? SpeechIdentification N Y IsQuestion? Declarative QuestionTransform N Y Declarative IsImperative? ImperativeTransform N SubjectReconstruction Revision Propositions A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 11

  12. A Cascade Model for Proposi*on Extrac*on Alice: Bob stopped by my o ffi ce and complained, ``Why is the company not launching the new Utterance of a speaker service?'' I think I have explained to him already. Bob stopped by Alice's o ffi ce and complained, ``Why is the company not AnaphoraResolution launching the new service?'' Alice think Alice have explained to Bob already. Anaphora- resolved utterance [L1] Bob stopped by Alice's o ffi ce and [L2] complained, ``Why is the LocutionExtraction company not launching the new service?'' Alice think [L3] Alice have explained to Bob already. Locution2 Locution1 Y [L2] complained, `` [L2'] Why is the company not launching the new service? '' Speech IsReportedSpeech? SpeechIdentification N Y [L2'] The company should launch the new service IsQuestion? Declarative QuestionTransform N Y Declarative IsImperative? ImperativeTransform N SubjectReconstruction Revision Propositions A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 12

  13. A Cascade Model for Proposi*on Extrac*on Alice: Bob stopped by my o ffi ce and complained, ``Why is the company not launching the new Utterance of a speaker service?'' I think I have explained to him already. Bob stopped by Alice's o ffi ce and complained, ``Why is the company not AnaphoraResolution launching the new service?'' Alice think Alice have explained to Bob already. Anaphora- resolved utterance [L1] Bob stopped by Alice's o ffi ce and [L2] complained, ``Why is the LocutionExtraction company not launching the new service?'' Alice think [L3] Alice have explained to Bob already. Locution2 Locution1 Y [L2] complained, `` [L2'] Why is the company not launching the new service? '' Speech IsReportedSpeech? SpeechIdentification N Y [L2'] The company should launch the new service IsQuestion? Declarative QuestionTransform N Y Declarative IsImperative? ImperativeTransform N [L2] Bob complained, ``Why is the company not launching the new service?'' SubjectReconstruction Revision Propositions A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 13

  14. A Cascade Model for Proposi*on Extrac*on Alice: Bob stopped by my o ffi ce and complained, ``Why is the company not launching the new Utterance of a speaker service?'' I think I have explained to him already. Bob stopped by Alice's o ffi ce and complained, ``Why is the company not AnaphoraResolution launching the new service?'' Alice think Alice have explained to Bob already. Anaphora- resolved utterance [L1] Bob stopped by Alice's o ffi ce and [L2] complained, ``Why is the LocutionExtraction company not launching the new service?'' Alice think [L3] Alice have explained to Bob already. Locution2 Locution1 Y [L2] complained, `` [L2'] Why is the company not launching the new service? '' Speech IsReportedSpeech? SpeechIdentification N Y [L2'] The company should launch the new service IsQuestion? Declarative QuestionTransform N Y Declarative IsImperative? ImperativeTransform N [L2] Bob complained, ``Why is the company not launching the new service?'' SubjectReconstruction [L3] Alice has explained to Bob already Revision Propositions A Cascade Model for Proposition Extraction in Argumentation Yohan Jo (yohanj@cs.cmu.edu) ! 14

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Tuning Ber*ni Cascade Model Parameters Dennis Wright 22

Tuning Ber*ni Cascade Model Parameters Dennis Wright 22

Tuning Ber*ni Cascade Model Parameters Dennis Wright 22 January 2016 Outline Model tuning benefits and risks Introduc*on to Ber*ni

888 views • 12 slides
A Recurrent Neural Cascade-based Model for Continuous-Time Diffusion Sylvain Lamprier LIP6 -

A Recurrent Neural Cascade-based Model for Continuous-Time Diffusion Sylvain Lamprier LIP6 -

A Recurrent Neural Cascade-based Model for Continuous-Time Diffusion Sylvain Lamprier LIP6 - Sorbonne Universit es 1 / 5 Cascade-based models for diffusion Information spreads from users to users in the network, following independent

143 views • 10 slides
Independent Cascade Model for Viral Marketing Wonyeol Lee Jinha Kim Hwanjo Yu Department of

Independent Cascade Model for Viral Marketing Wonyeol Lee Jinha Kim Hwanjo Yu Department of

0/19 CT-IC: Continuously activated and Time-restricted Independent Cascade Model for Viral Marketing Wonyeol Lee Jinha Kim Hwanjo Yu Department of Computer Science & Engineering , Korea ICDM 2012 CT-IC model for Viral Marketing 1/19

530 views • 31 slides
Influence Identification on Independent Cascade Model

Influence Identification on Independent Cascade Model

Influence Identification on Independent Cascade Model Part 1 Introduction Part 2 Related Work Part 3 Algorithm Part 4 Experiment Part 5 Conclusion Part 6 References

723 views • 33 slides
Loop O ( n ) model on random quadrangulations: the cascade of loop perimeters 12 Pascal

Loop O ( n ) model on random quadrangulations: the cascade of loop perimeters 12 Pascal

Loop O ( n ) model on random quadrangulations: the cascade of loop perimeters 12 Pascal Maillard 6 4 1 2 3 4 5 2 1 1 (Universit Paris-Sud / Paris-Saclay) 2 11 12 21 based on joint work with 1 1 Linxiao Chen and Nicolas

913 views • 63 slides
An Intranuclear Cascade Model for Cluster-Induced Reactions Monira J Kobra* and Yusuke Uozumi**

An Intranuclear Cascade Model for Cluster-Induced Reactions Monira J Kobra* and Yusuke Uozumi**

An Intranuclear Cascade Model for Cluster-Induced Reactions Monira J Kobra* and Yusuke Uozumi** *Rajshahi University, Bangladesh **Kyushu University, Japan Joint ICTP/IAEA workshop on nuclear structure and decay data 15-26 October, 2018

549 views • 32 slides
Beam energy scan using a viscous hydro+cascade model Yuriy KARPENKO Frankfurt Institute for

Beam energy scan using a viscous hydro+cascade model Yuriy KARPENKO Frankfurt Institute for

Beam energy scan using a viscous hydro+cascade model Yuriy KARPENKO Frankfurt Institute for Advanced Studies/ Bogolyubov Institute for Theoretical Physics FAIRNESS 2013, September 21, 2013 In collaboration with M. Bleicher, P . Huovinen, H.

470 views • 34 slides
Tree Recursion Announcements Order of Recursive Calls The Cascade Function (Demo) Each

Tree Recursion Announcements Order of Recursive Calls The Cascade Function (Demo) Each

Tree Recursion Announcements Order of Recursive Calls The Cascade Function (Demo) Each cascade frame is from a different call to cascade. Until the Return value appears, that call has not completed. Any statement can appear before

175 views • 15 slides
Study of Study of T ricyclic ricyclic Cascade Netw Cascade Networks using orks using Dynamic

Study of Study of T ricyclic ricyclic Cascade Netw Cascade Networks using orks using Dynamic

Study of Study of T ricyclic ricyclic Cascade Netw Cascade Networks using orks using Dynamic Optimization Dynamic Optimization Systems Biology interdisciplinary field that focuses on the systematic study of complex interactions in

588 views • 30 slides
Not the usual suspects: patient engagement and the HIV care cascade Malika Sharma, MD FRCPC MEd

Not the usual suspects: patient engagement and the HIV care cascade Malika Sharma, MD FRCPC MEd

Not the usual suspects: patient engagement and the HIV care cascade Malika Sharma, MD FRCPC MEd OHTN Endgame: Closing the gaps in the care cascade Disclosures Currently funded by the CIHR CTN/CANFAR Postdoctoral fellowship Previously

248 views • 23 slides
Update on Cascade Care procurement activities and proposed timeline Senior Citizens

Update on Cascade Care procurement activities and proposed timeline Senior Citizens

Update on Cascade Care procurement activities and proposed timeline Senior Citizens Foundation Fall Conference October 24, 2019 Agenda Introduction to Engrossed Substitute Senate Bill (ESSB) 5226: Cascade Care Health Care Authoritys

281 views • 10 slides
The Cascade High Productivity Language The Cascade High Productivity Language Brad Chamberlain

The Cascade High Productivity Language The Cascade High Productivity Language Brad Chamberlain

The Cascade High Productivity Language The Cascade High Productivity Language Brad Chamberlain Brad Chamberlain David Callahan David Callahan Hans Zima * Hans Zima * Chapel Team, Cascade Project Chapel Team, Cascade Project Cray Inc.,

713 views • 21 slides
Results of Cascade Road Diet: Background Before data set was collected on 1/18/18, and

Results of Cascade Road Diet: Background Before data set was collected on 1/18/18, and

Results of Cascade Road Diet: Background Before data set was collected on 1/18/18, and After data collected on 11/27/18 Cascade was re-striped on 8/15/18 This is the Total Traffic Volume Distribution Through ONE in January of

367 views • 4 slides
Why Do Cascade Sizes Follow a Power-Law? Andrzej Pacuk, Piotr Sankowski, Karol Wgrzycki , Piotr

Why Do Cascade Sizes Follow a Power-Law? Andrzej Pacuk, Piotr Sankowski, Karol Wgrzycki , Piotr

Why Do Cascade Sizes Follow a Power-Law? Andrzej Pacuk, Piotr Sankowski, Karol Wgrzycki , Piotr Wygocki University of Warsaw Paper: bit.ly/why-cascades WWW 2017 Information Cascade Cascades as Graphs Given a Social Network The process

1.02k views • 57 slides
2008 Voters approve PCC Bond $374,000,000 total $60,265,000 allocated to Cascade Campus

2008 Voters approve PCC Bond $374,000,000 total $60,265,000 allocated to Cascade Campus

2008 Voters approve PCC Bond $374,000,000 total $60,265,000 allocated to Cascade Campus CASCADE CAMPUS 2011 Guiding Principals finalized Based on outreach and feedback from Campus and Neighborhood communities Guiding Principals

690 views • 28 slides
Cascade-Correlation and Deep Learning Scott E. Fahlman Professor Emeritus Language Technologies

Cascade-Correlation and Deep Learning Scott E. Fahlman Professor Emeritus Language Technologies

Cascade-Correlation and Deep Learning Scott E. Fahlman Professor Emeritus Language Technologies Institute February 27, 2019 Two Ancient Papers Fahlman, S. E. and C. Lebiere (1990) "The Cascade-Correlation Learning Architecture,

633 views • 29 slides
Explicit and Implicit Discourse Relations: An Extrinsic Evaluation Peter Bourgonje and Manfred

Explicit and Implicit Discourse Relations: An Extrinsic Evaluation Peter Bourgonje and Manfred

Explicit and Implicit Discourse Relations: An Extrinsic Evaluation Peter Bourgonje and Manfred Stede Applied Computational Linguistics Universitt Potsdam Workshop on Coherence Relations Humboldt-Universitt zu Berlin January 17-18, 2020

342 views • 21 slides
Lear Learning M ning Multi ulti-Moda Modal l Grounded Lingu Grounded Linguistic istic

Lear Learning M ning Multi ulti-Moda Modal l Grounded Lingu Grounded Linguistic istic

March 2020 Lear Learning M ning Multi ulti-Moda Modal l Grounded Lingu Grounded Linguistic istic Semantics by Playing I Spy Tong Gao Introduction Early work on grounded language learning enabled a machine to map from

540 views • 24 slides
NumClaim: Investor's Fine-grained Claim Detection Chung-Chi Chen 1 , Hen-Hsen Huang 2,3 , Hsin-Hsi

NumClaim: Investor's Fine-grained Claim Detection Chung-Chi Chen 1 , Hen-Hsen Huang 2,3 , Hsin-Hsi

NumClaim: Investor's Fine-grained Claim Detection Chung-Chi Chen 1 , Hen-Hsen Huang 2,3 , Hsin-Hsi Chen 1,3 1 Department of Computer Science and Information Engineering , National Taiwan University, Taiwan 2 Department of Computer Science, National

358 views • 10 slides
+"!."'!M !"#$%#&'()*#((+&#&'+( ! !"#$%#&'()*,A,&!-, "

+"!."'!M !"#$%#&'()*#((+&#&'+( ! !"#$%#&'()*,A,&!-, "

+"!."'!M !"#$%#&'()*#((+&#&'+( ! !"#$%#&'()*,A,&!-, " 3.!1','+(*#(E*.!1#$$ " Q?"#$%! ! #((+&#&'+(*.!$'#@'$'&A ! F#33#*R #*.!$'#@'$'&A*-!&.'1*

353 views • 6 slides
Computational Pragmatics Autumn 2015 Raquel Fernndez Institute for Logic, Language &

Computational Pragmatics Autumn 2015 Raquel Fernndez Institute for Logic, Language &

Computational Pragmatics Autumn 2015 Raquel Fernndez Institute for Logic, Language & Computation University of Amsterdam Outline Today: Part 1: Speech act theory and dialogue acts Homework #2: dialogue acts in the Switchboard

598 views • 35 slides
INTO THE PIPELINE: THE LATEST IN PSYCHOPHARMACOLOGY Learning Objectives Describe the

INTO THE PIPELINE: THE LATEST IN PSYCHOPHARMACOLOGY Learning Objectives Describe the

INTO THE PIPELINE: THE LATEST IN PSYCHOPHARMACOLOGY Learning Objectives Describe the mechanisms of action of emerging and investigational treatments of mental illnesses Discuss implications of recent and novel data to clinical practice

683 views • 50 slides
Evaluation measures in NLP Zdenk abokrtsk 8th April 2020 NPFL124 Natural Language

Evaluation measures in NLP Zdenk abokrtsk 8th April 2020 NPFL124 Natural Language

Evaluation measures in NLP Zdenk abokrtsk 8th April 2020 NPFL124 Natural Language Processing Charles University Faculty of Mathematics and Physics Institute of Formal and Applied Linguistics unless otherwise stated Title of your

597 views • 32 slides
Dense matter QFT with the DS L=24 1e-08 LLR L=24 O( ) DS L=22 density-of-states DS L=20

Dense matter QFT with the DS L=24 1e-08 LLR L=24 O( ) DS L=22 density-of-states DS L=20

1 Dense matter QFT with the DS L=24 1e-08 LLR L=24 O( ) DS L=22 density-of-states DS L=20 DS L=18 DS L=16 LLR L=22 LLR L=20 method LLR L=18 LLR L=16 1e-16 0 0.5 1 1.5 2 Kurt Langfeld, Centre of Mathematical Sciences,

764 views • 53 slides

More recommend