Natural Language Argumentation Mining, Processing, and Reasoning over Textual Arguments Schloss Dagstuhl, 18–22 April 2016 1
Organizers Elena Cabrio University of Nice Sophia Antipolis 2
Organizers Elena Cabrio Serena Villata University of Nice CNRS Sophia Antipolis 2
Organizers Graeme Hirst University of Toronto 3
Adam Wyner University of Aberdeen 4
The schedule 5
Dagstuhl seminar 16161 — Natural Language Argumentation: Mining, Processing, and Reasoning over Textual Arguments — 18–22 April 2016 Revised 2016-04-17 Except where noted, all talks are 20 minutes, including questions, leaving 10–20 minutes at the end of the session for more questions and general discussion. Chairs will be asked to enforce time limits without favour or mercy! Monday 18 April Tuesday 19 April Wednesday 20 April Thursday 21 April Friday 22 April 8:50–10:20 1. Introduction 5. Legal Argumentation 9. From natural language to formal models of argumentati 11. Argumentation and humans 15. Discussion • Introduction. Chair: Francesca Toni Chair: Nancy Green [18-min talks] Chair: Smaranda Muresan Issues, challenges, ideas, prospects, next steps. • Vern Walker : The need for annotated corpora from • Pietro Baroni , Massimiliano Giacomin, and Beishui • Fabio Paglieri : A plea for sustainable argument • Tony Hunter , tutorial: Structured argumentation. legal documents, and for (human) protocols for creating Liao : Uncertainty and fuzziness from natural language to technologies them: The attribution problem argumentation models • Katarzyna Budzynska : Mining dialogical arguments and • Michał Araszkiewicz : Case-based reasoning structures • Adam Wyner , Anthony Hunter, and Tom Van Engers : ethos in statutory interpretation Working on the argument pipeline: Through flow issues • Serena Villata : Emotions and personality traits in • Leon van der Torre : Norms and arguments between natural language argument, instantiated argumentation: An empirical evaluation • Kevin Ashley : Putting argument mining to work: an arguments, and argumentation frameworks • Ariel Rosenfeld and Sarit Kraus : Strategical experiment in legal argument retrieval using the LUIMA • Leila Amgoud, Philippe Besnard, and Anthony Hunter : argumentative agent for human persuasion: A preliminary type system and pipeline Logical representation and analysis for RC-arguments report • Federico Cerutti : Interfaces to formal argumentation • Francesca Toni : Towards relation-based argumentation mining 10:20–10:35 Coffee break Coffee break Coffee break Coffee break Coffee break 10:35:12:15 2. Tutorials 6. Argument mining and the Web 10. Unshared Untask, session III 12. Argument mining for writing support 16. Discussion; closing • Claire Cardie , turorial: Sentiment and argument Chair: Laura Alonso Alemany Access dataset III for each group. Discuss, analyse, Chair: Simone Teufel Issues, challenges, ideas, prospects, next steps. • Henning Wachsmuth and Benno Stein : "PageRank" for • Christian Stab and Iryna Gurevych : Argumentative mining. and take notes. argument relevance writing support: structure identification and quality • Questions and discussions (both tutorials). • Giorgos Flouris, Antonis Bikakis, Theodore Patkos and assessment of arguments Dimitris Plexousakis : Argument extraction challenges in a • Diane Litman : Temporal argument mining for writing new web paradigm assistance • Smaranda Muresan : Argumentation mining in online • Robert Mercer : Locating and extracting key interactions: Opportunities and challenges components of argumentation from scholarly scientific • Elena Cabrio : Tweeties squabbling: Argument mining writing on social media • Nancy Green : On recognizing argumentation schemes in formal text genres 12:15–13:30 Lunch Lunch Lunch Lunch Lunch 13:30–15:30 3. Argument mining 7. Unshared Untask, session II 13. Unshared Untask, plenary wrap-up Chair: Jodi Schneider Access dataset II for each group. Discuss, analyse, Discussion with all groups about the data, the notes, • Andreas Peldszus and Manfred Stede : Joint prediction and take notes. and a preliminary sketch of the paper. in MST-style discourse parsing for argumentation mining • Patrick Saint-Dizier : Argument mining: The bottleneck of knowledge • Vincent Ng : Experiments on under-studied argumentation mining tasks: Argument strength scoring and reason classification • Orith Toledo-Ronen, Roy Bar-Haim and Noam Slonim : Expert stance graphs for computational argumentation • Mark Snaith, John Lawrence and Chris Reed : Mining arguments and protocols [ presented by Katarzyna Excursion to TBD Budzynska ] 15:30–16:00 Coffee and cake Coffee and cake Coffee and cake 16:00–17:40 4. Unshared Untask, session I 8. Argumentation mining: other applications 14. By request Ival Habernal and Adam Wyner — Introduction. Chair: Claire Cardie Chair: Elena Cabrio Formation of groups. Access dataset I for each group. • Graeme Hirst and Nona Naderi : Crowdsourced and This session is reserved for extra talks, new ideas, expert annotations for argument frame discovery Discuss, analyse, and take notes. and discussions that any participant may initiate. If • Jan Šnajder : Social media argumentation mining: The you'd like some time in this session, please ask Elena. quest for deliberateness in raucousness • Jodi Schneider and Richard D. Boyce : Medication safety as a use case for argumentation mining • Rizkiyanto, Courtney Schriek, Jan Martijn van der Werf and Floris Bex : Design reasoning and design rationale 18:00–19:00 Dinner Dinner Dinner Dinner
The schedule • Four sessions, starting 8.50 daily. • Meal and break times set by Dagstuhl. • Finish with lunch on Friday. • Box lunches available for early departures. • Wednesday afternoon excursion (optional) . 7
The schedule • Two tutorials. • UnShared UnTask (four sessions). • Seven sessions for talks and discussion. • One “by request” session. • Two all-discussion sessions. 8
Evenings • Evenings free for discussion, walks in the woods, etc. • Refreshments at about 20:00 in lounge / aufenthaltsraum (006). • Suggestions for evening events? 9
Wednesday excursion • Vote for destination: Völklinger Hütte or Mettlach 10
Völklinger Hütte
Völklinger Hütte • Ironworks, 1881–1986, now museum • UNESCO Heritage Site • Extensive contemporary gardens • Technology, rust, botany!
Mettlach
Mettlach • Cloef lookout point over the Saar loop. • Villeroy & Boch — museum, shop, “experience centre”, guided tour. • Small (modern) brewery. • Birkenstock outlet centre.
Wednesday excursion • Völklinger Hütte – 20 people ⇒ € 45, 30 people ⇒ € 35 (bus, entrance fee, guided tour (max 30 people)). • Mettlach – 20 people ⇒ € 23, 30 people ⇒ € 15 (bus only). V&B museum € 4, plus € 4–6 for tour (max 25 people). 15
UnShared UnTask • If you have not received a Google Drive invitation to collaborate this morning, we need your e-mail address. 16
Dagstuhl Reports • Collector: Alexis Palmer. • Will need your abstract soon after the seminar – LaTeX source. – To be submitted via Dagstuhl Reports website. 17
Administrative details • Do not cite or quote pre-submissions. • Sign your book in the library (always open). • Computers and printers in labs. • Wine, supplies, etc: Take what you want, pay on check-out. • Check out and pay on Friday before lunch. • Taxis can be arranged for departure. 18
Natural Language Argumentation Mining, Processing, and Reasoning over Textual Arguments 19
Natural Language Argumentation Mining, Processing, and Reasoning over Textual Arguments 19
Natural Language Argumentation Mining, Processing, and Reasoning over Textual Arguments 19
Natural Language Argumentation Mining, Processing, and Reasoning over Textual Arguments 19
Natural Language Argumentation Mining, Processing, and Reasoning over Textual Arguments 19
Goals What are we here for? • To understand better the specific kinds of tasks that NLP can carry out in argumentation. 20
Goals Interdisciplinarity • To build more connections between the research communities in NLP who work on argumentation and those in AI and formal argumentation theory. 21
Goals Domain-specificity • To understand better how computational argumentation tasks are tied, or not, to their specific domains … … such as scientific papers, legal argumentation, political discussions, and mathematical proofs (qua arguments) … … looking for new cross-domain generalizat- ions. 22
Goals Domain-specificity • To establish a set of domain-specific and cross-domain use-cases that will guide the direction of research in the field. 23
Goals Understand the challenges • To understand better the technical challenges to success in each of these tasks, and to discuss how the challenges can be addressed. • To develop and explicate specific challenge problems for the integration of argumentation theory and NLP … … that are beyond the state of the art (but not too much so), and in which success would have the greatest effect on the field. 24
Recommend
More recommend
