Discourse Processing ´18

Recent Advances in Discourse Processing

Seminar / in English

held by:        Prof. Vera Demberg, Dr. Frances Yung

time:            Wednesdays, starting on Oct 24, 2018,  12:15 - 13:45 hrs

location:        in building C7.1, room U15 (-1.15)

suitable for:   Master / Language Science and Technology

________________________________________________

All information regarding schedule and course materials can be found in the Moodle.

Language is the basic medium for people to communicate with each other. It provides a versatile system to encode meanings in numerous ways. However, in communicative discourse, people do not only use the literal meanings of language, meanings are also inferred from context.  Therefore, pragmatics is an important aspect in discourse processing.  Critically, the choice of linguistic expression among other alternatives is itself meaningful for the speaker and the listener.  This seminar introduces the recently proposed Rational Speech Act (RSA) model that quantifies this notion by Bayesian statistics.  We will go through the basics and variations of the RSA model, how the model provides a unified account for various linguistic/psycholinguistics phenomena, and how the model can be applied in combination with machine learning to improve the state of the art of NLP tasks.

READINGS:

Basics

Frank, Michael C., and Noah D. Goodman. "Predicting pragmatic reasoning in language games." Science 336.6084 (2012): 998-998.

Goodman, Noah D., and Michael C. Frank. "Pragmatic language interpretation as probabilistic inference." Trends in Cognitive Sciences 20.11 (2016): 818-829.

Further reading

Goodman, Noah D., and Andreas Stuhlmüller. "Knowledge and implicature: Modeling language understanding as social cognition." Topics in cognitive science 5.1 (2013): 173-184.

Frank, Michael C. "Rational speech act models of pragmatic reasoning in reference games." (2016).

Linguistics / Psycholinguistics

Degen, Judith, Michael Franke, and Gerhard Jager. "Cost-based pragmatic inference about referential expressions." Proceedings of the annual meeting of the cognitive science society. Vol. 35. No. 35. 2013.

Shafto, Patrick, Noah D. Goodman, and Thomas L. Griffiths. "A rational account of pedagogical reasoning: Teaching by, and learning from, examples." Cognitive psychology 71 (2014): 55-89.

Kao, Justine T., Roger Levy, and Noah D. Goodman. "The funny thing about incongruity: A computational model of humor in puns." Proceedings of the Annual Meeting of the Cognitive Science Society. Vol. 35. No. 35. 2013

Hawkins, Robert XD, Michael C. Frank, and Noah D. Goodman. "Convention-formation in iterated reference games." Proceedings of the 39th Annual Conference of the Cognitive Science Society. Cognitive Science Society. 2017.

Khani, Fereshte, Noah D. Goodman, and Percy Liang. "Planning, Inference and Pragmatics in Sequential Language Games." arXiv preprint arXiv:1805.11774 (2018).

Savinelli, K. J., Gregory Scontras, and Lisa Pearl. "Modeling scope ambiguity resolution as pragmatic inference: Formalizing differences in child and adult behavior." Proceedings of the Annual Meeting of the Cognitive Science Society. Vol. 39. 2017.

Applicational

Fried, Daniel, et al. "Speaker-Follower Models for Vision-and-Language Navigation." arXiv preprint arXiv:1806.02724 (2018).

Cohn-Gordon, Reuben, Noah Goodman, and Chris Potts. "Pragmatically Informative Image Captioning with Character-Level Reference." arXiv preprint arXiv:1804.05417 (2018).

Monroe, Will, et al. "Colors in context: A pragmatic neural model for grounded language understanding." arXiv preprint arXiv:1703.10186 (2017).

Andreas, Jacob, and Dan Klein. "Reasoning about pragmatics with neural listeners and speakers." arXiv preprint arXiv:1604.00562 (2016).

Last modified 2018-10-11 by gabi
Impressum - Haftungsausschluss