COMP 790.139 (Spr2018): Advanced Topics in NLP: Conversational Models

Instructor: Mohit Bansal
Units: 3
Office: SN 258
Lectures: Fridays 2:00pm-4:30pm, Rm SN-155
Office Hours: Fridays 4:30pm-5:30pm (by appointment) (SN 258)
Course Webpage: http://www.cs.unc.edu/~mbansal/teaching/nlp-seminar-spring18.html
Course Email: nlpcomp790unc -at- gmail.com


Syllabus Topics

This course will be an advanced topic seminar class on natural language processing for very diverse types of conversational models (rule-based, retrieval-based, neural generative models, grounded/visual, chit-chat vs. task-based, etc.) as well as important related aspects such as evaluation, deployment, etc.:

This will be a research-oriented grad-level seminar course, where we will read lots of interesting research papers, brainstorm about ideas on latest research topics, and code and write up fun and novel projects!
Please email me or drop by my office if you have any questions!


Prerequisites

Since this is an advanced topics NLP class, the student is expected to have equivalent experience to Dr. Bansal's' fall 2016 or fall 2017 regular NLP class.


Grading (tentative)

Grading will consist of:

Details in first class intro lecture slides. There will not be any exams. All submissions should be emailed to: nlpcomp790unc@gmail.com


Lateness Policy

Students are allowed 3 free late days for assignments over the semester. After that, late assignments will be accepted with a 20% reduction in value per day late.


Collaboration Policy

Paper summaries have to be written and submitted individually. Paper presentations have to be done individually. Projects can be individual (e.g., if it relates to your current research) or in pairs (but with proportional work, and with clearly outlined contributions from each team member).


Reference Books




Tentative Schedule

DateTopic Readings Discussion LeadersTodo's
Jan 17 (make-up class on Jan19 due to weather) Intro to Class (summary of different types of dialogue, models, and evaluation issues) SLP3 Chapters 29, 30; SLP2 Chapters 24 Mohit -
Jan 26 Rule-based Chatbots, Retrieval-based Chatbots Rule-based (Yichen):
(1) ELIZA: A computer program for the study of natural language communication between man and machine. Weizenbaum, J. (1966) ;
(2) PARRY: Artificial paranoia. Colby at al. (1971) ;
Retrieval-based (Han):
(3) Filter, rank, and transfer the knowledge: Learning to chat. Jafarpour et al. (2009) ;
(4) NPCEditor: Creating virtual human dialogue using information retrieval techniques. Leuski and Traum (2011) ;
(5) Cobot in LambdaMOO: A social statistics agent. Isbell et al. (2000)
Yichen, Han -
Feb 2 (no class due to travel; make-up class info will be emailed) TBD -
Feb 9 Frame-based and KB-based Agents Justin:
(1) GUS, A Frame-Driven Dialog System. Bobrow et al. (1977) ;
(2) Frame Based Dialog Agents. Jurafsky and Martin SLP3 Book Chapter 29.2 ;
Lisa+Tong:
(3) Frames: A Corpus for Adding Memory to Goal-Oriented Dialogue Systems. Asri et al. (2017) ;
(4) Learning Symmetric Collaborative Dialogue Agents with Dynamic Knowledge Graph Embeddings. He et al. (2017) ;
(5) Key-Value Retrieval Networks for Task-Oriented Dialogue. Eric and Manning (2017)
Justin, Lisa+Tong -
Feb 16 Neural Generative Models Haonan:
(1) Semantically Conditioned LSTM-based Natural Language Generation for Spoken Dialogue Systems. Wen et al. (2015) ;
(2) Conditional Generation and Snapshot Learning in Neural Dialogue Systems. Wen et al. (2016) ;
(3) Coherent Dialogue with Attention-based Language Models. Mei et al. (2016);
Hyounghun:
(4) A Hierarchical Latent Variable Encoder-Decoder Model for Generating Dialogues. Serban et al. (2016) ;
(5) A Diversity-Promoting Objective Function for Neural Conversation Models. Li et al. (2015) ;
(6) Adversarial Learning for Neural Dialogue Generation. Li et al. (2017)
Haonan, Hyounghun -
Feb 23 Personality and Emotion based Dialogue Eric:
(1) Improvising Linguistic Style: Social and Affective Bases for Agent Personality. Walker et al. (1997) ;
(2) PERSONAGE: Personality Generation for Dialogue. Mairesse and Walker (2007) ;
(3) Using Linguistic Cues for the Automatic Recognition of Personality in Conversation and Text. Mairesse et al. (2007);
Yicheng:
(4) A Persona-Based Neural Conversation Model. Li et al. (2016) ;
(5) Personalizing Dialogue Agents: I have a dog, do you have pets too? Zheng et al. (2018) ;
(6) Zara Returns: Improved Personality Induction and Adaptation by an Empathetic Virtual Agent. Siddique et al. (2017)
Eric, Yicheng -
Mar 2 Reinforcement Learning for Dialogue Policy Hao:
(1) Reinforcement learning for spoken dialogue systems. Singh et al. (1999) ;
(2) Automatic learning of dialogue strategy using dialogue simulation and reinforcement learning. Scheffler and Young (2002) ;
(3) Partially observable Markov decision processes for spoken dialog systems. Williams and Young (2007) ;
Ram:
(4) On-line Active Reward Learning for Policy Optimisation in Spoken Dialogue Systems. Su et al. (2016) ;
(5) Deep Reinforcement Learning for Dialogue Generation. Li et al. (2016) ;
(6) Composite Task-Completion Dialogue Policy Learning via Hierarchical Deep Reinforcement Learning. Peng et al. (2017)
Hao, Ram -
Mar 9 Grounded Dialogue (Visual, Speech, Gaze, etc.) Sweta:
(1) MATCH: An Architecture for Multimodal Dialogue Systems. Johnston et al. (2002) ;
(2) The ICSI meeting project: Resources and research. Janin et al. (2004) ;
(3) Eye Gaze for Spoken Language Understanding in Multi-Modal Conversational Interactions. Dilek Hakkani-Tür et al. (2014) ;
Jie:
(4) Visual Dialog. Das et al. (2017) ;
(5) GuessWhat?! Visual object discovery through multi-modal dialogue. De Vries et al. (2017) ;
(6) Image-Grounded Conversations: Multimodal Context for Natural Question and Response Generation. Mostafazadeh et al. (2017)
Sweta, Jie -
Mar 16 (Spring Break) TBD -
Mar 23 Midterm Project Presentations All Students Midterm project write-up's due Apr1
Mar 30 (univ holiday) -
Apr 6 Human-Robot Task Dialogue Yixin:
(1) Clarifying commands with information-theoretic human-robot dialog. Deits et al. (2012) ;
(2) Asking for help using inverse semantics. Tellex et al. (2014) ;
Haonan:
(3) Navigational Instruction Generation as Inverse Reinforcement Learning with Neural Machine Translation. Daniele et al. (2017) ;
(4) Back to the Blocks World: Learning New Actions through Situated Human-Robot Dialogue. She et al. (2014) ;
Justin:
(5) PLOW: A Collaborative Task Learning Agent. Allen et al. (2007) ;
(6) Learning to Interpret Natural Language Commands through Human-Robot Dialog. Thomason et al. (2015)
Yixin, Haonan, Justin -
Apr 13 Human-Robot Task Dialogue 2, Language Learning via Dialogue-based and Interactive Games**, Turn-taking/Gaze Hyounghun: (1) Embodied Collaborative Referring Expression Generation in Situated Human-Robot Interaction. Fang et al. (2015) ;
Yicheng: (2) Multi-Agent Cooperation and the Emergence of (Natural) Language. Lazaridou et al. (2017) ;
Yixin: (3) Emergent Communication in a Multi-Modal, Multi-Step Referential Game. Evtimova et al. (2017) ;
Jie: (4) Natural Language Does Not Emerge 'Naturally' in Multi-Agent Dialog. Kottur et al. (2017) ;
Eric: (5) Simon plays Simon says: The timing of turn-taking in an imitation game. Chao et al. (2011) ;
Han: (6) Generation of Nodding, Head Tilting and Eye Gazing for Human-Robot Dialogue Interaction. Liu et al. (2012)
Hyounghun, Yicheng, Yixin, Jie, Eric, Han -
Apr 20 Going Forward: Better Evaluation** and Real-World Deployment of Dialogue Systems** Ram: (1) How NOT To Evaluate Your Dialogue System: An Empirical Study of Unsupervised Evaluation Metrics for Dialogue Response Generation. Liu et al. (2016) ;
Tong: (2) Towards an Automatic Turing Test: Learning to Evaluate Dialogue Responses. Lowe et al. (2017)
Hao: (3) Why We Need New Evaluation Metrics for NLG. Novikova et al. (2017) ;
Yichen: (4) Adversarial Evaluation for Open-Domain Dialogue Generation. Bruni and Fernandez (2017) ;
Sweta: (5) Sounding Board – University of Washington’s Alexa Prize Submission. Fang et al. (2017) ;
Lisa: (6) Alana: Social Dialogue using an Ensemble Model and a Ranker trained on User Feedback. Papaioannou et al. (2017)
Yichen, Ram, Tong, Lisa, Sweta, Hao -
May 04 Final Project Presentations All Students Final project write-up's due May7


**Other papers on language emergence via multi-agent dialogue games:
(1) Emergence of Grounded Compositional Language in Multi-Agent Populations. Mordatch and Abbeel (2017)
(2) Emergence of Language with Multi-agent Games: Learning to Communicate with Sequences of Symbols. Havrylov and Titov (2017)
(3) Learning Language Games through Interaction. Wang et al. (2016)


**Other papers on better dialogue evaluation:
(1) A Review of Evaluation Techniques for Social Dialogue Systems. Curry et al. (2017)
(2) Learning and Evaluation of Dialogue Strategies for New Applications: Empirical Methods for Optimization from Small Data Sets. Rieser and Lemon (2011)


**Other papers on Alexa Prize systems:
(1) Alexa Prize Proceedings


Disclaimer

The professor reserves the right to make changes to the syllabus, including project due dates. These changes will be announced as early as possible.