pith. sign in

arxiv: 1907.10658 · v1 · pith:4Q4ZXMSAnew · submitted 2019-07-22 · 💻 cs.CL · cs.HC

SlugBot: Developing a Computational Model andFramework of a Novel Dialogue Genre

Kevin K.Bowden , Jiaqi Wu , Wen Cui , Juraj Juraska , Vrindavan Harrison , Brian Schwarzmann , Nick Santer , Marilyn Walker This is my paper

Pith reviewed 2026-05-24 17:42 UTC · model grok-4.3

classification 💻 cs.CL cs.HC
keywords dialogue modelAlexa Prizediscourse relationsontological resourceopen-domain dialogueconversational systemsSlugBotUniSlug
0
0 comments X

The pith

A new Discourse Relation Dialogue Model is developed and implemented in SlugBot to handle Alexa Prize open conversation beyond traditional models.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that task-oriented and search-oriented dialogue models cannot support the open conversational requirements of Alexa Prize systems. No prior computational theories define expected behaviors in this genre, so the team creates the Discourse Relation Dialogue Model. Implementing this model in a modular system tests and refines it while producing the UniSlug ontological resource. The structure of UniSlug then shapes content curation and the dialogue manager's operation.

Core claim

No existing computational theories circumscribe the expected system and user behaviors in the Alexa Prize conversational genre. A novel Discourse Relation Dialogue Model is therefore developed and paired with modular implementation in SlugBot to test and refine it, yielding the UniSlug ontological resource whose structure determines how content is curated and structured for the dialogue manager.

What carries the argument

The Discourse Relation Dialogue Model, which provides the theoretical framework for system and user behaviors and is tested via modular implementation leading to UniSlug.

If this is right

  • The model supports conversational capabilities that go beyond traditional task-oriented or search-oriented dialogue.
  • Modular implementation allows the theoretical model to be tested and refined through actual system use.
  • A novel ontological resource called UniSlug is produced as a direct outcome of the model.
  • UniSlug's structure dictates content curation and how the dialogue manager implements the model.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The iterative model-implementation loop could serve as a template for building dialogue systems in other open conversational domains.
  • UniSlug might be extended to incorporate external knowledge sources while preserving the discourse relation structure.
  • The approach highlights a general need to match theoretical models tightly to specific dialogue genres rather than reusing generic frameworks.

Load-bearing premise

No existing computational theories circumscribe the expected system and user behaviors in the Alexa Prize conversational genre.

What would settle it

An existing task-oriented or search-oriented dialogue model shown to fully support Alexa Prize open conversational behaviors without modification or extension would falsify the need for a new model.

Figures

Figures reproduced from arXiv: 1907.10658 by Brian Schwarzmann, Jiaqi Wu, Juraj Juraska, Kevin K.Bowden, Marilyn Walker, Nick Santer, Vrindavan Harrison, Wen Cui.

Figure 1
Figure 1. Figure 1: UniSlug is SlugBot’s large scale integrated ontology for Dialogue Management and Natural [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: SlugBot system architecture. do coreference resolution by mapping the coreference tags returned by CoreNLP to the data stored within our system. Entity Detection: We have also developed our own named entity recognizer SlugNERDS [9, 7] because the existing named entity recognizers were not fine-grained enough to support dialogue interaction. SlugNERDS is based on the Google Knowledge Graph1 and can take adv… view at source ↗
Figure 3
Figure 3. Figure 3: Content sources for SlugBot. UniSlug: In addition to these content sources, we host the UniSlug graph database illustrated in [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Major user moods. 2.5.1 Mixed Initiative Modules User what do you think of watchmen Agent I really like watchmen. In fact I think watchmen is action packed because there are lot’s of cool fight scenes. How about you, what’s your opinion? User do you like watchmen Agent Bot: Yes, I believe watchmen is really interesting. And you? What’s your opin￾ion? User What’s your favorite color? Agent I really like inf… view at source ↗
Figure 5
Figure 5. Figure 5: Sample flow. 1 User Precondition A 2 Agent Action A, Postcondition A 3 User Precondition C 4 Agent Action C, Postcondition C 5 User Precondition B 6 Agent Action B, Postcondition B 7 User No Precondition 8 Agent Exit Flow [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Our Web Application Architecture [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
read the original abstract

One of the most interesting aspects of the Amazon Alexa Prize competition is that the framing of the competition requires the development of new computational models of dialogue and its structure. Traditional computational models of dialogue are of two types: (1) task-oriented dialogue, supported by AI planning models,or simplified planning models consisting of frames with slots to be filled; or (2)search-oriented dialogue where every user turn is treated as a search query that may elaborate and extend current search results. Alexa Prize dialogue systems such as SlugBot must support conversational capabilities that go beyond what these traditional models can do. Moreover, while traditional dialogue systems rely on theoretical computational models, there are no existing computational theories that circumscribe the expected system and user behaviors in the intended conversational genre of the Alexa Prize Bots. This paper describes how UCSC's SlugBot team has combined the development of a novel computational theoretical model, Discourse Relation Dialogue Model, with its implementation in a modular system in order to test and refine it. We highlight how our novel dialogue model has led us to create a novel ontological resource, UniSlug, and how the structure of UniSlug determine show we curate and structure content so that our dialogue manager implements and tests our novel computational dialogue model.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper claims that traditional task-oriented (planning/frames) and search-oriented dialogue models are insufficient for the open conversational genre of Alexa Prize bots, asserts that no existing computational theories circumscribe expected system and user behaviors in this genre, and describes the development of a new Discourse Relation Dialogue Model implemented in a modular SlugBot system to test and refine it, resulting in the UniSlug ontology that structures content curation and dialogue management.

Significance. If the necessity of the new model were demonstrated via comparison to prior work and if the implementation were accompanied by independent evaluation metrics showing improved handling of open dialogue, the work could provide a useful framework and resource (UniSlug) for conversational systems beyond task-oriented domains. As presented, the absence of such grounding and results limits the contribution to a high-level description of an approach.

major comments (2)
  1. [Abstract] Abstract: The central premise that 'there are no existing computational theories that circumscribe the expected system and user behaviors in the intended conversational genre of the Alexa Prize Bots' is asserted without any literature review, comparison to existing models (e.g., information-state update, rhetorical structure theory applications, or open-domain dialogue frameworks), or concrete failure cases of task-oriented/search models on Alexa Prize requirements. This assertion is load-bearing for the novelty of the Discourse Relation Dialogue Model.
  2. [Abstract] Abstract: The paper states that the Discourse Relation Dialogue Model 'is tested and refined' through its implementation in the modular system leading to UniSlug, yet supplies no evaluation results, error analysis, independent metrics, or external benchmarks. Without these, the success criteria appear defined internally by the implementation choices, undermining claims that the model was validated.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments. We address each major point below and will revise the manuscript to strengthen the claims where possible.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The central premise that 'there are no existing computational theories that circumscribe the expected system and user behaviors in the intended conversational genre of the Alexa Prize Bots' is asserted without any literature review, comparison to existing models (e.g., information-state update, rhetorical structure theory applications, or open-domain dialogue frameworks), or concrete failure cases of task-oriented/search models on Alexa Prize requirements. This assertion is load-bearing for the novelty of the Discourse Relation Dialogue Model.

    Authors: We agree that the abstract's assertion would be strengthened by explicit comparison. The manuscript introduction contrasts task-oriented planning/frame-based models and search-oriented models with the open conversational requirements of Alexa Prize bots, but does not include a dedicated related-work section or concrete failure cases. We will add a related-work subsection that reviews information-state update approaches, RST applications to dialogue, and open-domain frameworks, along with specific examples of where those models are insufficient for sustained, non-task, non-search conversation. This directly addresses the load-bearing claim for novelty. revision: yes

  2. Referee: [Abstract] Abstract: The paper states that the Discourse Relation Dialogue Model 'is tested and refined' through its implementation in the modular system leading to UniSlug, yet supplies no evaluation results, error analysis, independent metrics, or external benchmarks. Without these, the success criteria appear defined internally by the implementation choices, undermining claims that the model was validated.

    Authors: The manuscript presents the iterative development process in which implementation challenges directly informed refinements to the Discourse Relation Dialogue Model and produced the UniSlug ontology. However, it is correct that no quantitative metrics, error analysis, or external benchmarks are reported. Because the contribution centers on the theoretical model and the resulting ontology rather than end-to-end system performance, we will expand the implementation section with concrete examples of model refinements triggered by system testing and any available internal qualitative observations. We cannot add new external benchmarks without additional experiments beyond the current scope. revision: partial

Circularity Check

0 steps flagged

No circularity; premise of novelty is asserted without self-referential reduction

full rationale

The paper asserts that traditional task-oriented and search-oriented models are insufficient and that 'there are no existing computational theories that circumscribe the expected system and user behaviors in the intended conversational genre of the Alexa Prize Bots,' then describes developing the Discourse Relation Dialogue Model and implementing it in UniSlug to test and refine it. This is a premise about external literature rather than a derivation, equation, or fitted parameter that reduces to its own inputs by construction. No self-citations, ansatzes, uniqueness theorems, or renamings appear in the provided text. The iterative model-then-implement process is standard and does not equate the claimed result to the input definitionally.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 2 invented entities

The central claim rests on the introduction of two new constructs without independent evidence or derivations from prior literature; no free parameters are mentioned but the model itself functions as an invented organizing structure.

axioms (1)
  • domain assumption Traditional task-oriented and search-oriented models are insufficient for the Alexa Prize conversational genre.
    Stated directly in the abstract as the motivation for developing a new model.
invented entities (2)
  • Discourse Relation Dialogue Model no independent evidence
    purpose: Computational theoretical model to circumscribe expected system and user behaviors in open conversational dialogue.
    Newly proposed in the paper with no prior citations or external validation described.
  • UniSlug no independent evidence
    purpose: Ontological resource whose structure determines content curation for the dialogue manager.
    Newly created resource introduced to support the model implementation.

pith-pipeline@v0.9.0 · 5769 in / 1251 out tokens · 26403 ms · 2026-05-24T17:42:47.946639+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

43 extracted references · 43 canonical work pages · 3 internal anchors

  1. [1]

    Allen and C

    James F. Allen and C. Raymond Perrault. Analyzing intention in utterances. Artificial Intelli- gence, 15:143–178, 1980

    work page 1980

  2. [2]

    Luke, i am your father: Dealing with out-of-domain requests by using movies subtitles

    David Ameixa, Luisa Coheur, Pedro Fialho, and Paulo Quaresma. Luke, i am your father: Dealing with out-of-domain requests by using movies subtitles. In Timothy Bickmore, Stacy Marsella, and Candace Sidner, editors, Intelligent Virtual Agents, pages 13–21, Cham, 2014. Springer International Publishing

    work page 2014

  3. [3]

    Logics of conversation

    Nicholas Asher and Alex Lascarides. Logics of conversation. Cambridge University Press, 2003

    work page 2003

  4. [4]

    Banchs and Haizhou Li

    Rafael E. Banchs and Haizhou Li. Iris: A chat-oriented dialogue system based on the vector space model. In Proceedings of the ACL 2012 System Demonstrations, ACL ’12, pages 37–42, Stroudsburg, PA, USA, 2012. Association for Computational Linguistics

    work page 2012

  5. [5]

    Large-scale personal assistant technology deployment: the siri experience

    Jerome R Bellegarda. Large-scale personal assistant technology deployment: the siri experience. In INTERSPEECH, pages 2029–2033, 2013

    work page 2029

  6. [6]

    M2D: Monolog to Dialog Generation for Conversational Story Telling

    Kevin K. Bowden, Grace I. Lin, Lena I. Reed, and Marilyn A. Walker. M2D: monolog to dialog generation for conversational story telling. CoRR, abs/1708.07476, 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  7. [7]

    arXiv preprint arXiv:1709.05411

    Kevin K. Bowden, Shereen Oraby, JiaQi Wu, Amita Misra, and Marilyn A. Walker. Combining search with structured data to create a more engaging user experience in open domain dialogue. CoRR, abs/1709.05411, 2017

    work page arXiv 2017

  8. [8]

    Slugbot: An Application of a Novel and Scalable Open Domain Socialbot Framework

    Kevin K Bowden, Jiaqi Wu, Shereen Oraby, Amita Misra, and Marilyn Walker. Slugbot: An application of a novel and scalable open domain socialbot framework. arXiv preprint arXiv:1801.01531, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  9. [9]

    Bowden, JiaQi Wu, Shereen Oraby, Amita Misra, and Marilyn A

    Kevin K. Bowden, JiaQi Wu, Shereen Oraby, Amita Misra, and Marilyn A. Walker. Slugnerds: A named entity recognition tool for open domain dialogue systems. InProceedings of the Eleventh International Conference on Language Resources and Evaluation, LREC 2018, Miyazaki, Japan, May 7-12, 2018., 2018

    work page 2018

  10. [10]

    Building a discourse-tagged corpus in the framework of rhetorical structure theory

    Lynn Carlson, Daniel Marcu, and Mary Ellen Okurowski. Building a discourse-tagged corpus in the framework of rhetorical structure theory. In Proc. of the Second SIGdial Workshop on Discourse and Dialog, Aalborg, Denmark, 2001

    work page 2001

  11. [11]

    Comparing system-response retrieval models for open-domain and casual conversational agent

    Franck Charras, Guillaume Dubuisson Duplessis, Vincent Letard, Anne-Laure Ligozat, and Sophie Rosset. Comparing system-response retrieval models for open-domain and casual conversational agent. In Second Workshop on Chatbots and Conversational Agent Technologies (WOCHAT@ IVA2016), 2016

    work page 2016

  12. [12]

    Purely corpus-based automatic conversation authoring

    Guillaume Dubuisson Duplessis, Vincent Letard, Anne-Laure Ligozat, and Sophie Rosset. Purely corpus-based automatic conversation authoring. In LREC, 2016

    work page 2016

  13. [13]

    Purely corpus-based automatic conversation authoring

    Guillaume Dubuisson Duplessis, Vincent Letard, Anne-Laure Ligozat, and Sophie Rosset. Purely corpus-based automatic conversation authoring. InProceedings of the Tenth International Conference on Language Resources and Evaluation LREC 2016, Portorož, Slovenia, May 23-28, 2016., 2016

    work page 2016

  14. [14]

    Dramabank: Annotating agency in narrative discourse

    David Elson. Dramabank: Annotating agency in narrative discourse. InLREC, pages 2813–2819, 2012

    work page 2012

  15. [15]

    Forsythand and Craig H

    Eric N. Forsythand and Craig H. Martell. Lexical and discourse analysis of online chat dialog. In Proceedings of the International Conference on Semantic Computing, ICSC ’07, pages 19–26, Washington, DC, USA, 2007. IEEE Computer Society. 17

    work page 2007

  16. [16]

    Cowan, Carl V ogel, and Nick Campbell

    Emer Gilmartin, Benjamin R. Cowan, Carl V ogel, and Nick Campbell. Chunks in multiparty conversation - building blocks for extended social talk. In IWSDS ’17, 2017

    work page 2017

  17. [17]

    Grosz and Candace L

    Barbara J. Grosz and Candace L. Sidner. Attention, intentions and the structure of discourse. Computational Linguistics, 12:175–204, 1986

    work page 1986

  18. [18]

    Active, a platform for building intelli- gent operating rooms

    Didier Guzzoni, Adam Cheyer, and Charles Baur. Active, a platform for building intelli- gent operating rooms. In Surgetica 2007 Computer-Aided Medical Interventions: tools and applications, pages 191–198. Sauramps Médical, 2007

    work page 2007

  19. [19]

    Probabilistic enrichment of knowledge graph entities for relation detection in conversational understanding

    Dilek Hakkani-Tür, Asli Celikyilmaz, Larry Heck, Gokhan Tur, and Geoff Zweig. Probabilistic enrichment of knowledge graph entities for relation detection in conversational understanding. In Proceedings of INTERSPEECH, 2014

    work page 2014

  20. [20]

    Towards an open- domain conversational system fully based on natural language processing

    Ryuichiro Higashinaka, Kenji Imamura, Toyomi Meguro, Chiaki Miyazaki, Nozomi Kobayashi, Hiroaki Sugiyama, Toru Hirano, Toshiro Makino, and Yoshihiro Matsuo. Towards an open- domain conversational system fully based on natural language processing. In Proceedings of COLING 2014, the 25th International Conference on Computational Linguistics: Technical Paper...

    work page 2014

  21. [21]

    Jerry R. Hobbs. Why is discourse coherent? Technical Report 176, SRI International, 333 Ravenswood Ave., Menlo Park, Ca 94025, 1978

    work page 1978

  22. [22]

    Hobbs and D

    J.R. Hobbs and D. Evans. Conversation as planned behaviour. Technical Report 203, Artificial Intelligence Center, SRI International, Menlo Park, CA, 1979

    work page 1979

  23. [23]

    Bowden, and Marilyn A

    Juraj Juraska, Panagiotis Karagiannis, Kevin K. Bowden, and Marilyn A. Walker. A deep ensemble model with slot alignment for sequence-to-sequence natural language generation. NAACL, 2018

    work page 2018

  24. [24]

    Evaluating spoken dialog systems for telecommunication services

    Candace Kamm, Shrikanth Narayanan, Dawn Dutton, and Russell Ritenour. Evaluating spoken dialog systems for telecommunication services. In Fifth European Conference on Speech Communication and Technology, 1997

    work page 1997

  25. [25]

    Edina: Building an Open Domain Socialbot with Self-dialogues

    Ben Krause, Marco Damonte, Mihai Dobre, Daniel Duma, Joachim Fainberg, Federico Fancellu, Emmanuel Kahembwe, Jianpeng Cheng, and Bonnie Webber. Edina: Building an open domain socialbot with self-dialogues. arXiv preprint arXiv:1709.09816, 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  26. [26]

    Redundancy localiza- tion for the conversationalization of unstructured responses

    Sebastian Krause, Mikhail Kozhevnikov, Eric Malmi, and Daniele Pighin. Redundancy localiza- tion for the conversationalization of unstructured responses. In Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue, pages 115–126, 2017

    work page 2017

  27. [27]

    Opensubtitles2018: Statistical rescoring of sentence alignments in large, noisy parallel corpora

    Pierre Lison, Jörg Tiedemann, and Milen Kouylekov. Opensubtitles2018: Statistical rescoring of sentence alignments in large, noisy parallel corpora. InProceedings of the Eleventh International Conference on Language Resources and Evaluation, LREC 2018, Miyazaki, Japan, May 7-12, 2018., 2018

    work page 2018

  28. [28]

    Plan recognition and discourse analysis: An integrated approach for understand- ing dialogues

    Diane Litman. Plan recognition and discourse analysis: An integrated approach for understand- ing dialogues. Technical Report 170, University of Rochester, 1985

    work page 1985

  29. [29]

    Using entity features to classify implicit relations

    Annie Louis, Aravind Joshi, Rashmi Prasad, and Ani Nenkova. Using entity features to classify implicit relations. In Proc. of the 11th Annual SIGdial Meeting on Discourse and Dialogue, Tokyo, Japan, 2010

    work page 2010

  30. [30]

    Lukin, Kevin Bowden, Casey Barackman, and Marilyn A

    Stephanie M. Lukin, Kevin Bowden, Casey Barackman, and Marilyn A. Walker. A corpus of personal narratives and their story intention graphs. In In Proceedings of the 10th International Conference on Language Resources and Evaluation (LREC), 2016

    work page 2016

  31. [31]

    Mann and Sandra A

    William C. Mann and Sandra A. Thompson. Rhetorical structure theory. Toward a functional theory of text organization. Text, 8(3):243–281, 1988. 18

    work page 1988

  32. [32]

    The stanford corenlp natural language processing toolkit

    Christopher Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven Bethard, and David McClosky. The stanford corenlp natural language processing toolkit. In Proceedings of 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations , pages 55–60. Association for Computational Linguistics, 2014

    work page 2014

  33. [33]

    Marcu and A

    D. Marcu and A. Echihabi. An unsupervised approach to recognizing discourse relations. In Proc. of the 40th Annual Meeting on Association for Computational Linguistics, pages 368–375, 2002

    work page 2002

  34. [34]

    Building up rhetorical structure trees

    Daniel Marcu. Building up rhetorical structure trees. In Proc. of AAAI/IAAI 1996, volume 2, pages 1069–1074, 1996

    work page 1996

  35. [35]

    Developing Non-goal Dialog System Based on Examples of Drama Television, pages 355–361

    Lasguido Nio, Sakriani Sakti, Graham Neubig, Tomoki Toda, Mirna Adriani, and Satoshi Nakamura. Developing Non-goal Dialog System Based on Examples of Drama Television, pages 355–361. Springer New York, New York, NY , 2014

    work page 2014

  36. [36]

    Prasad, A

    R. Prasad, A. Joshi, and B. Webber. Realization of discourse relations by other means: alternative lexicalizations. In Proc. of the 23rd International Conference on Computational Linguistics: Posters, pages 1023–1031. Association for Computational Linguistics, 2010

    work page 2010

  37. [37]

    The Penn Discourse TreeBank 2.0

    Rashmi Prasad, Nikhil Dinesh, Alan Lee, Eleni Miltsakaki, Livio Robaldo, Aravind Joshi, and Bonnie Webber. The Penn Discourse TreeBank 2.0. In Proc. of 6th International Conference on Language Resources and Evaluation (LREC 2008), 2008

    work page 2008

  38. [38]

    Interactive reinforcement learning for task-oriented dialogue management

    Pararth Shah, Dilek Hakkani-Tür, and Larry Heck. Interactive reinforcement learning for task-oriented dialogue management. In NIPS 2016 Deep Learning for Action and Interaction Workshop, 2016

    work page 2016

  39. [39]

    Bootstrapping a neural con- versational agent with dialogue self-play, crowdsourcing and on-line reinforcement learning

    Pararth Shah, Dilek Hakkani-Tur, Bing Liu, and Gokhan Tur. Bootstrapping a neural con- versational agent with dialogue self-play, crowdsourcing and on-line reinforcement learning. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 3 (Industry Papers), vo...

    work page 2018

  40. [40]

    Tonelli, G

    S. Tonelli, G. Riccardi, R. Prasad, and A. Joshi. Annotation of discourse relations for conversa- tional spoken dialogs. In Proc. of the Seventh International Conference on Language Resources and Evaluation (LREC 2010), Valletta, Malta, pages 2084–2090, 2010

    work page 2010

  41. [41]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. In Advances in Neural Informa- tion Processing Systems, pages 5998–6008, 2017

    work page 2017

  42. [42]

    M. A. Walker, D. Litman, C. A. Kamm, and A. Abella. PARADISE: A general framework for evaluating spoken dialogue agents. In Proc. of the 35th Annual Meeting of the Association for Computational Linguistics, ACL/EACL 97, pages 271–280, 1997

    work page 1997

  43. [43]

    Eliza - a computer program for the study of natural language communica- tion between man and machine

    Joseph Weizenbaum. Eliza - a computer program for the study of natural language communica- tion between man and machine. Commun. ACM, 9(1):36–45, January 1966. 19

    work page 1966