Abstract
Two new methods of analyzing dialogue interactions are outlined. One method depends on abstract representations of dialogue events as symbols in a formal language. This method invites analysis of the expressivity requirements of dialogue grammar, as well as distribution analysis of dialogue event symbol sequences. The method is presented in relation to a temporal construction from regular languages, one which supports increasingly fine granularity of temporal analysis. The other method proposed is also temporally oriented. It also depends on dialogue events and dialogue states, and proposes to analyze causal relations among dialogue events through survival analysis. These methods are suggested as additions to the extant repertoire of approaches to understanding the structure and temporal flow of natural dialogue. Additional methods of analysis of natural dialogue may contribute to deeper understanding of the phenomena. With deeper undertanding of natural dialogue one may hope to more fully inform the construction of believable artificial systems that are intended to engage in dialogue with a manner close to human interaction in dialogue.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
In this structure, the first element, S, is a special non-terminal symbol, the start symbol; the next element the remaining non-terminal symbols; the third element is a set of terminal symbols; the final element is a set of productions, here pointers to the productions. Each set in this structure is finite. The language generated is infinite.
- 2.
The representation \(s^i\) is adopted as a positive representation of silence, thinking of this as distinct from the usual representation of the empty string (\(\varepsilon \)). Shortly, representation of distinct sorts of silence will be introduced.
- 3.
On the first interpretation, \(v^k\) may represent an utterance such as, “the cat is on the mat.” On the second interpretation, \(v^k\) may represent the proposition, “‘the cat is on the mat’ is being uttered”. In the first case, \(v^k\) stands for the vocalization as a vocalization, and in the second case it stands for the proposition that the vocalization is under way.
- 4.
During \(\varDelta ^\dagger = \delta (\text {``Sam has''})\), \(\varDelta ^\dagger \) a proper sub-interval of \(\varDelta ^\ddagger = \delta (\text {``Sam has water''})\), the proposition that a vocalization of “Sam has baked a loaf of bread” is happening is not true.
- 5.
The symbol a is meant to be suggestive of the English onomatopoeic expression, “achoo”.
- 6.
I have not proven that the addition of clocks as proposed here does not increase the expressivity of the framework beyond the expressivity of regular languages. The intuition behind the argument that regular grammars with clocks of this sort remain regular is that the “grouping” of symbols within sets, in an initial grammar, only ever includes a single terminal symbol, never a non-terminal symbol; thus, the effect of bracket matching on either side of a recursive use of a non-terminal symbol (such as is the prototype of a context-free grammar rule) does not occur. The information encoded by the clock for a language may be captured in a constructed regular language without a clock by adding a terminal symbol for the starting time, and another symbol for subsequent “ticks” of the clock. Also, recall that regular languages are closed under both intersection and union [12].
- 7.
At greater levels of expressivity, context-free and higher, the inclusion problem is not decidable [12].
- 8.
In the present work, an overlap is equivalent to a component like \(\{v^1,v^2\}\).
- 9.
In order to subtract away a constant c from T using (5.36), it is sufficient to specify d as \(\frac{T-c}{T}\); to test whether \(T-\lnot L\) is less than some constant c, it suffices to specify d as \(\frac{c}{T}\).
- 10.
That is, this is true if laughter is close enough to the change of topic to be reasonably hypothesized as a discourse marker.
- 11.
Non-laughter, which terminates with the onset of laughter, is constructed as a dialogue event, so that one may examine the survival rate of non-laughter.
- 12.
Interestingly, given that dialogue participants may be imputed to all have a sense of when discussion of a topic is coming to an end, if this sensation does give release to laughter, or other signals, that correlate with topic end, then there is a case to be made for temporally inverted causation: the topic end may cause the laughter which precedes it.
- 13.
Laughter, it turns out, is not a precise signal of topic ending, but if laughter is present, it serves as a very good signal that the topic of discussion has not just started.
- 14.
Using the one-tailed asymptotic log-rank test \(p < 0.005\). See surv_test within the coin package within R (http://cran.r-project.org/web/packages/coin/index.html—last verified, July 2015) created by Torsten Hothorn, Kurt Hornik, Mark A. van de Wiel and Achim Zeileis.
- 15.
It is important to remember that the topic durations in the two representative plots relate to the same topics, but under distinct classifications.
- 16.
It is also necessary to apply this sort of analysis to additional dialogue corpora in order to understand the robustness of the effects found for the TableTalk corpus.
References
Bonin F, Campbell N, Vogel C (2014) Time for laughter. Knowl Based Syst 71:15–24
Bonin F, Vogel C, Campbell N (2014) Social sequence analysis: temporal sequences in interactional conversations. In: Péter B (ed) Proceedings of the 5th IEEE international conference on cognitive infocommunications, pp 403–406
Bouamrane MM, Luz S (2007) An analytical evaluation of search by content and interaction patterns on multimodal meeting records. Multimed Syst 13:89–102. doi:10.1007/s00530-007-0087-8
Bunt H, Alexandersson J, Chae JW, Fang AC, Hasida K, Petukhova O, Popescu-Belis A, Traum D (2012) ISO 24617–2: a semantically-based standard for dialogue act annotation. In: Proceedings of the LREC
Campbell N (2009) An audio-visual approach to measuring discourse synchrony in multimodal conversation data. In: Proceedings of the interspeech
Covington MA, He C, Brown C, Naçi L, McClain JT, Fjordbak BS, Semple J, Brown J (2005) Schizophrenia and the structure of language: the linguists view. Schizophr Res 77:85–98
Esposito A, Esposito AM, Vogel C (2015) Needs and challenges in human computer interaction for processing social emotional information. Pattern Recognit Lett (to appear). http://dx.doi.org/10.1016/j.patrec.2015.02.013
Fernando T (2004) A finite-state approach to events in natural language semantics. J Logic Comput 14(1):79–92. doi:10.1093/logcom/14.1.79
Fernando T (2007) Observing events and situations in time. Linguist Philos 30(5):527–550. doi:10.1007/s10988-008-9026-1
Gazdar G, Pullum G (1985) Computationally relevant properties of natural languages and their grammars. Technical Report CSLI-85-24, stanford: center for the study of language and information
Gilmartin E, Bonin F, Vogel C, Campbell N (2013) Laugher and topic transition in multiparty conversation. In: Fernández R, Isard A (eds) Proceedings of the SIGDIAL 2013 conference. Association for computational linguistics, Metz, pp 304–308. http://www.aclweb.org/anthology/W/W13/W13-4045
Hopcroft JE, Ullman JD (1979) Introduction to automata theory, languages, and computation. Addison-Wesley Publishing Co., Reading
Păun G (1984) Modelling the dialogue by means of formal language theory. In: Vaina L, Hintikka J (eds) Cognitive constraints on communication: representations and processes. Reidel, Dordrecht, pp 363–372
Petukhova V, Bunt H (2009) Towards a multidimensional semantics of discourse markers in spoken dialogue. In: Proceedings of the eighth international conference on computational semantics. Association for computational linguistics, Stroudsburg, IWCS-8 09, pp 157–168
Ramseyer F, Tschacher W (2010) Nonverbal synchrony or random coincidence? How to tell the difference. In: Esposito A, Campbell N, Vogel C, Hussain A, Nijholt A (eds) Development of multimodal interfaces: active listening and synchrony. LNCS. Springer, Berlin, pp 182–196
Reitter D, Moore J (2007) Predicting success in dialogue. In: Proceedings of the 45th annual meeting of the association of computational linguistics. Association for computational linguistics, pp 808–815
Reitter D, Keller F, Moore J (2006) Computational modeling of structural priming in dialogue. In: Proceedings of the human language technology conference of the North American chapter of the ACL. Association for computational linguistics, pp 121–124
Schegloff EA (2007) Sequence organization in interaction: a primer in conversational analysis, vol 1. Cambridge University Press, Cambridge
Vogel C (2013) Attribution of mutual understanding. J Law Policy 21(2):377–420
Vogel C, Behan L (2012) Measuring synchrony in dialog transcripts. In: Esposito A, Esposito AM, Vinciarelli A, Hoffmann R, Müller VC (eds) Cognitive behavioural systems, vol 7403. LNCS. Springer, Heidelberg, pp 73–88
Acknowledgments
I am grateful to Francesca Bonin, Nick Campbell, Anna Esposito and Emer Gilmartin. They have no responsibility for any wrong-headedness that appears here, though. This research is enhanced by supported from Science Foundation Ireland through the CNGL Programme (Grant 12/CE/I2267) in the ADAPT Centre (www.adaptcentre.ie) at Trinity College Dublin. The ADAPT Centre for Digital Content Technology is funded under the SFI Research Centres Programme (Grant 13/RC/2106) and is co-funded under the European Regional Development Fund.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Vogel, C. (2016). Communication Sequences and Survival Analysis. In: Esposito, A., Jain, L. (eds) Toward Robotic Socially Believable Behaving Systems - Volume II . Intelligent Systems Reference Library, vol 106. Springer, Cham. https://doi.org/10.1007/978-3-319-31053-4_5
Download citation
DOI: https://doi.org/10.1007/978-3-319-31053-4_5
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-31052-7
Online ISBN: 978-3-319-31053-4
eBook Packages: EngineeringEngineering (R0)