An ISU dialogue system exhibiting reinforcement learning of dialogue policies

Lemon, Oliver; Georgila, Kallirroi; Henderson, James; Stuttle, Matthew N.

doi:10.3115/1608974.1608986

Cited by 72 publications

(53 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Methods of analysing these asymmetries are summarized in the work of Coan and Allen [62]. Generally, most scientists today focus on the reduction in alpha band (8)(9)(10)(11)(12)(13) activity, for it is experimentally proved that there is an inverse relationship between alpha activity and brain activation when experiencing emotions. In other words -when emotions are experienced, we can observe a reduction in alpha band activity, which can be measured using EEG [61].…”

Section: Eegmentioning

confidence: 99%

“…Another type of approach is presented by Lemon et al [11], who proposed what they call an "ISU (Information State Update) Dialogue System", which uses reinforcement learning of dialogue strategies, and also has a fragmentary clarification feature. Primarily constructed in order to be able to collect data for Reinforcement Learning (RL) approaches to multimodal dialogue management, the ISU system can perform conversations within the in-car scenarios, focusing on providing the users with touristic information (e.g.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Evaluating Subjective Aspects of Hci on an Example of a Non-Task Oriented Conversational System

Dybała

Ptaszyński

Rzepka

et al. 2010

Int. J. Artif. Intell. Tools

View full text Add to dashboard Cite

The evaluation of subjective aspects of HCI, such as human-likeness, likeability or users' emotions towards computers is still quite a neglected issue, especially in the field of non-task oriented conversational systems (chatterbots). In this paper we try to bridge this gap by proposing a new methodology of evaluation. The methods presented were tested in our research on humor-equipped chatterbots. We describe them in details, discuss their drawbacks and usability. In one of the presented methods we used an emotiveness analysis system, which itself can be considered an AI tool, as it was used to detect users' emotions towards conversational systems, and to perform their automatic evaluation. We also propose some methods that we have not used yet, which, however, * Corresponding author: paweldybala@media.eng.hokudai.ac.jp (Pawel Dybala) Pawel Dybala, Michal Ptaszynski, Rafal Rzepka, and Kenji Araki 2 seem applicable in this field, such as brain scanning techniques. Finally, we give some ideas that should be addressed in the future.

show abstract

Section: Eegmentioning

confidence: 99%

mentioning

confidence: 99%

Evaluating Subjective Aspects of Hci on an Example of a Non-Task Oriented Conversational System

Dybała

Ptaszyński

Rzepka

et al. 2010

Int. J. Artif. Intell. Tools

View full text Add to dashboard Cite

show abstract

“…In general, for such research we require data that has either been generated and logged by context-tracking dialogue systems (Gabsdil and Lemon 2004;Lemon, Georgila and Henderson 2006a;Lemon et al 2006b) or that has been subsequently annotated (or a mixture of both). Both preliminary versions of our annotations and the version that we present here have been used successfully in Georgila et al (2005aGeorgila et al ( , 2006, Henderson et al (2005Henderson et al ( , 2008, Schatzmann et al (2005aSchatzmann et al ( , 2005b, Frampton and Lemon (2006).…”

Section: The 'Information State Update' Approachmentioning

confidence: 99%

“…The experiments were done using the 'TownInfo' multimodal dialogue system of Lemon et al (2006b). The learnt policy (trained on the Communicator data) was ported to the city information domain, and then evaluated with human subjects.…”

Section: Indirect Evaluation Via User Simulationsmentioning

confidence: 99%

“…As a proof of concept we modified the automatic annotation system in order to automatically annotate the TALK city information corpus. The TALK corpus consists of humanmachine dialogues in the city-information domain collected by having real users interact with the TALK 'TownInfo' dialogue system (Lemon et al 2006a(Lemon et al , 2006b). Users are asked to find a hotel, bar, or restaurant in a town.…”

Section: Porting To Other Domains and Applicationsmentioning

confidence: 99%

See 1 more Smart Citation

Automatic annotation of context and speech acts for dialogue corpora

Georgila¹,

Lemon²,

Henderson³

et al. 2009

Nat. Lang. Eng.

Self Cite

View full text Add to dashboard Cite

Richly annotated dialogue corpora are essential for new research directions in statistical learning approaches to dialogue management, context-sensitive interpretation, and contextsensitive speech recognition. In particular, large dialogue corpora annotated with contextual information and speech acts are urgently required. We explore how existing dialogue corpora (usually consisting of utterance transcriptions) can be automatically processed to yield new corpora where dialogue context and speech acts are accurately represented. We present a conceptual and computational framework for generating such corpora. As an example, we present and evaluate an automatic annotation system which builds 'Information State Update' (ISU) representations of dialogue context for the Communicator (2000 and 2001) corpora of human-machine dialogues (2,331 dialogues). The purposes of this annotation are to generate corpora for reinforcement learning of dialogue policies, for building user simulations, for evaluating different dialogue strategies against a baseline, and for training models for contextdependent interpretation and speech recognition. The automatic annotation system parses system and user utterances into speech acts and builds up sequences of dialogue context representations using an ISU dialogue manager. We present the architecture of the automatic annotation system and a detailed example to illustrate how the system components interact to produce the annotations. We also evaluate the annotations, with respect to the task completion metrics of the original corpus and in comparison to hand-annotated data and annotations produced by a baseline automatic system. The automatic annotations perform well and largely outperform the baseline automatic annotations in all measures. The resulting annotated corpus has been used to train high-quality user simulations and to learn successful dialogue strategies. The final corpus will be made publicly available.

show abstract