Les besoins croissants en formation et en entrainement au travail d’équipe ont motivé l’utilisationd’Environnements de réalité Virtuelle Collaboratifs de Formation (EVCF) qui permettent aux utilisateurs de travailler avec des agents autonomes pour réaliser une activité collective. L’idée directrice est que la coordination efficace entre les membres d’une équipe améliore la productivité et réduit les erreurs individuelles et collectives. Cette thèse traite de la mise en place et du maintien de la coordination au sein d’une équipe de travail composée d’agents et d’humains interagissant dans un EVCF.L’objectif de ces recherches est de doter les agents virtuels de comportements conversationnels permettant la coopération entre agents et avec l’utilisateur dans le but de réaliser un but commun.Nous proposons une architecture d’agents Collaboratifs et Conversationnels, dérivée de l’architecture Belief-Desire-Intention (C2-BDI), qui gère uniformément les comportements délibératifs et conversationnels comme deux comportements dirigés vers les buts de l’activité collective. Nous proposons un modèle intégré de la coordination fondé sur l’approche des modèles mentaux partagés, afin d’établir la coordination au sein de l’équipe de travail composée d’humains et d’agents. Nous soutenons que les interactions en langage naturel entre les membres d’une équipe modifient les modèles mentaux individuels et partagés des participants. Enfin, nous décrivons comment les agents mettent en place et maintiennent la coordination au sein de l’équipe par le biais de conversations en langage naturel. Afin d’établir un couplage fort entre la prise de décision et le comportement conversationnel collaboratif d’un agent, nous proposons tout d’abord une approche fondée sur la modélisation sémantique des activités humaines et de l’environnement virtuel via le modèle mascaret puis, dans un second temps, une modélisation du contexte basée sur l’approche Information State. Ces représentations permettent de traiter de manière unifiée les connaissances sémantiques des agents sur l’activité collective et sur l’environnement virtuel ainsi que des informations qu’ils échangent lors de dialogues.Ces informations sont utilisées par les agents pour la génération et la compréhension du langage naturel multipartite. L’approche Information State nous permet de doter les agents C2BDI de capacités communicatives leur permettant de s’engager pro-activement dans des interactions en langue naturelle en vue de coordonner efficacement leur activité avec les autres membres de l’équipe. De plus, nous définissons les protocoles conversationnels collaboratifs favorisant la coordination entre les membres de l’équipe. Enfin, nous proposons dans cette thèse un mécanisme de prise de décision s’inspirant de l’approche BDI qui lie les comportements de délibération et de conversation des agents. Nous avons mis en oeuvre notre architecture dans trois différents scénarios se déroulant dans des EVCF. Nous montrons que les comportements conversationnels collaboratifs multipartites des agents C2BDI facilitent la coordination effective de l’utilisateur avec les autres membres de l’équipe lors de la réalisation d’une tâche partagée. / Growing needs of educational and training requirements motivate the use of collaborative virtual environments for training (CVET) that allows human users to work together with autonomous agents to perform a collective activity. The vision is inspired by the fact that the effective coordination improves productivity, and reduces the individual and team errors. This work addresses the issue of establishing and maintaining the coordination in a mixed human-agent teamwork in the context of CVET. The objective of this research is to provide human-like conversational behavior of the virtual agents in order to cooperate with a user and other agents to achieve shared goals.We propose a belief-desire-intention (BDI) like Collaborative Conversational agent architecture(C2BDI) that treats both deliberative and conversational behaviors uniformly as guided by the goal-directed shared activity. We put forward an integrated model of coordination which is founded on the shared mental model based approaches to establish coordination in a human-agent teamwork. We argue that natural language interaction between team members can affect and modify the individual and shared mental models of the participants. Finally, we describe the cultivation of coordination in a mixed human-agent teamwork through natural language conversation. In order to establish the strong coupling between decision making and the collaborative conversational behavior of the agent, we propose first, the Mascaret based semantic modeling of human activities and the VE, and second, the information state based context model. This representation allows the treatment of semantic knowledge of the collaborative activity and virtual environment, and information exchanged during the dialogue conversation in a unified manner. This knowledge can be used by the agent for multiparty natural language processing (understanding and generation) in the context of the CEVT. To endow the communicative capabilities to C2BDI agent, we put forward the information state based approach for the natural language processing of the utterances. We define collaborative conversation protocols that ensure the coordination between team members. Finally, in this thesis, we propose a decision making mechanism, which is inspired by the BDI based approach and provides the interleaving between deliberation and conversational behavior of the agent. We have applied the proposed architecture to three different scenarios in the CVET. We found that the multiparty collaborative conversational behavior of C2BDI agent is more constructive and facilitates the user to effectively coordinate with other team members to perform a shared task.

Digitala läromedel blir ett vanligare inslag i skolgången då ny teknologi erbjuder tidigare okända pedagogiska möjligheter. Denna uppsats undersöker hur elever som använder ett digitalt läromedel i form av en virtuell lärmiljö för matematiklärande upplever denna lärmiljö. Dessutom undersöks elevernas prestation i lärmiljöns matematiska uppgifter. Skillnader mellan elever i olika årskurser samt elever som är matematiskt låg- eller högpresterande studeras. Matematisk prestation beskrivs utifrån Goods (1981) passivitetsmodell som innebär att lågpresterande elever är mindre risktagande i klassrumsmiljön. Elevernas upplevelse av digitala läromedel studerades i en virtuell lärmiljö bestående av två moduler, en spelmodul och en modul för skriven dialog. Upplevelsen av lärmiljön undersöktes genom att studera hur intressant eleverna tyckte att spelet var, huruvida eleverna tyckte att agenten gjorde att de brydde sig mer när de spelade, samt om de gillade den skrivna dialogen. Tidigare insamlad data från elever i årskurs 6-8 som använt den virtuella lärmiljön undersöktes med kvasiexperimentell metod och analyserades med ANOVA. Analysen påvisade en skillnad mellan låg- och högpresterande elever i hur mycket de tycker att en pedagogisk virtuell agent engagerar dem i en virtuell lärmiljö. Matematiskt högpresterande elever anser att agenten gör dem mer engagerade än vad matematiskt lågpresterande elever anser. Detta kan tyda på att lågpresterande elevers passivitet utöver traditionell klassrumspedagogik också påverkar elevernas upplevelse av digitala läromedel. I vidareutvecklingen av den virtuella lärmiljön och skapandet av andra virtuella lärmiljöer är det viktigt att se till att elever både lär sig och engageras av lärmiljön. Utvecklare bör också ta hänsyn till de skillnader som finns mellan låg- och högpresterande elevers upplevelse av lärmiljön. Detta kan exempelvis ske genom att den virtuella lärmiljön görs anpassningsbar för att passa elever oberoende av prestationsnivå. Detta är en viktig målsättning för att se förbättra lågpresterande elevers möjligheter i skolan, något som virtuella lärmiljöer och digitala läromedel i allmänhet kan utgöra ett kraftfullt medium för.

Cette thèse présente une série de mesures prises pour étudier l’apprentissage de représentations (par exemple, l’apprentissage profond) afin de mettre en place des systèmes de dialogue et des agents de conversation virtuels. La thèse est divisée en deux parties générales. La première partie de la thèse examine l’apprentissage des représentations pour les modèles de dialogue génératifs. Conditionnés sur une séquence de tours à partir d’un dialogue textuel, ces modèles ont la tâche de générer la prochaine réponse appropriée dans le dialogue. Cette partie de la thèse porte sur les modèles séquence-à-séquence, qui est une classe de réseaux de neurones profonds génératifs. Premièrement, nous proposons un modèle d’encodeur-décodeur récurrent hiérarchique ("Hierarchical Recurrent Encoder-Decoder"), qui est une extension du modèle séquence-à-séquence traditionnel incorporant la structure des tours de dialogue. Deuxièmement, nous proposons un modèle de réseau de neurones récurrents multi-résolution ("Multiresolution Recurrent Neural Network"), qui est un modèle empilé séquence-à-séquence avec une représentation stochastique intermédiaire (une "représentation grossière") capturant le contenu sémantique abstrait communiqué entre les locuteurs. Troisièmement, nous proposons le modèle d’encodeur-décodeur récurrent avec variables latentes ("Latent Variable Recurrent Encoder-Decoder"), qui suivent une distribution normale. Les variables latentes sont destinées à la modélisation de l’ambiguïté et l’incertitude qui apparaissent naturellement dans la communication humaine. Les trois modèles sont évalués et comparés sur deux tâches de génération de réponse de dialogue: une tâche de génération de réponses sur la plateforme Twitter et une tâche de génération de réponses de l’assistance technique ("Ubuntu technical response generation task"). La deuxième partie de la thèse étudie l’apprentissage de représentations pour un système de dialogue utilisant l’apprentissage par renforcement dans un contexte réel. Cette partie porte plus particulièrement sur le système "Milabot" construit par l’Institut québécois d’intelligence artificielle (Mila) pour le concours "Amazon Alexa Prize 2017". Le Milabot est un système capable de bavarder avec des humains sur des sujets populaires à la fois par la parole et par le texte. Le système consiste d’un ensemble de modèles de récupération et de génération en langage naturel, comprenant des modèles basés sur des références, des modèles de sac de mots et des variantes des modèles décrits ci-dessus. Cette partie de la thèse se concentre sur la tâche de sélection de réponse. À partir d’une séquence de tours de dialogues et d’un ensemble des réponses possibles, le système doit sélectionner une réponse appropriée à fournir à l’utilisateur. Une approche d’apprentissage par renforcement basée sur un modèle appelée "Bottleneck Simulator" est proposée pour sélectionner le candidat approprié pour la réponse. Le "Bottleneck Simulator" apprend un modèle approximatif de l’environnement en se basant sur les trajectoires de dialogue observées et le "crowdsourcing", tout en utilisant un état abstrait représentant la sémantique du discours. Le modèle d’environnement est ensuite utilisé pour apprendre une stratégie d’apprentissage du renforcement par le biais de simulations. La stratégie apprise a été évaluée et comparée à des approches concurrentes via des tests A / B avec des utilisateurs réel, où elle démontre d’excellente performance. / This thesis presents a series of steps taken towards investigating representation learning (e.g. deep learning) for building dialogue systems and conversational agents. The thesis is split into two general parts. The first part of the thesis investigates representation learning for generative dialogue models. Conditioned on a sequence of turns from a text-based dialogue, these models are tasked with generating the next, appropriate response in the dialogue. This part of the thesis focuses on sequence-to-sequence models, a class of generative deep neural networks. First, we propose the Hierarchical Recurrent Encoder-Decoder model, which is an extension of the vanilla sequence-to sequence model incorporating the turn-taking structure of dialogues. Second, we propose the Multiresolution Recurrent Neural Network model, which is a stacked sequence-to-sequence model with an intermediate, stochastic representation (a "coarse representation") capturing the abstract semantic content communicated between the dialogue speakers. Third, we propose the Latent Variable Recurrent Encoder-Decoder model, which is a variant of the Hierarchical Recurrent Encoder-Decoder model with latent, stochastic normally-distributed variables. The latent, stochastic variables are intended for modelling the ambiguity and uncertainty occurring naturally in human language communication. The three models are evaluated and compared on two dialogue response generation tasks: a Twitter response generation task and the Ubuntu technical response generation task. The second part of the thesis investigates representation learning for a real-world reinforcement learning dialogue system. Specifically, this part focuses on the Milabot system built by the Quebec Artificial Intelligence Institute (Mila) for the Amazon Alexa Prize 2017 competition. Milabot is a system capable of conversing with humans on popular small talk topics through both speech and text. The system consists of an ensemble of natural language retrieval and generation models, including template-based models, bag-of-words models, and variants of the models discussed in the first part of the thesis. This part of the thesis focuses on the response selection task. Given a sequence of turns from a dialogue and a set of candidate responses, the system must select an appropriate response to give the user. A model-based reinforcement learning approach, called the Bottleneck Simulator, is proposed for selecting the appropriate candidate response. The Bottleneck Simulator learns an approximate model of the environment based on observed dialogue trajectories and human crowdsourcing, while utilizing an abstract (bottleneck) state representing high-level discourse semantics. The learned environment model is then employed to learn a reinforcement learning policy through rollout simulations. The learned policy has been evaluated and compared to competing approaches through A/B testing with real-world users, where it was found to yield excellent performance.

Northvolt is a Swedish battery manufacturing company that specializes in the production of sustainable lithium-ion batteries for electric vehicles and energy storage systems. Established in 2016, the company has experienced significant growth in recent years. This growth has presented a major challenge for the IT Support team, as they face a substantial volume of ITrelated inquiries. To address this challenge and allow the IT Support team to concentrate on more complex support tasks, a question-answering chatbot has been implemented as part of this thesis project. The chatbot has been developed using the Microsoft Bot Framework and leverages Microsoft cloud services, specifically Azure Cognitive Services, to provide intelligent and cognitive capabilities for answering employee questions directly within Microsoft Teams. The chatbot has undergone testing by a diverse group of employees from various teams within the organization and was evaluated based on three key metrics: effectiveness (including accuracy, precision, and intent recognition rate), efficiency (including response time and scalability), and satisfaction. The test results indicate that the accuracy, precision, and intent recognition rate fall below the required thresholds for production readiness. However, these metrics can be improved by expanding the knowledge base of the bot. The chatbot demonstrates impressive efficiency in terms of response time and scalability, and its user-friendly nature contributes to a positive user experience. Users express high levels of satisfaction with their interactions with the bot, and the majority would recommend it to their colleagues, recognizing it as a valuable service solution that will benefit all employees at Northvolt in the future. Moving forward, the primary focus should be on expanding the knowledge base and effectively communicating the bot’s purpose and scope to enhance effectiveness and satisfaction. Additionally, integrating the bot with advanced AI features, such as OpenAI’s language models available within Microsoft’s ecosystem, would elevate the bot to the next level.

Intelligent conversational agents (ICAs) are artificial intelligence (AI)-enabled systems that can communicate with humans through text or voice using natural language. The first ICA, “Eliza,” appeared in 1966 to simulate human conversation using pattern matching. Commercial ICAs appeared on the AOL and MSN platforms in 2001 and aided in developing advanced AI and Human-Computer Interaction (HCI). Since then, ICAs have progressively appeared in consumer products and services. Their success depends on the user’s experience and attitude towards these services. This research examines customer attitudes towards ICAs through a theoretical framework of integrated Expectation Confirmation Theory (ECT) and Task Technology Fit Theory (TTF). By exploring user experience via an experiment that engages end-users with ICA’s different functions and tasks, this study examines user perception of ICA’s AI capabilities, such as Conversation Ability, Friendliness, Intelligence, Responsiveness, Task Performance, and Trust. This research investigates how customer satisfaction with ICA capabilities and perceived task technology fit influence their intention to use ICAs. A field survey of 380 Canadian end-users utilizing ICAs on the websites of five large Canadian telecom service providers enabled empirical testing of the model. / Thesis / Doctor of Philosophy (PhD)

[pt] Assistentes de Voz (AVs) trazem diversos benefícios para os usuários e estão se tornando progressivamente populares, mas algumas barreiras para adoção de AVs ainda persistem, como atitudes dos usuários, preocupações com privacidade e percepções negativas desses sistemas. Uma abordagem para mitigar os obstáculos e melhorar as interações pode ser entender os modelos mentais dos usuários de AVs, uma vez que estudos indicam que o entendimento dos usuários não é alinhado com as reais capacidades desses sistemas. Assim, considerando a importância de um modelo mental correto para as interações, explorar fatores geradores de percepções inadequadas e soluções para lidar com tal questão pode ser essencial. O objetivo desta pesquisa foi identificar fatores influentes para as percepções inadequadas de usuários e oferecer recomendações de design para alinhar os modelos mentais de usuários com as reais capacidades desses sistemas. Para alcançar esse objetivo, nós conduzimos uma revisão sistemática de literatura, entrevistas exploratórias com experts e um estudo Delphi de três rodadas com base em questionários. Os resultados indicam que os aspectos de design como a humanização dos AVs e a transparência em respostas do sistema são influentes para os modelos mentais. Apesar desses fatores terem sido indicados como causas para incorreções em modelos mentais, remover a humanização dos AVs e apresentar informações excessivas pode não ser uma solução imediada. Indica-se que designers devem avaliar o contexto de uso e os domínios de tarefa em que os AVs serão usados para guiar as soluções de design. Além disso, os designers devem entender os perfis e backgrounds dos usuários para ajustar as interações uma vez que as características dos usuários são influentes para sua percepção do produto. Finalmente, o time de desenvolvimento deve ter um entendimento correto e homogêneo do AVs, e deve possuir o conhecimento necessário para aplicar soluções corretamente. Esse último requisito é desafiador porque os AVs são produtos relativamente novos e podem demandar que os profissionais dominem novas habilidades e ferramentas. / [en] Voice Assistants (VAs) bring various benefits for users and are increasingly popular, but some barriers for VA adoption and usage still prevail, such as users attitudes, privacy concerns, and negative perceptions towards these systems. An approach to mitigating such obstacles and leveraging voice interactions may be understanding users mental models of VAs, since studies indicate that users understandings of VAs are unaligned with these systems actual capabilities. Thus, considering the importance of a correct mental model for interactions, exploring influential factors causing misperceptions and solutions to deal with this issue may be paramount. The objective of this research was to identify leading causes of users misperceptions and offer design recommendations for aligning users mental models of VAs with these systems real capacities. In order to achieve this goal, we conducted a systematic literature review (SLR), exploratory interviews with experts, and a questionnaire-based three-round Delphi study. The results indicate that design aspects such as VAs high humanness and the lack of outputs transparency are influential for mental models. Despite the indication that these drivers lead to users misperceptions, removing VAs humanness and excessively displaying information about VAs might not be an immediate solution. In turn, developers should assess the context and task domains in which the VA will be used to guide design decisions. Moreover, developers should understand the users profiles and backgrounds to adjust interactions, as users characteristics are influential for how they perceive the product. Finally, developing teams should have a correct and homogeneous understanding of VAs and possess the necessary knowledge to employ solutions properly. This latter requirement is challenging since VAs novelty might demand professionals to master new skills and tools.

Digitalization crosses all areas of life (Hess et al. 2014). Recent progress in artificial intelligence (AI) opens new potentials for further developments and improvements, with virtual coaching being a prime example. Virtual coaches (VCs) aim to optimize the user’s life by transforming cognition, affection, and behavior towards a stated goal. Since they emerged from the health and sports domain, a typical example are VCs in the form of digital avatars, which instruct physical exercises, shape health-related knowledge and provide motivational support to achieve the user’s goals (e.g., weight loss) (Ding et al. 2010; Tropea et al. 2019). Nonetheless, the application areas of VCs are versatile and exploring the potential areas (e.g., healthcare, work, finance, leisure, and environment) constitutes an essential topic of future research and development. According to Gartner’s hype cycle for human capital management technology, VCs are still in their infancy but are considered innovation triggers for the following years (Gartner, Inc. 2021). Specifically, VCs can be a replacement or complement for traditional human-to-human coaching scenarios and promise broad access to personalized coaching services independent of place and time (Graßmann and Schermuly 2021). As a result, VCs may contribute to solving challenges posed by an aging society and skilled labor shortage (European Commission 2016; Edwards and Cheok 2018). Last but not least, the recent COVID-19 pandemic additionally showcased the need for VCs as an alternative to traditional face-to-face interventions. Against this background and driven by the potential and promises of VCs, research has recently engaged in developing and understanding VC applications (Tropea et al. 2019; Lete et al. 2020; Graßmann and Schermuly 2021). To introduce the concept in information systems (IS) research and provide a basis for researchers and practitioners alike, this catchword aims at providing a holistic view on VCs. The structure of this paper is as follows. Section 2 elaborates a definition, delimits VCs from related system classes, and proposes a research framework. Section 3 aggregates existing research into the framework and concludes with an outlook on future IS research perspectives.

Turn-taking is a fundamental process that governs social interaction. When humans interact, they naturally take initiative and relinquish control to each other using verbal and nonverbal behavior in a coordinated manner. In contrast, existing approaches for controlling a robot's social behavior do not explicitly model turn-taking, resulting in interaction breakdowns that confuse or frustrate the human and detract from the dyad's cooperative goals. They also lack generality, relying on scripted behavior control that must be designed for each new domain. This thesis seeks to enable robots to cooperate fluently with humans by automatically controlling the timing of multimodal turn-taking. Based on our empirical studies of interaction phenomena, we develop a computational turn-taking model that accounts for multimodal information flow and resource usage in interaction. This model is implemented within a novel behavior generation architecture called CADENCE, the Control Architecture for the Dynamics of Embodied Natural Coordination and Engagement, that controls a robot's speech, gesture, gaze, and manipulation. CADENCE controls turn-taking using a timed Petri net (TPN) representation that integrates resource exchange, interruptible modality execution, and modeling of the human user. We demonstrate progressive developments of CADENCE through multiple domains of autonomous interaction encompassing situated dialogue and collaborative manipulation. We also iteratively evaluate improvements in the system using quantitative metrics of task success, fluency, and balance of control.