Les Agents Conversationnels Animés sont des personnages virtuels dont la fonction principale est d'interagir avec l'utilisateur. Ils sont utilisés dans différents domaines tels que l'assistance personnelle, l'entrainement social ou les jeux vidéo et afin d'améliorer leur potentiel, il est possible de leur donner la capacité d'exprimer des comportements similaires à ceux des humains. Les utilisateurs, conscient d'interagir avec une machine, sont tout de même capable d'analyser et d'identifier des comportements sociaux à travers les signaux émis par les agents. La recherche en ACA s'est longtemps intéressée aux mécanismes de reproduction et de reconnaissance des émotions au sein de ces personnages virtuels et maintenant l'intérêt se porte sur la capacité d'exprimer différentes attitudes sociales. Ces attitudes reflètent un style comportemental et s'expriment à travers différentes modalités du corps comme les expressions faciales, les regards ou les gestes par exemple. Nous avons proposé un modèle permettant à un agent de produire différents comportements non-verbaux traduisant l'expression d'attitudes sociales dans une conversation. L'ensemble des comportements générés par notre modèle permettent à un groupe d'agents animés par celui-ci de simuler une conversation, sans tenir compte du contenu verbal. Deux évaluations du modèle ont été conduites, l'une sur Internet et l'autre dans un environnement de réalité virtuelle, afin de vérifier que les attitudes étaient bien reconnues. / Embodied Conversational Agents are virtual characters which main purpose is to interact with a human user. They are used in various domains such as personal assistance, social training or video games for instance. In order to improve their capabilities, it is possible to give them the ability to produce human-like behaviors. The users, even if they are aware that they interact with a machine, are still capable of analyzing and identifying social behaviors through the signals produced by these virtual characters. The research in Embodied Conversational Agents has focused for a long time on the reproduction and recognition of emotions by virtual characters and now the focus is on the ability to express different social attitudes. These attitudes show a behavioral style and are expressed through different modalities of the body, like the facial expressions, the gestures or the gazes for instance. We proposed a model that allows an agent to produce different nonverbal behaviors expressing different social attitudes in a conversation. The whole set of behaviors produced by our model allows a goup of agents animated by it to simulate a conversation, without any verbal content. Two evaluations of the model were conducted, one on the Internet and one in a Virtual Reality environment, to verify that the attitudes produced are well recognized

Conversational agents are becoming increasingly common in our day to day lives. We can speak to our phones, our cars and our smart home devices. Despite these advances, the current conversational agents are still far from perfect. The complexities of language as well as the technologies that are used to enable conversational agents pose many challenges to the people designing and developing them. This study aims to bring light to how practitioners design and develop conversational agents that exist out there today. Interviews were conducted with 11 practitioners that have been creating conversational agents for various industries and use cases. The results show that practitioners face a range of challenges when creating conversational agents, including collecting data about the target users during the design process as well as integrating the conversational agent with other systems. The study suggests that practitioners may benefit from involving users early on in the design process. It also advocates for HCI educators to prepare future graduates for designing conversational agents, through training them in human conversation and communication. This study gives insights into how conversational agents are built today, the processes that are followed and the challenges that are faced by the people creating them. / Konverserande system blir allt vanligare i våra dagliga liv. Vi kan prata med våra telefoner, våra bilar och våra smarta hem enheter. Trots dessa framsteg är de nuvarande konverserande systemen fortfarande långt ifrån perfekta. Språkets komplexitet och de teknologier som används, ställer många utmaningar för de människor som designar och utvecklar dem. Denna studie syftar till att belysa hur yrkesarbetare designar och utvecklar konverserande system. Intervjuer genomfördes med 11 yrkesarbetare som har skapat konverserande system för olika branscher och användningsområden. Resultaten visar att yrkesarbetarna står inför en rad utmaningar när de skapar dessa system, bland annat att samla in data om användarna under designprocessen och att integrera konverserande system med andra system. Studien tyder på att utövare kan ha nytta av att involvera användare tidigt i designprocessen. Studien förespråkar också att utbildare inom människa-datorinteraktion bör förbereda studenter för att utforma konverserande system, genom att utbilda dem i mänsklig konversation och kommunikation. Denna studie ger insikter i hur konverserande system är uppbyggda idag, de processer som följs och de utmaningar som människorna som skapar dem står inför.

This thesis explores potential strategies to enhance user control and satisfaction within the movie selection process, with a particular focus on the utilization of conversational generative artificial intelligence, such as ChatGPT, for personalized movie recommendations. The study adopts a qualitative user-centered design thinking approach, aiming to compre-hensively understand user needs, goals, and behavior. In-depth interviews were conducted, utilizing the "Thinking aloud"method and trigger materials to elicit rich user feedback. Participants interacted with ChatGPT and various prototypes, providing valuable insights into their experiences. The study found that participants felt more in control when given the opportunity to specify wishes. In addition, the users found that the experience of receiving recommendations through ChatGPT was more satisfying than their usual way of receiving recommendations for movies. Furthermore, participants expressed a desire for additional information about recommended movies and more novel suggestions. The prototypes, designed as triggers for user feedback, were generally well-received, providing an engaging and fun user experience. Despite some participants expressing challenges in specifying movie choices based on an emotion, this new approach to movie selection was viewed positively. Despite limitations concerning the study’s validity, reliability, and testing situation, the findings suggest the potential of generative artificial intelligence conversational agents in enhancing the movie selection process. It is concluded that iterative design improvements and further research is necessary to fully leverage the potential of natural language processing technologies in recommendation systems. The study serves as a preliminary investigation into improving movie recommendations using generative artificial intelligence and offers valuable insights for future developments.

Teste de software é uma atividade do processo de desenvolvimento de software que busca revelar os defeitos presentes no software. Apesar dos constantes avanços e surgimento de diferentes modelos pedagógicos, este assunto continua sendo ensinado com o modelo tradicional de ensino. A comunidade de educação em computação vem investigando a aplicação de diferentes modelos pedagógicos e, nessa perspectiva, emerge o modelo pedagógico flipped classroom como uma interessante solução. Estudos mencionam sobre a necessidade de investigá-lo, buscando reconhecer seu potencial no ensino de um determinado conteúdo. Além disso, implementar o modelo exige uma intensificação de esforços para o professor, pois ele precisará planejar atividades e mecanismos de apoio ao ensino para cada um dos momentos previstos pelo modelo. Trabalhos relatam que os alunos durante o momento de estudo antes da aula podem ter dificuldade em entender aspectos teóricos sobre o conteúdo, pois esse pode ter conceitos que podem gerar dupla interpretação. Este trabalho de mestrado tem o propósito de contribuir com o ensino de teste de software, investigando o impacto do modelo flipped classroom no ensino de teste de software. O projeto também propõe uma abordagem de apoio ao planejamento e implantação desse modelo e analisa o uso de um agente conversacional como mecanismo de apoio ao flipped classroom. Estudos experimentais foram realizados visando entender o impacto do modelo no processo de aprendizado de teste de software. Na sequência, foram analisados conteúdos de teste de software ensinados em cursos de computação no Brasil, definidos objetivos de aprendizagem para esses conteúdos, selecionados mecanismos de apoio e atividades para contemplar os três momentos previstos pelo modelo e implementado um mecanismo para apoiar a implantação do modelo flipped classroom no ensino de teste. Por fim, um agente conversacional foi proposto e investigado no contexto de uma aula de teste de software em que o modelo flipped classroom foi utilizado. Os resultados deste trabalho indicam que o modelo flipped classroom permite que os alunos tenham melhores resultados no aprendizado, especialmente quando há realização de atividades práticas. Os alunos revelaram que preferem este modelo em detrimento ao tradicional. No entanto, o modelo produz uma carga de trabalho maior para o aluno. Por fim, o uso do agente conversacional como um mecanismo de apoio durante o momento antes da aula indica sua relevância para o flipped classroom. / Software testing is an activity of the development process that aims to reveal the faults present in the software. Despite of constant progress and emergence of different pedagogical models, this subject is still taught with strategies based mainly on the traditional model. Although software testing is considered a relevant subject, some students are not very motivated to learn it. The computer education community has investigated the use of different teaching models and, from this perspective, the Flipped Classroom pedagogical model has emerged as an interesting solution. Several studies mention that is required to investigate this model in order to identify its potential in teaching a specific subject and/or teaching area. In addition, the implementation of this model demands an intensification of efforts from the teacher because he/she needs to plan activities and support mechanisms for all phases of this model: before the class, during the class, and after the class. Results from the literature indicate that students may present difficulties in understand the theory during the phase before the classbecause studying alone can lead to many doubts and misunderstandings. Therefore, this masters thesis investigates the impact of the Flipped Classroom model on teaching of testing techniques and criteria. Experimental studies were initially undertaken aiming to understand the models impact in the learning process of software testing. Following, we analyzed what subjects of software testing are taught in Computer courses in Brazil. Based on the results, we defined learning objectives for these subjects, selected support mechanisms and activities to contemplate the three phases of the Flipped Classroom model. An approach, named FlipClass-STT, to support the implementation of this model was defined and experimentally evaluated. Finally, the conversational agent TOB-STT was proposed and evaluated in context of a software testing class, in which the Flipped Classroom model was employed. The results of this work indicate that Flipped Classroom allows the students to have better learning outcomes, especially when performing practical activities. The students have revealed that they prefer this model against the traditional one. However, it is also observed that model produces a greater workload for the student. The use of TOB-STT how a support mechanism during the phase before the class indicate its relevance and impact for learning software testing.

The technological progress has enabled a frequent use of videos by educators in the teaching-learning process. In order to conduct this work, a study on digital video (characteristics, formats, patterns, video conversions) was made, as well as a collection of data from different repositories of digital videos available. The analysis allowed the definition of improvements to be implemented in ReviMobile, and a study on conversational agents and m-learning was also conducted. This work consists in developing a repository of videos that adapts to the technological resource, that is, to the model of mobile device being used, performing all the necessary conversions (for example, from mp4 to wmv) in a transparent and automated way. This turns out to be a viable solution to the problems of execution of videos due to different video formats and languages of development, using the process of finding a conversational agent as information recommender. Therefore, the results benefit mainly teachers and students in the process of searching and viewing educational videos; that is, showing videos in the format that is compatible with the mobile device being used in a transparent and dynamic way. This work differs in relation to other video repositories in three factors: providing mobility, adaptation to different video formats of mobile devices and conversational agent as recommender of information. / O avanço tecnológico tem proporcionado utilização frequente de vídeos pelos educadores no processo de ensino-aprendizagem. Para viabilizar o presente trabalho, foi realizado um estudo sobre vídeos digitais (características, formatos, padrões, conversões de vídeo), bem como um levantamento de dados dos diferentes repositórios de vídeos digitais existentes. A análise permitiu definir melhorias a serem aplicadas no ReviMobile, sendo também realizado um estudo sobre agentes conversacionais e m-learning. Este trabalho consiste no desenvolvimento de um repositório de vídeos, que se adapta ao recurso tecnológico, isto é, ao modelo de dispositivo móvel utilizado pelo usuário realizando todas as conversões necessárias, (por exemplo, de wmv para mp4) de forma transparente e automatizada. Apresenta-se como uma solução viável aos problemas de execução de vídeos devido aos diferentes formatos e linguagens de desenvolvimento, utilizando-se no processo de busca de um agente conversacional como recomendador de informação. Logo, os resultados inferidos beneficiam principalmente professores e alunos no processo de busca e visualização de vídeos educacionais, ou seja, apresentando vídeos no formato compatível aos dispositivos móveis de forma transparente e dinâmica. Este trabalho diferencia-se em relação aos demais repositórios de vídeo por três fatores: propiciar mobilidade, adaptação aos diferentes formatos de vídeos dos dispositivos móveis e agente conversacional como recomendador de informações.

« Es-tu d’accord ? » En contexte numérique, cette question appelle un oui ou un non au traitement annoncé de nos données personnelles par une organisation; elle demande ainsi à l’utilisateur son consentement. Les agents conversationnels sont des systèmes de dialogue (ou bots) qui utilisent de nombreuses données personnelles pour parler avec des humains par interface textuelle (chatbot) ou vocale (assistant vocal, parfois appelé voicebot). Par une ethnographie au sein d’une start-up qui développe ce type d’agents, cette thèse explore comment le consentement se tisse dans les pratiques de design de cette organisation. Cette thèse se demande : que peut être une « bonne expérience » de consentement avec un agent conversationnel ? Elle s’ouvre par un chemin tracé au travers de littératures en design d’expérience conversationnelle et en éthique sur le consentement. Tandis qu’une bonne expérience conversationnelle est censée être fluide et agréable, celle du consentement demande des interruptions parfois inconfortables pour être suffisamment éclairé : comment ces deux versions de « bonne expérience » peuvent-elles être réunies sans que l’une ne blesse ou n’invalide l’autre ? Face aux nombreuses limites qui rendent le consentement numérique bien difficile à être valable, la tentation de le laisser de côté est palpable. Pourtant, son horizon éthique de respect vaut la peine d'en prendre soin. Par une perspective qui envisage le consentement comme un matter of caring, cette thèse développe ensuite un cadre théorique qui invite à prendre soin des situations de consentement. Il dote l'approche ventriloque, qui étudie la constitution de la réalité par la communication, d’une éthique du care : avec le concept de matter of caring, il s’agit de prendre soin de certaines préoccupations (matters of concern) pour améliorer des situations de consentement. En se faisant matter of caring, le consentement trouve une façon de n'être plus négligé comme une préoccupation qui ne parvient pas à compter, mais de participer à un changement dans les habitudes d'interaction. Cette participation au changement est également au cœur des pratiques ethnographiques qui constituent cette thèse : à partir de mon expérience comme participante active plutôt que de simple observatrice dans la startup mentionnée plus haut, j’invite à comprendre la pratique réflexive ethnographique par une approche relationnelle où l’ethnographe peut être activement engagée dans la constitution de l’organisation qu’elle étudie. C’est par cet engagement que s’est constitué le matériel de terrain que j’analyse tout au long de la thèse. Plus précisément, le dernier chapitre plonge dans mon terrain et ses tensions pour concevoir une bonne expérience de conversation et de consentement. En résistant à la tentation de laisser le consentement perdre de son importance, je montre que s’y accrocher comme un matter of caring amène à se concentrer sur les conditions dans lesquelles l’organisation demande le consentement : comment prévoit-on de traiter les informations personnelles et comment conçoit-on l’interaction par laquelle le bot demandera leur consentement à ses utilisateurs. Cette histoire n’est pas couronnée de succès mais dévoile plutôt une certaine vulnérabilité. Ainsi, cette thèse ne propose pas un modèle de consentement, ni des directives de design éthique d'un agent conversationnel. Elle s'attarde plutôt sur l'importance des conditions à permettre pour faire de la place aux interactions de consentement. Plutôt que de figer le consentement dans un état contraint de formalité, elle invite à penser le consentement comme une conversation, avec du respect et de l'épanouissement à l’horizon. / “Do you agree?” In the digital realm, this question calls for a yes or no answer to the processing of our personal data by an organization; that is, it asks for the user’s consent. Conversational agents are dialogue systems (or bots) that use many personal data to talk with humans through a textual interface (chatbot) or voice interface (voice assistant, sometimes called “voicebot”). Through an ethnography within a start-up that develops these kinds of agents, this doctoral dissertation explores the weaving of consent into this organization’s design practices. Thus, this dissertation asks: What does a “good experience” of consent with a conversational agent look like? It starts by reviewing the literature on design of conversational experiences and in the ethics of consent. While a good conversational experience is supposed to be smooth and enjoyable, the consent experience requires some interruptions to be sufficiently informed, which can be uncomfortable. So how can these two versions of a “good experience” be brought together without one hurting or invalidating the other? Faced with the many limitations for meaningful consent in a digital context, the temptation to put consent aside is palpable, yet its ethical horizon of respect is worth caring for. Then, with a perspective that views consent in terms of a matter of caring, this dissertation develops a theoretical framework that enables us to explore how people can care for consent situations. Centered on the concept of matter of caring, this framework enriches a ventriloquial approach to the study of the communicative constitution of reality with an ethics of care; it focuses attention on how certain matters of concern can be cared for in order to improve consent situations. By becoming a matter of caring, consent can no longer be neglected as a concern that fails to count; it rather participates in changing how human beings interact with each other. Participating in the bringing about of this change is at the core of the ethnographic methods that constitute this dissertation: Based on my experience with becoming an active participant in the mentioned start-up, rather than a mere participant observant, I explain how ethnographic reflexive practice can be viewed relationally; that is, how an ethnographer can be actively engaged in the constitution of the organization she is studying. This engagement shaped the fieldwork materials I analyze in-depth throughout the dissertation. More specifically, the last chapter of the dissertation delves into my fieldwork on the mentioned tension between designing a good conversational and consent experience. Resisting the temptation to let consent fade away, I show that holding on to it as a matter of caring makes us focus on the conditions under which the organization I studied asks for consent: how it plans to process personal information and how it designs the interaction through which the bot asks for consent from its users. This is not a success story, but rather a story of vulnerability. Thus, this dissertation does not propose a model for consent, nor does it suggest ethical design guidelines for a conversational agent. Instead, it highlights the importance of providing conditions to enable interactions in which consent becomes a matter of caring. Rather than freezing consent in a constrained state of formality, it invites us to think of consent as a conversation, with respect and flourishing on the horizon.

No description available.

Thesis (MSc)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: In this thesis we describe the creation of an embodied conversational agent which exhibits the behavioural traits of a child who has Asperger Syndrome. The agent is rule-based, rather than arti cially intelligent, for which we give justi cation. We then describe the design and implementation of the agent, and pay particular attention to the interaction between emotion, personality and social context. A 3D demonstration program shows the typical output to conform to Asperger-like answers, with corresponding emotional responses. / AFRIKAANSE OPSOMMING: In hierdie tesis beskryf ons die ontwerp en implementasie van 'n gestaltegespreksagent wat die gedrag van 'n kind met Asperger se sindroom uitbeeld. Ons regverdig die besluit dat die agent reël-gebaseerd is, eerder as 'n ware skynintelligensie implementasie. Volgende beskryf ons die wisselwerking tussen emosies, persoonlikheid en sosiale konteks en hoe dit inskakel by die ontwerp en implementasie van die agent. 'n 3D demonstrasieprogram toon tipiese ooreenstemmende Asperger-agtige antwoorde op vrae, met gepaardgaande emosionele reaksies.

The use of Conversational Agents (CAs) utilizing Artificial Intelligence Markup Language (AIML) has been studied in a number of disciplines. Previous research has shown a great deal of promise. It has also documented significant limitations in the abilities of these CAs. Many of these limitations are related specifically to the method employed by AIML to resolve ambiguities in the meaning and context of words. While methods exist to detect and correct common errors in spelling and grammar of sentences and queries submitted by a user, one class of input error that is particularly difficult to detect and correct is the malapropism. In this research a malapropism is defined a "verbal blunder in which one word is replaced by another similar in sound but different in meaning" ("malapropism," 2013). This research explored the use of alternative methods of correcting malapropisms in sentences input to AIML CAs using measures of Semantic Distance and tri-gram probabilities. Results of these alternate methods were compared against AIML CAs using only the Symbolic Reductions built into AIML. This research found that the use of the two methodologies studied here did indeed lead to a small, but measurable improvement in the performance of the CA in terms of the appropriateness of its responses as classified by human judges. However, it was also noted that in a large number of cases, the CA simply ignored the existence of a malapropism altogether in formulating its responses. In most of these cases, the interpretation and response to the user's input was of such a general nature that one might question the overall efficacy of the AIML engine. The answer to this question is a matter for further study.

Submitted by Silvana Teresinha Dornelles Studzinski (sstudzinski) on 2016-05-09T13:07:53Z No. of bitstreams: 1 Fabio Rodrigues dos Santos_.pdf: 7109958 bytes, checksum: 136a6e535dbd26994aa6dbd702767b78 (MD5) / Made available in DSpace on 2016-05-09T13:07:53Z (GMT). No. of bitstreams: 1 Fabio Rodrigues dos Santos_.pdf: 7109958 bytes, checksum: 136a6e535dbd26994aa6dbd702767b78 (MD5) Previous issue date: 2016-02-26 / IFRR - Instituto Federal de Educação Ciências e Tecnologia de Roraima / Um agente conversacional é capaz de interagir com usuários em linguagem natural, imitando o diálogo realizado entre seres humanos. Entretanto, a necessidade de trabalho manual por parte dos autores de conteúdo para a construção da sua base de diálogos não o torna muito atrativo para ser integrado a websites. Seria possível facilitar a criação de novos diálogos e automatizar a atualização dos mesmos usando como base os recursos disponibilizados pela Web Semântica para integrar o agente conversacional a um Sistema de Gerenciamento de Conteúdo Web (SGCW) de forma que seu conteúdo seja aproveitado. O trabalho aqui proposto descreve o modelo para utilização de informações de SGCW’s em agentes conversacionais com auxilio da web semântica, denominado Eduardo. O principal diferencial está na automatização da atualização dos diálogos, pois os mesmos são extraídos do SGCW para uma base de dados de triplas, que são consultadas para geração dos diálogos, caso haja alguma atualização na base do SGCW, basta refazer os procedimentos que já foram automatizados, para que os diálogos sejam atualizados, evitando assim o trabalho manual de edição dos arquivos escritos em linguagem AIML (Artificial Intelligence Markup Language). Foi desenvolvido um protótipo deste modelo, que permite a integração dinâmica de conteúdo com um Agente Conversacional. O protótipo foi avaliado quanto à capacidade de gerar as informações necessárias para os diálogos, acessando as páginas Web e representando seu conteúdo nos formatos RDFa e AIML. Em um segundo momento, ensaios de interação foram realizados com usuários nos componentes do Agente Conversacional, para avaliar sua funcionalidade, aceitação e demais aspectos uso. / A conversational agent is able to interact with users in natural language, imitating the dialogue held between humans. However, the need for manual labor by the authors content for the construction of dialogues base makes the not too attractive to be integrated into websites. It would be possible to facility the creation of new dialogues and automate the update them using as a basis the resources provided by the Semantic Web to integrate conversational agent to a Web Content Management System (WCMS) so that you content is used. The work proposed here describes the model for use of WCMS’s information in conversational agents with semantic web aid, called Eduardo. The main difference is in automating the update of dialogues, as they are extracted from WCMS for a triple database, which are referred to generation of dialogues, if there is any update on the basis WCMS, just redo the procedures that have been automated so that the dialogues are update, thus avoiding manual labor editing of files written in AIML (Artificial Intelligence Markup Language). It developed a prototype of this model which allows dynamic content integration with a conversational agent. The prototype was evaluated for ability to generate the information necessary for the dialogue, accessing the Web pages and their content representing in RDFa and AIML formats. In a second step, interaction assays were performed with users in the components of conversational agent to assess its functionality, acceptance and other aspects of use.