The requirements for virtual agents are more and more demanding. In order to manage the complex behavior of the agent, it's possible to take advantage of cognitive architectures which arised on the field neuroscience and artificial intelligence. This work examines PoJACTR library which links Pogamut library for developing intelligent agents in Unreal Tournament 2004 and jACT-R library which is Java implementation of one of the leading cognitive architectures ACTR. This work also studies certain agent implementation problems in PoJACTR and proposes a solution for them in form of debugging tools, which were subsequently implemented on an Eclipse IDE platform. In addition, it expands PoJACTR navigation and communication library modules for the game - Capture The Flag. As a validation, two agents (bots) were developed to play game, one in standard Pogamut and one in PoJACTR. When matched against each other in battle, PoJACTR bot had comparable performance to a Pogamut bot. The results showed that debugging tools facilitated development process of PoJACTR agents. Powered by TCPDF (www.tcpdf.org)

Este trabalho apresenta um framework que permite a programação de agentes capazes de perceberem seus próprios estados emocionais. O framework foi desenvolvido em Java com base na plataforma multi-agente Jason, estendendo a base de crenças de agentes Jason a fim de utilizar a ontologia afetiva desenvolvida. Além disso, o ambiente foi construído a partir de uma base de conhecimento que descreve rotinas em ambientes simulados. Um mecanismo de avaliação das emoções baseando-se nas anotações dos objetos foi construído apoiado por uma ontologia de preferência sobre essas anotações. Dessa forma, aplicações de entretenimento poderiam utilizar o sistema ou as bases de conhecimento apresentadas para diferentes propósitos. A criação de um mapa onde os personagens atuam, e a criação da rotina de cada personagem e suas preferências são alguns exemplos de utilizações. Para validação do framework desenvolvido, dois exemplos foram construídos. O primeiro utilizou a maior parte dos grupos afetivos da ontologia proposta, com a finalidade principal de demonstrar o modelo implementado. Já o segundo usa apenas um grupo emotivo e serve para demonstrar a utilização conjunta de todas as ontologias apresentadas. / This work presents a framework built on top of the Jason platform (BORDINI et al., 2004) to allow the development of software agents that have emotional states. The framework was developed in Java and extends the belief base of Jason agents so as to use an ontology for the OCC affective model (ORTONY; COLLINS; CLORE, 1988) that has been created as part of this work. The developed belief base allows an agent to perceive its own emotions throw inferring new beliefs based on the agent’s appraisal of the state of the environment. In addition, a model of agents’ routine tasks was defined, as was a model for agents’ preferences about aspects of environment, helping automate the ascription of emotional states. Finally, in order to validate the developed framework, two applications were developed. The first demonstrates the use of various different emotions from the affective model and the second uses in a single application all the ontologies and models developed as part of this work.

Este trabalho apresenta um framework que permite a programação de agentes capazes de perceberem seus próprios estados emocionais. O framework foi desenvolvido em Java com base na plataforma multi-agente Jason, estendendo a base de crenças de agentes Jason a fim de utilizar a ontologia afetiva desenvolvida. Além disso, o ambiente foi construído a partir de uma base de conhecimento que descreve rotinas em ambientes simulados. Um mecanismo de avaliação das emoções baseando-se nas anotações dos objetos foi construído apoiado por uma ontologia de preferência sobre essas anotações. Dessa forma, aplicações de entretenimento poderiam utilizar o sistema ou as bases de conhecimento apresentadas para diferentes propósitos. A criação de um mapa onde os personagens atuam, e a criação da rotina de cada personagem e suas preferências são alguns exemplos de utilizações. Para validação do framework desenvolvido, dois exemplos foram construídos. O primeiro utilizou a maior parte dos grupos afetivos da ontologia proposta, com a finalidade principal de demonstrar o modelo implementado. Já o segundo usa apenas um grupo emotivo e serve para demonstrar a utilização conjunta de todas as ontologias apresentadas. / This work presents a framework built on top of the Jason platform (BORDINI et al., 2004) to allow the development of software agents that have emotional states. The framework was developed in Java and extends the belief base of Jason agents so as to use an ontology for the OCC affective model (ORTONY; COLLINS; CLORE, 1988) that has been created as part of this work. The developed belief base allows an agent to perceive its own emotions throw inferring new beliefs based on the agent’s appraisal of the state of the environment. In addition, a model of agents’ routine tasks was defined, as was a model for agents’ preferences about aspects of environment, helping automate the ascription of emotional states. Finally, in order to validate the developed framework, two applications were developed. The first demonstrates the use of various different emotions from the affective model and the second uses in a single application all the ontologies and models developed as part of this work.

Abstract Background Since the emergence of digital banking, the financial sector has experienced a significant transformation in both how business is conducted and how services are provided to customers. Previous literature has examined how new technologies and the digitalization of banks' customer service affect customer loyalty. Although, since virtual agents acting as service providers in the banking sector is a relatively new phenomenon, there is limited research concerning the implications it will have on the bank-customer relationship. Hence, the novelty and relevance of the topic makes it interesting for further research.    Purpose Through the identified underlying factors affecting customer loyalty, the purpose of this study is to examine how customer loyalty will be affected by the implementation of virtual agents as service providers in major Swedish banks.   Method This is a qualitative study, and the empirical data were collected from semi-structured in-depth interviews with bankers at four major Swedish banks, as well as with ten highly-educated customers who are frequent users of bank services.   Findings The findings showed that virtual agents must affect customer service to a large extent to have a profound impact on customer loyalty. Virtual agents will be able to replace human bankers regarding simpler inquiries satisfyingly. On the other hand, the demand for personal interactions regarding more complex matters is found to be important.

Este trabalho apresenta um framework que permite a programação de agentes capazes de perceberem seus próprios estados emocionais. O framework foi desenvolvido em Java com base na plataforma multi-agente Jason, estendendo a base de crenças de agentes Jason a fim de utilizar a ontologia afetiva desenvolvida. Além disso, o ambiente foi construído a partir de uma base de conhecimento que descreve rotinas em ambientes simulados. Um mecanismo de avaliação das emoções baseando-se nas anotações dos objetos foi construído apoiado por uma ontologia de preferência sobre essas anotações. Dessa forma, aplicações de entretenimento poderiam utilizar o sistema ou as bases de conhecimento apresentadas para diferentes propósitos. A criação de um mapa onde os personagens atuam, e a criação da rotina de cada personagem e suas preferências são alguns exemplos de utilizações. Para validação do framework desenvolvido, dois exemplos foram construídos. O primeiro utilizou a maior parte dos grupos afetivos da ontologia proposta, com a finalidade principal de demonstrar o modelo implementado. Já o segundo usa apenas um grupo emotivo e serve para demonstrar a utilização conjunta de todas as ontologias apresentadas. / This work presents a framework built on top of the Jason platform (BORDINI et al., 2004) to allow the development of software agents that have emotional states. The framework was developed in Java and extends the belief base of Jason agents so as to use an ontology for the OCC affective model (ORTONY; COLLINS; CLORE, 1988) that has been created as part of this work. The developed belief base allows an agent to perceive its own emotions throw inferring new beliefs based on the agent’s appraisal of the state of the environment. In addition, a model of agents’ routine tasks was defined, as was a model for agents’ preferences about aspects of environment, helping automate the ascription of emotional states. Finally, in order to validate the developed framework, two applications were developed. The first demonstrates the use of various different emotions from the affective model and the second uses in a single application all the ontologies and models developed as part of this work.

La paralysie cérébrale est la pathologie motrice la plus commune dans l’enfance, elle engendre tous types de troubles principalement moteurs. La parole peut notamment être touchée affectant parfois profondément les capacités communicatives des enfants. Les thérapies pour pallier les troubles moteurs de la parole manquent d’efficacité. C’est dans ce cadre que le logiciel RePliCa a été développé afin de rendre leur rééducation plus efficiente. Une variable, encore peu documentée dans le cadre de la rééducation, est identifiée comme influençant grandement les progrès, il s’agitde l’engagement. L’engagement est considéré comme malléable et sensible aux changements environnementaux et pourrait donc être influencé par l’introduction d’un logiciel dans la rééducation. La première partie de cette thèse était consacrée à l’optimisation de cet outil. Dans cet objectif, une série d’études expérimentales a été menée auprès d’enfants au développement typique et a permis l’intégration d’un certain nombre de caractéristiques relatives à l’interface influençant positivement l’engagement. Une grille d’observation a ensuite été créée afin d’évaluer l’engagement des enfants paralysés cérébraux dans la rééducation lors du programme pilote avec le logiciel. L’étude clinique montre des résultats encourageants, l’engagement semblant se maintenir au cours des séances. Ces résultats sont mis en perspectives dans le cadre d’un modèle de l’engagement. / Cerebral Palsy is the most common physical disability in early childhood, resulting in multiple disorders, primarily of motor skills. Speech especially can be impacted and deeply affects children’s communicative abilities. The RePliCa software wasdeveloped to improve the efficiency of motor speech therapy. One variable, engagement, has not been fully researched in the field of rehabilitation but has been identified influential for children’s progress. Engagement is presumed to bemalleable and sensitive to environmental changes and therefore could be influenced by the introduction of software in rehabilitation. The first part of this thesis was devoted to its optimization. To this end, a series of experimental studies wereconducted, enabling the integration of a number of features on the interface in order to promote engagement. Once the software was implemented, engagement in rehabilitation was examined in a clinical study. Results showed encouraging results: engagement can be maintained over time. Results are put in perspective using an engagement model framework

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第19110号 / 情博第556号 / 新制||情||98(附属図書館) / 32061 / 京都大学大学院情報学研究科知能情報学専攻 / (主査)教授 西田 豊明, 教授 乾 敏郎, 教授 河原 達也 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DGAM

Virtual agents have been around since the 1960’s and theirusage is constantly increasing. They are used in manydifferent settings. At the same time, game engines are usedmore and more for various purposes. In this paper, the aim isto create two proof-of-concept virtual agents in the form of ahumanoid receptionist as part of a bigger project. The agentswere implemented using two of the most widely used gameengines, Unity and Unreal Engine. During the development,the different aspects of the engines and their suitability fordeveloping virtual agents were compared. It was found thatUnity was easier to use for someone inexperienced withgame engines, but that it in general is more of a matter ofpersonal preference.

There are dozens of formalisms specialized programming intelligent virtual agents (IVA). However, there is lack of an evaluation, which would provide the designer of a new agent with an overview of their usability in real-world applications. The goal of this study was evaluation of selected formalisms in order to identify their shortcomings and suggesting improvements. These improvements would allow the designer more accurate control of the IVA's behavior. Evaluation of formalisms was designed based on real-world scenarios and their implementation in different formalisms using the platform Pogamut 3. One of the formalisms is Jason (AgentSpeak language interpreter) and it was connected to that platform as a part of this thesis. During the implementation of the scenarios were identified some problems which make creating the behavior of intelligent virtual agents harder. For these problems was proposed solution in the form of action selection mechanism (ASM) prototype and the basic methodology for designing agents with this ASM. The results of this study could be a starting point for more advanced action selection mechanism, which would allows more precise control of the behavior of intelligent virtual agents. Powered by TCPDF (www.tcpdf.org)

Many contemporary computer games can be described as dynamic real-time simulations inhabited by autonomous intelligent virtual agents (IVAs) where most of the environment structure is immutable and navigating through the environment is one of the most common activities. Though controlling the behaviour of such agents seems perfectly suited for action planning techniques, planning is not widely adopted in existing games. This paper attempts to answer the question whether the current academic planning technology is ready for integration to existing games and under which conditions. The paper compares reactive techniques to classical planning in handling the action selection problem for IVAs in game-like environments. Several existing classical planners that occupied top positions in the International Planning Competition were connected to the virtual environment of Unreal Development Kit via the Pogamut platform. Performance of IVAs employing those planners and IVAs with reactive architecture was measured on a class of game like test environments under different levels of external interference. It was shown that agents employing classical planning outperform reactive agents only if the size of the planning problem is small or if the environment changes are either hostile to the agent or not very frequent.