Thesis (M.Phil.)--City University of Hong Kong, 2009. / "Submitted to Department of Computer Science in partial fulfillment of the requirements for the degree of Master of Philosophy." Includes bibliographical references (leaves [152]-163)

This thesis attempts to problematize and rethink the inter-related construction of the categories of "environment" and "fitness". It argues that environments are materially and discursively constructed through the mutually constitutive mobilization of networks of human and non-human actors by particularly powerful centers of translation, and that these processes increasingly involve the construction of environments configured to the requirements of an ideal of fitness - a fitness defined in terms of risk, flexibility, response-ability, responsibility, mobility, and consumption. In developing this argument particular attention is given to the relations between bodies and technologies as actors constitutive of the networks from which environments are constructed. As a specific illustrative example of this, the efforts of the fitness equipment manufacturer NordicTrack to mobilize and translate diverse networks of actors in the space of the home and then represent these hybrid networks as ontologically purified, meaningful and marketable environments are examined. The ontological and spatial ambiguity of the types of environments constructed by corporations such as NordicTrack is then discussed, this ambiguity being registered in the difficulty of positioning the boundaries between categories such as subject and object, nature and culture, human and machine, real and virtual. Finally, having illustrated that these ambiguous environments are perhaps constituted by communities of human and non-human actors, this thesis then suggests that such a recognition might open up space for critical geographical imaginations that are responsive to the possibility that political, ethical, and moral community and agency are co-constructions of humans and non-humans. / Master of Science

Traditional autonomy among robotic and other artificial agents was accomplished via automated planning methods that found a viable sequence of actions, which, if executed by an agent, would result in the successful completion of the given task(s). However, many tasks that we would like robotic agents to perform involve goals that are complex, poorly-defined, or hard to specify. Furthermore, significant amounts of data or computation are required for agents to reach reasonable performance. As a result, autonomous systems still rely on human operators to play a supervisory role to ensure that robotic operations are completed quickly and successfully. The presented work aims to improve the traditional methods of robot autonomy by developing an intuitive means for(human operators to adapt/mold the behaviors and decision making of autonomous agents) autonomous agents to leverage the flexibility and expertise of human end users. Specifically, this work shows the results of three machine learning-based approaches for modifying and extending established robot navigation behaviors and skills through human demonstration. Our first project combines Imitation learning with classical navigation software to achieve long-horizon planning and navigation that follows navigation rules specified by a human user. We show that this method can adapt a robot's navigation behavior to become more like that of a human demonstrator. Moreover, for a minimal amount of demonstration data, we find that this approach outperforms recent baselines in both navigation success rate and trajectory similarity to the demonstrator. In the second project, we introduce a method of communicating complex skills over a short-horizon task. Specifically, we explore using imitation learning to teach a robot the complex skill needed to safely navigate through negative obstacles in simulation. We find that this proposed method could imitate complex navigation behaviors and generalize to novel environments in simulation with minimal demonstration. Furthermore, we find that this method compares favorably to a classical motion planning algorithm which was modified to assign traversal cost based on the terrain slope local to the robot's current pose. Finally, we demonstrate a practical implementation of the second approach in a real-world environment. We show that the proposed method results in a policy that can generalize across differently shaped obstacles and across simulation and reality. Moreover, we show that the proposed method still outperforms the classical motion planning algorithm when tasked to navigate negative obstacles in the real world. / Doctor of Philosophy / With the rapid advancement of computing power and growing technical literacy of the general public, the tasks that robots should be able to accomplish have multiplied. Robots can, however, be limited by the human ability to effectively convey how tasks should be performed. For example, autonomous robot navigation to a specified path planning software suite that generates feasible and obstacle-free trajectories through a cluttered environment. While these modules can be modified to meet task-specific constraints and user preferences, current modification procedures require substantial effort on the part of an expert roboticist with a great deal of technical training. The desired tasks and skills are difficult to effectively convey in a machine legible format. These tasks often require technical expertise in multiple mechatronic disciplines and hours of hand tuning that the typical end user does not have. In this dissertation, we examine methods that directly leverage human users to teach robots how to perform tasks that are generally difficult to specify pragmatically. We focus on methods that allow human users to extend established robot navigation behaviors and skills by demonstrating their own preferred approaches. We evaluate the performances of our proposed approaches in terms of navigation success rate, adherence to the demonstrated behavior, and their ability to apply what they have learned to novel environments. Moreover, we showed that our approaches compare favorably to recent machine learning-based approaches to autonomous navigation, and classical navigation techniques with respect to these metrics.

Problem formulation: The problem that the thesis research relates to is the limitations of artificially intelligent chatbots as interlocutors. The emotional component of communication plays an essential role in the customer experience, but many users have a negative attitude toward chatbots due to their lack of humanity and empathy. The potential of the new ChatGPT in changing user attitudes toward chatbots is also being explored. However, the limited data available on recent versions of ChatGPT presents an additional challenge for research in this area.  Purpose: Our study aims to study people's emotional responses to human-like chatbots and their impact on user satisfaction. We also explore whether human likeness is a crucial driver of chatbot preference and how the new ChatGPT can change user attitudes toward them in a positive way.  Theoretical framework: The study's theoretical framework considers various aspects of using chatbots based on artificial intelligence (AI) in marketing. In this context, we observe ChatGPT as a revolutionary breakthrough in customer service, capable of improving customer experience and interaction with customer. We emphasise the emotional component of human-chatbot interactions, investigating customer emotions, attitudes, and trust, as well as the chatbot's capacity for empathy and human-like characteristics. Drawing from this theoretical exploration, we formulate four hypotheses to guide our research. Methodology: This quantitative study involves 79 respondents aged 18 years and over. The online survey was conducted using social media for dissemination. The empirical data obtained were coded and analysed using the SPSS program.  Empirical findings: Our study confirms the hypothesis of diverse emotional responses (H4) and a generally neutral emotional response during chatbot interactions (H3). We also find partial support for the presence of negative emotions (H2), but not for consistent positive emotions (H1). The data indicate a range of emotional responses, highlighting the complexity of human reactions to chatbots. Conclusion: Our research provides an overall picture of users' emotional responses to interactions with chatbots. Users show a variety of emotions, mostly neutral, which can change depending on the interaction. We also discovered the potential of the new ChatGPT in changing user attitudes towards chatbots in a more positive or neutral direction. The study also uncovers factors influencing users' emotional responses, such as age, attitudes, and past experiences. The results can be used to develop more effective marketing and business strategies for interacting with chatbots.

The aim of this thesis is to report on a transdisciplinary approach, regarding the complexity of thinking about human embodiment in relation to machine embodiment. A practical dimension of this thesis is to elicit some principles for the design and evaluation of virtual embodiment. The transdisciplinary approach suggests, firstly, that a single discipline or reality is, on its own, not sufficient to explain the complexity and dynamism of the embodied interaction between the human and machine. Secondly, the thesis argues for thinking of transdisciplinary research as a process of individuation, becoming or transduction, that is, as a process of mediation between heterogeneous approaches rather than perceiving research as a stabilized cognitive schema designed to accumulate new outcomes to the already-there reality. Arguing for going beyond the individualized approaches to embodiment, this thesis analyzes three cases where the problems that appear in one case are resolved through the analysis of the following one. Consisting of three phases, this research moves from objective scientific 'reality' to more phenomenological, subjective and complex realities. The first study employs a critical review of embodied conversational agents in human-computer interaction (HCI) in a learning context using a comparative meta-analysis. Meta-analysis was applied because most of the studies for evaluating embodiment are experimental. A learning context was selected because the number of studies is suitable for meta-analysis and the findings could be generalized to other contexts. The analysis reveals that there is no 'persona effect', that is, the expected positive effect of virtual embodiment on the participant's affective, perceptive and cognitive measures. On the contrary, it shows the reduction of virtual embodiment to image and a lack of consideration for the participant's embodiment and interaction, in addition to theoretical and methodological shortcomings. The second phase solves these problems by focusing on Mark Hansen's phenomenological account of embodiment in new media. The investigation shows that Hansen improves on the HCI account by focusing on the participant's dynamic interaction with new media. Nevertheless, his views of embodied perception and affection are underpinned by a subjective patriarchal account leading to object/subject and body/work polarizations. The final phase resolves this polarization by analyzing the controversial work of Alan Turing on intelligent machinery. The research provides a different reading of the Turing Machine based on Simondon's concept of individuation, repositioning its materiality from the abstract non-existent to the actual-virtual realm and investigating the reasons for its abstraction. It relates the emergence of multiple human-machine encounters in Turing's work to the complex counter-becoming of what it describes as 'the Turing Machine compound'.

L'authentification homme-machine est une problématique largement développée pour les télécommunications. Une authentification dans le sens "machine-homme" permettra d'assurer l'utilisateur humain assermenté du fonctionnement intègre d'une machine lors, par exemple, d'une session de vote électronique ou d'une vérification d'objet en traçabilité sécurisée. Cette thèse se focalise sur la traçabilité sécurisée sans accès (systématique) à un canal de communication. Nous décrivons différentes techniques d'authentification de produits manufacturés en nous concentrant sur une méthode de caractérisation de motifs imprimés. Pour effectivement authentifier un objet, nous montrons qu'un agent vérifieur doit s'assurer de l'intégrité du tiers et du système électronique utilisée pour la vérification. Cependant l'authenticité du système électronique lui-même reste à vérifier. La question que nous adressons alors est la suivante : comment un être humain peut-il se convaincre de l'intégrité et de l'authenticité d'un système embarqué dans un mode hors ligne ? Nous définissons deux familles de solutions. Dans la première, l'utilisateur fait appel, pour les calculs, à un dispositif auxiliaire tandis que dans la seconde l'utilisateur ne fait usage que d'un papier et d'un crayon. Pour chacune des deux familles, nous proposons un protocole d'authentification d'un système embarqué dont la puce, typiquement un FPGA ou un microcontrôleur, dépend de la configuration ou de la programmation d'une mémoire RAM / "Human-to-Machine" authentication is widely developed for modern telecommunications. A "Machine-to-Human" authentication will ensure the trusted human user about the integrity of the machine, for example during an electronic voting session or object verification in secure traceability. This work is focused on secure traceability without any systematic access to a communication network. We depict different technics for goods authentication and we focus on a method based on the characterization of printed patterns. To completely authenticate an object, we show that a human verifier has to be confident in the integrity of the third party and the electronic system involved in the verification phase. However, the authenticity of the electronic system itself has also to be verified. We address here the following question : how a human being can convince himself about the integrity and the authenticity of an embedded system in an off-line environment ? We propose two groups of solutions. In the first one, an auxiliary electronic device is used to perform computing operations. In the second one, the human capability (memory and computational abilities) is exploited. In each group, we propose a protocol to authenticate embedded systems for which the chip (typically an FPGA (Field Programmable Gate Array) or a microcontroller) is initialized according to the configuration or programming of its RAM memory

La traduction automatique (TA) a connu des progrès significatifs ces dernières années et continue de s'améliorer. La TA est utilisée aujourd'hui avec succès dans de nombreux contextes, y compris les environnements professionnels de traduction et les scénarios de production. Cependant, le processus de traduction requiert souvent des connaissances plus larges qu'extraites de corpus parallèles. Étant donné qu'une injection de connaissances humaines dans la TA est nécessaire, l'un des moyens possibles d'améliorer TA est d'assurer une collaboration optimisée entre l'humain et la machine. À cette fin, de nombreuses questions sont posées pour la recherche en TA: Comment détecter les passages où une aide humaine devrait être proposée ? Comment faire pour que les machines exploitent les connaissances humaines obtenues afin d'améliorer leurs sorties ? Enfin, comment optimiser l'échange: minimiser l'effort humain impliqué et maximiser la qualité de TA? Diverses solutions sont possibles selon les scénarios de traductions considérés. Dans cette thèse, nous avons choisi de nous concentrer sur la pré-édition, une intervention humaine en TA qui a lieu ex-ante, par opposition à la post-édition, où l'intervention humaine qui déroule ex-post. En particulier, nous étudions des scénarios de pré-édition ciblés où l'humain doit fournir des traductions pour des segments sources difficiles à traduire et choisis avec soin. Les scénarios de la pré-édition impliquant la pré-traduction restent étonnamment peu étudiés dans la communauté. Cependant, ces scénarios peuvent offrir une série d'avantages relativement, notamment, à des scénarios de post-édition non ciblés, tels que : la réduction de la charge cognitive requise pour analyser des phrases mal traduites; davantage de contrôle sur le processus; une possibilité que la machine exploite de nouvelles connaissances pour améliorer la traduction automatique au voisinage des segments pré-traduits, etc. De plus, dans un contexte multilingue, des difficultés communes peuvent être résolues simultanément pour de nombreuses langues. De tels scénarios s'adaptent donc parfaitement aux contextes de production standard, où l'un des principaux objectifs est de réduire le coût de l’intervention humaine et où les traductions sont généralement effectuées à partir d'une langue vers plusieurs langues à la fois. Dans ce contexte, nous nous concentrons sur la TA de revues systématiques en médecine. En considérant cet exemple, nous proposons une méthodologie indépendante du système pour la détection des difficultés de traduction. Nous définissons la notion de difficulté de traduction de la manière suivante : les segments difficiles à traduire sont des segments pour lesquels un système de TA fait des prédictions erronées. Nous formulons le problème comme un problème de classification binaire et montrons que, en utilisant cette méthodologie, les difficultés peuvent être détectées de manière fiable sans avoir accès à des informations spécifiques au système. Nous montrons que dans un contexte multilingue, les difficultés communes sont rares. Une perspective plus prometteuse en vue d'améliorer la qualité réside dans des approches dans lesquelles les traductions dans les différentes langues s’aident mutuellement à résoudre leurs difficultés. Nous intégrons les résultats de notre procédure de détection des difficultés dans un protocole de pré-édition qui permet de résoudre ces difficultés par pré-traduction. Nous évaluons le protocole dans un cadre simulé et montrons que la pré-traduction peut être à la fois utile pour améliorer la qualité de la TA et réaliste en termes d'implication des efforts humains. En outre, les effets indirects sont significatifs. Nous évaluons également notre protocole dans un contexte préliminaire impliquant des interventions humaines. Les résultats de ces expériences pilotes confirment les résultats obtenus dans le cadre simulé et ouvrent des perspectives encourageantes pour des tests ultérieures. / Machine Translation (MT) has made significant progress in the recent years and continues to improve. Today, MT is successfully used in many contexts, including professional translation environments and production scenarios. However, the translation process requires knowledge larger in scope than what can be captured by machines even from a large quantity of translated texts. Since injecting human knowledge into MT is required, one of the potential ways to improve MT is to ensure an optimized human-machine collaboration. To this end, many questions are asked by modern research in MT: How to detect where human assistance should be proposed? How to make machines exploit the obtained human knowledge so that they could improve their output? And, not less importantly, how to optimize the exchange so as to minimize the human effort involved and maximize the quality of MT output? Various solutions have been proposed depending on concrete implementations of the MT process. In this thesis we have chosen to focus on Pre-Edition (PRE), corresponding to a type of human intervention into MT that takes place ex-ante, as opposed to Post-Edition (PE), where human intervention takes place ex-post. In particular, we study targeted PRE scenarios where the human is to provide translations for carefully chosen, difficult-to-translate, source segments. Targeted PRE scenarios involving pre-translation remain surprisingly understudied in the MT community. However, such PRE scenarios can offer a series of advantages as compared, for instance, to non-targeted PE scenarios: i.a., the reduction of the cognitive load required to analyze poorly translated sentences; more control over the translation process; a possibility that the machine will exploit new knowledge to improve the automatic translation of neighboring words, etc. Moreover, in a multilingual setting common difficulties can be resolved at one time and for many languages. Such scenarios thus perfectly fit standard production contexts, where one of the main goals is to reduce the cost of PE and where translations are commonly performed simultaneously from one language into many languages. A representative production context - an automatic translation of systematic medical reviews - is the focus of this work. Given this representative context, we propose a system-independent methodology for translation difficulty detection. We define the notion of translation difficulty as related to translation quality: difficult-to-translate segments are segments for which an MT system makes erroneous predictions. We cast the problem of difficulty detection as a binary classification problem and demonstrate that, using this methodology, difficulties can be reliably detected without access to system-specific information. We show that in a multilingual setting common difficulties are rare, and a better perspective of quality improvement lies in approaches where translations into different languages will help each other in the resolution of difficulties. We integrate the results of our difficulty detection procedure into a PRE protocol that enables resolution of those difficulties by pre-translation. We assess the protocol in a simulated setting and show that pre-translation as a type of PRE can be both useful to improve MT quality and realistic in terms of the human effort involved. Moreover, indirect effects are found to be genuine. We also assess the protocol in a preliminary real-life setting. Results of those pilot experiments confirm the results in the simulated setting and suggest an encouraging beginning of the test phase.

No description available.