<p>Emerging technologies in the classroom are paving the way towards high-quality, hands-on distance learning. Augmented Reality (AR), which overlays virtual information into the physical world, provides a promising solution for the development and delivery of collaborative educational content. Frameworks such as ARkit, ARCore, have enabled AR experiences to become available to a wider audience. However, there are still several challenges to implementing an AR-based curriculum in classrooms, such as difficulty to create AR content, lack of an architecture capable of supporting collaboration between users, and questions about the user experience. This thesis introduces the MetaAR project, a series of solutions to enable instructors and designers to prototype AR experiences in collaborative and distant classrooms. We designed and tested interactive systems, each targeted towards solving a different problem: (1) MetaAR, an augmented reality authoring platform for instructors and students; (2) RobotAR, a robotics toolkit to create augmented reality-based makerspaces; (3) ColabAR, a toolkit for quick-prototyping of Tangible Augmented Reality (TAR) laboratories; (4) Grove-Blockly, a website with a STEAM curriculum involving IoTs, crafting and coding aimed at middle-schoolers; (5) Towards Modeling of Human Skilling for Electrical</p> <p>Circuitry using Augmented Reality Applications, which provides a model to cluster microskills found in AR (perceptual, cognitive, motor) and aligns them to educational content design for AR. Our preliminary results, obtained from user studies involving more than 120 participants, provide evidence of the sustainability and the positive reception of our prototypes in learning environments. We demonstrated an improvement in several of students’ key competencies and in the overall user experience for both instructors and students. Our hope is that this thesis provides a pathway towards more natural interactions and advances in our understanding of distance learning technology, which is becoming increasingly important in today's society.</p>

<div>Real-time social media data can provide useful information on evolving events and situations. In addition, various domain users are increasingly leveraging real-time social media data to gain rapid situational awareness. Informed by discussions with first responders and government officials, we focus on two major barriers limiting the widespread adoption of social media for situational awareness: the lack of geotagged data and the deluge of irrelevant information during events. Geotags are naturally useful, as they indicate the location of origin and provide geographic context. Only a small portion of social media is geotagged, however, limiting its practical use for situational awareness. The deluge of irrelevant data provides equal difficulties, impeding the effective identification of semantically relevant information. Existing methods for short text relevance classification fail to incorporate users' knowledge into the classification process. Therefore, classifiers cannot be interactively retrained for specific events or user-dependent needs in real-time, limiting situational awareness. In this work, we first adapt, improve, and evaluate a state-of-the-art deep learning model for city-level geolocation prediction, and integrate it with a visual analytics system tailored for real-time situational awareness. We then present a novel interactive learning framework in which users rapidly identify relevant data by iteratively correcting the relevance classification of tweets in real-time. We integrate our framework with the extended Social Media Analytics and Reporting Toolkit (SMART) 2.0 system, allowing the use of our interactive learning framework within a visual analytics system adapted for real-time situational awareness.</div>

Nowadays, increasing numbers of people do travel research on their smartphones. More precisely, digital maps provide locational information, which is important during the planning process. However, smartphones are restricted by their small screen size, resulting in fragmented information delivery; also, the design of digital maps lacks features. The aims of this study are to investigate users’ travel-planning behavior on smartphones, identify the pain points and missing contexts when using digital maps on smartphones, and provide design guidelines for future digital map design. The study was done by conducting a travel-planning activity and a codesign workshop to bring users into the design process, promote in-depth discussion, and explore a new design possibility for digital maps with users. The results showed that people’s goals when planning travel include reducing their workload, improving effectiveness, and ensuring flexibility. People use digital maps to support not only information searching but also information compiling, including saving locations and routes. In addition, several difficulties have been pointed out: cross-platform planning, information hierarchy, and retrieval on digital maps.

Evaluating the “success” of an information visualization (InfoVis) where its main purpose is communication or presentation is challenging. Within metrics that go beyond traditional analysis- and performance-oriented approaches, one construct that has received attention in recent years is “user engagement”. In this research, I propose Affective Engagement (AE)-- user's engagement in emotional aspects as a metric for InfoVis evaluation. I developed and evaluated a self-report measurement tool named AEVis that can quantify a user's level of AE while using an InfoVis. Following a systematic process of evidence-centered design, each activity during instrument development contributed specific evidence to support the validity of interpretations of scores from the instrument. Four stages were established for the development: In stage 1, I examined the role and characteristics of AE in evaluating information visualization through an exploratory qualitative study, from which 11 indicators of AE were proposed: Fluidity, Enthusiasm, Curiosity, Discovery, Clarity, Storytelling, Creativity, Entertainment, Untroubling, Captivation, and Pleasing; In stage 2, I developed an item bank comprising various candidate items for assessing a user's level of AE, and assembled the first version of survey instrument through target population and domain experts' feedback; In stage 3, I conducted three field tests for instrument revisions. Three analytical methods were applied during this process: Item Analysis, Factor Analysis (FA), and Item Response Theory (IRT); In stage 4, a follow-up field test study was conducted to investigate the external relations between constructs in AEVis and other existing instruments. The results of the four stages support the validity and reliability of the developed instrument, including: In stage 1, user's AE characteristics elicited from the observations support the theoretical background of the test content; In stage 2, the feedback and review from target users and domain experts provides validity evidence for the test content of the instrument in the context of InfoVis; In stage 3, results from Exploratory and Confirmatory FA, as well as IRT methods reveal evidence for the internal structure of the instrument; In stage 4, the correlations between total scores and sub-scores of AEVis and other existing instruments provide external relation evidence of score interpretations. Using this instrument, visualization researchers and designers can evaluate non-performance-related aspects of their work efficiently and without specific domain knowledge. The utilities and implications of AE can be investigated as well. In the future, this research may provide foundation for expanding the theoretical basis of engagement in the fields of human-computer interaction and information visualization.

Les outils actuels de gestion industrielle s'appuient généralement sur l'ingénierie concourante, qui implique la réalisation des étapes de gestion du cycle de vie des produits en parallèle et l'intégration des données techniques pour le partage entre les différents experts. Divers experts utilisent des logiciels spécifiques à leur domaine pour produire diverses données compilées via une maquette numérique. Ces experts multidisciplinaires ont tendance à travailler en collaboration pendant le développement de produits. Au cours d'activités de conception collaborative synchrones, telles que les revues de projet et la prise de décisions, les experts de différents domaines doivent dialoguer, négocier et choisir pour résoudre les différences multidisciplinaires. De nombreux domaines tels que la conception collaborative et de l’évaluation des produits avec multiples experts, ont une grande demande de nouveaux outils d'aide à la collaboration. Avec le développement des technologies de l'Interaction Homme-Machine (IHM), il est possible de concevoir des outils et des méthodes plus intuitifs pour améliorer la collaboration co-localisée entre les experts.Dans cette thèse, afin d'améliorer la collaboration avec des experts de différents domaines et communiquer via la maquette numérique, une IHM multi-utilisateurs avec des représentations différentes lors d'un travail collaboratif a été pris en considération et son influence sur la collaboration multidisciplinaire co-localisée est étudiée. Un schéma de la méthodologie d'évaluation de la contribution à un système multi-utilisateurs et des expérimentations sont proposées. Les résultats des expériences des résultats significatifs concernant l’efficacité de la réalisation de la tâche, l'utilisabilité de l'Interaction Homme-Machine et la performance de la collaboration lors de l'utilisation de l'interface multi-utilisateurs dans les scénarios de collaboration multidisciplinaires. Les contributions et apports de cette interface sont discutés. / The current industrial management tools generally rely on Concurrent Engineering, which involves conducting Product Lifecycle Management stages in parallel and integrating technical data for sharing across different experts. Various experts use domain-specific software to produce various data into Digital mock-up. These multidisciplinary experts have trends to work collaboratively during product development. During co-located synchronous collaborative design activities, such as project review and decision-making, experts from different domains must discuss, negotiate, and compromise to solve multidisciplinary differences. Many areas, such as early collaborative design and multi-expert product evaluation, have a great demand for new collaborative support tools. With the development of Human Computer Interaction, it is possible to devise more intuitive tools to enhance co-located collaboration across experts.In this thesis, to enhance the collaboration with experts on different domains to communicate with DMU, a multi-user interface across users with different representations during a collaborative work has been taken into consideration and its influence on co-located multidisciplinary collaboration is investigated. A schema of the methodology for evaluating the contribution to a multi-user system and the multiple users’ experiences is proposed. Results of experiments show the significances of the efficiency of task, the usability of interface, and the performance of collaboration during the use of multi-user CHI in multidisciplinary collaborative scenarios. The contributions of what multi-user interface brings to the design criteria of multi-user interface and multi-user co-located collaboration are discussed.

<p>The studio model of education incorporated in to many design-oriented HCI programs in the past two decades brings a number of objectives to programs implementing it. One objective is the building of a “bridge” between pedagogy and practice, preparing students for the differing realities between academia, and the constraints imposed in an organizational setting. The bridge also encourages the development of a student’s design philosophy, allowing them to acknowledge and understand their conceptions of design which influence decisions in project-processes, and the projected communities they may navigate towards in practice. This study addresses the dimensions of design philosophy held by students educated in these programs, and how such philosophies are engaged and shaped further in practice. Through a qualitative interview approach, this study presents 9 dimensions of design philosophy through the accounts of 10 students and practitioners, reflecting on their education and practice. Using existing work studying the flow of competence between practitioners and organizations, the discussion of the dimensions presented provides four ways in which the philosophies of practitioners may encounter tensions in practice. This research proposes future work on how the studio model in HCI pedagogy may better prepare students for enacting their philosophies, and further reflecting on the shaping of that philosophy through felt contrasts between education and practice. </p>

The researcher’s intent in this study was to understand users’ performance, specifically in terms of time, error and workload, in different display conditions while manipulating a space-time cube visualization. A convergent mixed-method design was applied to allow the researcher to better understand the research problems. In the study, time, error and perceived workload were investigated to test performance to detect if a display condition had a positive or negative influence on users’ abilities to perform a task. The qualitative data explored the differences in users’ experiences with the HoloLens and desktop<br>

The recent boom of Machine Learning Network Architectures like Generative Adversarial Networks (GAN), Deep Convolution Generative Adversarial Networks (DCGAN), Self Attention Generative Adversarial Networks (SAGAN), Context Conditional Generative Adversarial Networks (CCGAN) and the development of high-performance computing for big data analysis has the potential to be highly beneficial in many domains and fittingly in the early detection of chronic diseases. The clinical heterogeneity of one such chronic auto-immune disease like Systemic Lupus Erythematosus (SLE), also known as Lupus, makes it difficult for medical diagnostics. One major concern is a limited dataset that is available for diagnostics. In this research, we demonstrate the application of Generative Adversarial Networks for data augmentation and improving the error rates of Convolution Neural Networks (CNN). Limited Lupus dataset of 30 typical ’butterfly rash’ images is used as a model to decrease the error rates of a widely accepted CNN architecture like Le-Net. For the Lupus dataset, it can be seen that there is a 73.22% decrease in the error rates of Le-Net. Therefore such an approach can be extended to most recent Neural Network classifiers like ResNet. Additionally, a human perceptual study reveals that the artificial images generated from CCGAN are preferred to closely resemble real Lupus images over the artificial images generated from SAGAN and DCGAN by 45 Amazon MTurk participants. These participants are identified as ’healthcare professionals’ in the Amazon MTurk platform. This research aims to help reduce the time in detection and treatment of Lupus which usually takes 6 to 9 months from its onset.

This thesis introduces a new approach to designing a Message Handling Assistant (MA). It presents a model of an MA and an intention extraction function for text messages, such as emails and Newsgroups articles. Based on a speech act theory and the belief-desire-intention (BDI) theory of rational agency, we define a generic MA. By interpreting intuitive descriptions of the desired behaviours of an MA using the BDI theory and speech act theory, we conjecture that intentions of messages alone provide enough information needed to capture user models and to reason how messages should be processed. To identify intentions of messages written in natural language, we develop a model of an intention extraction function that maps messages to intentions. This function is modelled in two steps. First, each sentence in a message is converted into a tuple (performative, proposition) using a dialogue act classifier. Second, the sender's intentions are formulated from the tuples using constraints for felicitous human communication. As an investigation of the use of machine learning technologies for designing the intention extraction function, four dialog act classifiers are implemented and evaluated on Newsgroups articles. The thesis also proposes a semantic communication framework, which integrates the agent and Internet technologies for automatic message composing and ontology exchange services.

The tremendous growth of information on the Internet has been above our ability to process. A recommender system, which filters out useful information and generate recommendations, has been introduced to help users overcome the information overload problem and has been widely applied in an ever-increasing number of e-commercial websites. Collaborative filtering and content-based recommendation methods are two major approaches used in recommender systems. The collaborative filtering predicts items which a particular user prefers by using a database about the past preferences of users with similar interests. The content-based method analyses the content of the objects to generate a representative list of the user’s interests, and then compares the similarity of item descriptions. These two methods have some drawbacks in dealing with situations such as sparse data and cold start problems. Recently, hybrid methods combining collaborative filtering and content-based methods have been proposed to overcome these limitations. However, personalized recommender system attempt to penetrate people’s various demand and generate the tailored recommendations. A highly effective and personalised recommender system may still face new challenges including interestdrifting and multicriteria optimisation. For example, a user’s interest may change over time. They may no longer like a item which was strongly preferred. Another example is that a person’s preference is varying and always has multiple criteria. Classic collaborative filtering uses a single overall rating for prediction. It does not properly reflect the opinion on a item and the reason why people rated this item high or low. Unfortunately, the current recommender systems do not consider these important factors. First, we proposed a novel hybrid recommender system to overcome interest-drifting by embedding the time-sensitive functions into the recommendation process. The experimental results show that the intergraded approach with interest-drifting can constantly perform better and provide users with higher quality recommendations. Meanwhile, the experimental results on different size of training dataset show that our algorithm can boost the prediction accuracy for all configurations. The contributions of this proposed algorithm are in two main aspects. First, using time function to reflect users’ intersts changing in order to achieve higher quality of recommendations. Second, using intergraded methods to solve some problems such as sparsity and cold start. Then we developed a new technique to aggregate the multicriteria ratings for predicting more accurate recommendations. The results show that our algorithms outperforms the traditional collaborative filtering recommender system on both accuracy of predicting ratings and accuracy of recommendations. The one of contributions in this proposed method is that we introduced the multicriteria concept into recommender systems to reflect the users’ opinion more accurate. Another contribution is that we develop a linear method to aggregate multicriteria to single rating for higher quality of recommendations. Our experiments demonstrate that the recommendation achieved better performances when interest-drifting and multicriteria ratings were considered. The significance of our research study is that we consider incorporating interest-drifting, and multicriteria ratings into a recommender system to generate personalised and effective recommendations.