<div>As part of an increasing interest in a "Turn to Practice," HCI scholars have investigated the felt design complexities and ethical concerns in everyday technology practice, calling for practice-led research approaches. Given the ethical nature of technology design work, practitioners have to often negotiate and mediate their personal values, disciplinary notions of ethics, organizational policies and values, and societal impact of their design work. To tease apart and describe practitioner accounts of ethical aspects of their design work, I used three different approaches to investigate what practitioners from different professional roles communicate about and participate in (potentially) strengthening their ethical engagement in their everyday design work within and across role boundaries: survey, design of co-creation activities, and deployment/pilot of these co-creation activities. </div><div><br></div><div>In the survey study, I identify and describe the differences in disciplinary values, responsibilities, commitments, and alignment in relation to ethics and social responsibility through captured data from 256 technology and design practitioners from a range of professional roles.</div><div><br></div><div>As a part of the design phase of co-creation activities, I design, iterate, and prototype three co-creation activities (A: Tracing the Complexity; B: Dilemma Postcards; and C: Method Heuristics) and sequences of these activities to engage a range of different professional roles to communicate about their ethical action and (potentially) strengthen their ethical engagement in everyday design work. I define design vocabulary/Schemas: 1) <i>A.E.I.O.YOU model</i> to investigate the landscape of ethics in practice and 2) <i>Classifiers</i> to codify the activities and potential variants.</div><div><br></div><div>As a part of the deployment phase of these designed co-creation activities, I piloted four sequences of these activities with twelve practitioners with three different professional roles per sequence, engaging in approx. 23 hours of facilitation, artifact creation, and conversation. I present the results of deployment of the co-creation sessions where practitioners articulated that the co-creation activities helped <i>expand</i> their ethical horizons through self-awareness, <i>learn</i> new approaches to ethics vocabulary, <i>become (re-)aware </i>of their current practice, and <i>imagine</i> trajectories of change in their practice. Practitioners also identified a preliminary set of ethics-related practices that could be better supported such as tools for performance, leadership support, ethics education, and resources for ethical decision making. </div><div><br></div><div>Based on the results from these three approaches, I propose contributions to HCI and design audiences. For HCI researchers, practitioners, and educators, the survey results describe differences in professional notions and valence of ethics, framing the need for translation and transdisciplinary approach to ethics in a practice context. For design researchers, the designing of the co-creation activities is a methodological contribution where I propose and illustrate opportunities for creating novel ways to engage practitioners in co-creation work as a means of communicating their felt ethical concerns and practices. For co-creation researchers and professional ethicists, the engagement of practitioners in the co-creation reveal: 1) complexities to facilitate different disciplinary roles and design a space for ``representing'' a range of practitioners; and 2) gaps and potential synergies in supporting practitioners through practice-resonant ethics-focused methods. </div>

Crowdsourcing has become a popular choice for tackling problems that neither computers nor humans alone can solve with adequate speed, cost, and quality. However, instructing crowds to execute tasks in the manner expected by the requesters is challenging. It depends on not only requesters’ task design abilities but also workers’ understanding of the tasks. Task design bridges the communication gap between workers and requesters, which consists of instructions, payment, time limit on task, and the interface for workers to work on. It remains an underdeveloped but important topic that needs further exploration for improving crowdsourcing experience. My research studies task delivery from requesters to crowd workers. The goal is to improve the communication between the two and, in turn, increase accuracy of results and decrease variability due to differing interpretations and perspectives. Specifically, this dissertation presents a series of studies to show that high-quality results can be obtained from human workers through improved task design, by 1) designing incentives to recruit workers with the appropriate skills for given tasks, 2) designing unambiguous instructions to clearly express task requirements, 3) choosing the correct strategy to communicate the requisite task knowledge with workers, and 4) enhancing requesters’ ability to rapidly prototype Augmented Reality (AR) instructions. This dissertation demonstrates that crowdsourcing quality is improved when the tasks are communicated using mediums and structures that align with workers’ preference and utility

<div>Face redaction is used to deidentify images of people. Most approaches depend on face detection, but automated algorithms are still not adequate for sensitive applications in which even one unredacted face could lead to irreversible harm. Human annotators can potentially provide the most accurate detection, but only trusted annotators should be allowed to see the faces of privacy-sensitive applications. Redacting more images than trusted annotators could accommodate requires a new approach. </div><div>This dissertation leverages the characteristics of human perception of faces in median-filtered images in a human computation algorithm to engage crowd workers to redact faces—without revealing the identities. IntoFocus, a system I developed, permits robust face redaction with probabilistic privacy guarantees. The system's design builds on an experiment that measured the filter levels and conditions where participants could detect and identify faces. </div><div> Pterodactyl is a system that focuses on increasing the productivity of crowd-based face redaction systems. It uses the AdaptiveFocus filter, a filter that combines human perception of faces in median filtered images with a convolutional neural network to estimate a median filter level for each region of the image to allow the faces to be detected and prevent them from being identified.</div>

<p>The accelerated development of mobile computational platforms and artificial intelligence (AI) has led to increase in interconnected products with sensors that are creating smart environments. The smart environment expands the interactive spaces from limited digital screens, such as desktops and phones, to a much broader category that includes everyday objects, smart things, surrounding contexts, robots, and humans. The improvement of personal computing devices including smartphones, watches, and AR glasses further broadens the communication bandwidth between us and the ambient intelligence from the surrounding environment. Additionally in this smart environment people want to pursue personalization and are motivated to design and build their own smart environments and author customized functions.</p> <p> </p> <p>My work in this thesis focuses on investigating workflows and methods to support end-users to create personalized interactive experiences within smart environments. In particular, I designed the authoring systems by inspecting different interaction modalities, namely the direct input, spatial movement, in-situ activity and embodied interactions between users and everyday objects, smart things, robots and virtual mid-air contents. To this end, we explored 1) the software tools, hardware modules, and machines that support users to augment non-smart environments with digital interfaces and functions, and 2) the intelligence and context-awareness powered by the smart environments that deliver automatic and assistance during living and entertaining experiences. In this thesis, I mainly studied the following authoring workflows and systems: 1) customizing interactive interfaces on ordinary objects by surface painted circuits, 2) constructing a spatially aware environment for service robots with IoT modules, 3) authoring robot and IoT applications that can be driven by body actions and daily activities and 4) creating interactive and responsive augmented reality applications and games that can be played through natural input modalities.</p> <p> </p> <p>Takeaways from the main body of the research indicate that the authoring systems greatly lower the barrier for end-users to understand, control, and modify the smart environments. We conclude that seamless, fluent, and intuitive authoring experiences are crucial for building harmonious human-AI symbiotic living environments.</p>

<p>The increase in the use of digital devices, has vastly increased the amount of data used and consequently, has increased the availability and relevance of digital evidence. Typically, digital evidence helps to establish the identity of an offender by identifying the username or the user account logged into the device at the time of offense. Investigating officers need to establish the link between that user and an actual person. This is difficult in the case of computers that are shared or compromised. Also, the increasing amount of data in digital investigations necessitates the use of advanced data analysis approaches like machine learning, while keeping pace with the constantly evolving techniques. It also requires reporting on known error rates for these advanced techniques. There have been several research studies exploring the use of behavioral biometrics to support this user attribution in digital forensics. However, the use of the state-of-the-art XGBoost algorithm, hasn’t been explored yet. This study builds on previously conducted research by modeling user interaction using the XGBoost algorithm, based on features related to keystroke and mouse usage, and verifying the performance for user attribution. With an F1 score and Area Under the Receiver Operating Curve (AUROC) of .95, the algorithm successfully attributes the user event to the right user. The XGBoost model also outperforms other classifiers based on algorithms such as Support Vector Machines (SVM), Boosted SVM and Random Forest.</p>

<p>A role model may play an important role in an individual’s career, such as in the case of faculty of higher education. However, not much is known about how one perceives these role models in an academic setting. There is limited research into the unique attributes that distinguish between</p> <p>types of role models. Hence, this study attempts to better understand faculty role modelsand shed light onto those attributes which set them apart.</p> <p>The purpose of the study can be realized through two research questions, (1) What are some of the role models that academic faculty follow? and (2) What attributes do those role models possess? This study has iteratively developed surveys designed to elicit answers to these questions, and the survey responses will be used to promote a meaningful conversation about faculty role</p> <p>models. The results will contribute towards improving career development programs to create a positive impact on faculty effectiveness and success.</p>

<div>Camera view composing, such as photography, has become inseparable from everyday life. Especially with the development of drone technology, the flying mobile camera is accessible and affordable and has been used to take impressive photos. However, the process of acquiring the desired view requires manual searches and adjustments, which are usually time consuming and tedious. The situation is exacerbated with difficulty in the controlling of a mobile camera that has many Degree of Freedom. It becomes complicated to compose a well-framed view, because experience, timing, and aesthetic are all indispensable. Therefore, professional view composing with a mobile camera is not an easy task for most people. Powered by deep learning, recent breakthroughs in artificial intelligence have enabled machines to perform human-level automation in several tasks. The advances in automatic decision-making and autonomous control have the potential to improve the camera view composing process significantly. </div><div><br></div><div>We observe that (a) the human-robot interaction can be more intuitive and natural for photography tasks, and (b) the drone photography tasks can be further automated by learning professional photo taken patterns with data-driven methods. In this work, we present two novel frameworks for drone photography basing on the two observations. First, we demonstrate a multi-touch gesture-controlled gimbaled-drone photography framework-FlyCam. FlyCam abstracts the camera and the drone into a single flying camera object and supports the entire control intuitively on a single mobile device with simple touch gestures. Second, we present a region-of-interest based, reinforced drone photography framework-Dr$^{3}$Cam. Our full automation Dr$^{3}$Cam is built on top of state-of-the-art reinforcement learning research and enables the camera agent to seek for good views and compose visually appealing photos intelligently. Results show that FlyCam can significantly reduce the workload and increase the efficiency in human-robot interaction, while Dr$^{3}$Cam performs human-level view composing automation for drone photography tasks.</div>

With the interactivity and immersion of players into video games, rising development costs, and heightened expectations from AAA developers video games need to make sure they hit their target market more than ever. This is something that is less extreme in the educational game development space; but ultimately true with limited grant funding, limited development time within a student developer’s schedule, and how rapidly a recently leased student content creator will need to learn the space and needs of the client. When a student is brought on late into a development cycle, it can become troublesome when they are required to meet new developing features on a changing project. This paper looks over how one team approached this issue, with a focus on meeting the needs of a group of American high school teachers. Within this paper, the focus is how they tackled the issue, and how the teachers reacted to the end prototype, with some insight into the older prototypes of the project. Throughout it they had reinforced the ideas that communication, data validity, and set contract goals are important identifiers for project success. Teachers looking at video games care more about the data being valid and clearly communicated more than if a game is fun or laden with features and mini-games.

<div>Measurement of vital cardiovascular health attributes, e.g., pulse rate variability, and estimation of exertion level of a person can help in diagnosing potential cardiovascular diseases, musculoskeletal injuries and thus monitoring an individual's well-being. Cumulative exposure to repetitive and forceful activities may lead to musculoskeletal injuries which not only reduce workers' efficiency and productivity, but also affect their quality of life. Existing techniques for such measurements pose a great challenge as they are either intrusive, interfere with human-machine interface, and/or subjective in the nature, thus are not scalable. Non-contact methods to measure these metrics can eliminate the need for specialized piece of equipment and manual measurements. Non-contact methods can have additional advantages since they are potentially scalable, portable, can be used for continuous measurements, and can be used on patients and workers with varying levels of dexterity and independence, from people with physical impairments, shop-floor workers to infants. In this work, we use face videos and the photoplethysmography (PPG) signals to extract relevant features and build a regression model that can predict pulse rate, and pulse rate variability, and a classification model that can predict force exertion levels of 0%, 50%, and 100% (representing rest, moderate effort, and high effort), thus providing a non-intrusive and scalable approach. Efficient feature extraction has resulted in high accuracy for both the models.</div>

This research focuses on developing a series of mobile learning applications for future 'beyond' classroom learning environments. The thesis describes the general use pattern of the prototype and explores the key factors that could affect users‘ attitudes towards potential acceptance of the mobile learning applications. Finally, this thesis explores the user acceptance of the mobile learning applications; and investigates the mobility issue and the comparison of applying learning activities through mobile learning and e-learning.