Recommender systems (RS) are an efficient tool to reduce information overload when one has an overwhelming choice of resources. Embedding context-awareness into RS is found to increase accuracy and user satisfaction by allowing systems to consider users' current situation (context). Context-aware recommender system (CARS) has applications in various areas, including education, where it can help learners by suggesting learning resources, peers to collaborate with, and more. When CARS is used in a learning context, it adds to the issue of lack of trust in the information, source, and intention as one builds knowledge through it. Further, embedding context-awareness adds to the trust issue due to the additional layer of automated context detection and context interpretation without users' involvement. I investigate how to build trust in CARS in an educational setting. My investigation will be threefold (a) Understanding users' perceptions of CARS; (b) Investigating design interventions to build trust in CARS; (c) Designing and evaluating a multidimensional approach to build trust in CARS.

Educational recommender systems (RS) have become widely popular with the paradigm shift to online learning and the availability of a wide variety of learning resources. Educational RS in various education platforms use a wide variety of filtering techniques. This has led to the development of multiple types of RS. Context-aware recommender systems (CARS) are identified as an emerging type of RS that uses users' context for filtering recommendations, which makes recommendations more relevant to the user's current situation. CARS may face initial distrust compared to other RS due to the additional automation layer of context awareness and the use of more user data. Therefore, we conduct a survey-based study to find differences in user trust and perception between CARS and other RS. In the study, users viewed examples of CARS and RS. The results show that users have significantly lower trust in CARS compared to RS.

Learning through web browsing, often termed Search-as-Learning (SaL), can create information overload, due to thousands of search results. SaL can be made more efficient by developing context-aware tools that recommend items to the user and minimize information overload. However, to use context-aware recommender systems (CARS) users need to trust it. Literature has proposed explanations as a feature that helps to build trust. We investigate the impact of explanation on user trust and user experience for using CARS for SaL. Our study results show that people trust a CARS without explanation more during the first use, but for a CARS with explanations, user trust is significant only after multiple uses. Through interviews, we also uncovered the interesting paradox that even though users do not perceive that explanations add to their learning outcomes, they still prefer to use a CARS with explanations over one without.

Transitioning from the military to the civilian lifestyle, especially for military veterans who decide to pursue careers in the civilian workforce, is often a difficult experience. The job interview, a task in which the interviewees meet and discuss their skills and career goals with strangers in a position of authority, is the first step of assimilation into the civilian workplace, which might cause them to experience nervousness or anxiety. This feeling of excessive stress may compromise the interviewee's performance, therefore potentially impeding their successful transition to the workforce. Intelligent interview training technologies would benefit from automated stress detection systems that could assist interviewees in better understanding causes and antecedents of stressors during their interaction with the interviewer. This paper examines self-reported and bio-behavioral measures of stress experienced during mock job interviews conducted with 24 U.S. military veterans. Self-reported measures were captured via a global measure of stress reported by the participant at the conclusion of the interview, and a continuous moment-to-moment annotation of stress resulting from the retrospective inspection of the interview video recording. Bio-behavioral indices of stress include physiological reactivity measures captured via electrodermal activity and electrocardiogram signals, as well as acoustic measures extracted from speech. Results indicate that physiological reactivity measures exhibit moderate-to-strong correlation with self-reported measures of stress, and can be thus used to estimate the self-reported stress measures. Augmenting the feature space with demographic and psychological traits can further improve the accurate detection of stress during the interviews.

With advances in areas such as sensors and machine learning, wearable technologies will have increased potential to support our daily lives. Even though today’s landscape of smart wearable devices is highly varied, the real-world adoption of wearables has remained lukewarm. We propose that a key reason is that we currently only have a surface-level understanding of people’s interaction behaviors with wearable devices. A deeper understanding of user behaviors toward different wearable devices will help to inform wearable design for more seamless user experiences. We present an empirical study with 50 participants that explore people’s micro-behaviors toward five types of smart wearable devices (wristband, ring, clip, necklace, glasses) in a lab-based information-gathering context. A micro-analysis of participants’ session videos and interviews showed that people have different behaviors and attitudes in terms of affordances and functionality for different forms of wearables giving rise to a variety of design implications.

People use a significant amount of gestures when engaging in creative brainstorming. This is especially typical for creative workers who frequently convey ideas, designs, and stories to team members. These gestures produced during natural conversation contain information that is not necessarily conveyed through speech. This paper investigates the design of a system that uses people's gestures in natural communication contexts to produce external visualizations of their mental imagery, focusing on gestures that describe dimension-related information. While much psycholinguistics research address how gestures relate to the representations of concepts, little HCI work has explored the possibilities of harnessing gestures to support thinking.