HCI International 2023 – Late Breaking Posters

Artificial Intelligence (AI) technologies can act as persuaders when implemented in workplace tools and infrastructure. How users process and react to interacting with features of such AI technologies in the workplace remains ill-understood. Literature in human-AI interaction suggests that cues in the user interface can dictate how users process information communicated by an AI and how receptive they are to being persuaded to change or reinforce their behaviors. Literature from human-AI interaction and an existing systematic framework of the study and design of persuasive technology from human-computer interaction can be applied to examining how users interact with persuasive AI in workplace tools and infrastructure. This paper aims to illustrate the application of such a systematic framework for persuasive technology to the study of persuasive AI technologies in the workplace context. Adapted from the persuasive technology framework, an illustrative vignette of a widely used workplace AI-powered tool is offered to further demonstrate features and principles of systems that include a persuasive AI component.

AI and Deep Learning have led to the development of many tools for healthcare and medicine: image-based diagnostic tools, note-taking aids automatically transcribing speech, medical risk assessment and decision-support applications based on the patient parameters stored within Electronic Health Record (EHR) systems. The astonishing success of Large Language Model-based generative AI has further demonstrated a great potential for employing AI-based tools in many domains of human activity, including healthcare and public health. At this point, the majority of AI applications in healthcare are tools that work autonomously, and the physicians and medical personnel are called to use them as an input in their decision making outside of their use of EHR systems. We discuss the opportunities and challenges of employing AI-based capabilities within EHRs and outline a research roadmap for creating interactive collaborative integrated EHR assistant applications. We discuss parallels with our prior work in addressing usability of enterprise resource planning systems.

This paper describes an approach to integrating Design-Thinking (DT) and User-Centered Design Process (UCD) activities into a process that is facilitated by artificial intelligence (AI) for improved collaboration and data-driven decision-making at a faster pace, so as to improve the adoption. It also aims to identify if the AI can facilitate design thinking sessions and act as a collaborator to help the participants make decisions based on data faster. The proposed concept has been tested by developing an AI powered whiteboard software using Open AI’s APIs and a custom ML model on user-profile data to manage it, which was then run by a group of users for their Design-Thinking session for testing and accessing its success in enhancing the design process. The AI-facilitated design-thinking process produced desirable outcomes in significantly less time and helped speed up the Design-Thinking process.

This work demonstrates the variety of possible actions and reactions in the interplay of humans and AI when working on shared tasks. The coordination of human-AI task sharing has to take these varieties into account. In many instances the modes of reacting on each other are symmetrically distributed between humans and AI. Certain activities of coordinating the task sharing between human and AI are in a meta-relation to the interaction between humans and AI, some of these meta-activities may be reserved to humans.

The goal of eXplainable AI (XAI) is to increase the transparency and trustworthiness of AI algorithms to humans by clarifying their internal decision-making process. As AI technology continues to permeate various aspects of our daily lives, including the workplace, the importance of XAI has grown and some XAI methodologies have increased our understanding of AI algorithm’s inner logic. However, the achievement of explainability of AI models did not make them user-centered in real life and users still need to be aware of the internal mechanism of AI models to interact with them effectively. In this regard, the paper argues there is a need to move beyond XAI towards Human-Centered eXplainable AI (HCXAI) to ensure human-centered usage in our daily lives. As the steps heading for HCXAI, the paper suggests researching intuitive human behaviors first and then training AI models to comprehend these natural human behaviors. When users interact with AI models trained in this way, they will not need to think consciously about the model’s working mechanism anymore. The paper elaborates on this approach further with a practical case that can be encountered in our daily lives; the vision and language model for referring expression comprehension under the personal context.

This short research paper explores the ways in which ideas, theories, and techniques from the realm of safety research can be utilized in Human-Computer Interaction (HCI), specifically regarding the ethical aspects of artificial intelligence. Decades of dedicated effort have been invested in the domains of aviation, healthcare, and nuclear industry to enhance accident investigation and prevention methodologies. Safety-related issues can be considered as a subset of a significantly larger and more multidimensional group of ethical issues. Ultimately, both safety and ethical reflections are about avoiding unwanted consequences. A systems-theoretic causal model is one of the many theoretical approaches used in safety research. Systems-theoretic thinking that expands linear reasoning originates from engineering. By applying the systems-theoretic causal model it is possible to create causal mechanisms for explaining empirical sociotechnical phenomena. This makes it possible to perceive different types of interaction mechanisms at the system level and to compare different situations and system structures with each other. A potential benefit systems-theoretic modelling of ethical action is its ability to combine the terminology of engineering, moral philosophy and social sciences into the same framework. This allows for the development of new interdisciplinary research methods for the systematic analysis and design of ethical AI systems.

Large Language Models (LLMs) have significantly advanced the field of Natural Language Processing (NLP), demonstrating exceptional performance across diverse language tasks such as content summarization, sentiment analysis, and conversational AI. The advent of these models has profoundly impacted human-computer interaction research, marking the onset of a new era in the field. The pioneering model, ChatGPT, was introduced by OpenAI in November 2022, catalyzing the development of other commercial tools like Bing Chat and Google Bard. Subsequently, the emergence of open-source LLMs has democratized access to these powerful tools, enabling end-users to deploy them internally with relative ease. As the landscape of open-source LLMs continues to expand and evolve, researchers and practitioners are presented with a plethora of choices for NLP applications. This paper presents a comparative analysis for various open-source LLMs, assessing their unique features, strengths, and limitations. Our focus of comparisons will include licensing, training methods, computing resources needed, available application programming interfaces (APIs), robustness, and bias.

Text Style Transfer is a task concerned with modifying the attributes of a text while leaving its meaning unchanged. In recent years, this task has gained much attention due to the promising performance of deep learning models. However, progress is still slow due to an overreliance on large datasets and a lack of reliable evaluation methodology. To address the data challenge, the present research proposes the usage of prompting techniques which have shown very good performances on other natural language generation tasks in few-shot settings. Furthermore, considering the lack of conventional evaluation methods, this study explores various methods employed in previous research. This includes automatic metrics, but also human evaluation. Prompting outperforms fine-tuning according to computed metrics and to human assessment, but there seems to be little correlation between the human judgement and automatic metrics assessing then meaning conservation and the fluency of generated text.

Solving problems by human-AI configurations will likely become a pervasive practice. Traditional models of task allocation between human and machine must be revisited in light of the differences in the learning of humans versus intelligent machines; performance can no longer be the sole criterion for task allocation. We offer a new procedure for allocating tasks dynamically that begins with the determination of the desired level of machine autonomy.

Deep learning allows machines to have human-like learning and thinking capabilities, but its internals are often considered a black box about which we know very little. As AI (Artificial intelligence) penetrates deeper into life, the development of AI is maybe uncontrollable under the influence of human data. Questions about privacy, autonomy, fairness, and the potential for misuse of technology will become even more pressing.In the unstoppable process of technological development, art and design can often reveal science’s mysteries and even predict technology’s development. In this paper, through the research of algorithmic bias and AI ethics, we design an interactive installation with audience participation. Based on voice dialogue technology, the Installation uses visualization based on two-dimensional screens and physical Installation to demonstrate how AI constantly changes and generates a certain form in the learning process, hoping to trigger thinking about the subjectivity between humans and artificial intelligence. It summarizes what AI believes audiences think about AI on some issues.

This article explores the impact of artificial intelligence (AI) on human identity and raises ethical concerns regarding using biometric data for identification. As AI advances, digital avatars become more authentic and intelligent, leading to questions about human identity and its societal effects. The integration of 5G technology and the Internet of Things accelerates data transmission and storage, resulting in the convergence of the natural and virtual realms. The article proposes an interactive art installation called Human′ to address these issues. It employs AI to create digital avatars based on facial, social media, and personal data, enabling interaction with the audience through mobile devices. Human′ prompts reflections on ethical dilemmas arising from the fusion of human and digital identities. The installation involves design, prototyping, and development phases, including motion capture, interface design, front-end and back-end development, and integration with a WeChat applet. Exhibited in May 2021, the installation received positive feedback from the audience, who expressed curiosity about the future possibilities of AI while raising concerns about the societal impact of digital avatars.

Artificial intelligence, empowered by computing, has surpassed auxiliary design tools, fostering deep collaboration between human-machine intelligence in contemporary design. Advanced computing technologies and design trends shape the integration of design tools in the information age, influencing design methodology. The developmental history of design tools demonstrates that it undergoes four stages. This article analyzes the impact of calculation on design by studying tools, products, methods, processes, and concepts in each stage, exploring the evolving relationship between design and computation. Designers divide the transformation of the design-computation relationship, which progressively permeates design, transitioning from digitization and automation to parameterization and intelligence. It evolves from improving efficiency to assuming both subject and object roles in the design process and becoming an active partner in collaborative design.

This review paper presents an exploration of the deep learning-based music generation literature, designed to offer undergraduate researchers an initiation into the field. This study illustrates prevailing generative models and datasets currently influential in music generation. Four publications have been selected for detailed discussion, representing a spectrum of salient concepts in music generation and potential areas of further inquiry. By focusing on key studies and significant datasets, this review aspires to serve as a guide for undergraduate scholars keen on investigating the intersections of deep learning and music generation.

In this paper we are describing our preliminary studies and prototypes for evaluating a surfer’s performance on a stationary wave. We are briefly describing the sport’s environment and development and the current state-of-the-art of machine learning based single camera tracking approaches we have evaluated and applied. The main part of the paper deals with the first implementation, the tracking and the results of the evaluation of the movements of two surfers. We are closing with our lessons learned and our next steps.

Motion blur presents a significant challenge in visual SLAM. Since a satisfying performance highly relies on clear and feature-rich images, the system will easily fail to extract features and lose track due to motion blur. To address this issue, this study introduces a novel solution that estimates the camera motion using blur features extracted from images. The proposed approach models the relationship between camera motion and motion blur, creating a comprehensive motion blur dataset labeled with camera motion. By decoupling the motion estimation process into predicting magnitude and direction separately, a neural network is trained using blur and depth images as inputs to output the camera motion. Experimental results demonstrate that the proposed method successfully estimates motion from blur, even in long-term blur scenarios. This method can potentially serve as an auxiliary motion estimation module to enhance the robustness and accuracy of the visual odometer when motion blur is encountered.

Situation awareness has become a popular method for modelling human–automated process allocation and, more recently, applied to model collaboration between humans and autonomous artificially intelligent agents. However, the role of how to distribute decision-making within this new collaborative paradigm remains underexplored. In this manuscript paper, we propose an integration of situation awareness into the decision-making process by combining the Endsley Situation Awareness model with contextually tailored sequences of cognitive activity referred to as end-to-end decision-making loops. We posit that this approach benefits designers, implementers, and users of intelligent systems as it orchestrates a cognitive and perceptual alignment between human and non-human agents, thereby enabling the formation of meaningful end-to-end action loops.

Live streaming has become a prevalent form of online entertainment and commerce, where real-time interactions occur between streamers and their audiences. Currently, streamer assistants have some shortcomings in terms of personality and emotional expression. These shortcomings undermine the live streaming effect and audience experience, thereby damaging the streamer’s popularity and income. In this paper, we present the Intelligent Streamer Assistant with Personality, Emotion, and Memory (ISAPEM) framework, which aims to utilize playful animal avatars to establish a sense of belonging and trust proactively with the audience. Firstly, we determine the assistant’s personality. Subsequently, the assistant determines its emotions according to its personality and the danmaku (bullet chats/comments) context analysis, ranging from trust and joy to sadness. Next, the assistant displays matched expressions and actions, and then generates consistent dialogue using large language models (LLMs). For example, when faced with a challenging question in the danmaku, the assistant might appear perplexed, then reach for the corresponding danmaku, catch it, and swallow it. Finally, the assistant stores and analyzes danmaku interaction data to remember and understand the audience’s needs and preferences. Preliminary experimental findings indicate that the ISAPEM framework can create a warmer experience for the audiences and enhance their willingness to interact, which has the potential to foster a sense of belonging and trust among the audiences. This study proposes a novel design framework for streamer assistants that integrates cutting-edge anthropomorphic design cues (ADCs) with a danmaku-based physical interaction mode, expanding the application and interaction modes of novel ADCs and LLMs.

The coexistence of humans and Autonomous Mobile Robots (AMRs) in intralogistics is a growing reality. To enhance the usability of their interactions, AMRs can communicate their future trajectory to humans. This communication can be either implicit through their driving behavior, or explicit through additional signaling. We conducted a real-world participant study with 32 participants and a robot to compare two different communication tools: a floor projection as an explicit tool and a specific driving behavior as an implicit tool. We tested them in three scenarios: intersection, crossing, and bottleneck. We measured the interaction’s efficiency, legibility, and trust using quantitative data and questionnaires. We also asked participants to draw the expected trajectories of the AMR at the time of interaction. Our results showed no significant difference in the interaction time between the two communication tools. However, explicit communication increased the trust in the AMR and was perceived more easily by humans. On the other hand, explicit communication is more prone to misinterpretation by humans. Therefore, the design of explicit communication is crucial. The implemented implicit communication does not seem suitable for narrow corridor-like environments.

This work defines and delineates the concept of “EDGE Capabilities”, an acronym for Emergent, Dynamic, Global, and Evolutionary, as a way forward to development, deployment and command of Human-Autonomy Systems and capabilities. The concept emphasizes the coordination of kinetic and non-kinetic means of power across physical, information, and cognitive dimensions, can enable asymmetric exploitation of vulnerabilities and render irrelevant an opponent’s abilities, will, structures, and systems.The field of Human-Autonomy Systems (HAS) deals with how humans and automated intelligent artificial systems can work together to solve complicated tasks. This is particularly relevant in the domain of cyber-defence, where the speed and efficiency of response is of paramount importance. One solution that has been proposed is based on the Joint Cognitive Systems (JCS) body of research, defining and studying systems which are capable of jointly and independently detecting, identifying and engaging kinetic and non-kinetic threats and actors across physical, information, and cognitive dimensions.We identified five key characteristics of EDGE capabilities. First, they are highly dynamic and non-linear. Second, they require collaboration within and between different organizations and their cultures. This includes the third point, that they engage people with different backgrounds. The fourth characteristic is that EDGE capabilities are superior regarding managing and maintaining mission-critical availability, versatility and efficiency. Fifth and last, they are employed in a multi-domain perspective. EDGE capabilities can be characterized as emergent, enabling deployment with a high level of fluidity and flexibility, matching the variation of the operational environment.Every commander and every human and artificial agent must develop a capability for collective sensemaking and interaction to enable a comprehensive detailed system insight, leading to safe and efficient mission accomplishment. We propose formulating a future-oriented essence of EDGE Cognitive Command, with equal relevance for all agents constituting a Human-Autonomy System.

Intelligent systems support decision-makers in complex tasks by utilizing advanced technologies such as machine learning, artificial intelligence, or deep learning. Despite their potential to increase decision performance, several barriers exist to the widespread adoption of these complex systems. By examining the past IS adoption literature and the unique characteristics of intelligent systems, we propose the situation awareness model as a promising theoretical framework that can shed light on human-AI interaction effectiveness. Using this vantage point and our preliminary interview data with expert decision-makers, we identify several challenges to be addressed for effective human-AI collaborations.

Robots are increasingly being employed in educational environments to enhance learning experiences. Adolescents may interact with robots differently from younger children due to cognitive and perceptual maturation. This study examines the influence of robot ability, task complexity, risk, and self-construal on adolescents’ confidence and trust in robots, and decision-making. Six participants (aged 14–16) collaborated with the NAO robot in a length judgment task. Results showed that high-ability robots elicited more trust, confidence, and decision change. In complex tasks, trust was highest with low-ability robots, while trust was lowest in simple tasks. Participants were less inclined to share benefits but expect robots to bear losses. Interdependent individuals showed more decision changes. These findings increase the understanding of the way that adolescents collaborate with robots, especially in decision-making processes.

It is said that humans make their first impressions of a person 7% by verbal communication and 93% by visual and auditory communication. When robots intervene in human society and need to communicate with humans, the face is considered to be the most important interface. However, it is unclear whether differences can be recognized in the subtle facial expressions that appear in daily life. Therefore, in this study, we investigate the synchronization of emotions for knowing whether we can recognize such micro expressions. Emotional synchronization means that the feelings of a person A and a person B are communicated and emotionally coupled. For example, when the person A likes the person B, the person B feels the same emotion toward the person A at the same time, they are emotionally synchronized. We believe that the study of emotional synchronization will be useful for designing more friendly robots.

Experimental operations have always been a necessary link in popular science education. However, due to regional differences in educational resources, difficulty in operation for novice experimenters, the danger of some experiments, and the scarcity of consumables, virtual experiments for students are expected by the public. Based on flexible microfluidic sensing technology, this paper proposes an AR glove with a subtle force feedback effect, specially used for virtual experiments and teaching. In order to provide more accurate force feedback to the hand skin, we entered university laboratories, observed and analysed the standard hand movements of laboratory operators to collect data on force feedback points, and integrated them into the design of microfluidic AR gloves. In the end, 13 necessary sensing parts for each hand were obtained, which were integrated into the design to achieve a more accurate experimental force feedback effect.

In recent years, educational philosophies have slowly begun shifting to focus on differentiated and self-learning systems. In this regard, creating opportunities to further self-learning resources has become increasingly important. The creation of such platforms or opportunities for physical education, however, proves to be more difficult as individuals require continuous and precise feedback regarding the usage of their bodies. Accordingly, we have developed an augmented reality application that presents a platform for dance that focuses on differentiated and self-learning principles with accurate feedback. We built the AR app using the Swift programming language and used the Core ML action classification model to capture the body position. We also recognize the importance of social interaction in learning, such as its benefits to motivation or peer-to-peer support. As a result, we designed the app to allow its users to connect with their friends globally and dance together. We do this by converting the captured poses of various users from different locations during a live session and instantaneously converting them into AR avatars. This way, the users can dance with their friends live, emphasizing the undeniably social component of dance and learning while going through the learning process at their own pace.

Unlike connections of physical spaces typically observed through movement trajectory, connections of virtual spaces can be observed in high resolution through vision-based trajectory. Studies have explored user movements and their diffusion in virtual environments (VEs). However, frameworks are needed to objectively measure vision-based behaviors in VEs and characterize spatial networks. This paper proposes a spatial mapping approach for user data through visual-spatial attention, using cognitive psychology and spatial ecology concepts. The aim was to demonstrate a computational tool that can mathematically describe the connectivity of locations based on the trajectory of user attention. We first implemented VEs to represent specific places, which were then sampled into cells, collected user location and direction data within the VEs and calculated attention areas using the direction vector, constructed spatial graphs by creating links among cells within the attention area, and finally calculated the centrality of nodes within each spatial graph and performed community detection. The tool was tested on log data from two user studies on VEs. In the results, the centrality indicated the cells where user attention was focused, and the community detection identified cell clusters. By analyzing the features of cells with high centralities, such as buildings, lakes, and non-player characters, we can identify cells with similar features that will attract attention in VEs. Using the proposed tool, a quantitative description of attention was obtained without direct feedback from users. Practically, the spatial graph can provide guidelines for designing specific areas to attract attention and help managers cluster cells for management.

This study aims to examine the differences in User Experience (UX) in metaverse model houses, specifically in terms of interactive, informative, and sensorial experiences. After experiencing the metaverse model house on a tablet PC, 83 users were surveyed. A one-way ANOVA was performed to confirm the statistically significant differences by demographic characteristics. Consequently, a significant difference was identified in the sensorial and informative experiences between users in their 30s and those of other ages. Additionally, users in their 20s, 40s, and 50s used different methods of investigating information compared with users in their 30s. Users in their 20s were more familiar with the metaverse environment and comfortable exploring virtual apartment complexes. However, users in their 30s missed most of the content in the video clips and were less satisfied with the experience than the other age groups. Therefore, to improve the positive aspects of UX, the metaverse model house should be considered to provide better graphic elements, deliver information through various methods, and design intuitive environments considering the user's age.

We design a product concept named Hypersphere - an XR wearable that integrates neuro reality and virtual reality based on electroencephalogram (EEG) systems and head-mounted display (HMD). It aims to provide users with a more immersive, engaging, fluid, and safe experience by synthesizing visual, auditory, and haptic systems. In this paper, the design of the Hypersphere, including its structure, materials, function, and authentication process is discussed. The research contributes to the current understanding of XR wearable design concerning user experience in a synthetic world.

Users of virtual reality (VR) applications often must travel between locations in virtual spaces. An omnidirectional treadmill is a device that allows users to walk in all directions in a virtual space without physically moving the position of their body in the real world. The obvious advantage of an omnidirectional treadmill is that users can walk large distances in VR without concern they will collide with objects in their real physical space. Much research investigating the use of omnidirectional treadmills in VR has focused on comparing treadmills to other forms of navigation, and exploring various aspects of the user experience, such as symptoms of cybersickness. This paper describes our work in progress to develop and evaluate a VR application for the Infinadeck omnidirectional treadmill. Our goals are to create a VR application highlighting the benefits of the treadmill, develop the application in stages so that many students at our University can contribute, and increase understanding of how to develop pleasurable omnidirectional treadmill experiences.

New students often have trouble finding classrooms, laboratories, libraries, or other places inside a new study place. At Østfold University College (HIØF), they provide paper maps to the students to find locations inside the university which are difficult to understand and hard to use. To solve this problem, we have developed an app named “HIØF Easy Navigator”. The application guides new students to find the inside location here at HIØF, Norway. We also conducted a survey of eight students to check the user experience and opinions of paper maps and the HIØF Easy Navigator. Our findings demonstrate that the app is easier to use than paper maps to locate locations inside the university.

In early 2022, the UX Research team from Sidia’s Research and Development Institute, located at Manaus/Brazil, received a requirement from one of its projects to conduct a comparative study on the usage of different Software Development Kits (SDKs) for Mixed Reality (XR) applications. The study aims to compile developer’s necessities and difficulties while using common available SDKs in the market. This qualitative study is about the core SDK values pointed out by participants related to the development of XR applications. We have selected developers based on availability to conduct a study for a month. With the help of Microsoft Office Forms, we have structured a series of surveys about the main usage activities of five important SDK tools to develop XR experiences. The selected SDKs were: ARCore, AR Foundation, Vuforia, Spark AR (Scripting API), and A-Frame. The surveys have been structured obeying the flow of activities of an XR application development. The defined flow, based on interview with the candidates, consisted of main phases: Installation e Update, Development, Tests and Deploy. The survey, compiled in a detailed report, exposes the collected data and share some of the participant’s opinion to provide context and exemplify the exposed use cases. In this report, we observe that the characteristics and needs of the participants varies according to which phase of the activity flow they are executing. For some stages, users demanded agility and frictionless processes. At other times, the flexibility and customization of the SDK behaviors were widely punctuated by participants.

Digital accessibility is vital for ensuring equal access and usability for individuals with disabilities. However, addressing the unique challenges faced by individuals with disabilities in XR environments remains a complex task. This paper presents an ongoing accessibility framework designed to empower developers in creating inclusive XR applications. The framework aims to provide a comprehensive solution, addressing the needs of individuals with disabilities, by incorporating various accessibility features, based on XR accessibility guidelines, best practices, and state of the art approaches. The current version of the framework has focused on the accessibility of XR environments for blind or partially sighted users, enhancing their interaction with text, images, videos, and 3D artefacts. The proposed work lays the foundation for Extended Reality (XR) developers to easily encompass accessible assets. In this respect, it offers customizable text settings, alternative visual content text, and multiple user interaction control mechanisms. Furthermore, it includes features such as edge enhancement, interactive element descriptions with dynamic widgets, scanning for navigation, and foreground positioning of active objects. The framework also supports scene adaptations upon user demand to cater to specific visual needs.

The aim of the project Symotiv ( http://symotiv.de ) is to introduce people to classical music through interaction with a virtual orchestra. With the help of Virtual Reality, it does not only enable to immerse oneself in a three-dimensional, virtual concert hall, but also to experience images and sound from the perspective of a musician. This opened the possibility of experiencing and understanding what is happening from new spatial, but above all sonic perspectives. Therefore, we developed a workflow to capture the motions and sound of 50 musicians and the conductor by using a camera-based Machine Learning approach.

The purpose of the research is to determine the electronic word of mouth of a company dedicated to the sale of pharmaceutical products and toiletries, whose applied model was the one proposed by Matute, Polo & Utrilla (2015), the sample was 384 people surveyed, by what the approach applied was the quantitative and descriptive level, obtaining as a result a total score of 5.62 which expresses that the knowledge of the service through the eWOM is influencing the purchase decision and the amount of feedback of the eWOM applied in the company, so it is concluded that the eWOM has become a crucial component of the website, serving as its business card, this is due to the reviews and information provided by customers who have used the product in the past.

The purpose of the research is to determine the influence of social identity on electronic word of mouth (eWOM) of restaurant customers on Facebook, where the model proposed by Gonzáles, Marquina & Rodríguez (2020) who study identity variables was applied. Social and eWOM, applying the instrument that is the survey to 200 people who have a Facebook account, being evaluated by a 5-point Likert scale, for which the approach is quantitative, applied type, explanatory level and non-experimental design - cross-sectional, resulting in that the social identity has a significance of 0.000, influencing the eWOM, therefore, it is concluded that customers can interact with others on Facebook, who suggest trying new things, which arouses their curiosity about other restaurants through from the comments made by users, which are similar to the experience they had after eating in a restaurant. Because customers cooperated to show that they had a successful consumption, this encourages them to eat and comment on their dining experiences.

In this work, we compare the picture quality provided by two popular smartphone brands: Samsung and Apple. We use the top-end model of both brands in 2022: Galaxy S22 Ultra for Samsung and iPhone 13 Pro for Apple, and we built a dataset consisting of seven different environments. For each of these environments, we captured pictures with both devices under the same conditions in two camera categories: first, using the device’s native camera and then using the camera provided by the Instagram App. Since these pictures are in a compressed form, and therefore there is no reference image to compare them to, we use “No-Reference Image Quality Assessment” (NR-IQA) metrics to provide values to measure the quality of its pictures. Experimental results with some of the most used NR-IQA metrics show that Samsung device has best images for most of the environments when using the device’s native camera, but Apple provides best results for the Instagram App.

Recently, platforms that service digital contents such as movies, dramas, and webtoons are switching from downloading methods to streaming methods, and illegal content servicing methods are also changing from downloading to streaming. These streaming-based copyright infringement crimes avoid the blocking and investigation by law enforcement agencies through methods such as using overseas servers or disguising the IP of operating servers using cloud services, etc., which requires an appropriate measures to respond to them. In this paper, we propose a signature information generation method through which the illegality of distributed digital contents can be quickly checked. Using this method, there is an advantage in quickly finding illegally distributed digital contents based on a cloud server, which can prompt to block illegal distribution sites.

Food design is an emerging field of research in the past decade, with interdisciplinary attributes that integrate culture, ecology, health, and society. Especially in the post-epidemic and post-carbon context, the importance of food design as a pathway to sustainable development has been gradually highlighted. By summarizing the food design research from the perspectives of eating, cooking, experience design of food, as well as ecology, agriculture, and system design of food, the article defines the research scope of sustainable food design. Based on this, the author proposes a four-dimensional transformation theory of food design in the post-carbon context and analyzes how to translate strategy into action with case studies. The four-dimensional transformation are from Human-Centered to Life-Centered, from Object-Centered to Hyper-object Centered, from Experience-Economy to Post-Carbon Economy, from design for consumption to design for Crisis. In this way, the article builds a new framework and guideline for sustainable food design and propose strategies to deal with the crisis of the times.

This paper introduces the concept of an intelligent greenhouse and intelligent seedbed as innovative solutions to enhance precision agriculture in small-scale cultivation settings. The core motivation behind this project lies in the need to improve cultivation practices, by reducing resource input and adopting an overall more efficient approach, while also making the system accessible both locally as well as remotely, through a variety of interfaces, to different stakeholders. While smart agriculture solutions already exist for large-scale farming, this paper has its main focus on the specific context of greenhouse cultivation in relatively smaller, scattered plots, as commonly found in Crete; the proposed intelligent greenhouse and seedbed aim to provide automated monitoring, customized cultivation support, and accessible information to various stakeholders, regardless of their level of expertise or professional involvement in farming. Despite some technical challenges, initial tests with both systems yielded promising results, helping to effectively monitor and manage the cultivation process; furthermore, users with varying levels of agricultural knowledge found the system generally easy to control. Some of the challenges faced include maintaining sub-system reliability, interference from external factors, as well as sensor limitations.

Government applications and technologies enabled by cloud computing, mobile computing, and the Internet of Things have resulted in an increase in the volume of data that must be processed, stored, and transferred. The complexity of data centers can be difficult to comprehend, but comprehension is critical to their success. The investigation employs a qualitative approach, with graphs and tables visualized using the NVivo software tool, to detect the reconfiguration of each government actor involved in the implementation of the Electronic-Based Government System. To achieve an integrated and comprehensive Electronic-Based Government System (SPBE), Central Agencies and Local Governments must work together closely. Cloud computing enables data center infrastructure to be supplied in an adaptable, load-sensitive, and integrated manner.

Wish is an online shopping platform that differs by offering a “discovery based” shopping experience. The majority of the transactions on Wish originate without a search query, which is different from most ecommerce platforms as they are search focused. Through insights gathered via qualitative and quantitative research conducted with Wish buyers, the authors summarize in this paper four main factors that influenced Wish users to embrace the discovery based shopping experience on Wish - mobile optimization, gamified deals, personalized content, and unique inventory.

Nowadays, a company’s website has become one of the typical touchpoints of interaction enabling customers to gain more valuable knowledge about the provided products and services. However, websites tend to contain an abundant amount of information that might overwhelm customers and disturb their willingness to purchase. One of the solutions to facilitate the search process on the website is an Online Recommendation Agent (RA). As a result, this study employs user usability testing and the technology acceptance model (TAM) to investigate the user perception and acceptance of the online RA for enhanced product search on the Danish manufacturing company’s webpage. The study results present factors influencing consumers’ intention to adopt online RA, and challenges the company has to tackle for successful integration of the product finder to grant customers a better tailored and consistent shopping experience.

In the digital economy, companies increasingly shift from physical to online marketing, leveraging the ability to transcend geographical boundaries and rapidly disseminate brand information. As a powerful online communication platform with a vast user base, WeChat plays a crucial role in mobile payment, information transfer, and emotional communication through Wechat Red Packets. When consumers engage with Wechat red envelope covers, they undergo five stages of attention, interest, search, action, and sharing, enabling the secondary communication of brand information. This study analyzes the communication influence mechanism of brand information in Wechat red envelope covers using the AISAS model. It also identifies factors influencing consumers' willingness to use Wechat red envelope covers through a survey, establishes a model for understanding these factors, and examines their impact on users' behavior from a brand communication perspective. Hypotheses are proposed based on the model and tested using PLS 3.0 software. The findings highlight the significant and positive influence of entertainment experience, effort expectation, social influence, convenience, intention to use, and user behavior on the information experience. Wechat red envelope covers hold substantial business value as a marketing method. Leveraging the festive season as a marketing hotspot and consistently conveying the brand message while sending and receiving Wechat red envelope covers will enhance brand awareness and influence.

Introduction: Video analytics has become an important tool for improving business performance, personnel management, and security in different environments, including shopping centres. However, its use also poses a threat to users’ privacy. Aim: In this context the aim of the article is to determine the architecture of video analytics in shopping malls as a business marketing strategy and the vulnerability of users’ privacy. Method: A descriptive methodology with a qualitative approach was used to design a prototype focused on the optimisation of customer business processes and decision making in shopping centres in Ecuador, from problem identification and information requirements management to prototype design, implementation, and training of the model with accurate data. Results: The technological solution is based on the OMIA platform, which generates relevant Key Performance Index (KPI) based on the automatic processing of images from video surveillance systems. Conclusion: It is important that the implementation of the model is trained, supervised, optimised, and regulated to guarantee the privacy and security of information and users under the Organic Law on Data Protection in Ecuador.

E-textiles involve seamlessly embedding electronic components into fabric, enabling textiles to sense, react, and communicate with the environment. However, the development of embedded textiles requires collaboration across multidisciplinary research fields. This study aims to address the need for large-scale textile-electronic research and development methods that can be applied in the textile industry. By bridging the gap between industries, this research aims to facilitate the creation of innovative e-textiles by showing the potential of technical embroidery, lamination, conductive screen-printing and modular, flexible electronic components through the showcase of four demonstrators.

Despite the increasing popularity of the online Customer-to-Customer (C2C) second-hand market, factors that affect consumers’ willingness to engaged in online C2C second-hand transactions are not fully investigated. Therefore, we take an approach to identify the factors influencing consumers’ intention to purchase on C2C second-hand platforms in China, verifying the applicability of existing achievements in China's second-hand trading context and exploring new factors based on the characteristics of C2C second-hand transactions. Based on the literature review and characteristics of C2C second-hand transactions, eight consumer factors were introduced into the research model: disposition to trust, familiarity with buying, familiarity with selling, frugality, environmentalism, dematerialism, fashion consciousness and hygiene consciousness. Among them, familiarity with selling and hygiene consciousness are new research variables that never appeared in previous studies. A questionnaire containing 38 questions with 10 constructs and demographic data was designed and distributed to 425 participants, where 377 valid responses were finally obtained. The results provided evidence that consumers’ familiarity with buying, frugality, dematerialize, fashion consciousness and hygiene consciousness all significantly affected their intentions to purchase in C2C second-hand transactions. Also consumers’ familiarity with selling and disposition to trust had positive impacts on purchase intention through the mediation effect of trust. Another important finding revealed that the impact of environmentalism on second-hand purchase intention was not significant, showing a difference in the role of environmentalism in China and developed countries. The findings provided insight into marketing and design strategies for C2C second-hand platforms to promote C2C second-hand consumption.

This paper addresses the global issues of food scarcity and ecological imbalances, resulting in hunger, malnutrition, and health complications. To foster sustainable development and harmonize human-ecological balance, an interdisciplinary approach utilizing web3.0 technology is adopted to promote the integration of agriculture and urban environments. The research explores the web3.0 technology to select crops and plans manual work by utilizing the internet of things and blockchain. This facilitates real-time monitoring and recording of environmental, climate, and soil information, allowing for the selection of suitable crops in specific areas, improving agricultural efficiency, and ensuring food safety. The study also investigates the destination of agricultural products, selling them directly to the public using web3.0 technology to provide green ecological food while ensuring the income source of low-income farmers. This strategy promotes economic growth and achieves a virtuous circle of economy and ecology. In addition, the study explores the recycling of agricultural plants, aiming to identify the most optimal recycling program through the utilization of web3.0 technology. By implementing this approach, we can facilitate the sustainable development of agriculture and rural areas while mitigating the impact on the environment. At last, a sustainability model based on trust mechanisms, cooperation and benefit mechanisms, and coordination mechanisms is developed, mobilizing multiple sectors to ensure sustainable integration of Web3.0 agriculture and ecological systems for economic growth and ecological balance. To effectively address the food and ecological crisis, the study puts forward a web3.0 technology-assisted sustainability model and multi-participation sustainability model based on web3.0 technology, which can promote the benign interaction between cities and agriculture and build a more beautiful and livable city, and im-prove food security and environmental preservation.

Identifying individual driving strategies often relies on theoretical task models, arbitrary group divisions, or somewhat untransparent evaluations by instructors. We propose using cluster analysis as an exploratory, data-driven approach to categorize drivers based on their driving and scanning behavior. Therefore, we analyzed a combination of variables regarding longitudinal vehicle guidance, lateral vehicle guidance, and gaze behavior when approaching an intersection. Data stemmed from a driving simulator study including drivers with normal vision, simulated, and pathological visual field loss. They performed 32 intersections that varied concerning complexity and the availability of an auditory scanning assistant. The total sample comprised 2145 data points. K-means on two dimensions of a prior Principal Component Analysis yielded the best results with two clusters that can be interpreted as high acter and low acter, referring to the extent and earliness of gaze shifts as well as the duration of the intersection approach. These two strategy clusters were rated based on performance criteria to check the effectiveness of these strategies for the different driver groups and situations. While high acters were more frequent under complex conditions, this strategy failed more frequently in these cases. Future developments for this promising approach to cluster strategies in driving-related areas are discussed.

Driving cessation in old age is linked to negative mental and physical health outcomes. Autonomous vehicles offer an opportunity to maintain independence and mobility for older adults. However, vehicle automation can compromise driver engagement, posing risks to drivers and the public. The concern is particularly significant for older and novice drivers who are already overrepresented in accident statistics. Driving simulators provide a safe environment to evaluate performance and assess driver skills and engagement. This study employed a specialized protocol to accommodate older drivers and minimize the adverse effects of simulated driving. The complete number of participants included older (n = 13) and novice drivers (n = 9), who underwent manual and semi-autonomous driving simulations. Results included positive impacts of the simulator modifications with only two dropouts (both from the older group) resulting in a total dropout rate of 9% (15% of older drivers). Driver engagement varied between driving conditions, with superior performance in responding to surprise events during manual driving condition.

As part of our research on the perception of risk by aircraft mechanics, our objective is to evaluate how they perceive risk when exposed to different types of hazards. For this purpose, we chose to construct a taxonomy of hazards which will allows us to: 1) Meet the requirements of the International Civil Aviation Organization (ICAO), which mentions the importance of identifying hazards before assessing risks [1]. This step is also crucial to address risk perception [2]. 2) Be able to study the impact of a hazard category on mechanics’ risk perception.First, two concepts must be distinguished: hazard and risk. Risk refers mainly to the probability and severity of consequences [3]. Hazard is a subjective property of an object or situation [4]. A hazard becomes a risk for the person who is exposed to it [5]. In our study, we consider the hazard as: “Anything that can injure the mechanic and/or damage the aircraft and is a result of the design of the aircraft, the GSEs and tools”. This in order to be able, based on a literature review, to list the hazards, to select those that meet the definition, and finally to classify the hazards into categories and create a taxonomy.

At SAE Level 4, the Automated Driving System (ADS) handles all driving tasks, including emergencies, within a defined Operational Design Domain. This shift in driving responsibility has enabled the development of Unmanned Automated Taxi (UAT) services that can operate without a human driver. Since UAT services rely on system-user interactions, the role of the Human-Machine Interface (HMI) is crucial to improve the user experience. Therefore, this study aims to derive the mobile phone HMI (Human-Machine Interaction) interface for improving the UAT user experience by creating use cases in UAT services and proposing mobile phone User Interface (UI) examples. First, we created the user interaction use cases in the seven stages of the UAT service user journey (Lim and Hwangbo, 2021). And we categorized the information that can be provided in each use case based on the In-Vehicle-Information (IVI) classification criteria (Choi, 2017). A UAT mobile phone UI example was designed using the classified information. These use cases and UI examples can be utilized for future HMI interface research to improve UAT user reliability and satisfaction. In future studies, we plan to evaluate the effectiveness of the proposed UI using vehicle simulator-based experiments.

This study aims to develop a comprehensive model that accurately represents the perceptual system of drivers and provides insights into their behavior. Existing models, namely the Skill-Rule-Knowledge model and the dynamic driving task model, were integrated. This integrated model successfully elucidates the sequential processes involved in sensation reception, recognition, judgment, and control. Furthermore, the study investigates the drivers’ behavior during takeover situations. Specifically, the model’s sequence was determined by considering two distinct scenarios: unplanned operational design domain (ODD) exit resulting from automated driving system (ADS) failure and planned ODD exit during highway driving. Empirical data was gathered using a simulator with 36 participants in the ADS failure situation and 40 participants in the highway exit situation. Two takeover situations were compared based on drivers’ perception time and perception-reaction time. The results revealed that both the perception time and perception-reaction time were significantly longer in the highway exit situation compared to the ADS failure situation. These findings have important implications for understanding the differences in drivers’ cognitive processes and reaction times in varying driving contexts. It is worth noting that this model will undergo further refinement, validation, and application to other features of ADS or different driving scenarios in future research.

With current airspace infrastructure being stretched to their limits by the increase in demand for air traffic, Flow Centric Operations (FCO) have been touted to be a possible solution to efficiently manage airspace capacity to support more traffic. However, the novelty of such a concept (assigning flights to Air Traffic Controllers (ATCOs) based on traffic flows) necessitates human factors evaluation in this environment. Particularly, ATCO situational awareness rose to the forefront as a crucial human factor in this arena due to the difference in mental models adopted in this concept. Three fundamental traffic flows, with varying levels of complexity, as determined by a subject matter expert were evaluated for ATCO Situational Awareness. The behavioral results obtained using SAGAT revealed that Situational Awareness for more complex flows were higher, even though participants self-indicated that their situational awareness was lower during the complex flows. This mismatch could be a result of paradoxical complacency during less complex flows, and increased focus during the more complex flows.

The nature of ergonomics performs a unique role in protecting human health and preventing health risks. Improvements in productivity indices in the systems can be seen, resulting in better work conditions for people employed in production and services systems. To manage the processes of Ergonomics seeking to continuously adapt the working conditions to the psycho-physiological characteristics of the workers, by means of techniques, methods, and process controls, becomes the problematic adduced in this work, when there is no data to facilitate integration to the other processes of the company. The purpose of this descriptive study aims to demonstrate how the formation of homogeneous groups for ergonomics keeps the process dynamic through the generation of data that allows easy integration with other areas. This integration makes it possible to support the operational areas in risk management and consequently in the prevention of occupational diseases, reduction of absenteeism and incidents, providing improvements in the environment aiming at comfort, increased productivity, and better quality of life.

Privacy is a Fundamental Human Right and data anonymization is an essential process that aims to render data unidentifiable, therefore, private. The aim of this work is to test the anonymity of the first data collected via AnyMApp, a platform to anonymously perform online usability tests. Due to time and resource constraints, the authors opted to identify open-source free tools, which were evaluated recently for their usability and functionalities. Even being a highly ranked anonymization tool (ARX), the use of this tool in our data setting was not straightforward and required guidance from tutorials. In the end, it was not possible to run any of the anonymization models available in ARX in our dataset because there were no quasi-identifier attributes within the collected data. There were sensitive data collected though, with medical diagnosis information, but not able to cause privacy concerns when added with all the other non-identifiable and non-sensitive attributes. Despite our results for this use-case, we will keep verifying/anonymizing collected data from AnyMApp because we will have different use-cases and health-related data that can be more complex and comprise quasi-identifiers and sensitive attributes that we nee to guarantee that are anonymized.

Furry fandom becomes an influential sub-culture in social media among the young generation. More and more people love anthropomorphic animal characters and like to engage in magical thinking, the fantasy activities like a furry animal. However, the furry content or activities are mainly based offline, on the website, or scatter on social media. It is hard to find a mobile-based platform for furry fans to create, share, and exchange. In addition, the blind box industry develops quickly, which grabs the imagination of young consumers and raises incredible social excitement. However, the physical version of the blind box has various limitations, such as delivery, space occupation, expense, etc. Inspired by the needs of furry fans and the concept of a blind box, we develop a platform named ‘FURRY’ - a creative, fantasy, and interactive platform for the creation, sharing, and monetization of the animal-themed virtual blind box based on augmented reality technology and social media. Users can collect random parts of a virtual animal from the blind box and wear them to become an anthropomorphic animal character they prefer. In this paper, the whole structure of the platform is discussed. The four sections of FURRY including obtaining, playing, building fursona, and mine are introduced in detail. Then the markerless AR applied in the project is discussed. Two stages preprocessing and real-time processing are introduced to create an interactive user experience. The results of the present research are expected to help vitalize and expand the practice of AR-based culture content.

Goth is a music-based subculture that has a dark worldview and has influenced fashion, fiction, and art. In Japan, the Goth subculture was merged with Lolita clothing to create Gothic Lolita, which has been popular for over 20 years. Although previous research has shown that wearing Gothic Lolita can have positive emotional effects on the wearers. However, these studies mainly rely on questionnaires and interviews, which can be influenced by various factors such as the possibility of mixing the intentions of participants into the results. To address this issue, we applied a method that applied EEG signals and HRV indexes to evaluate the emotions of the participants. Since the Gothic Lolita is the darkest style within Lolita clothing, in this study, we did a preliminary experiment to collect physiological information while wearing Lolita clothing. We applied time-series analysis to collected data. As the result, the collected EEG signals and HRV indexes were more positive when wearing Lolita clothing than when wearing ordinary clothing. Both questionnaires and physiological information evaluations showed that wearing Lolita clothing increased the participants’ arousal, valence, and dominance.

The aim of this study is to develop a comparative evaluation system of the lower-limb exoskeletons. First, the comparative evaluation factors were classified into five categories: ease of wearing, safety, sense of wearing, effectiveness, and ease of use, based on the characteristics of the lower-limb exoskeletons. Then, a total of 15 questions were designed for 20 participants who had no experience with musculoskeletal disorders in this study. All participants tested the three lower-limb exoskeletons (CEX, Archelis, and Chairless Chair) and scored their subjective ratings on an 11-point scale. The results showed that the CEX was rated the highest for ease of wearing and effectiveness, whereas Archelis received the highest score for safety and ease of use. Based on the results, it was noted that the back support of the lower-limb exoskeletons should be removed, and the knee angle should be adjustable for various ranges to fit the users. In addition, exoskeleton developers would be better focused on ease of wearing and ease of use to improve the lower-limb exoskeleton. Based on this study, it is expected to be used as important and fundamental data for users’ requirements for more convenient and safer exoskeletons’ design.

Interactive jewelry and kinetic jewelry are emerging concepts in the field of jewelry design which overlaps with each other. However, interactive jewelry lacks specific parameters and a standard definition, which hinders early research progress and designers’ idea construction. To address this challenge, based on the idea of taking ambiguity as a design strategy, the fuzzy definition boundary between kinect jewelry and interactive jewelry makes the design features of kinect jewelry have certain enlightening significance for the interactive jewelry. This paper aims to collect design pictures related to kinetic jewelry using a Python model, analyze the collected pictures, and obtain current mainstream design parameters to reference for interactive jewelry design. The data will show details such as the interaction mode, wearing part, and materials used of kinect jewelry, thus providing valuable suggestions for interactive jewelry design. Additionally, the paper considered the possibility of how to leverage ambiguity in design as a resource.

The AnyMApp platform enables the testing of mock-ups from web or mobile applications, anonymously, and online, with the integration of three main parts: 1) survey-like questions regarding demographics and information about the project and use-case; 2) the mock-up interfaces of the use-case functionalities to be tested; 3) and survey-like questions regarding satisfaction and experience. The aim of this work is to present the preliminary results of the first use-case tested within the AnyMApp platform. Results show that the integration of diverse data can help define a comprehensive overview and improvement strategy for the tested application. All participants agree the AnyMApp platform is useful and simplifies and quickens usability testing, with bigger samples.

Based on the theory of Kansei Engineering, the relationship between user’s perceptual elements and the design elements of electric toothbrush modeling is quantified through Kansei Engineering to provide new ideas for the innovative design of electric toothbrush. The existing products were researched based on users’ perceptual elements, and representative product samples were selected to dismantle the modeling elements of electric toothbrushes through morphological analysis. Collect and establish the user’s perceptual vocabulary, quantify the user’s perceptual vocabulary by using the Semantic Difference method, use the Quantitation Theory Type I and SPSS software to count and analyze the correlation between the perceptual vocabulary and the modeling elements of the electric toothbrush, and derive the user’s perceptual imagery preference. Summarizing the design points and proposing modeling design options of the electric toothbrushes, which makes the electric toothbrush better meet the user’s emotional needs and promote the human-computer interaction between the user and the electric toothbrush product.

In recent years, the development of food culture and transportation has increased our chances of meeting unknown dishes. However, it is not easy to choose these new dishes without enough information about them. Although there is much research on menu choice and new dining experiences, but only some research is based on the behavior analysis of menu choice at a restaurant. Moreover, there have yet to be approached to propose a menu choice system based on this analysis. This study aims to survey users’ behavior when choosing a menu at Kaiten-Sushi restaurants and clarify users’ dissatisfaction and factors to interrupt the choice of new dishes using the service blueprint method and the customer journey map method. Moreover, the proposal of a new support system to choose a menu based on the survey result is the final purpose of this research. Therefore, we visited three Kaiten-sushi restaurants in Hakodate and investigated dining experiences there. As a result, we could extract the following three points; 1) There is an active type of menu choice and a passive one at Kaiten-sushi restaurants. 2) There are different types of expressions of allergy information, easy to understand or not. 3) Different languages in the menu information depending on the restaurant. Based on these results, we proposed a new Kansei support system. The Kansei support system has two main features; 1) Allergy information for specific products is expressed on each dish information page of the system. 2) The new expression of dish information which users can be interested to learn new information about the dish.

Micro-service architecture (MSA) refers to a structure in which several software with different roles are bundled into one service, rather than one or a small number of software when building a service. A lot of software is used to build a service with an MSA structure, and recently, open-source software is being actively used. Open-source software is not always available free of charge as each license exists. However, since commercialized MSA-structured cloud-based software does not provide execution code to consumers, it is difficult to check whether open-source software is used without permission. Therefore, even if the licenses granted by open-source software are violated or open-source software contains potential risks, it is very difficult for consumers to identify these issues. To solve this problem, a graph modeling language (GML) can be used as a method for analyzing the structure of software operating based on the cloud MSA environment and determining the open-source license included in the software. However, since GML contains a lot of information such as software structure and information, it can be abused by analyzing the software structure when GML is leaked. To solve this problem, in this study, we conduct research on how to encrypt GML and perform analysis in the encrypted state.