Context: Chatbots are intelligent agents that mimic human behavior to carry on meaningful conversations. The conversational nature of chatbots poses challenges to designers since their development is different from other software and requires investigating new practices in the context of human-AI interaction and their impact on user experience. Since human dialogue involves several variables beyond verbalizing words, it is vital to design well-thought dialogues for chatbots to provide a humanized and optimal interaction. Objective: The main objective of this work is to unveil textual, visual, or interactive design practices from text-based chatbot interactions and how they can potentiate or weaken some perceptions and feelings of users, such as satisfaction, engagement, and trust, for the creation of the Guidelines for Chatbot Conversational Design (GCCD) guide. Method: We used multiple research methods to generate and validate the guide. First, we conducted a Systematic Literature Review (SRL) to identify conversational design practices and their impacts. These practices were inserted into the GCCD guide through qualitative analysis and coding of SLR results. Then, the guide was validated quantitatively through a survey and qualitatively through a case study. The survey aimed to assess the guide’s clarity and usefulness based on the reading of the guide by the participants and their responses to a questionnaire adapted from the Technology Acceptance Model. The case study aimed to assess the guide’s usefulness based on its practical application by participants in a situation that simulates a real scenario and follow-up interviews. Results: The survey showed that software developers with different levels of experience strongly agreed that the guide could induce greater user satisfaction and engagement. Furthermore, they also strongly agreed that the guide is clear, understandable, flexible, and easy to use. Although participants suggested some improvements, they reported that the guide’s main strengths are objectivity and clarity. The case study confirmed the survey findings, as participants reported positive feelings toward the guide and an intention to use it. Their extensive perceptions given through the conducted interviews unveiled that their previous experiences with chatbots and in specific software development positions influenced their design and adoption of practices. Conclusion: The guide proved to be useful for developers with different levels of knowledge, with the potential to become a strong ally for developers in the conversational design process.
Music's profound emotional impact extends beyond traditional listening experiences, playing a critical role in shaping user engagement in entertainment contexts, including digital games. Music visualizations - graphical representations of musical elements such as rhythm, melody, and harmony - are becoming increasingly significant for enriching emotional depth in these contexts. Despite the potential of music to enhance digital media experiences, there remains a gap in the intentional integration of musical visualizations with specific emotional narratives. In this paper, we propose Thunder, a design process for creating musical visualizations that prioritize conveying the emotional communication of music. Our approach builds on an existing method for rapid prototyping of animated musical visualizations and comprises four iterative phases. By applying this process in creating two music visualizations, we demonstrate that our process aligns closely with the designed emotional content of the music. The results showed that our alternative visualizations effectively communicated intended negative emotions, such as "sadness" and "shame", and were well-received by the public, highlighting the importance of synchronized visual and musical elements. With this work, we hope to offer valuable insights for game designers and entertainment app developers, providing a comprehensive design process for crafting musical visualization experiences that emphasize the emotional richness of music and pave the way for more immersive and emotionally resonant digital media experiences.
Due to data production and availability growth, professionals in several disciplines have been facing an increasing need to explore and understand data, obtain insights, and communicate them effectively. Many visualization systems have been developed commercially and within the research community to support non-expert analysts. We can consider at least three challenges these tools aim to face: support the selection of appropriate visualizations and decide on the visual mappings, extract and communicate factual information from the visualizations, and use visualizations in data-rich narratives. In response to these challenges, we developed VisStoryMaker, a visualization tool that supports both exploration and communication about data. To aid users in exploring and understanding data, VisStoryMaker recommends visualizations through system-generated questions and data facts. To support communicating about data, the system recommends visual annotations of data facts and provides a story-building module, allowing analysts to use the generated charts and facts as a blueprint for a data story. We have conducted empirical studies to compare VisStoryMaker’s features with existing applications: chart recommendations with Voyager 2, storytelling construction with Flourish, and data facts and chart annotations with Tableau. Our findings indicate that the system-generated questions and data facts supported non-expert analysts in exploratory analysis. They perceived visual data facts annotations as useful and supported them in raising hypotheses about the data, understanding data, and leading to insights, thus enhancing data analysis. Participants perceived the visual annotations and StoryMaker as helpful in organizing the system-generated pieces of information and incorporating them into comprehensive narratives and presentations.
The number of individuals with depression is gradually increasing worldwide. In the university population, the prevalence of depression is even higher than in the general population. Health and computing researchers have sought to create computational solutions that can support diagnosis and create interventions for people with a possible depressive profile, analyzing their behavior through the collection of digital phenotyping data, such as heart rate and sleep quality, coming from sensors present in mobile and wearable devices, such as smartphones and smartwatches. Research in this area is multidisciplinary, and Human-Computer Interaction (HCI) researchers have a major role in supporting the design of this technology. In this sense, the goal of this paper is mapping, among others, how final users and health professionals have been involved in this type of research. We performed a Systematic Mapping focusing on the university environment. We extracted data about the methodological approach adopted, sensor data collected, and human involvement. The results show which data is being collected and correlated with depression indicators, the absence of multidisciplinary teams, and the lack of adoption of protocols to manage mental health emergencies. We also discussed possibilities for human-in-the-loop approaches and some design implications for HCI teams.
Visual impairment is a condition that is very present in society and that has been growing due to a series of factors. This condition can affect self-esteem when some processes are limited to visual aspects. Make-up, widely used for correction, enhancement and beautification, is a practice that is very present in people’s lives, however, it still has some barriers to use for those with visual impairments, and need to be offered proposals that facilitate the use of products by people with this condition, in order to include them in this branch, as it ensures accessibility. The prototype presented here, entitled “K.eyes”, seeks to provide visually impaired women with technological support capable of communicating with them during the self-makeup process, based on data that would be obtained with computer vision. For requirements gathering, a study was carried out from the analysis of an accessible online channel, interviews with the target audience were done and an exploratory research was done on a self-makeup teaching channel for a visually impaired person. As a result, a prototype developed in a mobile version is presented, which was evaluated through a usability test with 6 blind or low vision women. The analysis of the results shows that the application is of great interest to women with visual impairments and has great potential to help in the self-makeup process.
Context: Web chats are widely used in corporate and social settings, making inclusivity in web-based communication platforms crucial. However, visually impaired individuals often encounter obstacles in digital conversations due to the dynamic nature of web chat elements, which poses challenges for screen reader compatibility. This paper addresses these challenges by adapting an open-source web chat designed for chatbot interactions to be fully compatible with screen readers. Our solution involves modifying the web chat’s code to enhance the user experience for visually impaired individuals. Key modifications include integrating semantic HTML tags, Accessible Rich Internet Applications (ARIA) landmarks, and other accessibility-focused techniques. We employed a combination of Proof of Concept and Experimentation, starting with an examination of the existing web chat, followed by the code modifications. Validation was conducted through execution and analysis of two use case scenarios. Results show significant improvements in user experience, as evidenced by screen reader transcriptions. Notable enhancements include improved navigation and the overall announcement of chat updates. This research contributes to Human Factors in Computing Systems by addressing web chat accessibility gaps. The adapted open-source web chat demonstrates the positive impact of minor code modifications on usability for users with visual impairments, exemplifying inclusive design in digital communication platforms.
The following paper describes the results regarding an extension course carried out with elderly people in a public university, about fake news and disinformation. The proliferation of misinformation and fake news in digital platforms and media represents a growing societal challenge, especially for the elderly, who may be more susceptible to the spread and manipulation of fake news. In this experience report, we describe the implementation of a course for this audience, which focuses on developing skills to identify and verify online information, combat misinformation, and strengthen the digital autonomy of this demographic. Thirty-four elderly people were present and pointed out positive satisfaction with the course.
The VLibras Mobile Application is a translation tool from Brazilian Portuguese to Libras (Brazilian Sign Language) using an Avatar in 3D animation. The tool provides deaf people a means of accessing textual content based on animation in Libras. Furthermore, VLibras can be used as a communication tool between deaf people, Libras interpreters, and people in general (interested in learning Libras, for example). In this context, having access is not the same as being accessible. Thus, it is essential to improve the deaf user experience through the compliance of VLibras Mobile with ABNT NBR 17060 Standard. The assessment was conducted for all VLibras Mobile screens and features, covering the totality of criteria of the standard. Analogous to the WCAG standard, the NBR 17060 Standard also considers three accessibility levels (A: mandatory, AA: minimally accessible, and AAA: optional), as well as the four principles (Perceptible, Operable, Understandable, and Robust). It also included the standard categories: Perception and Comprehension, Control and Interaction, Media, and Coding. The goal was to compile a general overview of VLibras Mobile compliance with the standard to know this quality regarding accessibility. Also, the analysis allowed identifying an ensemble of non-compliance criteria, both located on a specific screen or across the application. These were transcribed as 12 issues with their respective severity level and reported to developers to improve VLibras Mobile accessibility. Also, the specific audience of VLibras, deaf people, led to classify a group of criteria as non-applicable. The results revealed that the most important features of VLibras, Traduction and Home, demonstrated higher compliance rates. However, the screen Customization showed significant non-compliance. Overall, the issues reported are of low severity, and their correction has a transversal impact on the application. This text is an experience report of the assessment conducted, its challenges, results, and lessons learned.
Sleep is a vital physiological need for human health and well-being, and its quality plays a crucial role in our daily lives. Inadequate sleep can lead to a range of health problems, including chronic fatigue, lack of concentration, cognitive impairment, and an increased risk of cardiovascular and metabolic diseases. Amidst an increasingly aware society of the benefits of well slept nights, there is a growing interest in monitoring and optimizing sleep. This trend is amplified by the continuous advancement of technology, particularly the increase of the Internet of Things (IoT) and E-Health technologies. In this context, the development of consumer sleep technologies emerges alongside the increasingly easier methods to utilize them and the growing awareness of the importance of sleep for health. These technologies offer a variety of solutions, ranging from mobile applications to smart wearable devices. These tools can empower individuals to take steps to improve their sleep hygiene and, consequently, their quality of life. However, some user reviews on the Google Play platform raises issues related to their user experience across all consumer sleep technologies. This article aims to evaluate the User Experience of consumer sleep technologies based on users reviews on Google Play by using the extended MALTU instrument. The evaluation consists in collecting reviews on the platform to analyse and categorize them with consistency and real user reviews. A bibliographical research about consumer sleep technologies was also carried out. The technologies under evaluation include Oura, a sensor-equipped ring, Whoop, a sensor-equipped band and the Sleep as Android app. The results revealed issues regarding interface, synchronization and metrics.
Information Visualization (InfoVis) provides techniques and methods for interactive visual representations to enhance human understanding of data. This facilitates monitoring and makes relevant information accessible to all users, regardless of their experience in data analysis. Software professionals in mobile development have shown interest in using visual representations (e.g., charts, tables) to support the analysis of user reviews of applications available in app stores. However, analysis tools often focus on data mining resources and do not provide appropriate visualizations for the target audience (i.e., mobile software professionals). Considering the context previously described, this paper presents the main results of a dissertation project. In this project, a set of 16 guidelines was proposed to support the design of UX data visualizations. The guidelines were conceived from a Systematic Literature Mapping (SLM) that explored 21 papers from academic literature. These papers were analyzed to extract the guidelines in the form of recommendations. The guidelines, along with Shneiderman’s seminal work ‘Visual Information-Seeking Mantra’, were utilized to develop a proposed UX data visualization named User eXperience Reviews Information Visualization (UX-RIVIS). Shneiderman’s mantra allowed for the introduction of levels of navigation, which produced visualizations ranging from data overview to detailed data. UX-RIVIS visualizations were evaluated by 23 software professionals using the method of Intermediate Semiotic Inspection (i.e., a qualitative approach for assessment). The evaluation provided results that examined the communicability and perceptiveness of software professionals regarding the ease of use of the visualizations. The results indicated the usefulness of the visualizations provided by UX-RIVIS for software professionals in data analysis. It was also noted that these professionals preferred certain visualizations, such as tables and pie charts. Moreover, participants showed interest in using the visualizations for their work activities.
The use of Web Accessibility standards is fundamental as a first step in evaluating the accessibility of digital resources. The Web Content Accessibility Guidelines (WCAG) have become the leading worldwide reference for accessibility standards and are used by many governmental regulations in numerous countries. However, many studies have revealed difficulties novice evaluators encounter in conducting accessibility inspections using WCAG. In Brazil, there is ongoing work to adopt WCAG 2.2 as the basis of a national standard by the Brazilian Technical Standards Association (Associacao Brasileira de Normas Tecnicas - ABNT), which will include a checklist to simplify the use of the guidelines. This experience report describes the experiences of two groups of students engaged in an outreach digital accessibility project, aimed at providing detailed inspection reports to entities representing people with disabilities when filing official complaints concerning digital accessibility. The first group of students evaluated websites using the WCAG 2.2 documentation alone. The second group had their first experience conducting accessibility inspections using the checklist and guidance from the ABNT checklist under construction. The report provides insights into the main difficulties encountered by the students in both scenarios and the strategies they used to overcome them. The experience report provides a detailed account of the types of success criteria that presented the most difficulties and how the new checklist provided better guidance to them. The reflections in the report provide lessons learned concerning using the ABNT checklist and possibilities for improvements for both the checklist and future versions of WCAG.
Narrative data visualizations are proposed to present data in a story-like form that may improve its understanding. Besides this, the end-users of visualizations may have varying needs and preferences in how data is presented or organized, which makes end-user development a promising topic to be blended into narrative data visualization research. This paper presents a doctoral research journey to define an approach to assist visualization designers in planning narrative data visualizations with customization features called customizable narrative data visualizations. We defined a four-stage methodology to cope with our research goals, ranging from understanding the problem to applying the solution. Throughout our research, we conducted different studies, including systematic reviews and user studies like focus groups and workshops, to outline and analyze our process. The process was considered useful and promising in assisting designers in identifying narrative elements from data and defining how the end-users can customize a narrative visualization. Besides the process itself, the studies presented valuable contributions to the HCI area, such as grasping the perspectives of designers and end-users regarding customizable narrative visualizations. We believe our results are important findings that help set the path for researchers and professionals interested in creating narrative data visualizations in which end-users have agency over the content displayed.
This Grand Challenge highlights that the HCI field is going through a time when it needs to revisit and find new theoretical and methodological foundations to account for the phenomena it encompasses. Thus, it addresses philosophical aspects of the field, encompasses studies that aim to inform and enrich the understanding of human-computer interaction, and highlights the need to bring, experiment with, and conceive new ways of understanding HCI and acting in it. In addition to the application of multidisciplinary theories from anthropology, linguistics, psychology, design, sociology, and other areas, this Grand Challenge exposes the need to investigate other philosophical positions that have been little explored until now, such as those of a phenomenological nature, the creation of theoretical models, the application of existing theories in new ways, or the synthesis of several theories from a new perspective aiming to describe, explain, or inform the field. Consequently, it includes new methods and techniques for data collection and analysis, design approaches, or evaluation strategies that can be applied to HCI research and practice. This paper is presented bilingually in English and Portuguese as a testament to its situated and universal implications.
Technological advances in the last decades have allowed acquiring, storing, analyzing and manipulating big volumes of data, contributing to the development of increasingly advanced, dynamic and intelligent software. Looking beyond the advantages Big Data and Machine Learning (ML) solutions have brought, it is necessary to consider some of their challenges, such as the complexity of dealing with the data itself and the understanding of how the algorithms used in such solutions work. User-Centered Design (UCD) comprises a set of processes that focus on the users’ perspective when designing solutions, be it through interviews, surveys and evaluations, or even by actively participating in the design activities. Understanding users and their needs, and helping them understand and trust the products they use are key factors when designing Big Data and ML software; fields such as Explainable Artificial Intelligence (XAI) help develop solutions that meet these goals. Thus, this paper presents a systematic mapping covering 54 papers presenting approaches and methods that incorporate UCD into developing Big Data and ML solutions, with results highlighting an interest in usability factors such as perceived utility and learnability. This systematic mapping study sheds light on existing challenges and opportunities for enhancing these processes.
E-Commerce moves trillions of dollars worldwide every year, connecting buyers and sellers. However, users, increasingly attentive, view the process of paying for and receiving products and services over the Internet with suspicion, whether due to reports of negative experiences from other users or even their own. Cases of fraud are damaging to a company's image. In this article we analyze a large volume of user reviews of different marketplace-type e-commerce applications (apps) with the aim of mapping the user experience behind the interface for the transaction mediation service between buyers and sellers offered by such apps. Our results show that the impression of insecurity occurs mainly due to two elements: 1) Interface complexity - the interfaces of these applications tend to have many usability elements that may end up hindering the clarity of the process; and, 2) Communication of responsibility - when users often do not know that they are dealing with a third party or they do know, but are unaware of the limitations of the platform's mediation service. The results of the analysis carried out here demonstrate that the loss in incremental profit, when the user uninstalls the application, tends to be much greater than the losses caused directly by fraud, highlighting the need to not only provide security, but also to communicate it.
Despite their popularity, chatbots frequently struggle to meet user’s expectations by providing appropriate responses, accurately understanding inquiries, and engaging in natural dialogue. Previous research has shown that incorporating linguistic and conversational strategies, particularly those regarding politeness, can significantly enhance user satisfaction and interaction quality. However, current conversation design practices often lack well-defined approaches to support such strategies and force designers to rely on personal preferences. To improve the process of designing polite chatbots, scholars have examined politeness parameters, which aid in understanding the manifestation of politeness in chatbot dialogues through language function variability and the use of indirectness. This study aims to assess the effectiveness of these politeness parameters in supporting designers during conversational design scenarios. Three scenarios were created, each with dialogues between a chatbot and users featuring distinct conversational intentions and politeness parameters. Participants with diverse backgrounds evaluated each dialogue by identifying strategies and intentions and assessing their level of politeness. The main findings indicate that participants could identify different strategies, evaluate politeness, and comprehend the designer’s intent. However, the parameters proved to be impractical for designers without a linguistic background. Therefore, the study introduces a card-based approach to facilitate and guide the design of polite conversational agents.
Discussions about the relationship between technologies in social contexts have been present in the Brazilian Human-Computer Interaction (HCI) community for years, including in the “Grand Challenges of Research in Human-Computer Interaction in Brazil” from 2012-2022. The challenge presented here seeks to continue these discussions, including and expanding through Plurality, the intersections between diversity, equity, accessibility, and inclusion. This challenge also proposes a critical analysis of the methods and theories used in HCI research and productions, seeking a decolonial perspective of scientific and technological processes in our Brazilian and Latin American territorial reality, which often originated in the global north and replicated in other regions. Besides geographical positions, it also considers how body and identity can generate exclusion or inclusion. We also discuss that the decoloniality process corroborates plurality and thus influences the creation of our theories, methodologies, actions, practices, technologies, interactive computational systems, and, consequently, the people who use them. This paper is presented bilingually in English and Portuguese as a testament to its situated and universal implications.
The employment of Machine Learning (ML) and Deep Learning models across diverse domains has grown exponentially in recent years. These models undertake critical tasks spanning medical diagnoses, criminal sentencing, and loan approvals. Consequently, there is a need for these models to be interpretable, enabling users to grasp the rationale behind predictions and engendering trust. Equally vital is the capability of developers to pinpoint and rectify any erroneous behaviors. In this context emerges the field of Explainable AI (XAI), which aims to develop methods to make ML models more interpretable while maintaining their performance level. Various methods have been proposed, many leveraging visual explanations to elucidate model behavior. However, a notable gap remains: a lack of rigorous assessment regarding the effectiveness of these explanations in enhancing interpretability. In this paper, we evaluate whether the SHAP method, widely used in the XAI field, achieves its objective of making a model more interpretable. We conducted a study based on the concept of simulatability. We asked participants who have mathematical or statistical backgrounds and intermediate to advanced knowledge of machine learning and model interpretability to classify some given instances of a dataset, presented first without the explanations provided by the SHAP method and then with them. The goal was to assess whether the explanation made the participants better at classifying the instances. Our findings reveal that the visualizations can be confusing even for users with a mathematical background. Additionally, we argue that there is a need for XAI researchers to work collaboratively with visual analytics experts to develop these visualizations, as well as test the visualizations with users of various backgrounds.
The popularization of computing devices and internet access, especially in developing countries, has driven attention to an increasing number of software users with varied cultural and linguistic backgrounds. Human-Computer Interaction (HCI) research has thus far considered language communities as static and isolated entities, generally associating them with a specific country. Linguists know, however, that cultures interact and exchange ideas, languages spread across state borders, and individuals often speak two or more languages. For an effective HCI design, developers should consider these social and individual aspects of multilingualism. In this study, we aim to investigate technology users’ behavior and difficulties in different multilingual contexts. An online survey has been conducted with multilingual participants from two universities: one in Switzerland and the other in Brazil. Analysis of 196 valid responses confirms a high level of multilingualism in general, but higher among participants in Switzerland. The types of tasks and the language-related issues reported by the participants varied between the two contexts, indicating that the level of multilingualism impacts users’ behavior online. A final discussion summarizes challenges and opportunities for future studies aiming to propose HCI design approaches with a focus on multilingualism.
This paper addresses sociocultural aspects as a Grand Challenge for Research in Human-Computer Interaction in Brazil for 2025-2035. The challenge considers the sociocultural diversity involved in the design, implementation, evaluation, and use of interactive computational artifacts. The current scenario reveals a gap between technological solutions and the real needs of various sociocultural groups. The challenge herein described includes the creation or adaptation of tools that respect sociocultural plurality, impacting both the creation of interactive computational artifacts and the education of conscious and critical professionals. This challenge arises from the need for technologies that do not discriminate against, exclude, or harm social groups, and that are developed in a responsible, conscious, ethical, and sustainable manner. Failing to meet this challenge could lead to the formation of technological "bubbles" that promote oppressive views, diminish human capabilities, and prioritize technocapitalist interests over human concerns. In response, we suggest short-, medium-, and long-term actions, emphasizing awareness, research, and continuous monitoring. Therefore, the challenge of considering sociocultural aspects in Human-Computer Interaction aims to transform the relationship between interactive computational artifacts and society, promoting inclusion and positive impact through technological solutions that respect and represent human diversity.
