Recent advances in artificial intelligence (AI), including generative models and socially interactive agents, are reshaping the design of interactive systems and raising new challenges for Human–Computer Interaction (HCI). While AI research has traditionally focused on improving performance and scalability, HCI emphasizes usability, transparency, and the broader social implications of technology, all aspects that imply the application of proper human-centered design methods. Bridging these perspectives is increasingly critical as AI moves from backend functionality to a central role in user interaction. This paper reports on the workshop Future Horizons in Human–AI Interaction: Joint Perspectives from HCI and AI, held at AVI 2026. It synthesizes key insights on emerging challenges and opportunities in designing human-centered AI systems, emphasizing the importance of integrating perspectives from HCI and AI, to advance the design of interactive AI systems that are both technically robust and aligned with human values.

While markerless motion capture provides rich biomechanical data for clinical assessments, its integration into everyday orthopedic practice is limited by the abstract nature of kinematic signals and the opacity of standard machine learning models. To bridge this interpretability gap, we present a framework that embeds explainable AI directly into a real-time movement classification engine. Our primary aim is to achieve a dual-interface architecture that transforms model outputs into role-specific visual narratives: (i) a Clinician Interface provides analytical dashboards, while a (ii) Patient Interface translates technical metrics into engaging and accessible visualizations. This framework highlights the need for a human-centered paradigm to leverage complex biomechanical data and their explanations in clinical decision-making and trust.

The dominant paradigm for interacting with the Web has long relied on menu-driven navigation and complex graphical structures that can impose a high cognitive load on users. Recent advances in Large Language Models (LLMs) enable flexible NL understanding, yet purely conversational systems lack the stability and visual affordances required for complex tasks. This paper introduces Mirage, a framework that bridges these paradigms by making graphical user interface (GUI) components interpretable and orchestrable by LLMs, enabling on-the-fly construction of simplified UIs tailored to user requests. Developers declare application functions as reusable components with semantic descriptions and typed schemas; Mirage maps user goals, expressed in natural language, to the most relevant components and renders them as interactive interfaces. We implemented Mirage as a Next.js framework and validated it in a user study comparing the new paradigm to the navigation of a traditional website. Results show that Mirage significantly reduces task completion time, lowers cognitive workload, and achieves higher usability without compromising accuracy. Our approach also highlights new design directions toward supporting on-the-fly building of goal-driven interfaces.

In this paper, we present a study aimed at evaluating how ChatGPT-4 understands end-users’ natural language instructions to express automation rules for smart home applications and how it translates them into Python code ready to be deployed. Our study used 34 natural language instructions written by end users who were asked to automate scenarios presented as visual animations. The results show that ChatGPT-4 can produce coherent and effective code even if the instructions present ambiguities or unclear elements, understanding natural language instructions and autonomously resolving 94% of them. However, the generated code still contains numerous ambiguities that could potentially affect safety and security aspects. Nevertheless, when appropriately prompted, ChatGPT-4 can subsequently identify those ambiguities. This prompts a discussion about prospective interaction paradigms that may significantly improve the immediate usability of the generated code.

This paper explores how a library of conversational patterns defined for Web browsing can be adopted to improve the accessibility of prompt interfaces for Large Language Models. It reports preliminary findings from a formative evaluation with blind and visually impaired users. It also discusses how these findings support transferring the benefits of voice-based Web browsing to the inclusivity of prompt-based interfaces.

This article explores Human-Centered Artificial Intelligence (HCAI) in medical cytology, with a focus on enhancing the interaction with AI. It presents a Human-AI interaction paradigm that emphasizes explainability and user control of AI systems. It is an iterative negotiation process based on three interaction strategies aimed to (i) elaborate the system outcomes through iterative steps (Iterative Exploration), (ii) explain the AI system's behavior or decisions (Clarification), and (iii) allow non-expert users to trigger simple retraining of the AI model (Reconfiguration). This interaction paradigm is exploited in the redesign of an existing AI-based tool for microscopic analysis of the nasal mucosa. The resulting tool is tested with rhinocytologists. The article discusses the analysis of the results of the conducted evaluation and outlines lessons learned that are relevant for AI in medicine.

A B S T R A C T Micro-electronics tools, coupled with card-based tools, are employed for prototyping smart devices with nonexperts.Lately, researchers have started investigating what tools can actively engage people with intellectual disabilities (ID) in their prototyping.This paper posits itself in this line of work.It presents a toolkit for ID people to rapidly prototype together their own ideas of smart things, for their own shared environment.It analyses and discusses engaging or disengaging features of the toolkit in light of the results of two workshops with eight ID participants.Lessons of broad interest for the design of similar toolkits are drawn from the literature and study findings.

Recent developments in accessible electronic-making toolkits have opened up avenues for individuals with intellectual disabilities(ID) to actively participate in creating their own smart objects based on the Internet of Things (IoT) technology. These toolkits present a novel opportunity to foster the inclusion of this often-under-considered community in the development of personalized solutions that can impact their autonomy and well-being. However, understanding IoT encompasses comprehending the inter-object connection paradigm at the core of this technology, an aspect that is not adequately covered by the existing accessible toolkits. To fill this gap, this paper illustrates the design and evaluation of MakeNodes, a tangible toolkit to involve people with ID in making smart-thing networks in a guided, collaborative, and engaging way. The toolkit comprises a series of input and output nodes that can be paired to make any object or surface smart. The paper illustrates how MakeNodes empowers people with ID through IoT-making activities that promote collaborative work to address personal needs. Key findings are discussed in the form of lessons learned concerning the importance of physical and visual elements, hands-on exploration, and interaction robustness in improving the accessibility of IoT-making toolkits for people with ID.

Conversational agents are emerging as channels for a natural and accessible interaction with digital services. Their benefits span across a wide range of usage scenarios and address visual impairments and any situational impairments that may take advantage of voice-based interactions. A few works highlighted the potential and the feasibility of adopting conversational agents for making the Web truly accessible for everyone. Yet, there is still a lack of concrete guidance in designing conversational experiences for browsing the Web. This paper illustrates a human-centered process that involved 26 blind and visually impaired people to investigate their difficulties when using assistive technology for accessing the Web, and their attitudes and preferences on adopting conversational agents. In response to the identified challenges, the paper introduces patterns for conversational Web browsing. It also discusses design implications that can promote Conversational AI as a technology to enhance Web accessibility.

Digital well-being traditionally means limiting the effects on individuals of technology abuses.However, in a broader perspective, it can be crucial to consider the pervasiveness of technology, and the effect it can have not only on individuals but also on their peers in the context of diverse everyday-life situations.Within this view, which emphasises the social side of digital well-being, the paper argues the need of educating young generations to participate in the making of technology for a social goal and have a reflective attitude towards technology and its impact on society.It, therefore, presents a design toolkit as a means to (i) engage young generations to become active in design for social digital well-being and, thanks to the exposure to how technology works, (ii) reflect deeply on the pros and cons of technology in use in their everyday life.By presenting the results of a study with 24 high-school pupils and their teachers, the paper discusses how a phygital toolkit, which structures the design process, engages them in the rapid prototyping of their own smart things, and how it acts as a proxy for soliciting their own reflections around technology and social digital well-being.

We have witnessed astonishing progress in mobile and ubiquitous technology in the past two decades. Our interaction-supporting devices have become increasingly capable and diverse, while the supporting infrastructure, including heterogeneous networks and cloud computing, has matured to enable our vision of a ubiquitous computing environment. As we navigate this new frontier, we find ourselves pondering the words of Bill Buxton: "Now that we can do anything, what should we do?" and, more importantly, "How should it work?" Buxton's thought-provoking statement urges us to shift our focus towards methodological considerations and, crucially, ethics. The MobileHCI community, as a whole, recognizes the importance of ethical practices in shaping our field. While our work has often touched upon privacy-maintaining interactions and service provision, it is time to delve deeper into the ethical dimensions of our research and design choices.

Smart things are everywhere: smart watches, smart jewellery, smart art installations. All have in common the capability of sensing with input devices (e.g., buttons), computing, reacting with output devices (e.g., light matrices), and connecting with other smart things. Engaging their end users in their design can be challenging and yet give end users unique empowering opportunities. Toolkits are fundamental for engaging end users and enabling them to design. This paper presents the IoTgo phygital toolkit for smart-thing design with end users who are not experts, e.g., artists and school learners, which results from co-design actions and interventions in the field.

Designing smart things, so as to ideate and program them, is a complex process and an empowerment opportunity. Toolkits can help non-experts engage in their design, that is, in their ideation and programming. This paper presents the IoTgo toolkit, made of card-based material and digital components. Playing with IoTgo can help understand the design context, ideate smart things for it, make them communicate and program them with patterns. In recent years, IoTgo was used and evolved at a distance with few participants. This paper presents a study with it, for the first time in presence, in a high-school. Results are discussed and suggest what role IoTgo or similar toolkits can play in the rapid design of smart things with non-seasoned designers or programmers.

The Internet of Things is a pervasive technology widely adopted in several contexts ranging from smart homes to automotive cars. In this context, the End User Development is gaining momentum: different solutions support non-technical users to define the smart objects behavior to better satisfy their needs. This workshop aimed to stimulates participants in providing discussions in line with the workshop themes and goals, for example, user mental models, acceptance of EUD solutions, the role of AI in EUD tools, humanoid robots were discussed.

Children's role in the design of new technology has been widely investigated. Recently, the research focus has shifted, from the technology they help create, towards what children gain by participating in design workshops. This paper intercepts this line of research. It reports on a design workshop with 27 children, aged from 11 to 14 years old, ideating, programming and prototyping smart objects for their town park. Data were gathered in relation to children's beliefs, before and after the workshop, and in relation to their understanding of design, after the workshop. The analysis of the gathered data suggests that the workshop positively affected children's beliefs and understanding of design, giving indications for future work concerning design as means of empowerment.

Smart Interactive Experiences (SIEs) are usage situations enabled by the Internet of Things that empower users to interact with the surrounding environment. The goal of our research is to define methodologies and software environments to support the design of SIEs; more specifically, we focus on design paradigms suitable for experts of given domains, who however might not be experts in technology. In this context, this paper discusses some trade-offs that we identified between six different dimensions that characterise the quality of software environments for SIE design. The trade-offs emerged from the analysis of data collected in an experimental study that compared three different design paradigms to understand in which measure each paradigm supports the creative process for SIE design. After reporting on the study procedure and the data analyses, the paper illustrates how the resulting trade-offs led us to identify alternatives for SIE design paradigms, and to structure on their basis a modular architecture of a software platform where the strengths of the three paradigms can be exploited flexibly, i.e. depending on the constraints and the requirements characterising specific design situations.

Abstract Automation in computing systems has always been considered a valuable solution to unburden the user. Internet of Things (IoT) technology best suits automation in different domains, such as home automation, retail, industry, and transportation, to name but a few. While these domains are strongly characterized by implicit user interaction, more recently, automation has been adopted also for the provision of interactive and immersive experiences that actively involve the users. IoT technology thus becomes the key for Smart Interactive Experiences (SIEs), i.e., immersive automated experiences created by orchestrating different devices to enable smart environments to fluidly react to the final users’ behavior. There are domains, e.g., cultural heritage, where these systems and the SIEs can support and provide several benefits. However, experts of such domains, while intrigued by the opportunity to induce SIEs, are facing tough challenges in their everyday work activities when they are required to automate and orchestrate IoT devices without the necessary coding skills. This paper presents a design approach that tries to overcome these difficulties thanks to the adoption of ontologies for defining Event-Condition-Action rules. More specifically, the approach enables domain experts to identify and specify properties of IoT devices through a user-defined semantics that, being closer to the domain experts’ background, facilitates them in automating the IoT devices behavior. We also present a study comparing three different interaction paradigms conceived to support the specification of user-defined semantics through a “transparent” use of ontologies. Based on the results of this study, we work out some lessons learned on how the proposed paradigms help domain experts express their semantics, which in turn facilitates the creation of interactive applications enabling SIEs.

This article illustrates ABBOT, a pervasive interactive game for children at the early years of primary school that aims to stimulate exploration of outdoor environments. ABBOT combines a smart tangible object to play outdoors, with a mobile app to access new content related to the discovered natural elements. The tangible object helps children capture images of the elements they find interesting in the physical environment. Through simple interactive games on a tablet, at home children can continue to interact with the collected digital materials and can also access new related content. The article illustrates the design of ABBOT; it also reports on an exploratory study with 160 kids of a preschool and a primary school that helped us assess the attitude of kids towards the game.

Research on the Internet of Things (IoT) has devoted many efforts to technological aspects. Little social and practical benefits have emerged so far. IoT devices, so-called smart objects , are becoming even more pervasive and social, leading to the need to provide non-technical users with innovative interaction strategies for controlling their behavior. In other words, the opportunities offered by IoT can be amplified if new approaches are conceived to enable non-technical users to be directly involved in “composing” their smart objects by synchronizing their behavior. To fulfill this goal, this article introduces a model that includes new operators for defining rules combining multiple events and conditions exposed by smart objects, and for defining temporal and spatial constraints on rule activation. The article also presents the results of an elicitation study that was conducted to identify possible visual paradigms for expressing composition rules. Prototypes implementing the resulting visual paradigms were compared during a controlled experiment and the one that resulted most relevant for our goals was used in a study that involved home-automation experts. Finally, the article discusses some design implications that came out from the performed studies and presents the architecture of a platform supporting rule definition and execution.

The EFESTO platform allows the creation of interactive workspaces supporting end users in the exploration and seamless composition of heterogeneous data sources. By means of a visual paradigm implemented within a Web composition environment, the end users dynamically create "live" mashups where relevant information, extracted from different types of data sources including the Linked Open Data, and functions that can be performed on it can be flexibly shaped-up at runtime.

In recent years, models, composition paradigms, and tools for mashup development have been proposed to support the integration of information sources, services and APIs available on the Web. The challenge is to provide a gate to a “programmable Web,” where end users are allowed to construct easily composite applications that merge content and functions so as to satisfy the long tail of their specific needs. The approaches proposed so far do not fully accommodate this vision. This article, therefore, proposes a mashup development framework that is oriented toward the End-User Development. Given the fundamental role of user interfaces (UIs) as a medium easily understandable by the end users, the proposed approach is characterized by UI-centric models able to support a WYSIWYG (What You See Is What You Get) specification of data integration and service orchestration. It, therefore, contributes to the definition of adequate abstractions that, by hiding the technology and implementation complexity, can be adopted by the end users in a kind of “democratic” paradigm for mashup development. This article also shows how model-to-code generative techniques translate models into application schemas, which in turn guide the dynamic instantiation of the composite applications at runtime. This is achieved through lightweight execution environments that can be deployed on the Web and on mobile devices to support the pervasive use of the created applications.

There is a huge and ever increasing amount of data sources available on the Web, which provide content through programmatic interfaces. Unfortunately, such data sources are accessible only through programming and therefore it is difficult for non-technical users to take advantage of such enormous data assets. The need therefore arises for paradigms to let laypeople, i.e., users without expertise in programming, explore and compose data sources. This paper discusses mechanisms for data source exploration and integration, which emerged from a study where laypeople were involved in discussions to gather their requirements about accessing and composing services. The paper also describes the prototypes that we defined to respond to the requirements highlighted by end users.

Currently, there is a lack of empirically validated usability evaluation methods that can properly be integrated during the early stages of Web development processes. This has motivated us to propose a usability inspection method called WUEP that can be integrated into different model-driven Web development processes. In previous work, we presented the operationalization and validation of WUEP in a specific process based on the Object-Oriented Hypermedia (OO-H) method. In this paper, we present further analysis of the empirical validation of the operationalization of WUEP into WebML, which is one of the most well-known industrial model-driven Web development process. The effectiveness, efficiency, perceived ease of use, and satisfaction of WUEP was evaluated in comparison to Heuristic Evaluation. The results show that WUEP is more effective and efficient than heuristic evaluation in the detection of usability problems. The inspectors were also satisfied when applying WUEP, and found it easier to use than heuristic evaluation.

While business process definition is becoming more and more popular as an instrument for describing human activities, there is a growing need for software tools supporting business process abstractions to help users organize and monitor their desktop work. Tools are most effective when they embed some knowledge about the process, e.g., in terms of the typical activities required by the process, so that users can execute the activities without having to define them. Tools must be lightweight and flexible, so as to enable users to create or change the process as soon as there is a new need. In this article, we first describe an application-independent approach to flexible process support by discussing the abstractions required for modeling, creating, enacting, and modifying flexible processes. Then, we show our approach at work in the context of project-centered learning. In this application, learners are challenged to perform concrete tasks in order to master specific subjects; in doing so, they have to conduct significant projects and cope with realistic (or even real-life) working conditions and scenarios. Often, students are geographically dispersed or under severe timing constraints, because these activities intertwine with their normal university activity. As a result, they need communication technology in order to interact and workflow technology in order to organize their work. The developed platform provides a comprehensible, e-learning-specific set of activities and process templates, which can be combined through a simple Web interface into project-centered collaboration processes. We discuss how the general paradigm of flexible processes was adapted to the learning concept, implemented, and experienced by students.

The development of user interfaces (UIs) is one of the most time-consuming aspects in software development. In this context, the lack of proper reuse mechanisms for UIs is increasingly becoming manifest, especially as software development is more and more moving toward composite applications. In this paper we propose a framework for the integration of stand-alone modules or applications, where integration occurs at the presentation layer. Hence, the final goal is to reduce the effort required for UI development by maximizing reus.

Context-aware, multi-channel Web applications are more and more gaining consensus among both content providers and consumers, but very few proposals exist for their conceptual modeling. This article illustrates a conceptual framework that provides modeling facilities for context-aware, multichannel Web applications; it also shows how high-level modeling constructs can drive the application development process through automatic code generation. Our work stresses the importance of user-independent, context-triggered adaptation actions, in which the context plays the role of a “first class” actor , operating independently of users on the same hypertext the users navigate. Modeling concepts are based on WebML (Web Modeling Language), an already established conceptual model for data-intensive Web applications, which is also accompanied by a development method and a CASE tool. However, given their general validity, the concepts of this article shape up a complete framework that can be adopted independently of the chosen model, method, and tool.

As the amount of information on the World Wide Web continues to grow, efficient hypertext navigation mechanisms are becoming crucial. Among them, effective history mechanisms play an important role. We therefore decided to provide a new method to access users' navigation histories, called xMem (Extended Memory Navigation), building on semantic-based and associative accesses, so as to imitate some of the features of the human memory. Such a memory may give users better understanding of the context of their searches, intermixing semantic aspects with the temporal dimension.The paper presents the experimental study conducted on the xMem approach to revisit the Web interaction history. Two controlled experiments have been performed with the aim to evaluate the effectiveness of the xMem history mechanism with respect to traditional Web browser histories. The results from the first experiment show a clear advantage, in terms of the time needed to complete a retrieving task, for the subjects that used the xMem prototype. Accordingly, users found retrieving previously visited pages with xMem more satisfying than using Web interaction histories sorted by the only time dimension. The results from the second experiment show the relevance in the process of information retrieval of clusters and keywords semantically related to the context of the search.

memorie semantiche per la navigazione web centrata sull'utente

It is indubitable that usability inspection of complex hypermedia is still an "art," in the sense that a great deal is left to the skills, experience, and ability of the inspectors. Training inspectors is difficult and often quite expensive. The Systematic Usability Evaluation (SUE) inspection technique has been proposed to help usability inspectors share and transfer their evaluation know-how, to simplify the hypermedia inspection process for newcomers, and to achieve more effective and efficient evaluation results. SUE inspection is based on the use of evaluation patterns, called abstract tasks, which precisely describe the activities to be performed by evaluators during inspection. This article highlights the advantages of this inspection technique by presenting its empirical validation through a controlled experiment. Two groups of novice inspectors were asked to evaluate a commercial hypermedia CD-ROM by applying the SUE inspection or traditional heuristic evaluation. The comparison was based on three major dimensions: effectiveness, efficiency, and satisfaction. Results indicate a clear advantage of the SUE inspection over the traditional inspection on all dimensions, demonstrating that abstract tasks are efficient tools to drive the evaluator's performance.

Abstract This paper introduces a model‐driven approach to the design of collaborative Web‐based applications, i.e. applications in which several users play different roles, in a collaborative way, to pursue a specific goal. The paper illustrates a conference management application (CMA), whose main requirements include: (i) the management of users profiles and access rights based on the role played by users during the conference life cycle; (ii) the delivery of information and services to individual users; (iii) the management of the sequence of activities that lead to the achievement of a common goal. The presented approach is based on WebML, a conceptual modelling language for the Web. The paper also highlights some general properties—as understood by the practical experience of CMA development—that a Web modelling language should feature in order to fully support the development of collaborative applications. Copyright © 2003 John Wiley & Sons, Ltd.

SUE inspection is a novel usability evaluation method for hypermedia, which falls into the category of inspection methods. Its primary goal is to favor a systematic usability evaluation (SUE), for supplying hypermedia usability inspectors with a structured flow of activities, allowing them to obtain more reliable, comparable, and cost-effective evaluation results. This is obtained especially due to the use of evaluation patterns, called abstract tasks (AT), which describe in details the activities that evaluators must carry out during inspection. AT helps share and transfer evaluation know-how among different evaluators, thus making it easier to learn the SUE inspection method by newcomers. A further notable advantage provided by the SUE inspection over other existent approaches is that it focuses on navigation and information structures, making evident some problems that other "surface-oriented" approaches might not reveal. This paper describes the SUE inspection for hypermedia. It also reports on a validation experiment, involving 28 evaluators, that has shown the major effectiveness and efficiency of the SUE inspection for hypermedia, with respect to traditional heuristic evaluation techniques.

In the usability inspection of complex hypermedia a great deal is left to the skills, experience, and ability of the inspectors. The SUE inspection technique has been proposed to help usability inspectors share and transfer their evaluation know-how, make it easier the hypermedia inspection process for newcomers, and achieve more effective and efficient evaluations. The SUE inspection is based on the use of evaluation patterns, called Abstract Tasks, which precisely describe the activities to be performed by evaluators during inspection. This paper presents an empirical validation of this inspection technique: two groups of novice inspectors have been asked to evaluate a commercial hypermedia CD-ROM applying the SUE inspection or the traditional heuristic evaluation technique. Results have shown a clear advantage of the SUE inspection over the heuristic evaluation, demonstrating that Abstract Tasks are efficient tools to drive evaluator's performance.

This paper presents a systematic approach to the heuristic evaluation of hypermedia that specifically addresses the peculiar features of this class of systems. We propose a set of hypermedia-specific usability attributes and define general "patterns of evaluation activities" called abstract tasks that can be performed by usability experts to check such attributes systematically. The usage of abstract tasks makes application inspection more effective, since it guides the work of evaluators and supports standardization across different evaluators. Our approach is model-based since we use a hypermedia design model (HDM) to identify the constituents of an application that represent the "focus of interest" for the evaluation, to decompose general usability principles into hypermedia-specific attributes, and finally to formulate abstract tasks in a precise way. To show the effectiveness of our method, we discuss the most significant usability problems discovered in eight commercially available hypermedia CD-ROM's.