Graphical user interface (GUI) customization relies on predefined configuration options and settings, constraining diverse individual needs and preferences within predetermined boundaries and often requiring technical expertise. To address these limitations, this work introduces MorphGUI, a framework leveraging Large Language Models (LLMs) to enable interface customization through natural language. By allowing users to express desired changes using their own words and harnessing the generative capabilities of LLMs, MorphGUI mitigates the limitations of predefined options and reduces the need for technical expertise. The framework translates functional and stylistic requests into either modifications of existing application components or generation of new ones. Through a use case implementation with a calendar application and a user study (n=18), where participants were tasked with modifying interfaces towards a target goal, we investigate if MorphGUI can enable effective natural language-driven interface customization for non-expert users through both functional and visual modifications. Results show that participants successfully customized interfaces using natural language. Users found the system intuitive and achieved good performance regardless of technical background, we report analysis of optimal prompt length, challenges in separating functional and visual instructions in structured templates, correlation between LLM experience and success, and learning effects. The study revealed opportunities for enhanced guidance, examples, and scaffolding to help users structure their customization requests more effectively. • MorphGUI enables real-time GUI customization through natural language. • Dual-input approach reduces ambiguity in customization requests. • Users effectively achieve intended interfaces regardless of complexity level. • MorphGUI system achieves acceptable usability (SUS=68) for non-experts.

Intelligent Tutoring Systems (ITSs) have shown great potential in enhancing how education is delivered. Many existing ITSs leverage Reinforcement Learning (RL) to optimize the sequence of exercises proposed to the learner. These systems adapt content based on the student's performance on previous exercises, addressing knowledge gaps while advancing through mastered concepts. However, they typically operate at the whole-exercise level, without visibility into the intermediate steps. In reality, learners may fail to solve an exercise because they encounter difficulties with specific sub-steps. Existing ITSs rely on datasets that do not include exercise decomposition in steps. To overcome this limitation, in this paper, we employ GPT-o3-mini to generate synthetic step-by-step solutions for mathematics exercises from the Junyi Academy dataset. To evaluate if these synthetic steps are useful in reaching the final solution, we use three models of varying size from the Llama family to simulate students of different knowledge levels (i.e., low, medium, high) and verify if the step-by-step guidance increases their problem-solving capabilities. By comparing direct answers for exercises to answers that leverage an incremental step guidance strategy, models successfully solve up to 42% more exercises. This evaluation serves as a foundation for creating synthetic step-by-step solutions that can be employed to develop next-generation step-aware ITSs tailored to students' specific knowledge gaps.

Recent advancements in AI, particularly Generative AI (GenAI) and Large Language Models (LLMs), have facilitated the integration of AI techniques into digital wellbeing applications, i.e., digital tools that aim at helping people’s wellbeing as a sum of mental and emotional wellness. These AI-powered systems hold the potential to foster healthier habits by collecting and analyzing user behavioral data to provide personalized and dynamic solutions tailored to each user’s needs and lifestyle, therefore improving the efficacy with respect to traditional non-AI interventions. Yet, their development presents significant challenges, including ethical concerns, privacy risks, and the potential for over-reliance on automated interventions. In this paper, we conduct a systematic literature review to examine the key characteristics, challenges, and opportunities in the existing research about AI-powered digital wellbeing tools. Based on our findings, we propose a design framework that outlines 6 critical dimensions and 23 sub-dimensions, spacing from user data and privacy to intervention strategies and personalization, offering practical guidance for researchers and practitioners developing AI-powered digital wellbeing applications. The framework emphasizes the importance of developing tailored and adaptive user-centered interventions adhering to scientific principles, psychological models and responsible data collection. We discuss the applicability and utility of our framework in evaluating and guiding the integration of AI in digital wellbeing applications.

Manually translating web designs into code is a costly and time-consuming process, particularly due to the frequent iterations and refinements between designers and developers. Deep learning techniques, which possess the capability to automatically translate designs into functional code using an encoder-decoder architecture, have emerged as a promising solution to enhance this tedious process. However, many current methods depend on simplistic datasets that do not capture the diversity of components found in modern websites. Additionally, the potential of transformer-based models, which have enabled significant progress in vision and language modeling tasks due to their scalability and ability to handle cross-modal relationships, has not been investigated in this context. Addressing these limitations, this paper contributes with: 1) a web scraping methodology to automatically collect and process a diverse dataset of real-world websites with reduced noise and complexity, 2) a synthetic dataset of webpage mockups along with their sketched conversions, and 3) an evaluation of two recent multimodal transformer architectures on these proposed datasets. Results on synthetic and sketch-based datasets demonstrate the architectures potential as effective design-to-code automation solutions, while identifying remaining challenges in modeling real-world website complexity.

The HCI research community has traditionally considered digital wellbeing an end-user responsibility, designing tools for digital self-control that support them to self-regulate their usage of apps and Web sites. Yet, these attempts are often ineffective in the long term, as many tech companies still adopt “attention-capture” designs that compromise users’ sense of agency and self-control. Taking a complementary perspective, this article presents a set of eight heuristics to create user interfaces that preserve and respect user attention by design . The heuristics stem from a systematic literature review and are grounded in the three fundamental psychological needs defined by the self-determination theory, i.e., autonomy, competence, and relatedness. In addition to being informed by theory and research, each heuristic is accompanied by practical strategies and real-world examples, offering designers actionable guidelines to value people’s attention in user interfaces.

Recent research has identified the detrimental consequences stemming from the pervasive and excessive use of digital devices, prompting the emergence of the concept of digital wellbeing. This workshop serves as a platform for both researchers and practitioners to convene and delve into discussions surrounding the imperative task of educating young generations on digital wellbeing. Participants will engage in a collaborative exploration of innovative strategies and tools aimed at fostering a more mindful and conscious engagement with digital platforms. Through shared insights and collective expertise, the workshop aims at paving the way for a healthier and more balanced relationship with technology among today’s teenagers.

User interfaces heavily rely on attention-capture design patterns, e.g., infinite scroll and other variable-rewarding mechanisms, that erode users’ sense of autonomy and undermine their digital wellbeing. Instead of having users rely on external self-regulation tools, this paper advocates that tech companies and designers should prioritize users’ digital wellbeing by design. To take a first step in this direction, we present Digital Wellbeing Lens, a Figma plugin that guides designers in creating user interfaces that respect and preserve user time and attention. The plugin allows the continuous evaluation of prototypes against attention-capture patterns, calculating a digital wellbeing score and suggesting suitable design alternatives. Besides introducing the plugin, we demonstrate its practical application through a use case involving the design of a mobile social media app, and we report on the results of a first exploratory study with four designers, discussing the opportunities and challenges of embracing this paradigm shift.

Smartphones have become integral to everyday life, and despite the many advantages they bring, they also raise privacy and security concerns, particularly regarding nontransparent, unauthorized, or malicious data collection risks. While users are aware of these issues, the means the smart-phone’s operating system provides to manage the applications’ permissions and protect personal data often prove challenging or overwhelming for non-technical users. In this context, this article introduces Privacy Manager, a mobile application designed for Android devices that relies on an End-User Development (EUD) approach to offer personalized data protection and privacy security measures to bridge the gap for regular users. The application facilitates effective management of permissions granted to installed applications through user-friendly interfaces and customizable settings, enhancing user awareness and control over their data. A validation study was carried out with 8 participants over a week. The study involved an initial questionnaire followed by a week-long period of real-world application usage. It concluded with a final questionnaire with an assessment using the System Usability Scale (SUS).

Abstract Smart speakers are entering our homes and enriching the connected ecosystem already present in them. Home inhabitants can use those to execute relatively simple commands, e.g., turning a lamp on. Their capabilities to interpret more complex and ambiguous commands (e.g., make this room warmer) are limited, if not absent. Large language models (LLMs) can offer creative and viable solutions to enable a practical and user-acceptable interpretation of such ambiguous commands. This paper introduces an interactive disambiguation approach that integrates visual and textual cues with natural language commands. After contextualizing the approach with a use case, we test it in an experiment where users are prompted to select the appropriate cue (an image or a textual description) to clarify ambiguous commands, thereby refining the accuracy of the system’s interpretations. Outcomes from the study indicate that the disambiguation system produces responses well-aligned with user intentions, and that participants found the textual descriptions slightly more effective. Finally, interviews reveal heightened satisfaction with the smart-home system when engaging with the proposed disambiguation approach.

Many tech companies exploit psychological vulnerabilities to design digital interfaces that maximize the frequency and duration of user visits. Consequently, users often report feeling dissatisfied with time spent on such services. Prior work has developed typologies of damaging design patterns (or dark patterns) that contribute to financial and privacy harms, which has helped designers to resist these patterns and policymakers to regulate them. However, we are missing a collection of similar problematic patterns that lead to attentional harms. To close this gap, we conducted a systematic literature review for what we call ‘attention capture damaging patterns’ (ACDPs). We analyzed 43 papers to identify their characteristics, the psychological vulnerabilities they exploit, and their impact on digital wellbeing. We propose a definition of ACDPs and identify eleven common types, from Time Fog to Infinite Scroll. Our typology offers technologists and policymakers a common reference to advocate, design, and regulate against attentional harms.

This review aims to analyze the recent literature to find how the preattentive visual process is currently used in information visualization, particularly to improve the cognitive process in chart comprehension (i.e., perceptual effectiveness). The purpose of our literature review is to provide an overview of how concepts related to the preattentive process are used pragmatically in recent research. We searched different bibliography sources between 2010 and 2021, getting 29 articles that fit the review focus. In general, we discovered that the research work on exploiting the preattentive process in information visualization is currently not thoroughly explored. The main contribution of the paper is the analysis of the papers, from which we identified two categories of research directions according to the primary uses of preattentive concepts: “preattentive attributes as design components,” and “the preattentive process as a measuring tool,” with a gap between these two approaches. The review also highlighted two limitations in the current research literature: most works tend to focus on a particular chart type, only, with difficult to generalize results, and the manipulation of preattentive attributes is done implicitly, without providing to the graph designer the suitable awareness over the design decisions impact. We finally present a proposal about how to use the knowledge about the preattentive process in the first stages of design in information visualization to start closing the gap mentioned above with the graph designer.

Public media and researchers in different areas have recently focused on perhaps unexpected problems that derive from an excessive and frequent use of technology, giving rise to a new kind of psychological “digital” wellbeing. Such a novel and pressing topic has fostered, both in the academia and in the industry, the emergence of a variety of digital self-control tools allowing users to self-regulate their technology use through interventions like timers and lock-out mechanisms. While these emerging technologies for behavior change hold great promise to support people’s digital wellbeing, we still have a limited understanding of their real effectiveness, as well as of how to best design and evaluate them. Aiming to guide future research in this important domain, this article presents a systematic review and a meta-analysis of current work on tools for digital self-control. We surface motivations, strategies, design choices, and challenges that characterize the design, development, and evaluation of digital self-control tools. Furthermore, we estimate their overall effect size on reducing (unwanted) technology use through a meta-analysis. By discussing our findings, we provide insights on how to (i) overcome a limited perspective that exclusively focuses on technology overuse and self-monitoring tools, (ii) evaluate digital self-control tools through long-term studies and standardized measures, and (iii) bring ethics in the digital wellbeing discourse and deal with the business model of contemporary tech companies.

In the contemporary attention economy, tech companies design the interfaces of their digital platforms by adopting attention-capture dark patterns to drive their behavior and maximize time spent and daily visits. Two popular examples are viral recommendations and content autoplay on social networks. As these patterns exploit people’s psychological vulnerabilities and may contribute to technology overuse and problematic behaviors, there is the need of promoting the design of technology that better align with people’s digital wellbeing. This workshop seeks to advance this timely and urgent need, by inviting researchers and practitioners in interdisciplinary domains to engage in conversation around the design of interfaces that allow people to take advantage of digital platforms in a meaningful and conscious way.

In the development projects implemented by novices, the usefulness of the documentation in the form of comments on the final working code is minimal to guide future implementations. Such documentation does not account for novices’ development process, including their choices, the errors they faced, the solutions they found, the sources they consulted, the lessons learned, and the advice to remember or give to someone else. Indeed, novices do not usually rely on their documentation to keep track of the successes and errors they find during the development process. Nevertheless, if enabled to capture various moments of the process seamlessly, novices can produce documentation that has the potential to become a valuable asset for them and other developers. This paper presents Dear Diary, a tool to support non-expert programmers in straightforwardly creating documentation artifacts directly from the IDE.

Nowadays, end users can exploit end-user development (EUD) platforms to personalise their Internet of Things (IoT) ecosystems, typically through trigger–action rules. Unfortunately, within such platforms, users are forced to adopt a unique, vendor-centric abstraction: to define triggers and actions, they must specifically refer to every single device or online service needed to execute the intended behaviours. As a consequence, little social and practical benefits of EUD in this domain have emerged so far. In this paper, we build on the idea that other abstractions besides the vendor-centric one are possible, and that the growth of end-user personalisation in the IoT may depend on their identification. Specifically, we report on the results of a 1-week-long diary study during which 24 participants were free to collect trigger–action rules arising during their daily activities. First, we demonstrate that users would adopt different abstractions by personalising devices, information and people-related behaviours, where the individual is at the centre of the interaction. Then, we show that the adopted abstraction may depend on different factors, ranging from the user profile, e.g. their programming experience, to the context in which the personalisation is introduced. While users are inclined to personalise physical objects in the home, for example, they often go ‘beyond devices’ in the city, where they are more interested in the underlying information. Finally, we discuss the retrieved results by identifying new design opportunities to improve the relationship between users and the IoT.

While Digital Self-Control Tools (DSCTs) mainly target smartphones, more effort should be put into evaluating multi-device ecosystems to enhance digital wellbeing as users typically use multiple devices at a time. In this paper, we first review more than 300 DSCTs by demonstrating that the majority of them implements a single-device conceptualization that poorly adapts to multi-device settings. Then, we report on the results from an interview and a sketching exercise (N=20) exploring how users make sense of their multi-device digital wellbeing. Findings show that digital wellbeing issues extend beyond smartphones, with the most problematic behaviors deriving from the simultaneous usage of different devices to perform uncorrelated tasks. While this suggests the need of DSCTs that can adapt to different and multiple devices, our work also highlights the importance of learning how to properly behave with technology, e.g., through educational courses, which may be more effective than any lock-out mechanism.

Several tools have been developed to support novices learning to program. Most of them focus on the code and provide features regarding the visualization of the data structures or the debugging. However, in introductory programming courses, students are typically given exercises in the form of a problem written in natural language; and the first challenge they face is understanding the problem, identifying the relevant information, and then translating that information into code. To our knowledge, little attention has been paid to proposing tools targeted at supporting this problem-solving step, even though it is crucial for deriving a correct solution. In this paper, we present an IDE to encourage novices to understand the problem before start coding, decompose it down into subproblems, explore alternative implementations for each subproblem, and arrange these implementations to build a general solution. Finally, the adopted problem-solving approach is discussed.

People, especially young adults, often use their smartphones out of habit: They compulsively browse social networks, check emails, and play video-games with little or no awareness at all. While previous studies analyzed this phenomena qualitatively , e.g., by showing that users perceive it as meaningless and addictive, yet our understanding of how to discover smartphone habits and mitigate their disruptive effects is limited. Being able to automatically assess habitual smartphone use, in particular, might have different applications, e.g., to design better “digital wellbeing” solutions for mitigating meaningless habitual use. To close this gap, we first define a data analytic methodology based on clustering and association rules mining to automatically discover complex smartphone habits from mobile usage data. We assess the methodology over more than 130,000 phone usage sessions collected from users aged between 16 and 33, and we show evidence that smartphone habits of young adults can be characterized by various types of links between contextual situations and usage sessions, which are highly diversified and differently perceived across users. We then apply the proposed methodology in Socialize, a digital wellbeing app that (i) monitors habitual smartphone behaviors in real time and (ii) uses proactive notifications and just-in-time reminders to encourage users to avoid any identified smartphone habits they consider as meaningless. An in-the-wild study with 20 users (ages 19–31) demonstrates that Socialize can assist young adults in better controlling their smartphone usage with a significant reduction of their unwanted smartphone habits.

In the Internet of Things era, users are willing to personalize the joint behavior of their connected entities, i.e., smart devices and online service, by means of trigger-action rules such as “IF the entrance Nest security camera detects a movement, THEN blink the Philips Hue lamp in the kitchen.” Unfortunately, the spread of new supported technologies makes the number of possible combinations between triggers and actions continuously growing, thus motivating the need of assisting users in discovering new rules and functionality, e.g., through recommendation techniques. To this end, we present , a semantic Conversational Search and Recommendation (CSR) system able to suggest pertinent IF-THEN rules that can be easily deployed in different contexts starting from an abstract user’s need. By exploiting a conversational agent, the user can communicate her current personalization intention by specifying a set of functionality at a high level, e.g., to decrease the temperature of a room when she left it. Stemming from this input, implements a semantic recommendation process that takes into account ( a ) the current user’s intention , ( b ) the connected entities owned by the user, and ( c ) the user’s long-term preferences revealed by her profile. If not satisfied with the suggestions, then the user can converse with the system to provide further feedback, i.e., a short-term preference , thus allowing to provide refined recommendations that better align with the original intention. We evaluate by running different offline experiments with simulated users and real-world data. First, we test the recommendation process in different configurations, and we show that recommendation accuracy and similarity with target items increase as the interaction between the algorithm and the user proceeds. Then, we compare with other similar baseline recommender systems. Results are promising and demonstrate the effectiveness of in recommending IF-THEN rules that satisfy the current personalization intention of the user.

In the Internet of Things era, users are willing to personalize the joint behavior of their connected entities, i.e., smart devices and online service, by means of IF-THEN rules. Unfortunately, how to make such a personalization effective and appreciated is still largely unknown. On the one hand, contemporary platforms to compose IF-THEN rules adopt representation models that strongly depend on the exploited technologies, thus making end-user personalization a complex task. On the other hand, the usage of technology-independent rules envisioned by recent studies opens up new questions, and the identification of available connected entities able to execute abstract users' needs become crucial. To this end, we present HeyTAP, a conversational and semantic-powered trigger-action programming platform able to map abstract users' needs to executable IF-THEN rules. By interacting with a conversational agent, the user communicates her personalization intentions and preferences. User's inputs, along with contextual and semantic information related to the available connected entities, are then used to recommend a set of IF-THEN rules that satisfies the user's needs. An exploratory study on 8 end users preliminary confirms the effectiveness and the appreciation of the approach, and shows that HeyTAP can successfully guide users from their needs to specific rules.

We are witnessing unprecedented datafication of the society we live in, alongside rapid advances in the fields of Artificial Intelligence and Machine Learning. However, emergent data-driven applications are systematically discriminating against many diverse populations. A major driver of the bias are the data, which typically align with predominantly Western definitions and lack representation from multilingually diverse and resource-constrained regions across the world. Therefore, data-driven approaches can benefit from integration of a more human-centred orientation before being used to inform the design, deployment, and evaluation of technologies in various contexts. This workshop seeks to advance these and similar conversations, by inviting researchers and practitioners in interdisciplinary domains to engage in conversation around how appropriate human-centred design can contribute to addressing data-related challenges among marginalised and under-represented/underserved groups.

Nowadays, users can personalize the joint behavior of their connected entities, i.e., smart devices and online service, by means of trigger-action rules. As the number of supported technologies grows, however, so does the design space, i.e., the combinations between different triggers and actions: without proper support, users often experience difficulties in discovering rules and their related functionality. In this paper, we introduce TAPrec, an End-User Development platform that supports the composition of trigger-action rules with dynamic recommendations. By exploiting a hybrid and semantic recommendation algorithm, TAPrec suggests, at composition time, either a) new rules to be used or b) actions for auto-completing a rule. Recommendations, in particular, are computed to follow the user's high-level intention, i.e., by focusing on the rules' final purpose rather than on low-level details like manufacturers and brands. We compared TAPrec with a widely used trigger-action programming platform in a study on 14 end users. Results show evidence that TAPrec is appreciated and can effectively simplify the personalization of connected entities: recommendations promoted creativity by helping users personalize new functionality that are not easily noticeable in existing platforms.

In information visualization (InfoVis), the Visualizers (graph designers and creators) have to consider multiple parameters, such as colors and graphic symbols, to obtain a chart that correctly represents a data set. Along with this, visualizers must adequately select the combination of these range of parameters to drive the observers' attention to the relevant data. When a visualizer drives the attention to relevant aspects of the information, she is providing a starting point to read the graph; this focus point might help the observer to complete the task faster and more efficiently, minimizing distraction from unimportant information. Contemporary tools for InfoVis help visualizers to a certain extent, but most of them currently do not provide insights or suggestions about the modifications needed to drive data attention. This article presents the preliminary results of an exploratory approach to draw the attention to some specific data subset selected by the graph creator, through a systematic variation of some preattentive attributes (i.e., color, texture and orientation). As a first simple method to validate the feasibility of the approach, a set of charts is created from the same source data, with exhaustive variations on preattentive attributes. All generated charts are then automatically evaluated using a salience map algorithm for data analysis images, to identify their focus attention point. After that, the algorithm chooses the chart that best emphasizes the data subset initially specified by the visualizer. To validate our approach, we have implemented a prototype tool, and preliminary results confirm that it is possible to systematically change the attention area of a chart.

End users can program trigger-action rules to personalize the joint behavior of their smart devices and online services. Trigger-action programming is, however, a complex task for non-programmers and errors made during the composition of rules may lead to unpredictable behaviors and security issues, e.g., a lamp that is continuously flashing or a door that is unexpectedly unlocked. In this paper, we introduce EUDebug, a system that enables end users to debug trigger-action rules. With EUDebug, users compose rules in a web-based application like IFTTT. EUDebug highlights possible problems that the set of all defined rules may generate and allows their step-by-step simulation. Under the hood, a hybrid Semantic Colored Petri Net (SCPN) models, checks, and simulates trigger-action rules and their interactions. An exploratory study on 15 end users shows that EUDebug helps identifying and understanding problems in trigger-action rules, which are not easily discoverable in existing platforms.

As smartphone use increases dramatically, so do studies about technology overuse. Many different mobile apps for breaking "smartphone addiction" and achieving "digital wellbeing" are available. However, it is still not clear whether and how such solutions work. Which functionality do they have? Are they effective and appreciated? Do they have a relevant impact on users' behavior? To answer these questions, (i) we reviewed the features of 42 digital wellbeing apps, (ii) we performed a thematic analysis on 1,128 user reviews of such apps, and (iii) we conducted a 3-week-long in-the-wild study of Socialize, an app that includes the most common digital wellbeing features, with 38 participants. We discovered that digital wellbeing apps are appreciated and useful for some specific situations. However, they do not promote the formation of new habits and they are perceived as not restrictive enough, thus not effectively helping users to change their behavior with smartphones.

Semantic technologies and Linked Data are increasingly adopted as core application modules, in many knowledge domains and involving various stakeholders: ontology engineers, software architects, doctors, employees, etc. Such a diffusion calls for better access to models and data, which should be direct, mobile, visual and time effective. While a relevant core of research efforts investigated the problem of ontology visualization, discovering different paradigms, layouts, and interaction modalities, a few approaches target mobile devices such as tablets and smartphones. Touch interaction, indeed, has the potential of dramatically improving usability of Linked Data and of semantic-based solutions in real-world applications and mash-ups, by enabling direct and tactile interactions with involved knowledge objects. In this paper, we move a step towards touch-based, mobile interfaces for semantic models by presenting an ontology browsing platform for Android devices. We exploit state of the art touch-based interaction paradigms, e.g., pie menus, pinch-to-zoom, etc., to empower effective ontology browsing. Our research mainly focuses on interactions, yet providing support to different visualization approaches thanks to a clear decoupling between model-level operation and visual representations. Presented results include the design and implementation of a working prototype application, as well as a first validation involving habitual users of semantic technologies. Results show a low learning curve and positive reactions to the proposed paradigms, which are perceived as both innovative and useful.

Nowadays, end users can personalize their smart devices and web applications by defining or reusing IF-THEN rules through dedicated End-User Development (EUD) tools. Despite apparent simplicity, such tools present their own set of issues. The emerging and increasing complexity of the Internet of Things, for example, is barely taken into account, and the number of possible combinations between triggers and actions of different smart devices and web applications is continuously growing. Such a large design space makes end-user personalization a complex task for non-programmers, and motivates the need of assisting users in easily discovering and managing rules and functionality, e.g., through recommendation techniques. In this article, we tackle the emerging problem of recommending IF-THEN rules to end users by presenting RecRules , a hybrid and semantic recommendation system. Through a mixed content and collaborative approach, the goal of RecRules is to recommend by functionality : it suggests rules based on their final purposes, thus overcoming details like manufacturers and brands. The algorithm uses a semantic reasoning process to enrich rules with semantic information, with the aim of uncovering hidden connections between rules in terms of shared functionality. Then, it builds a collaborative semantic graph, and it exploits different types of path-based features to train a learning to rank algorithm and compute top-N recommendations. We evaluate RecRules through different experiments on real user data extracted from IFTTT, one of the most popular EUD tools. Results are promising: they show the effectiveness of our approach with respect to other state-of-the-art algorithms and open the way for a new class of recommender systems for EUD that take into account the actual functionality needed by end users.

The usage of smartphone-based virtual assistants (e.g., Siri or Google Assistant) is growing, and their spread has generally a positive impact on device accessibility, e.g., for people with disabilities. However, people with dysarthria or other speech impairments may be unable to use these virtual assistants with proficiency. This paper investigates to which extent people with ALS-induced dysarthria can be understood and get consistent answers by three widely used smartphone-based assistants, namely Siri, Google Assistant, and Cortana. We focus on the recognition of Italian dysarthric speech, to study the behavior of the virtual assistants with this specific population for which no relevant studies are available. We collected and recorded suitable speech samples from people with dysarthria in a dedicated center of the Molinette hospital, in Turin, Italy. Starting from those recordings, the differences between such assistants, in terms of speech recognition and consistency in answer, are investigated and discussed. Results highlight different performance among the virtual assistants. For speech recognition, Google Assistant is the most promising, with around 25% of word error rate per sentence. Consistency in answer, instead, sees Siri and Google Assistant provide coherent answers around 60% of times.

Most engineering courses include fundamental practice activities to be performed by students in computer labs. During lab sessions, students work on solving exercises with the help of teaching assistants, who often have a hard time for guaranteeing a timely, optimized, and "democratic" support to everybody. This paper presents a learning environment to improve the experience of the lab sessions participants, both the students and the teaching assistants. In particular, the environment was designed, implemented, and experimented in the context of a database course. The application designed to support the learning environment stores all the events occurring during a SQL practice lab, i.e., task progression, query submissions, error feedback, assistance requests and interventions, and it provides information useful both for use on-the-fly and for later analysis. Thanks to the analysis of these data, the application dynamically provides teaching assistants with a graphical interface highlighting where assistance is most needed, by considering different factors such as the progression rate, the percentage of correct solutions, and the difficulties in solving the current exercise. Furthermore, the stored data allow teachers later on to analyze and to interpret the behavior of the students during the lab, and to have insights on their main mistakes and misconceptions. After describing the environment, the interfaces, and the approaches used to identify the students' teams that need timely assistance, the paper presents the results of different analyses performed using the collected data, to help the teacher better understand students' educational needs.

Administrators and operators of next generation cities will likely be required to exhibit a good understanding of technical features, data issues, and complex information that, up to few years ago, were quite far from day-to-day administration tasks. In the smart city era, the increased attention to data harvested from the city fosters a more informed approach to city administration, requiring involved operators to drive, direct, and orient technological processes in the city more effectively. Such an increasing need requires tools and platforms that can easily and effectively be controlled by nontechnical people. In this paper, an approach for enabling “easier” composition of real-time data processing pipelines in smart cities is presented, exploiting a visual and block-based design approach, similar to the one adopted in the Scratch programming language for elementary school students. The proposed approach encompasses both a graphical editor and a sound methodology and workflow, to allow city operators to effectively design, develop, test, and deploy their own data processing pipelines. The editor and the workflow are described in the context of a pilot of the ALMANAC European project.

During the last decade, network providers are faced by a growing problem regarding the distribution of bandwidth and computing resources. Recently, the mobile edge computing paradigm was proposed as a possible solution, mainly in consideration of the provided possibility of transferring service demands at the edge of the network. This solution heavily relies on the dynamic allocation of resources, depending on the user needs and network connection, therefore it becomes essential to correctly predict user movements and activities. This paper proposes an unsupervised methodology to define meaningful user locations from noninvasive user information, captured by the user terminal with no computing or battery overhead. The data is analyzed through a conjoined clustering algorithm to build a stochastic Markov chain to predict the users’ movements and their bandwidth demands. Such a model could be used by network operators to optimize network resources allocation. To evaluate the proposed methodology, we tested it on one of the largest public community’s labeled mobile and sensor dataset, developed by the “CrowdSignals.io” initiative, and we present positive and promising results concerning the prediction capabilities of the model.

This paper presents an exploratory study aimed at identifying the pain points that novice programmers experience, from the software engineering perspective, when developing and deploying smart and distributed systems, that may be classified as Ambient Intelligence (AmI) systems. The exploratory study was conducted among undergraduate students, that worked in groups for developing AmI projects during a university course. Based on their own experiences, individually and as a group, the pain points were identified and prioritized over a common architecture and a set of software development activities. The quantification of the pain points was based on the difficulty level that the students perceived on the development activities and the time they spent completing them. Results represent a starting point for the design of tools and methodologies targeted at overcoming the complexity that novice programmers face when developing AmI systems.

Presents the introductory welcome message from the conference proceedings. May include the conference officers' congratulations to all involved with the conference event and publication of the proceedings record.

In the field of Ambient Assisted Living a limited amount of research aims at supporting caregivers that work with people with disabilities in assisted living facilities (ALFs). In fact, research activities on healthcare support systems in AAL mainly focus on improving the quality of life for people in their own homes or supporting nurses and doctors in hospitals. This paper explores and applies the Internet of Things paradigm in the ALFs context. In particular, we present the design, the implementation, and the experimental evaluation of a system capable of supporting the daily activities of healthcare assistants that operate in ALFs for people with physical or cognitive disabilities. The solution combines wearable and mobile technologies to improve assistance requests and anomaly detection. With this healthcare support system, caregivers can be automatically alerted of potentially hazardous situations that happen to the inhabitants while these are out of sight. Furthermore, inhabitants can require assistance instantly and from any point of the facility. We evaluated the system in two ways. We performed a functional test with two professional caregivers, and we deployed the system in an ALF in Italy for 36 hours, collecting the opinions of the involved caregivers and inhabitants.

The Social Web is quickly becoming a way of life: millions of people, everywhere, use social network sites to stay connected with their friends, discover new people and activities, and share user-created contents. Moreover, the Social Web phenomenon experiments an astoundingly rapid growth that is not likely to slow down in the near future. At the same time, the borderline between social networks and social media is more and more blurred. This complex and evolving scenario requires a new generation of computer scientists and engineers that understand how to properly design software for supporting and fostering social interactions. This paper describes a university-level experience started four academic years ago with a Social Web course. The course uses a multidisciplinary and active learning approach by requesting the students to design and prototype a Social Web application, and the teachers provide an active support and follow-up along the semester. The paper presents the adopted teaching strategies and analyzes the attained learning outcomes, both from the qualitative and quantitative point of views.

Due to the increasing spread of location-aware applications, developers interest in user location estimation has grown in recent years. As users spend the majority of their time in few meaningful places (i.e., groups of near locations that can be considered as a unique place, such as home, school or the workplace), this paper presents a new energy efficient method to estimate user presence in a meaningful place. Specifically, instead of using commonly used but energy hungry methods such as GPS and network positioning techniques, the proposed method applies a Machine Learning algorithm based on Decision Trees, to predict the user presence in a meaningful place by collecting and analyzing: a) user activity, b) information from received notifications (receipt time, generating service, sender-receiver relationship), and c) device status (battery level and ringtone mode). The results demonstrate that, using 20 days of training data and testing the system with data coming from 14 persons, the accuracy (percentage of correct predictions) is 89.40% (standard deviation: 8.27%) with a precision of 89.04% and a recall of 89.40%. Furthermore, the paper analyzes the importance of each considered feature, by comparing the prediction accuracy obtained with different combinations of features.

This paper presents two one-switch games designed for children with severe motor disabilities, based on the GNomon framework. These mini games demonstrate that it is possible to make dynamic video games with time-dependent game mechanics and flexible layout configurations while being accessible and playable with a single switch. The games were designed in close collaboration with a team of speech therapists, physiotherapists, and psychologists from one of the Local Health Agencies in Turin, Italy. Moreover, the games have been already evaluated with a group of children with different motor impairments through a series of trials with encouraging results.

Nowadays home automation, with its increased availability, reliability and with its ever reducing costs is gaining momentum and is starting to become a viable solution for enabling people with disabilities to autonomously interact with their homes and to better communicate with other people. However, especially for people with severe mobility impairments, there is still a lack of tools and interfaces for effective control and interaction with home automation systems, and general–purpose solutions are seldom applicable due to the complexity, asynchronicity, time dependent behavior, and safety concerns typical of the home environment. This paper focuses on user–environment interfaces based on the eye tracking technology, which often is the only viable interaction modality for users as such. We propose an eye-based interface tackling the specific requirements of smart environments, already outlined in a public Recommendation issued by the COGAIN European Network of Excellence. The proposed interface has been implemented as a software prototype based on the ETU universal driver, thus being potentially able to run on a variety of eye trackers, and it is compatible with a wide set of smart home technologies, handled by the Domotic OSGi Gateway. A first interface evaluation, with user testing sessions, has been carried and results show that the interface is quite effective and usable without discomfort by people with almost regular eye movement control.