GoodIT '23: Proceedings of the 2023 ACM Conference on Information Technology for Social Good

The marine ecosystem faces a significant threat due to the release of human waste into the sea. One of the most challenging issues is identifying and removing small particles that settle on the sand. These particles can be ingested by local fauna or cause harm to the marine ecosystem. Distinguishing these particles from natural materials like shells and stones is difficult, as they blend in with the surroundings. To address this problem, we utilized the Litter On The Sand (LOTS) dataset, which comprises images of clean, dirty, and wavy sand from three different beaches. We established an initial benchmark on this dataset by employing state-of-the-art Deep Learning segmentation techniques. The evaluated models included MultiResU-Net, Half MultiResU-Net, and Quarter MultiResU-Net. The results revealed that the Half MultiResU-Net model outperformed the others for most types of sand analyzed, providing valuable insights for future efforts in combating marine litter and preserving the health of our marine ecosystems.

In this work, it is explored whether real-time EEG (Electroencephalography) can adjust the difficulty in a serious game focused on engagement, attention, and learning about plastic pollution in our oceans. Using EEG to balance the game around the players’ affective state by measuring brain activity in real time, it is aimed to better fit the player's skill level, enabling a stable flow state. The experimental study included 34 participants with an experimental group (n=17), and a control group (n=17). The experimental group played the game about the plastic pollution in our oceans with an adaptive difficulty adjustment (DDA) based on changes in their levels of attention and calm measured by EEG. The evaluation is based on a user engagement questionnaire, structured interviews, the EEG data, and a knowledge test. The results revealed high engagement in the game from both the experimental group and the control group. However, the participants in the control group were more attentive while playing the game and scored higher on all questions in the knowledge test compared to the experimental group. In conclusion, our study cannot provide evidence for using EEG-DDA to increase the engagement, attention, and learnings about pollution in the oceans in a serious game. However, there are still advantages for including EEG in game related research, and much future research is needed in how to provide optimal learning in serious games.

Stablecoins have emerged as a solution to the legal status and volatile exchange rate issues surrounding cryptocurrencies like Bitcoin. Terra is a blockchain protocol that enables the creation of stablecoins pegged to fiat currencies, including UST and KRT, which have gained popularity for their use in DeFi applications. However, in May 2022, TerraUSD (UST) lost its peg to the USD, collapsing to a value of 0.091 USD due to a massive withdrawal of UST from the Anchor Protocol, leading to a sharp drop in price. In this paper we analyze people’s opinions and thoughts about the topic using Twitter data and we propose a method to identify the accounts that acted as opinion leaders. We observed an increase in negativity in the month leading up to the collapse and identified the United States as the most prolific country in terms of tweets; we identified opinion leaders accounts and we showed that the number of followers is not an important metric to identify opinion leaders. The failure of the Terra project highlights the intrinsic fragility of algorithmic stablecoins, which can fail due to sudden fluctuations in demand and supply, and their potential for success and stability remains uncertain.

Yearly, more than 200 million malaria cases are recorded worldwide. Most of these cases are witnessed in less developed countries as the environments are not well-maintained, which forms breeding places for mosquitoes. Female mosquito-anopheles is responsible for malaria infection, dengue, chikungunya, and zika. Developing countries struggle to fight diseases; malaria still claims more than 400,000 lives annually. One current way to keep away anopheles mosquitoes is using commercially available electric liquid mosquito repellents, which can adversely affect the human body when used for extended periods. Furthermore, energy and sprays are wasted as they constantly work even without the presence of anopheles mosquitoes. We propose a low-cost IoT-based TinyML model that intelligently discharges the mosquito repellent when an anopheles mosquito is in the room. First, we prove the concept by exploring two lightweight deep learners with a 1D Convolution Neural Network (1D-CNN) and 2D Convolution Neural Network (2D-CNN) to classify raw sounds from mosquito wingbeats. We adopted a Leaky ReLU in building the 1D-CNN to speed up training and improve classification performance. Furthermore, we adopted batch normalization to avoid degradation and vanishing gradient problems. We implemented the experiments in an Edge impulse platform. Each of the CNN models recorded stable classification performance during the proof of concept study, while the 1D-CNN took less time and computing resources in training, validation, and testing. As we aimed to propose a low-cost solution, we evaluated the performance of the 1D-CNN-based prototype in the actual deployment by playing mosquito wingbeat sounds on a laptop which we placed next to it in intervals of 0.5, 1.0, 1.5, 2.0, 2.5, and 3 meters. The model showed promising results across distances and thus could be used to chase away mosquitoes in a room of small to medium size.

Geographical location is a crucial element of humanitarian response, outlining vulnerable populations, ongoing events, and available resources. Latest developments in Natural Language Processing may help in extracting vital information from the deluge of reports and documents produced by the humanitarian sector. However, the performance and biases of existing state-of-the-art information extraction tools are unknown. In this work, we develop annotated resources to fine-tune the popular Named Entity Recognition (NER) tools Spacy and roBERTa to perform geotagging of humanitarian texts. We then propose a geocoding method FeatureRank which links the candidate locations to the GeoNames database. We find that not only does the humanitarian-domain data improves the performance of the classifiers (up to F1 = 0.92), but it also alleviates some of the bias of the existing tools, which erroneously favor locations in the Western countries. Thus, we conclude that more resources from non-Western documents are necessary to ensure that off-the-shelf NER systems are suitable for the deployment in the humanitarian sector.

Young children's interactions with media influence their social play, which impacts their development. Seven years ago, at the start of a project developing a system to support preschool children's creative, collaborative roleplay, we identified a need for children's media specifically intended to set up play. In this paper, we present best practices (e.g., designing a balanced cast of characters with complementary skills) and lessons learned (e.g., to proactively counter themes that may encourage stereotypical play) over the course of this project. We also describe our ongoing and planned future work which aims to provide additional scaffolding for children who need extra support to engage with their peers and to create abstractions based on what we have found works well so others can leverage that knowledge as they create their own content.

Video games have many potential uses beyond pure entertainment, including their use in educational contexts. Yet, it remains really challenging to put together guidelines to design effective game-like interventions in educational contexts. This study examines existing work relating to gamification, game-based learning, and serious games, and finds there is still limited qualitative work concerning the student perspective and limited work developing pedagogical guidelines for developers wishing to develop effective game-based learning experiences. The study focuses on the perception of students in regard to game-based learning activities in the context of a BSc Computer Science online degree. Students enrolled in the online degree were invited to fill in an online survey after their experience with a selection of game-based learning activities in the online degree. Reflexive Thematic Analysis was used to evaluate the open-ended responses from 55 participants. First, quantitative and qualitative results revealed insightful information along with four overarching themes (“Complementary to lectures on topics that are usually hard or too abstract to teach”, “Allow students to take on identities and learn from different angles and perspectives”, “Balanced challenge and context relevance to minimise students wasting their time”, and “Reward players for their effort with meaningful rewards and provide a safe space for failure”), suggesting that game-based learning interventions offer more than just motivation and engagement. Second, technical and pedagogical principles emerged from the data analysis, proposing guidelines for future designers of game-based learning activities in similar educational contexts. Finally, the study provides a selection of twelve open-source and browser-based game-based learning activities, the ones students encountered in the BSc Computer Science online degree.

Automation and Artificial Intelligence (AI) are continuously advancing decision-making in public administrations. Research focuses on investigating benefits and challenges of AI agents that make decisions without human-intervention, namely AI-systems that independently exercise administrative discretion. However, little is known about citizens’ perspective of such autonomous systems. To address this gap, we investigated citizens’ fairness and satisfaction perceptions of an AI-based self-service kiosk that performs administrative discretion. Through a controlled Wizard of Oz experiment, we assessed how citizens perceived the kiosk’s decision to either impose or waive a fine. Precisely, we measured how engaging citizens in the decision-making process through a dialogue affect their perceptions. We contribute novel citizen-oriented views on how to apply AI-based discretion in public administration. Our results revealed that while citizens do not generally oppose delegating discretionary power to fully autonomous systems, engaging them in the decision-making process through a dialogue will positively alter their perceptions. Moreover, we provide insights on how other factors (i.e., decision justification and citizens’ self-gain) influence citizens’ perception of AI-based administrative discretion.

Over the past three decades it has become increasingly common to include social and professional issue topics within undergraduate computing programs. A decade ago, a consensus had emerged around best practices in teaching these topics: namely by covering a professional code of conduct along with a handful of ethical theories and then applying them to computing or workplace dilemmas and choices. Yet despite the successful wide-scale inclusion of ethics instruction within most computing programs, the perception persists that the societal harms of computing remain undiminished. This paper argues that ethics was never the answer to this problem. Addressing the social consequences of computing requires recognizing that computing is deeply enmeshed in political issues, and that the route to addressing this in our curricula is to integrate political topics within them. We can learn effective ways for doing so by making use of pedagogical approaches already pioneered within digital literacy and citizenship education which prioritize questions around justice, equity, and participation. These approaches also focus on engendering critical perspectives towards the students’ digital and non-digital ecosystems as well as encouraging democratic activism and civic engagement with their communities. These citizenship approaches can help our computing curricula better achieve the goals that initially motivated the inclusion of ethics: to help our students play a part in constructing a better world.

Access to the Internet is a crucial enabler for many of the Sustainable Development Goals (SDGs) of the United Nations. Unfortunately, a significant part of the world’s population is left behind due to the lack of access to a reliable and affordable Internet connection. Satellites have the potential to impact the current market of Internet services significantly. In particular, Low Earth Orbit (LEO) satellites promise high-bandwidth without compromising latency. They can be employed in 5G Non-Terrestrial Networks (e.g., IoT connectivity, connected autonomous driving, communication in rural areas, and more). Smart farming and precise agriculture (even remotely controlled), especially in underdeveloped areas, are compelling use cases for LEO satellites. In these scenarios, high bandwidth and low latency are required to facilitate both quick transmission of images/videos and prompt remote control of drones, tractors, actuators, etc. This study compares different TCP protocols based on their performance over satellite communication in a smart farming case study. It also proposes and analyzes a solution leveraging on a limited buffer size to maintain a high throughput while lowering per-packet delays.

This research fits into the scenario of gender disparity in STEM disciplines and aims to identify problems, stereotypes, and gender biases, as well as to highlight solutions to promote gender equality within the Bachelor’s degree course in Computer Science (CS) at the University of Padua (Italy), considering both the opinions of female and male students. Data collection was carried out through an online questionnaire addressed to students enrolled or previously enrolled in the Bachelor’s degree course in CS. The final sample included 167 volunteer participants. The results highlight how girls believe more than boys in the stereotype that women are disadvantaged in the CS field and that their work is not recognized as equal to that of their male colleagues. However, there are encouraging results regarding the decreasing belief that CS is “a man’s thing”, with a perception of less diffusion among the population. In situations of gender bias, such as feeling devalued and ignored or experiencing sexist comments from fellow students and professors, females report feeling emotionally uncomfortable and not having control of the situation, which could lead them to not react to the discrimination they face. In this scenario, it is important to increase the communicative effectiveness and psychological well-being of female, but also male, students. Finally, several strategies were proposed to promote a peaceful and inclusive university social environment.

In recent years, machine learning models have evolved, and the training of these models requires large amounts of data. However, the training data often contains sensitive information, raising privacy concerns. Federated Learning has been proposed as a solution to mitigate privacy risks. Despite its advantages, Federated Learning still faces challenges such as the aggregator being a single point of failure, the existence of malicious participants, and the lack of incentives. Combining Federated Learning with blockchain technology could potentially address these challenges. In this study, we propose a new method for asynchronous Federated Learning using blockchain smart contracts. Our proposed method operates autonomously and in a decentralized manner without the need to trust any central organization, making it trustless. We propose an algorithm that motivates workers to submit high-quality models as quickly as possible. Workers’ behaviors are driven by incentive mechanisms. We deployed a smart contract on a local Ethereum blockchain and executed multiple workers. Our evaluation results demonstrate that learning converges and achieves accuracy comparable to conventional Federated Learning, indicating the effectiveness of our proposed method.

As web designers may deliberately adopt design patterns to hook users’ attention, researchers and practitioners have innovated several tools for supporting users’ digital self-control, hoping to help users self-regulate technology use – especially social networks and video streaming platforms – and achieve digital wellbeing. Unfortunately, these tools often restrict usage, e.g., through self-imposed timers and blockers, limiting interaction possibilities. This paper describes the design, development, and evaluation of two alternative strategies for digital self-control targeting the Facebook and YouTube websites. Specifically, we implemented a Chrome extension that a) highlights when the user is scrolling infinitely by progressively darkening the background (nudging strategy), and b) redesigns the homepages isolating guilty pleasure recommendations and proposing a minimalistic interface (redesign strategy). We compared the two strategies in a three-week field study with 14 participants, finding that both strategies promoted intentional use and allowed participants to decrease time spent and passive scrolling. In particular, participants liked the nudging strategy more as it supported conscious use without changing the overall user experience. We conclude with design implications for moving from traditional digital self-control tools to diverse approaches that may better support digital wellbeing in the long term.

Pervasive and ubiquitous applications provide novel and exciting services leveraging on a multitude of data obtained from people’s devices, adapting the computation to the context in which the user currently is. This improves the service quality of these applications, which can provide a more tailored configuration of the application itself depending on the user context and needs. In these scenarios privacy is of paramount importance, since users must be also be protected against the misuse of their personal data. Analyzing ubiquitous systems in terms of service quality and privacy issues is however a challenging task, due to the heterogeneity of the possible attacks, which makes it difficult to compare two applications. In this paper we propose a novel methodology to jointly evaluate the service quality and the privacy issues in ubiquitous applications in an extensible and comparable way, building on the data available in each part of the system to be analyzed, and defining service qualities and privacy issues so that they can be easily re-used in other analyses. Our evaluation on a candidate application highlights the benefits of our proposal, showing the dependency between privacy levels and service quality, and paving the way for a novel methodology for the definition of these scenarios.

Games have been used for a few decades in research, formal education, and training of children and adults, and digital games are no strangers to educational uses. After all, everyone likes to play games, so it should stand to reason that educational digital games are going to be a hit. Unfortunately, this is not the case. In fact, educational digital games are often criticized for being too focused on educational content and not enough on engaging, challenging, and entertaining players. Making games for entertainment is difficult and requires multidisciplinary expertise. Making educational games that are engaging and entertaining is also difficult and requires additional input from educators and domain experts, and rigorous evaluation methodologies, all of which must revolve around the players. In this position article, we introduce the early stage “EduGames: Play to Learn” research project aimed at supporting the public in acquiring Critical and Computational Thinking skills to tackle the problem of detecting misinformation, and supporting the game development and research communities in creating and evaluating games that are entertaining and educational. As part of this project, we call for more, and more structured, synergy between academia, educators, and the game development industry.

In the next few years, people will be called upon to try to slow climate change and achieve carbon neutrality collectively. The use of persuasive digital tools and engaging mechanisms can play an important role in matching such objectives. In our research, we explore the usage of a Multimodal Conversational Agent embedded in a Smart Mirror and connected to home automation appliances to help users reduce their energy consumption. The agent employs a variety of gamification techniques to encourage short and long-term sustainable behavior, and it is designed to be an enjoyable and non-intrusive experience. It informs householders about their energy consumption and encourages them regularly to reduce and optimize their electric usage. In order to keep the user engaged, the mirror contains visually appealing components and recommendations for daily challenges.

Advancements in robotics and automation technologies have the potential to enable breakthrough innovations in a variety of industries, and the pharmaceutical sector is no exception. The preparation of galenic formulations, involving the compounding and dispensing of medications, when personalized medicines are needed, e.g., to overcome allergy problems, is a critical process in the field of small scale pharmaceutical manufacturing. Traditionally, this process has relied solely on human expertise of pharmacists and their manual labor, which can be time-consuming, prone to errors, and subject to variations in quality. To overcome these limitations, the use of collaborative robots is envisaged in our project. A collaborative robot can in fact work with the pharmacist synergistically, by improving accuracy and increasing productivity. However, the main challenge is providing the pharmacists with an interactive system that supports them in robot programming. In this paper, we analyze the problem from the users’ point of view and propose preliminary low-fidelity prototypes of an interactive system suitable to pharmacists’ needs and skills.

Mobile crowdsensing has rapidly become an interesting and useful methodology to collect data in modern smart cities, thanks to the pervasiveness of users mobile devices. Although there are many different proposals, opportunistic and participatory mobile crowdsensing are the most popular ones. They share a common goal, but require a different effort from the user, which often results in increased costs for the service provider. In this work we forecast user participation in mobile crowdsensing by leveraging a large dataset obtained from a real world application, which is key to understand whether there are areas in a city which need additional data obtained through raised incentives for participants or by other means. We then build a custom regressor trained on the dataset we have, which spans across several years in different cities in Italy, to predict the amount of reports in a given area at a given time. This allows service providers to preventively issue participatory tasks for workers in areas which do not meet a minimum number of measurements. Our results indicate that our model is able to predict the number of reports in an area with an average mean error depending on the precision needed, in the order of 10% for areas with a low number of reports.

Interactive digital narratives (IDNs) have the potential to adequately and effectively represent the highly relevant and complex issue of climate change. The interactivity in interactive digital narratives (IDNs) can increase narrative engagement, as IDNs require active participation. Such narrative engagement, in turn, is a well-known mediator of narrative persuasion. One recent IDN is the award-winning interactive Cli-Fi (‘Climate Fiction’) documentary ‘De eeuw van mijn opa’ (DEVMO; ‘Grandfather's century’), by filmmaker Sam van Zoest. Comparing the original interactive version with a non-interactive version created by the researchers, we used a between-subjects experiment (n=62) to test whether interactivity (yes/no) had an effect on narrative engagement and narrative persuasion. Perceived effectance and perceived autonomy were included as control variables. The results showed that both versions of the documentary had a persuasive effect when comparing scores before versus after exposure. However, the interactive version was not significantly more persuasive compared to the non-interactive version. Furthermore, no evidence was found of narrative engagement as a mediating factor, although narrative engagement did positively affect narrative persuasion. Surprisingly, no differences were found in control variables perceived effectance and perceived autonomy between the conditions with and without interactivity. We discuss several explanations for our findings relating to the study's power and the operationalization of interactivity in ‘De eeuw van mijn opa’.

Digital wellbeing (DWB) became a prominent topic since the use of technology evoked a concern over its impact on users’ mental health and wellbeing. Hence, raising designers' interest and recall of DWB is necessary. As nudges approved their effectiveness in changing individuals’ behavior in different domains including users’ addictive behaviors online, they have not, yet, been used to change designers’ behaviors towards wellbeing-sensitive designs. In this study, we conducted a co-design workshop with digital designers to identify what type of nudges they prefer to ensure a more inclusive and wellbeing sensitive design. Basic psychological needs of autonomy, competence and relatedness were used to represent wellbeing. We performed template analysis on the workshop outputs where 24 nudge ideas were produced by three mixed groups of designers, Human Computer Interaction researchers and psychologists. We identified three themes representing nudge characteristics which are encouraging reflective thinking, facilitating actions, and stimulating heuristic behavior, and six subthemes representing design categories. Our results show that nudges encouraging reflective thinking represent the majority of the suggestions. This result indicates that designers have ownership for reminding themselves to incorporate basic human needs within the design through transparent nudges where the intention of behavior change is clearly identified.

One of the most dangerous problems facing humanity is air pollution. According to GBD estimates, poor air quality and indoor air pollution cause nearly 2 million premature deaths in India. Air quality monitoring stations are expensive to install. These issues require a cost-effective resolution. India’s energy infrastructure requires a low-power solution in order to prevent new issues while resolving old ones. Image-based air pollution detection using artificial intelligence has become a popular option. Nevertheless, two issues remain: There are few image-based air pollution data sets. Existing methods utilize a model with numerous parameters, which requires a great deal of processing power. Based on that, we developed Eff-AQI, a reliable artificial intelligence model with 1.9 million parameters. The proposed model could obtain the following results: 9.56 RMSE, 0.99 R2, 89.92% balanced accuracy, and 89.38% accuracy for AQI estimation; 14.62 RMSE, 0.99 R2, 90.56% balanced accuracy, and 91.83% accuracy for PM2.5 estimation; and 14.40 RMSE, 0.98 R2, 96.25% balanced accuracy, and 95.42% accuracy for PM10 estimation. The proposed model outperformed DOViT, the model with the highest accuracy among all surveyed SoTAs, by 2.64 points, and it has 46.32 times smaller parameters compared to DOViT. The proposed model can achieve the same R2 score with 54.16 times smaller parameters than Ensemble DNN. We also make available on Kaggle the novel air pollution image data with the corresponding labels: AQI, PM2.5, PM10, O3, CO, SO2, and NO2. The investigation revealed that the majority of SoTAs could utilize the dataset to enhance performance. The proposed model is more accurate and has fewer parameters than SoTAs. Environmental sustainability and reduced pollution are also involved. Increasing society’s or stakeholders’ high-confidence understanding of air pollution situations in order to develop effective and efficient mitigation solutions. This initiative is beneficial to AI for the social good and the Sustainable Development Goals, especially SDG 3, "Good Health and Well-Being," and SDG 11, "Sustainable Cities and Communities."

This paper proposes a low-cost system to help users perform exercises at home, monitoring muscle imbalances. Muscle imbalance occurs when there is a strength disparity between muscles on one side of the body or one side of a joint.

The proposed device detects asymmetries in exercises involving the lower limbs, employing two slim insoles equipped with only three pressure sensors positioned along the foot, one on the heel and two in the front. The data collected by these sensors are wirelessly transmitted to a user-friendly interface, which serves as a guide, assisting users in achieving more balanced and symmetrical training, important for maintaining musculoskeletal health and enhancing overall physical well-being. The paper presents an analysis of relevant literature, introduces the device and its characteristics, and presents the physical prototype and its experimental results.

Labour exploitation in the Taiwan Distant Water Fishing (DWF) industry has been a persistent issue for many years. Fishermen working on these vessels are often subjected to long working hours, low salaries, and poor living conditions. These conditions can lead to physical and mental health problems, exploitation, and abuse. To address this issue, a system has been developed with two modules. The first module collects data from three sources: CCTV footage from DWF vessels, Global fishing watch (GFW) open data, and Mobile Face Verification System (MFVS) interface for collecting data from fishermen and captains. The second module uses the collected data to identify and recognize instances of labour exploitation on DWF vessels. Our proposed system research shows that by combining different data sources, including MFVS, GFW, and Transfer-Learning, the You Only Look Once v7 (TL-YOLOv7) model can effectively identify and recognize labour exploitation. The proposed model aligns with Sustainable Development Goals by promoting decent work. It also improves working conditions to safeguard fishermen’s physical and mental health. The TL-YOLOv7 model achieves a higher mean average precision (mAP) value of 0.835 than the Pre-trained model of 0.691. This implies that the TL-YOLOv7 model exhibits higher accuracy in object detection. TL-YOLOv7 model achieves a lower RMSE of 0.44 compared to 5.24 for the GFW model, indicating a reduced overall deviation from the actual working hours. The system can help identify exploitation instances and promote better working conditions for fishermen in Taiwan’s DWF industry.

IoT (Internet-of-Things) powered devices can be exploited to connect vehicles to a smart city infrastructure and thus allow vehicles to share their intentions while retrieving contextual information about diverse aspects of urban viability. Such a complex system is aimed at improving our way of living in the city by mitigating the effect of traffic congestion, and consequently stress and pollution. We place ourselves in a transient scenario in which next generation vehicles that are able to communicate with the surrounding infrastructure coexist with traditional vehicles with limited or absent IoT-capabilities. In this work we focus on intersection management and, in particular, on reusing existing traffic lights empowered by a new management systems. We propose an auction based system in which traffic lights are able to exchange contextual information with vehicles and the nearby traffic lights with the aim of reducing average waiting times at intersections and consequently, overall trip times. We evaluate our proposal using the well known MATSim transport simulator, by using a synthetic Manhattan map and a new map we build on an urban area located in our town, in Norther Italy. In such an area, instrumentation through IoT devices has been set up as part of an European research project. Results show that the proposal is better performing than the classical Fixed Time Control system currently adopted for traffic lights, and then auction strategies that do not exploit coordination among nearby traffic lights.

Fall detection poses significant challenges for researchers due to the potential for severe injuries, such as femoral neck fractures, brain hemorrhages, and skin burns, which can cause considerable pain. Furthermore, undetected falls can lead to deteriorating health over time, resulting in a painful end-of-life scenario or even death. It is vital to efficiently detect falls in order to promptly notify relevant individuals, such as nurses, for timely intervention. This study presents a solution for fall detection specifically designed for healthcare institutions, with a focus on individuals aged 65 and over. Ultra-wideband (UWB) radars are employed as the primary technology for this purpose. Two distinct approaches were implemented: one involving feature extraction, dimensionality reduction, and a Random Forest classifier, and the other utilizing a simple Convolutional Neural Network (CNN) architecture. The results consistently demonstrate that the deep learning approach outperforms the classical machine learning approach by an average margin of 7.5% and 8.1% when all raw data are filtered either by Butterworth or Type 1 Chebychev filters, respectively. The superior performance of the deep learning approach can be attributed to the effective filtering process applied to the UWB radar data. Importantly, a leave-one-subject-out strategy was employed to validate the fall detection performance. This strategy is not commonly used for validation in other research studies present in the literature.

Chronic patient self-management is crucial for maintaining physical and psychological health, reducing pressure on healthcare systems, and promoting patient empowerment. Digital technologies, particularly chatbots, have emerged as powerful tools for supporting patients in managing their chronic conditions. Large language models (LLMs), such as GPT-4, have shown potential in improving chatbot-based systems in healthcare. However, their adoption in clinical practice faces challenges, including reliability, the need for clinical trials, and privacy concerns. This paper proposes a general architecture for developing an LLM-based chatbot system that supports chronic patients while addressing privacy and security concerns. The architecture is designed to be independent of specific technologies and health conditions, focusing on data protection legislation compliance. A prototype of the system has been developed for hypertension management, demonstrating its potential for motivating patients to monitor their blood pressure and adhere to prescriptions.

Highly skilled professionals’ forced migration from Ukraine was triggered by the conflict in Ukraine in 2014 and amplified by the Russian invasion in 2022. Here, we utilize LinkedIn estimates and official refugee data from the World Bank and the United Nations Refugee Agency, to understand which are the main pull factors that drive the decision-making process of the host country. We identify an ongoing and escalating exodus of educated individuals, largely drawn to Poland and Germany, and underscore the crucial role of pre-existing networks in shaping these migration flows. Key findings include a strong correlation between LinkedIn’s estimates of highly educated Ukrainian displaced people and official UN refugee statistics, pointing to the significance of prior relationships with Ukraine in determining migration destinations. We train a series of multilinear regression models and the SHAP method revealing that the existence of a support network is the most critical factor in choosing a destination country, while distance is less important. Our main findings show that the migration patterns of Ukraine’s highly skilled workforce, and their impact on both the origin and host countries, are largely influenced by pre-existing networks and communities. This insight can inform strategies to tackle the economic challenges posed by this loss of talent and maximize the benefits of such migration for both Ukraine and the receiving nations.

Recent advancements of blockchain technologies ensure security and trustability of Community Currency Systems (CCSs), enabling their increasingly widespread adoption. These systems aim at empowering the local economies by virtue of a medium of exchange whose governance and circulation are local. Smart contracts enable the enforcement of token economy policies, which facilitate the experimentation of radically new economic models. Recent studies investigated blockchain-based CCSs. Still, to the best of our knowledge, this is the first study analyzing a CCS providing a token-based Universal Basic Income (UBI). We evaluate the Circles UBI decentralised application utility in delivering an unconditional income to its users, focusing on its main pilot project running in Berlin. We analyse the structural changes in the network, especially in relation to a subsidy program, involving local businesses. We also identify prominent users based on centrality measures, and investigate how the UBI was effectively spent. We adopt a method agnostic to the economic context to identify optimal aggregation windows for the temporal network of CCS transactions based on the Causal Fidelity (CF) index. This aims to provide static representations as accurate as possible in terms of sequential order of edges, which aspect was not considered in previous research on CCSs. Our findings suggest that the pilot project sustained the expansion of the economic network and the system facilitated trade in urban communities in Berlin. Future research is needed to identify methods to ensure sustainability of self-organised CCSs adopting a UBI issuance scheme and to further decentralise their governance.

The narrative sense-making of video games relies on different sources of information. Among them is the multimodal system through which games are made sensorially perceivable, the sensorimotor experiences they afford and require, and the mnemonic recollection through which they are made understandable. These sources show strong feedback loops and give rise to an overall meaning that is more than the sum of its parts. As such, narrative-driven video games can be considered complex in themselves. In order to test this theoretical model, a custom-made video game has been developed. By employing simple mechanics and simple graphics, the game will be the basis for conducting a think-aloud session. The session will give insights into the actual cognitive mechanisms of players, to investigate how their sense-making works.

This paper introduces Locale, a novel system that enables cost-effective fidelization processes in local communities while offering enhanced security guarantees compared to traditional methods. By leveraging the unique properties of blockchain technology, Locale provides a solution that is tailored to the specific needs and dynamics of local communities. Locale has a new time-stamping mechanism based on the pay-to-contract method which embeds the commitments inside the public key instead of a separate field in the transaction. This ensures both authenticity and local privacy since the commitment is not visible by external blockchain observers. Nonetheless, this mechanism offers the same security guarantees of traditional timestamping mechanisms. This innovation has independent relevance beyond the context of Locale and opens up new avenues for research and development in the field. By bridging the gap between blockchain technology and local community development, this research contributes to the advancement of practical and efficient solutions for the common good.

In recent years, the delivery of government services via mobile applications has increased significantly, with public assistance being a prime example. Digital government services often reflect the value-laden nature of the political context they emerge from, and limited consideration of end-user values in the design of social safety net apps can cause decreased use and interruptions of public aid. To fill this gap, we adopted the lens of Value-Sensitive Design to investigate the embodiment of end-user values in existing food assistance mobile apps. We adopted the public value-based digital government framework and conducted a thematic analysis of mobile app reviews and Reddit posts. This study reveals the reflection and violation of public values and the underrepresentation of end-user dignity in current food aid mobile apps. We provide design implications for more human-centric mobile apps for social safety nets that preserve the dignity of beneficiaries.

This paper focusses on creating accessible alternatives for the graphical representation of mathematical functions, with particular attention to students with visual impairments and children with Attention Deficit/Hyperactivity Disorder (ADHD). We design a virtual reality (VR) application which can stimulate different senses by immersing students in a virtual reality environment where they can explore the graphical representation of a function and hear how it is played. The purpose of the application is to improve immersion and inclusion in mathematics. The results of the tests show that this type of tool is useful both for learning through play and for students with different learning needs. The use of audio is an effective way to make graphs accessible, but the problem is that there are no universal design principles that apply to various graphs that encode different data types. Additionally, users with ADHD could appreciate the flexibility, speed, cost effectiveness and greater measure of independence provided by the system; however, more research is needed to justify this statement.

This paper employs a participatory research approach to develop a digital solution for supporting People with Intellectual Disabilities (PwID) during their transition from a ‘protected’ training environment into the labour market. The study takes place in a context where social integration for PwID is not yet a reality, making this transition period emotionally complex and challenging. Over two years, this research has engaged stakeholders and potential users in ideating a solution, resulting in the creation of SuperGuide, a co-designed digital tool that supports PwID in the practical and emotional aspects of their journey. This paper describes and reflects on the inclusive ideation process, explores the possibilities considered and presents preliminary requirements for the tool’s development. By documenting this research, we aim to contribute to the understanding of inclusive design methodologies and the practical implementation of digital solutions for supporting PwID in an effective and holistic way.

Social media has become an integral part of our daily lives, with billions of users accessing these platforms mainly from their smartphones. Existing literature suggests that social media has an impact on mental health, but it is not clear whether this impact is positive or negative. This raises concerns about the relationship between social networks and public health, which require further investigation. In response, this doctoral consortium paper proposes a research study that investigates the impact of visualizing the frequency and nature of social media use on mental health and family functioning. The research study consists of two main user studies and four supporting surveys, aimed at reflecting on and adding to the existing literature. Furthermore, the study sampled participants from different ethnographic backgrounds, including the United Kingdom, Saudi Arabia, and Malaysia. This approach is beneficial because it allows cross-cultural comparisons, which can help identify similarities and differences in the impact of social media on mental health and family functioning across different cultures. It is important to examine the impact of social media on mental health and family functioning across cultures because social norms, values, and attitudes may differ between cultures, leading to different results. The analysis includes in-group comparisons between different cultures and gender, with an emphasis on identifying negative impacts. This research addresses a literature gap, providing evidence of the negative impact of social media on mental health and family functioning and the effectiveness of visualizing this effect.

In a world where sustainable and collaborative behavior is increasingly important due to climate change, environmental concerns, and social engagement, individual willpower may not be enough to sustain positive behavior for long-term sustainability. To encourage collaborative behavior, many blockchain-based applications are emerging that provide an incentive in the form of fungible tokens, non-fungible tokens (NFTs), or reputation points. Existing services address specific solutions such as waste disposal, peer-to-peer energy management, and sustainable mobility. However, the tokens issued by these services generally can be used only by the services themselves and are not interchangeable with other tokens. This paper proposes a platform that aggregates different blockchain-based services, and that exploits a conversion mechanism enabling the user to convert a given service token with other service tokens. The conversion is not based on a monetary value, rather, it relies on the amount of saved CO2 a service token represents. The platform provides a token (the “sCO2” token) anchored to a fixed amount of saved CO2, that may be converted for token of other services or used to get a discount on various municipal services such as waste tax, parking, public transport, etc. The proposed system aims to increase the engagement and awareness among citizens and end-users, provide an accountable and transparent way to track people’s sustainable behavior, and issue certificates to organizations based on how much CO2 their services have helped save.

Diseases transmitted by mosquito vectors, such as malaria, dengue, and Zika virus, pose significant healthcare challenges worldwide. Accurately estimating mosquito populations is vital for understanding transmission risks. Previous studies have explored the use of mosquito wingbeats for population surveys and control efforts. However, these methods require extensive data collection and annotation, which is time-consuming and resource-intensive. Additionally, laboratory-collected datasets lack biodiversity information from wild mosquitoes. To overcome these limitations, we propose an IoT system that automates the collection of mosquito wingbeat sounds in the field. Our system integrates with the Biogents BG-Counter 2 smart mosquito trap and incorporates a cost-effective acoustic sensing device. Initial assessments indicate successful integration, seamless data transmission, and satisfactory audio quality for classifying mosquito wingbeats. This solution offers an efficient alternative for gathering wingbeat data from species naturally present in the field.

A major problem in blockchain-based supply chain management is the potential unreliability of digital twins when considering digital representations of physical goods. Indeed, the use of blockchain technology to trace goods is obviously ineffective if there is no strong correspondence between what is physically exchanged and the digital information that appears in blockchain transactions.

In this work, we propose a model for strengthening the supply chain management of physical goods by leveraging blockchain technology along with a digital-twin verification feature. Our model can be instantiated in various scenarios and we have in particular considered the popular case of food traceability. In contrast to other models known in the literature that propose their own ad-hoc properties to assess the robustness of their supply chain management systems, in this work, we use the formalism of secure computation, where processes are described through generic and natural ideal functionalities.

The negative effects of climate change are calling for new approaches to promote energy efficiency and the use of renewable energy sources at multiple scale levels. As virtual assistants are becoming a common household item, recent studies have looked at integrating IoT and virtual assistants for energy management purposes. Despite the prominence of these works, a critical gap in the current body of research is the almost absence of real-world implementations covering different sectors of society. To address this gap, we developed the PowerShare Virtual Assistant (VA), a voice-based eco-feedback system. The paper presents results from the real-world deployment of the PowerShare VA in three distinct sectors - 1) residential, 2) commerce, and 3) industry. By looking at the human response to our system in different daily life scenarios, we aim to contribute to future research on using VA in the context of energy efficiency.

Digital accessibility is considered an important aspect to allow all people, including those with permanent or temporary disabilities, to access the continuously increasing number of digital services. This raises the need for tools able to provide support for monitoring the level of accessibility of a large number of websites in order to understand their actual level of accessibility, and identify the areas that need more interventions for their improvement. We present how we have extended a tool for accessibility validation for this purpose, and the results that we obtained in the validation of about 2.7 million Web pages of Italian public administration Web sites.

Mitigating the negative effects of climate change remains a big challenge. Changes in individual human behaviors are recognized to be strategic for climate change mitigation and scientists agree that solutions must be devised which guide people towards sustainable behaviors. In this paper, we present EcoDrive, a mobile application, which allows users to analyze their driving style in real-time, and correct it accordingly, in a nonintrusive way, using the concept of eco-feedback. As a distinguished aspect, in EcoDrive eco-feedback is accompanied by the motivational technique of gamification. The app was designed to use gamification in a non-intrusive manner, in order to preserve driver’s safety. An experimental study proves that higher engagement is achieved when eco-feedback is combined with gamification mechanisms. 4 lessons learned from it, showing how the concept of gamification can be a mechanism to encourage citizens to adopt a more sustainable driving style.

Human activity recognition (HAR) using ambient sensors has emerged as a promising approach to telemonitoring daily activities and enhancing the elderly quality of life. Deep learning models have demonstrated competitive performance in HAR on real-world datasets. However, acquiring large amounts of annotated sensor data for extracting robust features is costly and time-consuming. To overcome this limitation, we propose a novel model based on the self-supervised learning framework, SimCLR, for daily activity recognition using ambient sensor data. The core component of the model is the encoder module, which consists of two convolutional layers followed by a long short-term memory (LSTM) layer. This architecture allows the model to capture both spatial and temporal dependencies in the sensor data, enabling the extraction of informative features for downstream tasks. Through extensive experiments on three CASAS smart home datasets (Aruba-1, Aruba-2, and Milan), we showcase the superior performance of the model in semi-supervised learning and transfer learning scenarios, surpassing state-of-the-art approaches. The findings highlight the potential of self-supervised learning in extracting valuable information from unlabeled sensor data, reducing costly annotation efforts for real-world HAR applications.

The growing popularity of online social media (OSM) has led to the creation of a wide amount of social media platforms. In this context, the increasing competition among platforms and the emergence of decentralized alternatives such as Blockchain Online Social Media (BOSM), have led to more frequent user migrations: individuals tend to switch platforms in search of improved features, content, or communities. Therefore there has been increasing interest in user migration studies modeling and predicting user migration. However, user migration, especially in blockchain-based platforms remains an understudied problem. Existing methods rely on user activity to derive interaction graphs and then address the user migration prediction problem as a node classification task, where user decisions are encoded as node labels. While the performance look promising, there are currently two important research gaps: i) there is no work using graph neural networks, the state-of-the-art in machine learning on graphs; and ii) there is a lack of methods designed to improve prediction performance in the case of class imbalance, i.e. the presence of dominant behavior among the ones to predict. In this paper, we propose a machine learning pipeline utilizing graph neural networks (GNNs) to predict user migration in BOSM. We model the data as a directed temporal multilayer graph, capturing social and monetary interactions among users. To address the problem of class imbalance in node classification, we introduce a data-level balancing technique following an undersampling approach. The evaluation, conducted on data describing user migration across blockchain online social media platforms, shows that graph neural networks are a suitable machine learning approach to perform user migration prediction. Furthermore, the proposed undersampling approach improves predictive power on severely imbalanced data. These results highlight how graph neural networks are effective in predicting user migration, without the need for manual feature engineering and in the absence of user information. Our methodology holds potential for applications beyond user migration, such as fraud detection and bot detection, and opens up venues for further research in other prediction tasks in online social networks and blockchain-based systems.

The rising occurrence of natural and human-made disasters emphasises the urgent need for effective training of medical first responders (MFRs). Virtual Reality (VR) has recently been used to enhance traditional MFR training. However, to ensure that VR training improves disaster preparedness, it is not only crucial for MFR stakeholders to actively participate in the design process. It may also be beneficial to place the co-design process in VR so that novice co-designers establish a profound, hands-on understanding of VR as a training tool. Thus, we introduce the Collaborative Scenario Builder (CSB), a prototype for MFRs without technical and designerly expertise with which to co-design scenarios for virtual simulation training in VR. An evaluation with 33 MFR participants indicates that CSB is usable and provides participants with an embodied understanding of VR, leading to new perspectives in their collaborative design considerations. Thus, CSB contributes to a co-design workflow with MFR co-designers that ensures that created VR training tools are needed and beneficial for MFRs so that they can provide better aid to people in the face of disasters.

This research aims at exploring how culture-specific narratives influence the imagination of future AI technology. Design professionals play a crucial role in this research context as they not only take part in the development of technology applications but are also involved in shaping public narratives about socio-technical relations as agents of public imagination. We examined how designers envision future societies and how future narratives as a result of design processes are reflected upon by an external audience. In Study 1, we rolled out three speculative design workshops in Japan and Germany with young design professionals. The analysis of the resulting design representations (artifacts) reveals that the relationship between human autonomy and AI-driven automation represents a central theme in the imagination of futures. Visions created in Japan draw from (pop-)culture-specific narratives, but also from governmental agendas. In projects from Germany, corporations lead the deployment of AI systems that alter society and social interactions. In a follow-up Study 2, we analyze the perceptions of created artifacts through a survey with 174 students from computer science. We identify loss of privacy, ethical concerns, governmental control, surveillance, and loss of free speech as main themes of reflection, which is in line with themes raised by designers. Students focus on societal impacts of use and aims of depicted applications rather than on AI as technical facet. Visualizations created in Study 1 appear to have a strong influence on the audience’s impression of the depicted technological applications in Study 2. Although cultural differences regarding style of visualization seem to play a decisive role for creation in Study 1, they do not seem to have a major effect on participants’ reflections in Study 2. Overall, findings from both studies contribute to advancing the understanding of design speculations in creation and perception.

In recent months, the impact of Artificial Intelligence (AI) on citizens’ lives has gained considerable public interest, driven by the emergence of Generative AI models, ChatGPT in particular. The rapid development of these models has sparked heated discussions regarding their benefits, limitations, and associated risks. Generative models hold immense promise across multiple domains, such as healthcare, finance, and education, to cite a few, presenting diverse practical applications. Nevertheless, concerns about potential adverse effects have elicited divergent perspectives, ranging from privacy risks to escalating social inequality. This paper adopts a methodology to delve into the societal implications of Generative AI tools, focusing primarily on the case of ChatGPT. It evaluates the potential impact on several social sectors and illustrates the findings of a comprehensive literature review of both positive and negative effects, emerging trends, and areas of opportunity of Generative AI models. This analysis aims to facilitate an in-depth discussion by providing insights that can inspire policy, regulation, and responsible development practices to foster a citizen-centric AI.

The demand for large-scale diverse datasets is rapidly increasing due to the advancements in AI services impacting day-to-day life. However, gathering such massive datasets still remains a critical challenge in the AI service engineering pipeline, especially in the computer vision domain where labeled data is scarce. Rather than isolated data collection, crowdsourcing techniques have shown promising potential to achieve the data collection task in a time and cost-efficient manner. In the existing crowdsourcing marketplaces, the crowd works to fulfill consumer-defined requirements where in the end consumer gains the data ownership and the crowd is compensated with task-based payment. On the contrary, this work proposes a blockchain-based decentralized marketplace named Vision Sovereignty Data Marketplace (ViSDM), in which the crowd works to fulfill global requirements & holds data ownership, the consumers pay a certain data price to perform a computing task (model training/testing), the data price is distributed among the crowd in a one-to-many manner through smart contracts, thus allowing the crowd to gain profit from each consumer transaction occurring on their data. The marketplace is implemented as multiple smart contracts and is evaluated based on blockchain-transaction gas fees for the stakeholder interaction & by running scenarios-based simulations. Furthermore, discussions address the challenges included in maintaining data quality and the future milestones towards deployment.

In 2015, the United Nations put forward 17 Sustainable Development Goals (SDGs) to be achieved by 2030, where data has been promoted as a focus to innovating sustainable development and as a means to measuring progress towards achieving the SDGs. In this study, we propose a systematic approach towards discovering data types and sources that can be used for SDG research. The proposed method integrates a systematic mapping approach using manual qualitative coding over a corpus of SDG-related research literature followed by an automated process that applies rules to perform data entity extraction computationally. This approach is exemplified by an analysis of literature relating to SDG 7, the results of which are also presented in this paper. The paper concludes with a discussion of the approach and suggests future work to extend the method with more advanced NLP and machine learning techniques.

This PhD study project researches people with health conditions affecting cognition (from here referred to as PHCAC) as stakeholders who help codesign assistive technology, devices that aid in day-to-day life (a variety of which ranges from special designed wheelchairs, to digital software to aid accessibility). It focuses on the aspects of how to improve employment prospects by improving the usability of assistive technologies. It also looks into specific elements that see little research in the development of assistive technologies, such as utilising social context, and longitudinal deployment with accessibility in mind. This is with the intent to see how they impact the design of assistive technologies and, ergo, it’s impact on it’s usability.

So far this project has conducted research by assisting in a couple smaller projects that were looking into similar but different topics of research. These topics were things such as "digital meaningful making", and "Enabling digital first" with stakeholders similar to this study (people with acquired brain injuries and sight-impaired respectively) [16]. These were done during the COVID-19 pandemic due to the severe effects that it had on the ability to conduct social research. They also allowed for the principal investigator to gain experience and understand the area around the research topic in more first-hand methods.

After these were done, some bespoke work for the project was done to flesh out the beginning of the PhD, a literature review was conducted into the research topic itself. This was done to contextualise all the data that had been collected from the smaller projects. Another study that was conducted looked into the generation of social context data and what it can mean for the development of assistive technology, and it utilised a method of qualitative data generation called a World Café. This was a foundational study to be used as a stepping stone for the technologically based component of the PhD.

The next steps of the research detail a shift to focus on the technology itself and its development. After this the research moves onto directly tackling the implementation of automated action taken on bespoke feedback, with the possibilities of how these can be achieved.

A clinical trial is a study that evaluates new biomedical interventions. To design new trials, researchers draw inspiration from those current and completed. In 2022, there were on average more than 100 clinical trials submitted to ClinicalTrials.gov every day, with each trial having a mean of approximately 1500 words [1]. This makes it nearly impossible to keep up to date. To mitigate this issue, we have created a batch clinical trial summarizer called CliniDigest using GPT-3.5. CliniDigest is, to our knowledge, the first tool able to provide real-time, truthful, and comprehensive summaries of clinical trials. CliniDigest can reduce up to 85 clinical trial descriptions (approximately 10,500 words) into a concise 200-word summary with references and limited hallucinations. We have tested CliniDigest on its ability to summarize 457 trials divided across 27 medical subdomains. For each field, CliniDigest generates summaries of μ = 153, σ = 69 words, each of which utilizes of the sources. A more comprehensive evaluation is planned and outlined in this paper.

Sea surface temperature (SST) is uniquely important to the Earth’s atmosphere since its dynamics are a major force in shaping local and global climate and profoundly affect our ecosystems. Accurate forecasting of SST brings significant economic and social implications, for example, better preparation for extreme weather such as severe droughts or tropical cyclones months ahead. However, such a task faces unique challenges due to the intrinsic complexity and uncertainty of ocean systems. Recently, deep learning techniques, such as graphical neural networks (GNN), have been applied to address this task. While such techniques achieve certain levels of success, they often have significant limitations in exploring dynamic spatio-temporal dependencies between signals. To solve this problem, this paper proposes a novel graph convolution network architecture with static and dynamic learning layers for SST forecasting. Specifically, two adaptive adjacency matrices are firstly constructed to respectively model the stable long-term and short-term evolutionary patterns hidden in the multivariate SST signals. Then, a personalized convolution layer is designed to fuse these information. The developed network can be learned in an end-to-end manner. Our experiments on real SST datasets demonstrate the state-of-the-art performances of the proposed approach on the forecasting task.

The Non-Fungible Token (NFT) market is developing in tandem with the growth of the cryptocurrency market and the advancement of blockchain technologies. This has fostered a swiftly prospering NFT market, which subsequently entered a period of decline. Nevertheless, the overall rise procedure of the NFT market has not been well understood. We consider that evolving social media communities, accompanying market growth, offer valuable insights into market behaviours. Our research primarily focused on the NFT Twitter communities, conducting two experiments to gauge their impact on NFT price movements. Using a Granger causality test on tweet counts and NFT prices, we found that for most of the top 19 original projects, tweet volume positively influenced the price, or vice versa. This trend was seldom observed for copycat projects. To assess price movement predictability, we experimented with forecasting the Markov-normalized NFT price, indicative of price shift direction and magnitude, using tweet-extracted features. Social media words, as predictors, achieved testing accuracy above the baseline for all 19 top projects. Furthermore, both market-related and NFT event-related words notably contributed to price movement predictions. We summarized the characteristics of categorization and sentiment for the words with the most and least feature importance.

The advent of metaverse platforms has caused a revolution in the way people interact and engage with digital content creation and consumption. These platforms employ advanced technologies like blockchain and augmented reality to provide users with a fully decentralized environment. Among the metaverse platforms, Decentraland stands out as a popular blockchain-based virtual reality platform that enables users to monetize their content creation. However, a major criticism of these platforms is the low number of active users. Existing methodologies for measuring user traffic and engagement in these metaverse platforms are often unclear and inconsistent, and recent analyses do not provide a correct evaluation of the active users. In this paper, we propose a methodology for evaluating user traffic on decentralized metaverse platforms. Our approach employs a traffic monitor that links traffic with transactions to analyze user behavior. Decentraland has been used as a case study due to its decentralized nature and transparent provision of information about user traffic. The study includes an analysis of user behavior using graph analysis, as well as an examination of user traffic and transactions. We use user traffic to measure community engagement and economic activity, and we assess value exchange through parcel transaction prices. Our analysis indicates a lower average node degree in the interaction graph compared to traditional social media, no correlation between user traffic and parcel transaction prices, the exhibition of patterns of close proximity between the dates of traffic and transactions, and the presence of four distinct user clusters based on travel patterns, with the finding that a small proportion of highly engaged users contribute significantly to the total distance travelled. This study is the first to focus on user traffic and transaction activity within a metaverse platform, and the proposed methodology can be adapted for use on comparable metaverse platforms as well.

The metaverse represents now one of the most promising innovations that lie under the umbrella of Web3-powered technologies. Its societal impact has important repercussions both on social good and on social computing. Among the technologies that are driving the metaverse, we find cryptocurrencies and Non-Fungible Tokens (NFTs) that are used to represent the property of goods and assets in the virtual worlds. In particular, NFTs are used to represent the parcels in which the metaverse is divided, with unknown consequences on the metaverse economy. In this paper, we explore the social impact of employing NFTs in the metaverse and propose an analysis of parcel sales taking Decentraland and The Sandbox as a case study. Our analysis uncovers that prices of the metaverse parcels are increasing over time and that acquiring parcels is becoming prohibitive. Additionally, we identify that the factors that drive the price of sales in the metaverse are the proximity to specific landmarks, like roads, squares, districts, or parcels owned by influencers or enterprises. We argue that this could be a factor in reducing the social good of the metaverse because it fuels a speculative market similar to the one of the physical estate.

Older adults continue to be targeted by cybersecurity attacks: a trend which shows no signs of slowing, and one that has become even more problematic given that many older adults adopted new digital technologies during the Covid-19 lockdowns. Yet there remains a scarcity of solutions designed to help older adults protect themselves online. In part, this is due to a lack of understanding of the specific needs of older adults, who are the fastest growing, and arguably most technologically diverse population on the internet. This study draws upon recent qualitative research to identify key dimensions which are likely to influence older adult cybersecurity behaviour and subsequent vulnerability. We show how these dimensions can be used, for example, to develop a wide range of personas that help illustrate the range of abilities and attitudes in the older adult population. The dimensions outlined here can be used to help researchers, designers, and developers better understand the diverse needs of older adult users when developing digital or security solutions for this population.

In the last decades, Unmanned Aerial Vehicles (UAVs) are finding more and more fields of application. Their flexibility and cost-efficiency make them useful to support complex operations in agriculture, remote sensing or construction, just to name a few. In the Labyrinth project we aim at investigating the applicability of UAV usage to critical scenarios like air, water and road traffic control or emergency, with a strict focus on safety, security and efficiency. This involves also the cybersecurity aspect, which is the main focus of this work. UAVs used in critical applications are in fact potentially exposed to a wide set of cyber threats. The NIST cybersecurity framework [17] defines five different security functions which are: identify, protect, detect, respond and recover. In this paper we address the identify and detect functions with an approach involving threat analysis and anomaly detection. Firstly, we identify which threats pose a significant risk to the Labyrinth use case, for instance leading to the collision of UAVs in case an attacker is successful. Secondly, we present a machine learning-based pipeline aimed at detecting deviations in the position reportings of the drone, to support the detect function during flight operations. The pipeline is tailored to the Labyrinth system reporting needs and is based on unsupervised machine learning to overcome the lack of labeled data. Anomalous points, i.e., points deviating from a coherent path, potentially because of a cyber-attack or a failure, are visually separated from the coherent ones and marked as noise. To prove its robustness, we test the pipeline introducing artificial perturbations in the data.

Today, access to the Internet provides access to various forms of knowledge like free online lecture series offered by prestigious universities, massive open online courses, films and books, and Wikipedia. In addition, it is possible to join online communities on any topic of interest, get to know people with common interests, exchange thoughts and participate in debates. To enable access to these unprecedented knowledge bases, it is crucial to be able to translate texts into any language known by users. For this reason, Machine Translation has been a very active research field for the last thirty years.

In this paper, we investigate the task of Chinese-Italian translations by exploiting Neural Machine Translation approaches. We trained several deep neural networks starting from two already available datasets containing Chinese-Italian parallel corpora. Then, we compared their performance against some of the most common machine translation services freely available online. In particular, we take advantage of Microsoft Translator, Google Translate, DeepL, and ModernMT.

Society 5.0 envisions a more resilient, sustainable, and human-centered society fostered by ever-evolving cooperation and knowledge sharing among the many digital systems already shaping our daily lives. However, the current state of smart cities often consists of siloed systems, with different actors and stakeholders managing their services and assets independently. This phenomenon is evident in both technological and operational domains, posing challenges to seamless collaboration. In this context, new cloud computing models and technologies like event and service mesh promise to reduce the burden associated with the development and integration of solutions. In the attempt to pave the way for more integrated IT environments, we propose a practical architecture that combines service and event mesh technologies, enabling the seamless exploitation of service invocation and composition based on event distribution and direct service calls. Our proposal allows applications to remain transparent of the underlying technology, facilitating various optimizations on the network and management plane, necessary to meet the diverse operational requirements of complex and heterogeneous applications. We validate our proposal in a real-use case scenario implementation, discussing the tradeoffs that emerge.

The positioning accuracy of UWB-based mobile Internet of Things (IoT) devices is frequently impacted by the complicated indoor environment, which is a common application for automated following mobile IoT devices. To address the issue of abnormal value errors such as high noise and UWB jitter value when tracking and locating mobile IoT devices in complicated indoor environments, this paper proposes to use a hybrid filtering weighted following algorithm based on UWB, which combines the benefits and drawbacks of Gaussian, median, and average filtering techniques, introduces the residual value of ranging, and combines geometric positioning to determine the ideal following value. The experimental results show that the proposed algorithm can effectively filter out the UWB error under multi-factor interference and finally estimate the UWB value closest to the actual value, thereby improving the stability and sensitivity of the following process and obtaining a better follow effect.

This paper presents a study with kindergarten teachers to assess the advantages, challenges and opportunities of commercial robots to teach computational thinking to young children. Recent studies have highlighted the potential benefits of introducing CT concepts at an early stage. Robots are an engaging and effective educational tool for teaching CT to young children, providing hands-on and interactive learning experiences. Entirely tangible robotic environments have successfully connected the abstract world of CT with the concrete world of preschoolers. Children can program robots by pressing buttons, drawing the path or using code cards. However, there is limited research on the use of commercial robots in preschool classrooms. This research aims to address this gap by investigating preschool teachers’ perspectives on the advantages, challenges, and opportunities associated with using commercial robots in the context of kindergarten classrooms. We contribute with a list of practical, pedagogical and motivational aspects that should be taken into account while evaluating robots and design considerations to build robotic environments for kindergarten classrooms.

Plastic pollution is one of the most severe environmental issues we are facing. Plastics are everywhere, from the human body to the most isolated locations on Earth. We present the 8th Continent, a gamified immersive art experience that provides a space for community dialogue. The game brings participatory expression into the digital realm to alert the public to the current extent of plastic pollution. The work enables the participants’ engagement as content creators to form an inclusive digital space for knowledge sharing and community expression. The paper focuses on the design methodology and development of the work for a large extended reality display.

Information visualization is a powerful tool to communicate complex social/political/environmental phenomena. In spite of that, the complexity of these phenomena and the data involved can result in complex visualization or simplification of the visuals and a consequent data loss. To tackle this topic, we designed an interactive visualization to facilitate the understanding and awareness of the interdependencies and consequences of a complex social/political contemporary phenomenon, such as the war ravaging Ukraine. To this end, we engaged the public through a visualization that presents the interdependencies between the Ukrainian war and climate change in terms of environment, food and raw materials produced and exported, people (refugees and displaced), and support (military, humanitarian, and economic) provided to Ukraine. We evaluated our prototype and found that our visualization increased the users’ awareness of war-affected aspects that directly or indirectly influence climate change, especially for food and people dimensions.

The neurodivergent population with Autism Spectrum Disorder (ASD) often lacks social competence. Numerous educational and clinical approaches have been utilized to teach social interactions to individuals with ASD. Game-driven interventions have shown promise in improving social interactions and communication abilities. This paper introduces "Emotion Adventure," a game-driven digital therapeutics prototype designed to teach the concept of the Theory of Mind, a crucial skill for social cognition. The study assesses the prototype’s usability and acceptability among children with ASD, focusing on game performance, satisfaction, difficulty, and interests. Performance data from the acceptance test were collected and analyzed. The findings indicate positive reception and user experience of the prototype, suggesting its potential to enhance social skills in children with ASD through video games. The study aims to identify potential biomarkers for treating and intervening with children with ASD based on the data analysis.

Air pollution is currently a hot topic and a significant threat that impacts not only climate change but also the health of individuals. As a matter of fact, it was estimated that it causes 6.7 million premature deaths annually, and, in 2021, 97% of the urban population was exposed to particulate matter concentrations above the World Health Organization’s health-based guideline. Generally, individuals have no perception of the air quality around them and in the city where they live, and their knowledge of the subject depends on what they learn from newscasts or newspapers. However, increasing awareness of the topic can help them make more conscious choices. To partly tackle this problem, we developed a web-based prototype exploiting two modalities to communicate air quality data: data sonification (through audio) and data visualization (through animated video). With the aim of investigating the best communication modality that, eventually, can raise awareness on the topic, we performed a preliminary study. To anticipate some findings, we found out that the videos were considered less mentally demanding and less frustrating, while the sound was considered more pleasant. At the same time, while the videos required less time to be understood and communicated a more precise level of pollution, the audios were considered, on hand, more involving, making the users also feel more immersed in the experience, and, on the other hand, gave the possibility to concentrate on something else while experiencing the data.

The purpose of this research concerns the integration between Human-Computer Interaction (HCI) and interactive digital tools to increase users’ awareness about sustainability in the context of Citizen Science communities. In particular, this research project aims to design and develop meaningful and effective interactive data storytelling and interactive digital narratives based on HCI methodologies for Citizen Science. Specifically, we will investigate how cultural elements can be implemented in the design process to improve significance and sense of ownership of technological tools for scientific communication. Such tools can be critical for scientific communication in the context of the local communities of citizen scientist, as the knowledge that they produce tend to remain in the niche of their group. Interactive digital tools can convey simple, yet meaningful narrations to the public and enhance both awareness and decision-making processes about sustainability.

Sustainability is currently a hot topic also inside universities. Every year, several universities publish an open report based on their efforts to achieve the 17 Sustainable Development Goals (SDGs). However, the university community is often unaware of these reports and does not know how an individual can contribute to sustainability inside and outside the university premises. To partly tackle that, we exploited a three-step process composed of interviews, data analysis, and co-design to understand which data are relevant in this context and what elements an interactive data visualization system should have to raise awareness of the actions performed by the university, and of what can be done by the university community to improve the overall sustainability of the campuses. In this paper, we present our experience in co-designing solutions with students of the University of Bologna. We finally extract some guidelines that can be exploited in a similar context to make university sustainability reporting an action toward a more sustainable future.

Solution of multi-objective optimization in the logistics sector have become an integral important part of the Intelligent Transportation System (ITS). In this work we focus on the intelligent and sustainable transportation processes through the design of the multi-objective model for the logistic route-order dispatching system. We consider transportation costs, emissions, order importance and risks for failures, for the logistic route-order dispatching system. We present an Integer Linear Programming (ILP) optimization model and apply state-of-the-art techniques as a part of SCIP framework to solve pilot problem instances and evaluate the performance of the model. We obtain results of solving the model on a single monolithic Google Cloud Compute (GCP) to estimate the time complexity of the solving process in relation to the various problem sizes. The results from the experiments show low complexity of the problems of various sizes. Therefore scalability of the model looks promising for the applicability in various industry-related scenarios and computing environments. In particular, using hybrid-cloud systems and state-of-the-art optimization frameworks such as IBM CPLEX or Gurobi.

While online dating applications (dating apps) have become prevalent, dating violence that causes various harm, including physical, psychological, and emotional damage, still exists. This dating violence has been more serious for marginalized users such as females and LGBTQ. Even though these risks and threats constantly happen, many dating apps have been criticized for not providing proper interventions to prevent those abuse and create a safe and inclusive environment for users. This study evaluated the safety and inclusiveness interventions of two popular dating apps, Tinder and Bumble, by adopting Responsible Social Media (RSM) Guidelines. An expert review with six industry experts was conducted using RSM Guidelines, and the following discussion sessions were analyzed to generate design implications to improve safety and inclusiveness interventions in the two dating apps. The two apps were found to lack interventions to protect users’ privacy and control possible abuse, such as disenabling location tracking and information verification processes. Based on the findings, this study suggests design implications for building safe and inclusive dating apps. Further, it discusses developing domain-specific RSM Guidelines for online dating app practitioners and researchers.