Search

In this paper, we present a model for supporting user‐specific explanations of AI systems. We then discuss a user study that was conducted to gauge whether the decisions for adjusting output to users with certain characteristics was confirmed to be of value to participants. We focus on the merit of having explanations attuned to particular psychological profiles of users, and the value of having different options for the level of explanation that is offered (including allowing for no explanation, as one possibility). Following the description of the study, we present an approach for mining data from user participant responses in order to determine whether the model that was developed for varying the output to users was well‐founded. While our results in this respect are preliminary, we explain how using varied machine learning methods is of value as a concrete step toward validation of specific approaches for AI explanation. We conclude with a discussion of related work and some ideas for new directions with the research, in the future. [ABSTRACT FROM AUTHOR]

Caffeine is a chemical that is commonly ingested by people daily to modify their behavior. Its physiological and psychological effects have been studied extensively for theoretical and applied reasons. We briefly review the current information on caffeine's physiological effects. We then review caffeine's effect on cognition and summarize these effects as changes in cognitive architectures (a fixed set of mechanisms to explain cognition), which provide a unified way to represent the changes. Modeling the effects of caffeine on an individual's physiology, as well as their cognitive function, is a logical addition to cognitive architectures because caffeine moderates cognitive performance. Cognitive architectures have recently been connected with physiological simulators, allowing physiological variables to interact with cognition. This combination provides a natural way to represent caffeine in current cognitive architectures and model how cognition and physiology interact, and use such models in system design. Our review notes how caffeine influences several aspects of users' capabilities that will influence system performance. It also notes gaps in the caffeine literature needed to improve models of users, including studies on the distribution of half-life, the need for the use of dosages vs. doses, and task-based effect studies. KEY POINTS: Caffeine is commonly consumed by individuals in varying environments, whether learning, operating high-stakes equipment, or going about normal work activities, to improve their performance. Caffeine consumption leads to physiological effects as well as effects on cognition. These effects can be summarized as a set of changes to cognition that can be implemented in a cognitive and physiological architecture that can be used to predict performance for use in system design. There are several details that future studies must consider to accurately measure these changes in a way that they can be implemented in a model. [ABSTRACT FROM AUTHOR]

Abstract Human digital twins are computational models of the human actions involved in interacting and operating technical artifacts. Such models provide a conceptual and practical tool for artificial intelligence designers when they seek to replace human work with intelligent machines. Indeed, digital twins have long served as models of technical and cyber-physical processes. Human digital twins have such models as their foundations but also include models of human actions. As a result, human digital twin models enable technology designers to model how people interact with intelligent technical artifacts. Yet, development of human digital twins is associated with certain conceptual problems. To clarify the basic idea, we constructed a human digital twin for Minsky’s M-Machine. The abstract conceptual structure of this machine and its generality allowed us to analyze the general properties of human digital twins, their design, and their use as tools in designing intelligent technologies.

Human digital twins are computational models of the human actions involved in interacting and operating technical artifacts. Such models provide a conceptual and practical tool for artificial intelligence designers when they seek to replace human work with intelligent machines. Indeed, digital twins have long served as models of technical and cyber-physical processes. Human digital twins have such models as their foundations but also include models of human actions. As a result, human digital twin models enable technology designers to model how people interact with intelligent technical artifacts. Yet, development of human digital twins is associated with certain conceptual problems. To clarify the basic idea, we constructed a human digital twin for Minsky's M-Machine. The abstract conceptual structure of this machine and its generality allowed us to analyze the general properties of human digital twins, their design, and their use as tools in designing intelligent technologies. [ABSTRACT FROM AUTHOR]

User modeling is a key topic in many applications, mainly social networks and information retrieval systems. To assess the effectiveness of a user modeling approach, its capability to classify personal characteristics (e.g., the gender, age, or consumption grade of the users) is evaluated. Due to the fact that some of the attributes to predict are multiclass (e.g., age usually encompasses multiple ranges), assessing fairness in user modeling becomes a challenge since most of the related metrics work with binary attributes. As a workaround, the original multiclass attributes are usually binarized to meet standard fairness metrics definitions where both the target class and sensitive attribute (such as gender or age) are binary. However, this alters the original conditions, and fairness is evaluated on classes that differ from those used in the classification. In this article, we extend the definitions of four existing fairness metrics (related to disparate impact and disparate mistreatment) from binary to multiclass scenarios, considering different settings where either the target class or the sensitive attribute includes more than two groups. Our work endeavors to bridge the gap between formal definitions and real use cases in bias detection. The results of the experiments, conducted on four real-world datasets by leveraging two state-of-the-art graph neural network-based models for user modeling, show that the proposed generalization of fairness metrics can lead to a more effective and fine-grained comprehension of disadvantaged sensitive groups and, in some cases, to a better analysis of machine learning models originally deemed to be fair. The source code and the preprocessed datasets are available at the following link: . [ABSTRACT FROM AUTHOR]

These days, due to the increasing amount of information generated on the web, most web service providers try to personalize their services. Users also interact with web-based systems in multiple ways and state their interests and preferences by rating the provided items. In this paper, we propose a framework to predict users' demographic based on ratings registered by users in a system. To the best of our knowledge, this is the first time that the item ratings are employed for users' demographic prediction problem, which has extensively been studied in recommendation systems and service personalization. We apply the framework to Movielens dataset's ratings and predict users' age and gender. The experimental results show that using all ratings registered by users improves the prediction accuracy by at least 16% compared with previously studied models. Moreover, by classifying the items as popular and unpopular, we eliminate ratings belong to 95% of items and still reach an acceptable level of accuracy. This significantly reduces update cost in a time-varying environment. Besides this classification, we propose other methods to reduce data volume while keeping the predictions accurate. [ABSTRACT FROM AUTHOR]

Accurate news representation is of crucial importance in personalized news recommendation. Most of existing news recommendation model lack comprehensiveness because they do not consider the higher-order structure between user–news interactions, relevance between user clicks on news. In this paper, we propose graph neural news recommendation based on multi-view representation learning which encodes high-order connections into the representation of news through information propagation along the graph. For news representations, we learn click news and candidate news content information embedding from various news attributes. And then combine obtained structure-based representations with representations from news content. Besides, we adopt a candidate-aware attention network to weight clicked news based on their relevance with candidate news to learn candidate-aware user interest representation for better matching with candidate news. The performance of the model has been improved in common evaluation metric. Extensive experiments on benchmark datasets show that our approach can effectively improve performance in news recommendation. [ABSTRACT FROM AUTHOR]

Soft skills, such as personality characteristics, communication skills and leadership, affect personal career performance greatly. Therefore, predicting the soft skills of interviewees can provide interviewers with a strong reference for the decision of hiring. Nowadays, as asynchronous video interviews have gradually become a popular form of interviews, automatic interview evaluation of soft skills has attracted widespread attention from researchers. However, existing automatic evaluation methods have two significant drawbacks. First, most of them model the problem as multi-modal fusion of long-term sequences, while ignoring the consistency of multi-modal expression in short-time windows, which is a key attribute of the interview scene. Second, without embedding of professional knowledge in the interview field, the interpretability of the model is relatively weak. To address the above problems, we propose a novel Multi-modal Window-Consistency Fusion network, namely MWCF, to capture the expression consistency of different modalities in a short-time window and re-weight the language signals to enhance important portions in verbal clues. Meanwhile, in order to enhance the interpretability of the evaluation model, we introduce the professional knowledge of interviewers by proposing a topic generation module based on question attention, and embedding the most representative keywords under different soft skills into the model. Furthermore, a real-world interview dataset is built by developing an asynchronous interview platform, and extensive experiments are conducted to show the superior performance of our proposed model. [ABSTRACT FROM AUTHOR]

In this article, we study the problem of personalized news headline generation, which aims to produce not only concise and fact-consistent titles for news articles but also decorate these titles as personalized irresistible reading invitations by incorporating readers' preferences. We propose an approach named PNG (Personalized News headline Generator) by utilizing distant supervision in readers' past click behaviors to resolve. First, user preference representations are learned through a knowledge-aware user encoder that comprehensively captures the genuine, sequential, and flash interests of users reflected in their historical clicked news. Then, a user-perturbed pointer-generator network is devised to accomplish the headline generation in which the learned user representations implicitly affect the word prediction. The proposed model is optimized by reinforcement learning solvers where indicators on factual, personalized, and linguistic aspects of the generated headline are regarded as rewards. Extensive experiments are conducted on the real-world dataset PENS,1 which is a large-scale benchmark collected from Microsoft News. Both the quantitative and qualitative results validate the effectiveness of our approach. [ABSTRACT FROM AUTHOR]

Creating an accurate model of a user's skills is an essential task for Intelligent Tutoring Systems (ITS) and robotic tutoring systems. This allows the system to provide personalized help based on the user's knowledge state. Most user skill modeling systems have focused on simpler tasks such as arithmetic or multiple-choice questions, where the user's model is only updated upon task completion. These tasks have a single correct answer and they generate an unambiguous observation of the user's answer. This is not the case for more complex tasks such as programming or engineering tasks, where the user completing the task creates a succession of noisy user observations as they work on different parts of the task. We create an algorithm called Time-Dependant Bayesian Knowledge Tracing (TD-BKT) that tracks users' skills throughout these more complex tasks. We show in simulation that it has a more accurate model of the user's skills and, therefore, can select better teaching actions than previous algorithms. Lastly, we show that a robot can use TD-BKT to model a user and teach electronic circuit tasks to participants during a user study. Our results show that participants significantly improved their skills when modeled using TD-BKT. [ABSTRACT FROM AUTHOR]

Augmented reality (AR) has gained significant attention in recent years for its applications in training and assistance in various industrial settings. Yet, a less understood question is: How can AR systems, coupled with artificial intelligence (AI) capabilities, adaptively tailor instructions and feedback interventions to the specific needs of users, their cognitive states, and levels of expertise during task execution? This paper addresses this question by conducting a systematic review that delves into three specific research areas: the state-of-the-art of AR-based systems for industrial applications in terms of features and training/assistance capabilities, the existing gaps in transforming AR into an “intelligent companion” that adapts to both the work context and the user’s needs, and how these sources of multimodal data captured by AR headsets, wearables, and IoT sensors can be harnessed to interpret, predict, and guide task performance and learning through AR. To this end, this paper synthesizes recent studies in the field of industrial AR, summarizing their main findings, contributions, and associated limitations when integrating AI capabilities into AR. The results suggest that AR can effectively tackle key industry challenges associated with training and upskilling, process improvement, and error prevention. However, limitations remain in integrating multimodal data-driven capabilities into AR to effectively tailor AR guides to how individual workers learn and perform complex industrial tasks. The paper concludes with a framework as well as several research directions and examples to realize intelligent AR systems enhanced with advanced AI capabilities for activity understanding, user modeling, and interventions, serving as adaptive and personalized companions for industrial workers.

The rapid development of the internet and smartphones has enabled people to access numerous information systems and large volumes of data. User profiling technology can meet the dual challenge of analyzing user characteristics, interests, or preferences and recommending corresponding resources. Nevertheless, the insufficiency and isolation of data in traditional information systems limit the effect of user profiling, and social networks can compensate for this deficiency. With massive quantities of data in text, images, videos, and relationships in social networks, user profiling can achieve highly accurate analytical results. This review comprehensively discusses user profiling for social networks, defines its criteria, and expounds on the entire process, from data collection to the profiling model and performance evaluation. It includes various models with advantages and disadvantages and corresponding application scenarios. Additionally, considering that technology serves humans, this review provides users with multiple applications in the industry for user profiling based on social networks. Furthermore, it discusses the ethical and legal issues associated with user profiling. Finally, this review highlights possible future research directions in this field. Overall, this review can help researchers enhance their understanding of the current state of research in the field of user profiling and gain ideas for further study.

Creating an accurate model of a user’s skills is an essential task for Intelligent Tutoring Systems (ITS) and robotic tutoring systems. This allows the system to provide personalized help based on the user’s knowledge state. Most user skill modeling systems have focused on simpler tasks such as arithmetic or multiple-choice questions, where the user’s model is only updated upon task completion. These tasks have a single correct answer and they generate an unambiguous observation of the user’s answer. This is not the case for more complex tasks such as programming or engineering tasks, where the user completing the task creates a succession of noisy user observations as they work on different parts of the task. We create an algorithm called Time-Dependant Bayesian Knowledge Tracing (TD-BKT) that tracks users’ skills throughout these more complex tasks. We show in simulation that it has a more accurate model of the user’s skills and, therefore, can select better teaching actions than previous algorithms. Lastly, we show that a robot can use TD-BKT to model a user and teach electronic circuit tasks to participants during a user study. Our results show that participants significantly improved their skills when modeled using TD-BKT.

AbstractIn response to the current problem of single sports plan and lack of long-term motivation in recommendation systems, a more intelligent personalized sports health recommendation system was designed by introducing Q-Learning (Quality Learning) algorithm. Firstly, user sports health data was collected, and the user model was constructed to track user sport preferences and historical behavior. Secondly, the sports environment was defined, including different types of sports activities, venues, and weather. Then, the reward function was formulated to reward and punish users based on their sports activities and goals, in order to maximize long-term health benefits. Finally, the Q-Learning algorithm was implemented to continuously iteratively learn and optimize user recommendation models to provide the best personalized sports recommendations. For personalized accuracy, indicators such as precision, recall, F1 value, MAE (Mean Absolute Error), and RMSE (Root Mean Square Error) were used to evaluate, while the system’s participation in sports, user satisfaction, long-term incentive effects, and overall health improvement were collected. The results showed that the average precision of the recommendation system on 10 different datasets was 88%, and the average AUC (Area Under Curve) was 96%, which was 6.7% higher than the SVD (Singular Value Decomposition) algorithm. The user’s sports persistence rate was improved by 25%, and the health score was improved by about 13.3%. These data not only reflect the superior performance of the recommendation system but also highlight its positive impact on long-term user motivation and overall health levels. The results indicate that the proposed personalized exercise health recommendation system, assisted by the Q-Learning algorithm, has significantly improved accuracy. Moreover, it offers users more intelligent and personalized exercise suggestions, effectively increasing long-term participation in physical activities and overall health levels. [ABSTRACT FROM AUTHOR]

The article focuses on addressing the challenges related to user heterogeneity in digital experiments, specifically in detecting and analyzing heterogeneous treatment effects (HTEs). The topics covered include proposing a framework for HTE detection, and demonstrating the prevalence and significance of user characteristics in influencing experiment outcomes, emphasizing the importance of considering user heterogeneity in experimentation analysis.

The lack of cybersecurity awareness among everyday users is a significant issue that can have detrimental effects on individuals and organizations alike. Traditional training methods such as slideshows and presentations have proven to be ineffective and can cause trainees to feel overwhelmed, overloaded, confused, or bored. To address this issue, the development of an adaptive serious game that teaches cybersecurity in an effective, engaging, and personalized manner is proposed. Serious games provide an immersive and simulated experience that can help users determine how they might act in real-life scenarios. However, existing cybersecurity serious games often measure effectiveness outside of the game using surveys, tests, and interviews, which can lessen immersion and the simulated experience. Therefore, measuring improvement within the game itself can provide more meaningful data and derive truer conclusions about the usefulness of serious games in teaching cybersecurity. The goal of this research study is to develop such a game and measure its effectiveness in a way that can inform future cybersecurity training programs. By providing an engaging and personalized experience, serious games can improve cybersecurity awareness and reduce the risk of cyber threats. The results show that 79% of the participants admitted that they learned new things by playing the game, 84% said that they were engaged by the background story, 68% agreed that they had fun while playing the game, and 84% would recommend the game to others. [ABSTRACT FROM AUTHOR]

This study investigated the use of affect and physiological signals of heart rate, electrodermal activity, pupil dilation, and skin temperature to classify advertising engagement. The ground truth for the affective and behavioral aspects of ad engagement was collected from 53 young adults using the User Engagement Scale. Three gradient-boosting classifiers, LightGBM (LGBM), HistGradientBoostingClassifier (HGBC), and XGBoost (XGB), were used along with signal fusion to evaluate the performance of different signal combinations as predictors of engagement. The classifiers trained on the fusion of skin temperature, valence, and tiredness (features n = 5) performed better than those trained on all signals (features n = 30). The average AUC ROC scores for the fusion set were XGB = 0.68 (0.10), LGBM = 0.69 (0.07), and HGBC = 0.70 (0.11), compared to the lower scores for the set of all signals (XGB = 0.65 (0.11), LGBM = 0.66 (0.11), HGBC = 0.64 (0.10)). The results also show that the signal fusion set based on skin temperature outperforms the fusion sets of the other three signals. The main finding of this study is the role of specific physiological signals and how their fusion aids in more effective modeling of ad engagement while reducing the number of features. [ABSTRACT FROM AUTHOR]

Researchers typically leverage side information, such as social networks or the knowledge graph, to overcome the sparsity and cold start problem in collaborative filtering. To tackle the limitations of existing user interest modeling, we propose a knowledge-enhanced user multi-interest modeling for recommender systems (KEMIM). First, we utilize the user-item historical interaction as the knowledge graph’s head entity to create a user’s explicit interests and leverage the relationship path to expand the user’s potential interests through connections in the knowledge graph. Second, considering the diversity of a user’s interests, we adopt an attention mechanism to learn the user’s attention to each historical interaction and each potential interest. Third, we combine the user’s attribute features with interests to solve the cold start problem effectively. With the knowledge graph’s structural data, KEMIM could describe the features of users at a fine granularity and provide explainable recommendation results to users. In this study, we conduct an in-depth empirical evaluation across three open datasets for two different recommendation tasks: Click-Through rate (CTR) prediction and Top-K recommendation. The experimental findings demonstrate that KEMIM outperforms several state-of-the-art baselines.

Persuasive technologies are designed to change human behavior or attitude using various persuasive strategies. Recent years have witnessed increasing evidence of the need to personalize and adapt persuasive interventions to various users and contextual factors because a persuasive strategy that works for one individual may rather demotivate others. As a result, several research studies have been conducted to investigate how to effectively personalize persuasive technologies. As research in this direction is gaining increasing attention, it becomes essential to conduct a systematic review to provide an overview of the current trends, challenges, approaches used for developing personalized persuasive technologies, and opportunities for future research in the area. To fill this need, we investigate approaches to personalize persuasive interventions by understanding user-related factors considered when personalizing persuasive technologies. Particularly, we conducted a systematic review of 72 research published in the last ten years in personalized and adaptive persuasive systems. The reviewed papers were evaluated based on different aspects, including metadata (e.g., year of publication and venue), technology, personalization dimension, personalization approaches, target outcome, individual differences, theories and scales, and evaluation approaches. Our results show (1) increased attention toward personalizing persuasive interventions, (2) personality trait is the most popular dimension of individual differences considered by existing research when tailoring their persuasive and behavior change systems, (3) students are among the most commonly targeted audience, and (4) education, health, and physical activity are the most considered domains in the surveyed papers. Based on our results, the paper provides insights and prospective future research directions.

Background: As of 2022, people are getting better at learning how to coexist with the Covid-19 global pandemic. In Saudi Arabia, many attempts have been made to raise public health awareness. However, most health awareness campaigns are generic and might not influence the desired behavior among individuals. Objectives: This study aims to apply geospatial intelligence and user modeling to profile the districts of the city of Jeddah. This customized map can provide a baseline for a customized health awareness campaign that targets the locals of each district individually based on the virus spread level. Methodology: It is ongoing research, which has resulted in the creation of a health messages library in the first phase [1]. This paper focuses on a second phase of the research study, which aims to provide a customized baseline for this campaign by applying the geospatial artificial intelligence technique known as space-time cube (STC). STC was applied to create a local map of the Saudi city of Jeddah, representing three different profiles for the city’s districts. The model is built using valid COVID-19 clinical data obtained from one of Jeddah’s general hospitals. Results and implications: When applied, STC displays three profiles for the districts of Jeddah city: high infection, moderate infection, and low infection. To assess the geo-intelligent map, a new instrument was created and validated. The usability and practicality of this map were quantitatively evaluated in a cross-sectional survey using the goal-question-metric measurement framework, and a total of 43 participants filled out the questionnaire. The results indicate that the geo-intelligent map is suitable for everyday use, as evidenced by the participants’ responses. We argue that the developed instrument can also be used to assess any geo-intelligence map. This research provides a legitimate approach to customizing health awareness messages during pandemics.

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

As recommender systems become increasingly important in daily human decision-making, users are demanding convincing explanations to understand why they get the specific recommendation results. Although a number of explainable recommender systems have recently been proposed, there still lacks an understanding of what users really need in a recommendation explanation. The actual reason behind users' intention to examine and consume (e.g., click and watch a movie) can be the window to answer this question and is named as self-explanation in this work. In addition, humans usually make recommendations accompanied by explanations, but there remain fewer studies on how humans explain and what we can learn from humangenerated explanations. To investigate these questions, we conduct a novel multi-role, multi-session user study inwhich users interact with multiple types of system-generated explanations as well as human-generated explanations, namely peer-explanation. During the study, users' intentions, expectations, and experiences are tracked in several phases, including before and after the users are presented with an explanation and after the content is examined. Through comprehensive investigations, three main findings have been made: First, we observe not only the positive but also the negative effects of explanations, and the impact varies across different types of explanations. Moreover, human-generated explanation, peer-explanation, performs better in increasing user intentions and helping users to better construct preferences, which results in better user satisfaction. Second, based on users' self-explanation, the information accuracy is measured and found to be a major factor associated with user satisfaction. Some other factors, such as unfamiliarity and similarity, are also discovered and summarized. Third, through annotations of the information aspects used in the human-generated selfexplanation and peer-explanation, patterns of how humans explain are investigated, including what information and how much information is utilized. In addition, based on the findings, a human-inspired explanation approach is proposed and found to increase user satisfaction, revealing the potential improvement of further incorporating more human patterns in recommendation explanations. These findings have shed light on the deeper understanding of the recommendation explanation and further research on its evaluation and generation. Furthermore, the collected data, including human-generated explanations by both the external peers and the users' selves, will be released to support future research works on explanation evaluation. [ABSTRACT FROM AUTHOR]

The generation of large amounts of personal data provides data centers with sufficient resources to mine idiosyncrasy from private records. User modeling has long been a fundamental task with the goal of capturing the latent characteristics of users from their behaviors. However, centralized user modeling on collected data has raised concerns about the risk of data misuse and privacy leakage. As a result, federated user modeling has come into favor, since it expects to provide secure multi-client collaboration for user modeling through federated learning. Unfortunately, to the best of our knowledge, existing federated learning methods that ignore the inconsistency among clients cannot be applied directly to practical user modeling scenarios, and moreover, they meet the following critical challenges: (1) Statistical heterogeneity. The distributions of user data in different clients are not always independently identically distributed (IID), which leads to unique clients with needful personalized information; (2) Privacy heterogeneity. User data contains both public and private information, which have different levels of privacy, indicating that we should balance different information shared and protected; (3) Model heterogeneity. The local user models trained with client records are heterogeneous, and thus require a flexible aggregation in the server; (4) Quality heterogeneity. Low-quality information from inconsistent clients poisons the reliability of user models and offsets the benefit from highquality ones, meaning that we should augment the high-quality information during the process. To address the challenges, in this paper, we first propose a novel client-server architecture framework, namely Hierarchical Personalized Federated Learning (HPFL), with a primary goal of serving federated learning for user modeling in inconsistent clients. More specifically, the client trains and delivers the local user model via the hierarchical components containing hierarchical information from privacy heterogeneity to join collaboration in federated learning. Moreover, the client updates the personalized user model with a fine-grained personalized update strategy for statistical heterogeneity. Correspondingly, the server flexibly aggregates hierarchical components from heterogeneous user models in the case of privacy and model heterogeneity with a differentiated component aggregation strategy. In order to augment high-quality information and generate high-quality user models, we expand HPFL to the Augmented-HPFL (AHPFL) framework by incorporating the augmented mechanisms, which filters out low-quality information such as noise, sparse information and redundant information. Specially, we construct two implementations of AHPFL, i.e., AHPFL-SVD and AHPFL-AE, where the augmented mechanisms follow SVD (singular value decomposition) and AE (autoencoder), respectively. Finally, we conduct extensive experiments on real-world datasets, which demonstrate the effectiveness of both HPFL and AHPFL frameworks. [ABSTRACT FROM AUTHOR]

Socially assistive robotics (SAR) aims at designing robots capable of guaranteeing social interaction to human users in a variety of assistance scenarios that range, e.g., from giving reminders for medications to monitoring of Activity of Daily Living, from giving advices to promote an healthy lifestyle to psychological monitoring. Among possible users, frail older adults deserve a special focus as they present a rich variability in terms of both alternative possible assistive scenarios (e.g., hospital or domestic environments) and caring needs that could change over time according to their health conditions. In this perspective, robot behaviors should be customized according to properly designed user models. One of the long-term research goals for SAR is the realization of robots capable of, on the one hand, personalizing assistance according to different health-related conditions/states of users and, on the other, adapting behaviors according to heterogeneous contexts as well as changing/evolving needs of users. This work proposes a solution based on a user model grounded on the international classification of functioning, disability and health (ICF) and a novel control architecture inspired by the dual-process theory. The proposed approach is general and can be deployed in many different scenarios. In this paper, we focus on a social robot in charge of the synthesis of personalized training sessions for the cognitive stimulation of older adults, customizing the adaptive verbal behavior according to the characteristics of the users and to their dynamic reactions when interacting. Evaluations with a restricted number of users show good usability of the system, a general positive attitude of users and the ability of the system to capture users personality so as to adapt the content accordingly during the verbal interaction. [ABSTRACT FROM AUTHOR]

Federated learning (FL) is a novel distributed machine learning paradigm. It can protect data privacy in distributed machine learning. Hence, FL provides new ideas for user behavior analysis. User behavior analysis can be modeled using multiple data sources. However, differences between different data sources can lead to different data distributions, i.e., non-identically and non-independently distributed (Non-IID). Non-IID data usually introduce bias in the training process of FL models, which will affect the model accuracy and convergence speed. In this paper, a new federated learning algorithm is proposed to mitigate the impact of Non-IID data on the model, named federated learning with a two-tier caching mechanism (FedTCM). First, FedTCM clustered similar clients based on their data distribution. Clustering reduces the extent of Non-IID between clients in a cluster. Second, FedTCM uses asynchronous communication methods to alleviate the problem of inconsistent computation speed across different clients. Finally, FedTCM sets up a two-tier caching mechanism on the server for mitigating the Non-IID data between different clusters. In multiple simulated datasets, compared to the method without the federated framework, the FedTCM is maximum 15.8% higher than it and average 12.6% higher than it. Compared to the typical federated method FedAvg, the accuracy of FedTCM is maximum 2.3% higher than it and average 1.6% higher than it. Additionally, FedTCM achieves more excellent communication performance than FedAvg. [ABSTRACT FROM AUTHOR]

The widespread use of temporal aspects in user modeling indicates their importance, and their consideration showed to be highly effective in various domains related to user modeling, especially in recommender systems. Still, past and ongoing research, spread over several decades, provided multiple ad-hoc solutions, but no common understanding of the issue. There is no standardization and there is often little commonality in considering temporal aspects in different applications. This may ultimately lead to the problem that application developers define ad-hoc solutions for their problems at hand, sometimes missing or neglecting aspects that proved to be effective in similar cases. Therefore, a comprehensive survey of the consideration of temporal aspects in recommender systems is required. In this work, we provide an overview of various time-related aspects, categorize existing research, present a temporal abstraction and point to gaps that require future research. We anticipate this survey will become a reference point for researchers and practitioners alike when considering the potential application of temporal aspects in their personalized applications. [ABSTRACT FROM AUTHOR]

The present paper introduces a novel approach to studying social media habits through predictive modeling of sequential smartphone user behaviors. While much of the literature on media and technology habits has relied on self-report questionnaires and simple behavioral frequency measures, we examine an important yet understudied aspect of media and technology habits: their embeddedness in repetitive behavioral sequences. Leveraging Long Short-Term Memory (LSTM) and transformer neural networks, we show that (i) social media use is predictable at the within and between-person level and that (ii) there are robust individual differences in the predictability of social media use. We examine the performance of several modeling approaches, including (i) global models trained on the pooled data from all participants, (ii) idiographic person-specific models, and (iii) global models fine-tuned on person-specific data. Neither person-specific modeling nor fine-tuning on person-specific data substantially outperformed the global models, indicating that the global models were able to represent a variety of idiosyncratic behavioral patterns. Additionally, our analyses reveal that individual differences in the predictability of social media use were not substantially related to differences in the frequency of smartphone use in general or the frequency of social media use, indicating that our approach captures an aspect of habits that is distinct from behavioral frequency. Implications for habit modeling and theoretical development are discussed. • The study introduces a novel conceptualization of social media habits based on the predictability of app use behaviors. • LSTM and transformer neural networks can predict momentary social media use from app log sequences. • There are pronounced individual differences in the person-level predictability of social media use. • Predictability was only weakly associated with behavioral frequency, indicating that it captures a distinct aspect of habits. [ABSTRACT FROM AUTHOR]

Abstract Gamification refers to the attempt to transform different kinds of systems to be able to better invoke positive experiences such as the flow state. However, the ability of such intervention to invoke flow state is commonly believed to depend on several moderating factors including the user’s traits. Currently, there is a dearth of research on the effect of user traits on the results of gamification. Gamer types (personality traits related to gaming styles and preferences) are considered some of the most relevant factors affecting the individual’s susceptibility to gamification. Therefore, in this study we investigate how gamer types from the BrainHex taxonomy (achiever, conqueror, daredevil, mastermind, seeker, socializer and survivor) moderate the effects of personalized/non-personalized gamification on users’ flow experience (challenge-skill balance, merging of action and awareness, clear goals, feedback, concentration, control, loss of self-consciousness and autotelic experience), enjoyment, perception of gamification and motivation. We conducted a mixed factorial within-subject experiment involving 121 elementary school students comparing a personalized version against a non-personalized version of a gamified education system. There were no main effects between personalization and students’ flow experience, perception of gamification and motivation, and enjoyment. Our results also indicate patterns of characteristics that can lead students to the high flow experience (e.g., those who prefer to play multiplayer have a high flow experience in both personalized and non-personalized versions). Based on our results, we provided recommendations to advance the design of gamifed educational systems.