Research indicates that smartphone users reuse the same applications throughout the day. This study aimed to forecast a list of probable applications to be launched on a smartphone based on prior usage patterns, without the use of contextual information. We proposed a ranked-based algorithm that considers the sequential behavior of application usage history and presents usage sessions as "application chains". We evaluated the algorithm using datasets from 397 users, comprising 433,663 application chains with a minimum of three applications and 174 applications for the longest chains, averaging 40.91 ± 18.76 applications per chains, recorded over varying time periods and across multiple countries. Our results indicate that the proposed algorithm outperforms alternative approaches, achieving a significantly higher F1 score of 62 ± 6% without the use of contextual information. The ability to predict application launch can enable the provision of additional services such as digital wellness and improved application’s Quality of Experience.

Human behaviour is a dense longitudinal multi-featured measure that directly impacts the health of individuals in the short and long terms. Therefore, issues usually emerge from the insistence on performing risky behaviours, such as smoking or eating fast foods, which continuously increase the gap between current and beneficial health states. This paper introduces the term "health debt" as an economic metaphor to represent the quantification of this gap in domains such as sleep, contributing to physical and mental health states. Then, we present a theoretical framework that relies on behaviour change recommendations to quantify this debt. The practical instantiation of this framework relies on passively assessed sleep related data via personal wearable devices, and uses of an attention-based predictive model as the fitness function of a genetic algorithm that acts as a recommender. We evaluate this proposal by means of a case example aimed at improving the sleep duration of individuals. Results show, for example, that the use of individual rather than generic datasets produces more accurate models. At the same time, the use of constraints on the variability of behaviours features generates more feasible recommendations. These foundations open new research opportunities to support the adoption of preventive medicine based on longitudinal wearable passive data analysis.

Wearable devices are a useful and widely used source of continuous and temporal dependant data. In contrast to the traditional clinical environment, these devices allow time series data collection in an individual’s daily living environment. However, missing data can occur while using them. Many techniques have been applied to solve these data gaps; nonetheless, missing time series data poses extra challenges, such as maintaining the temporal dependency. In this article, we addressed the forecast of sleep trackers data (sleeping heart rate (HR) and time asleep) for 2 main reasons: (1) to design models capable of accurately forecasting missing data from those devices, and (2) to apply those models to empower sleep interventions that may increase its quality, by forecasting future sleep events. We collected wearables data over 290 days (per individual) from 12 participants using a smartwatch and made this dataset publicly available. We then explored several hyperparameters of 2 Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU). We further elaborated and compared the performance of 3 approaches to training those RNNs. Although similar performance, slightly more accurate results were obtained after training a GRU network on an entire population’s dataset, which was able to forecast the average, minimum, and maximum sleeping HR with a root-mean-squared error (RMSE) of 4.4 (± 1.4), 4.9 (± 2.6), and 12.1 ( 4.0) beats per minute, respectively. However, the total time ±asleep was impossible to forecast with low error.

In recent years, research on the Quality of Experience (QoE) of smartphone applications has received attention from both industry and academia due to the complexity of quantifying and managing it. This paper proposes a smartphone-embedded system able to quantify and notify smartphone users of the expected QoE level (high or low) during their interaction with their devices. We conducted two in the wild studies for four weeks each with Android smartphones users. The first study enabled the collection of the QoE levels of popular smartphone applications' usage rated by 38 users. We aimed to derive an understanding of users' QoE level. From this dataset, we also built our own model that predicts the QoE level for application category. Existing QoE models lack contextual features, such as duration of the user interaction with an application and the user's current physical activity. Subsequently, we implemented our model in an Android application (called expectQoE) for a second study involving 30 users to maximize high QoE level, and we replicated a previous study (2012) on the factors influencing the QoE of commonly used applications. The expectQoE, through emoji-based notifications, presents the expected application category QoE level. This information enable the user's to make a conscious choice about the application to launch. We then investigated whether if expectQoE improved the user's perceived QoE level and affected their application usage. The results showed no conclusive user-reported improvement of their perceived QoE due to expectQoE. Although the participants always had high QoE application usage expectations, the variation in their expectations was minimal and not significant. However, based on a time series analysis of the quantitative data, we observed that expectQoE decreased the application usage duration. Finally, the factors influencing the QoE on smartphone applications were similar to the 2012 findings. However, we observed the emergence of digital wellbeing features as facets of the users' lifestyle choices.

Ontologies are commonly used as a strategy for knowledge representation. However, they are still \n presenting limitations to model domains that require broad forms of temporal reasoning. This study is part of \n the Onto-mQoL project and was motivated by the real need to extend static ontologies with diverse time \n concepts, relations and properties, which go beyond the commonly used Allen´s Interval Algebra. Therefore, \n we use the n-ary relations as the basis for temporal structures, which minimally modify the original ontology, \n and extend these structures with a generic set of time concepts (moments and intervals), time concept \n properties (precise and uncertain), time relations (interval-interval, interval-moment, and moment-moment), \n and time relation properties (qualitative and quantitative). We divided the scientific contribution of this study \n into three parts. Firstly, we present the ontological temporal model (classes and properties) and how it is \n integrated into static ontologies. Secondly, we discuss the creation of axioms that give the semantics for \n precise temporal elements. Finally, as our main contribution, these ideas are extended with axioms for \n uncertain time. All these elements follow the Ontology Web Language (OWL) standards, so this proposal is \n still compatible with the main ontology editors and reasoners currently available. A case example \n demonstrates the use of this approach in the nutrition assessment domain.

In this paper, we present a hybrid method (qualitative and quantitative) to model and predict the Quality of Experience (QoE) of mobile applications used on WiFi or cellular network. Our 33 living lab participants rated their mobile applications' QoE in various contexts for four weeks resulting in a total of 5663 QoE ratings. At the same time, our smartphone logger (mQoL-Log) collected background information such as network information, user activity, battery statistics and more. We focused this study on frequently used and highly interactive applications including Google Chrome, Google Maps, Spotify, Instagram, Facebook, Facebook Messenger and WhatsApp. After pre-processing the dataset, we used classical machine learning techniques and algorithms (Extreme Gradient Boosting) to predict the QoE of the application usage. The results showed that our model can predict the user QoE with 94±0.77 accuracy. Surprisingly, after the following top three features: session length, battery level and network QoS, the user activity (e.g., if walking) and intended action to accomplish with the app were the most predictive features. Longer application use sessions often have worse QoE than shorter sessions.

The increasing presence of stress in people' lives has motivated much research efforts focusing on continuous stress assessment methods of individuals, leveraging smartphones and wearable devices. These methods have several drawbacks, i.e., they use invasive external devices, thus increasing entry costs and reducing user acceptance, or they use some of privacy-related information. This paper presents an approach for stress assessment that leverages data extracted from smartphone sensors, and that is not invasive concerning privacy. Two different approaches are presented. One, based on smartphone gestures analysis, e.g., `tap', `scroll', `swipe' and `text writing', and evaluated in laboratory settings with 13 participants (F-measure 79-85 percent within-subject model, 70-80 percent global model); the second one based on smartphone usage analysis and tested in-the-wild with 25 participants (F-measure 77-88 percent within-subject model, 63-83 percent global model). Results show how these two methods enable an accurate stress assessment without being too intrusive, thus increasing ecological validity of the data and user acceptance.

Socially assistive agents, be it virtual avatars or robots, need to engage in social interactions with humans and express their internal emotional states, goals, and desires. In this work, we conducted a comparative study to investigate how humans perceive emotional cues expressed by humanoid robots through five communication modalities (face, head, body, voice, locomotion) and examined whether the degree of a robot's human-like embodiment affects this perception. In an online survey, we asked people to identify emotions communicated by Pepper - a highly human-like robot and Hobbit - a robot with abstract humanlike features. A qualitative and quantitative data analysis confirmed the expressive power of the face, but also demonstrated that body expressions or even simple head and locomotion movements could convey emotional information. These findings suggest that emotion recognition accuracy varies as a function of the modality, and a higher degree of anthropomorphism does not necessarily lead to a higher level of recognition accuracy. Our results further the understanding of how people respond to single communication modalities and have implications for designing recognizable multimodal expressions for robots.

In this paper, we study the influence of video stalling on QoE. We provide QoE models that are obtained in realistic scenarios on the smartphone, and provide energy-saving approaches for smartphone by leveraging the proposed QoE models in relation to energy. Results show that approximately 5J is saved in a 3 minutes video clip with an acceptable Mean Opinion Score (MOS) level when the video frames are skipped. If the video frames are not skipped, then it is suggested to avoid freezes during a video stream as the freezes highly increase the energy waste on the smartphones.

Mobile applications such as VoIP, (live) gaming, or video streaming have diverse QoS requirements ranging from low delay to high throughput. The optimization of the network quality experienced by end-users requires detailed knowledge of the expected network performance. Also, the achieved service quality is affected by a number of factors, including network operator and available technologies. However, most studies measuring the cellular network do not consider the performance implications of network configuration and management. To this end, this paper reports about an extensive data set of cellular network measurements, focused on analyzing root causes of mobile network performance variability. Measurements conducted on a 4G cellular network in Germany show that management and configuration decisions have a substantial impact on the performance. Specifically, it is observed that the association of mobile devices to a point of presence (PoP) within the operator's network can influence the end-to-end performance by a large extent. Given the collected data, a model predicting the PoP assignment and its resulting RTT leveraging Markov chain and machine learning approaches is developed. RTT increases of 58% to 73% compared to the optimum performance are observed in more than 57% of the measurements. Measurements of the response and page load times of popular websites lead to similar results, namely, a median increase of 40% between the worst and the best performing PoP.

Mobile applications such as VoIP, (live) gaming, or video streaming have diverse QoS requirements ranging from low delay to high throughput. The optimization of the network quality experienced by end-users requires detailed knowledge of the expected network performance. Also, the achieved service quality is affected by a number of factors, including network operator and available technologies. However, most studies focusing on measuring the cellular network do not consider the performance implications of network configuration and management. To this end, this paper reports about an extensive data set of cellular network measurements, focused on analyzing root causes of mobile network performance variability. Measurements conducted over four weeks in a 4G cellular network in Germany show that management and configuration decisions have a substantial impact on the performance. Specifically, it is observed that the association of mobile devices to a Point of Presence (PoP) within the operator's network can influence the end-to-end RTT by a large extent. Given the collected data a model predicting the PoP assignment and its resulting RTT leveraging Markov Chain and machine learning approaches is developed. RTT increases of 58% to 73% compared to the optimum performance are observed in more than 57% of the measurements.

Widespread acceptance and use of personalized mobile devices facilitates the provision of the ubiquitous mobile communications and computing applications that enable the human Quality of Life (QoL) improvement. However, there are multiple human factors influencing these applications, stemming from their users' needs, expectations, and ways of maximizing their experience and the impact on their QoL. For a successful adoption of these applications, our mobile technologies lab, mQoL, employs an iterative, user-centric, Living Lab approach for the applications' design and evaluation. This paper introduces the concept of mobile technologies for QoL specifically aiming to first understand the individual's QoL from information available in mobile devices, i.e., assess physical, psychological, social or environmental aspects of a daily life of an individual, and, based on this understanding, provide services to improve this individual's QoL. This paper also explains the methodological aspects of our research, including transdisciplinarity aspects, and delineates the research challenges for a Living Lab approach at large.

Software Defined Network (SDN) solutions bring larger flexibility to network providers and enable improvement of the users' experience for diverse multimedia networked services. Following this successful approach, we propose a Software Defined Health (SDH) solutions, bringing flexibility to health services providers and enabling an improvement of health state of their users. In this paper we coin the term and explain a set of research challenges that need to be addressed in order to fulfil the SDH vision.

mHealth solutions enable connected patients to take care of their health and care needs while `on the move'. There exist a variety of emerging innovative technologies supporting them in their goal of symptoms tracking, medication adherence, or connecting with others, either peers suffering from the same disease or expert practitioners supervising the health status from away. In this article we first present and categorize some examples of such technologies for performance maximization, health maintenance, disease prevention and disease management. Then we discuss these technologies in terms of their design space supporting the patient's self-care, supporting the peer efforts in care, or supporting the patient-practitioner interaction. We discuss the technologies' strengths and weaknesses in terms of the barriers for adoption and scaling. We especially focus on discussion on the evidence of mHealth as an intervention, for its safety, efficiency, efficacy and effectiveness and provide some guidelines for future work areas to be completed, such that evidence can be provided with a sufficient power and emerging mHealth technologies can enable the support for patient health self-management.

Localization has been an important research issue in ubiquitous computing and wireless sensor networks (WSNs). From location-based services to autonomous mobile devices, location is prevalent in a wide range of applications. Localization methods for a mobile node vary from using GPS to localizing with respect to beacons with known locations, or using sensors like accelerometers and compasses. All of these solutions either require additional sensors in a node that reduce its battery life or require some kind of infrastructure and access to a database of anchor locations. Additionally, most solutions are for networks where nodes can communicate indiscriminately with each other. In this paper we introduce ReNLoc, a minimalistic anchor-free multilateration algorithm for 2D space (extendable to 3D space) with a centralized and a distributed version made for networks where the mobile nodes can only get range measurements to nodes with an unknown but fixed position that we call base nodes. The main assumption is that there is a minimum of three base nodes. ReNLoc takes advantage of geometric constraints that arise from the range measurements and represents them as sets on which we perform minimization over known geometric properties. We show that ReNLoc outperforms the commonly used multidimensional scaling (MDS) algorithm in a purely indirect ranging setup.

The emerging interactive mobile applications rely the effectiveness of their delivery and their users quality of experience (QoS) on the performance of the underlying heterogeneous networking infrastructures, i.e., their quality of service (QoS). However, the performance that the users will be provided with is a priori unknown, depending on their specific context (e.g., location, time, operator). Moreover, some interactive applications require a particular QoS to be provided, in order to assure their users QoE, e.g., a conversational voice (VoIP) and video-session requires Round Trip Time (RTT) a no higher than 150 ms. In this paper we report the results from a six-months field study involving 15 users of three Swiss network operators where we quantified their networks QoS and context on a minute basis. We then analyse the duration of time intervals and context in which the users are provided with `VoIP-friendly RTT', thus emulating VoLTE. On average, in daily life environments, a mobile user would be able to use a VoIP for 8 minutes for all operators and would need to wait for 3 minutes in between the calls. These numbers reduce to 2-3 minutes each while user is a highly mobile state; only 20% of the VoIP calls over 3G/4G would be able to last 19 minutes. We discuss the implications for mobile network operators focusing on their intend for the deployment of VoLTE services, implications for users themselves, as well as implications for service providers striving to maximise users interactive mobile applications experience.

Use of public transportation is on a rise, enabling greener cities and improvement of the quality of life of the city residents. The challenge for the public transportation system is that its demands are non linear and the passengers' experience depends on many human factors. UnCrowdTPG is a mobile service and a research platform that has been developed for understanding the mobility patterns of the users of public transportation services in Geneva (TPG), as well as their experience with respect to the crowds perceived aboard the vehicles. UnCrowdTPG models algorithmically this experience using crowdsourcing techniques and it provides to the individual estimations of the future expected experiences aboard, thus enabling them to better plan their journeys with respect to the crowds. In this paper we position the UnCrowdTPG service as a research platform. Additionally we discuss the research questions in areas like local governance and public healthcare supported by the data analysed collectively from all UnCrowdTPG users. We also discuss the challenges for implementing the truly smart transportation system in a small-sized European city like Geneva.

Advances in miniaturized computing, storage and communication resources for personal wearable electronics devices, as well as the availability of diverse sensors for state assessment enable the development of a wide variety of wearable body area network (BAN) systems for psychophysiological measurements. These systems pave the way for acquisition of quality data relevant for research studies, amongst others, on the autonomic nervous system (ANS) activation in emotion. We present a high-level overview of BAN and its features, and we review 173 publications that report research studies on emotion activation and particularly 15 BAN systems employed in these studies. We discuss each BAN in terms of its capacity for ambulatory, i.e., out of the laboratory, assessment of the ANS activation in emotion. Finally, we highlight the design challenges to be addressed to make BAN systems effective for a wide range of applications to support users wellbeing and overall Quality of Life improvement. This paper provides knowledge to those interested in (ambulatory) assessment of the ANS activation on the set of systems currently used in research, and it aims to highlight opportunities for scientists and practitioners in, amongst others, the affective computing domain, enabling them to reflect upon their BAN requirements and study designs.

The five papers in this special section cover a wide range of topics, from mobile health systems to service-oriented decision support in medical systems and service quality evaluation.

Mobile devices with ever-increasing functionality and the ubiquitous availability of wireless communication networks are driving forces behind innovative mobile applications enriching our daily life. One of the performance measures for a successful application deployment is the ability to support application-data flows by heterogeneous networks within certain delay boundaries. However, the quantitative impact of this measure is unknown and practically infeasible to determine at real time due to the mobile device resource constraints. We research practical methods for measurement-based performance evaluation of heterogeneous data communication networks that support mobile application-data flows. We apply the lightweight Comparative Output-Input Analysis (COIA) method estimating an additional delay based on an observation interval of interest (e.g., 1 second) induced on the flow. An additional delay is the amount of delay that exceeds nonavoidable, minimal end-to-end delay caused by the networks propagation, serialization, and transmission. We propose five COIA methods to estimate additional delay, and we validate their accuracy with measurements obtained from the existing healthcare and multimedia streaming applications. Despite their simplicity, our methods prove to be accurate in relation to an observation interval of interest, and robust under a variety of network conditions. The methods offer novel insights into application-data delays with regards to the performance of heterogeneous data communication networks.

People-centric sensing using people's smartphones offers new research opportunities for large case studies. It presents many challenges, e.g., efficient capture of person's mobility, understanding of context changes and preservation of user privacy. We propose an accurate and energy-efficient method able to capture user's mobility, thus the context changes, while preserving his/her privacy. Our solution can be applied to systems that aim to efficiently sense context on smartphones to study large scale phenomena or perform location management.

Much research in ubiquitous computing assumes that a user's phone will be always on and at-hand, for collecting user context and for communicating with a user. Previous work with the previous generation of mobile phones has shown that such an assumption is false. Here, we investigate whether this assumption about users' proximity to their mobile phones holds for a new generation of mobile phones, smart phones. We conduct a data collection field study of 28 smart phone owners over a period of 4 weeks. We show that in fact this assumption is still false, with the within arm's reach proximity being true close to 50% of the time, similar to the earlier work. However, we also show that smart phone proximity within the same room (arm+room) as the user is true almost 90% of the time. We discuss the reasons for these phone proximities and the implications of this on the development of mobile phone applications, particularly those that collect user and environmental context, and delivering notification to users. We also show that we can accurately predict the proximity at the arm level and arm+room level with 75 and 83% accuracy, respectively, with features simple to collect and model on a mobile phone. Further we show that for several individuals who are almost always within the arm+room level, we can predict this level with over 90% accuracy.

The emergence of high bandwidth public wireless networks and miniaturized personal mobile devices gives rise to new mobile healthcare services. To this end, the COPD24 system provides highly customizable vital signs tele-monitoring and tele-treatment for Chronic Obtrusive Pulmonary Disease (COPD) patients based on a body area network (BAN) and a mobile health care (m-health) service platform utilizing next generation public wireless networks. The developed system allows the incorporation of diverse medical sensors via wireless connections, and the live transmission of the measured vital signs to healthcare providers as well as real-time feedback to the patient. The system also incorporates environmental state sensors, e.g., air quality sensors deployed in access points of Future Internet (FI) infrastructure, enabling to derive high-granularity, up to date information about the state of the air the patient is in and thus enabling him to make an informed choice on how to manage his health. In this paper we present the COPD system case study, and we derive its design implications for m-health and FI technologies.

Mobile devices with ever increasing functionality are an important driving force behind innovative mobile applications that enrich our daily life. The ubiquitous availability of wireless data communication networks is an additional driving force. Their ability to support application data flows is one of the performance criteria for successful mobile application deployment. Nevertheless, the quantitative impact of this performance is unknown and practically infeasible to determine at real-time at the application-level due to mobile device resource constraints. We research practical methods for measurement-based application-level performance evaluation of data communication networks that support mobile application data flows. In this paper, we apply the lightweight Comparative Output-Input Analysis (COIA) method that estimates additional delay at observation time scale of interest (e.g. 1 s) induced on the application data flow. The additional delay is the amount of delay exceeding non-avoidable, minimal end-to-end data delay caused by data communication network propagation and transmission delays. We propose five COIA methods to estimate the additional delay. We validate their accuracy with measurements obtained from an m-health application, namely health telemonitoring provided by the MobiHealth system. Despite their simplicity, our methods prove to be accurate in relation to the observation time scale of interest, and robust under a variety of network conditions. The methods offer novel insights into the impact of data communication network performance on the application data flow delay behaviour.

The advances in the area of location based computing will soon make it feasible to predict the availability of wireless networks and their application level Quality of Service (QoS) characteristics along the location and time dimensions. Such predictions are referred to as QoS predictions which are provided by the QoS Context Source (QoSCS) hosted in the fixed network. Herewith we present a network selection mechanism on the mobile device which requests QoS predictions as needed and uses these predictions combined with the current location and time to select the wireless network which satisfies QoS requirements of the services hosted on a mobile device. We evaluate the performance of the proposed scheme using simulations and compare the results with the other mechanism that only uses information locally available on the mobile device for the network selection. The results obtained from the simulations encourage further system development and prototyping.

Emerging healthcare applications rely on personal mobile devices to monitor and transmit patient vital signs to hospital-backend servers for further analysis. However, these devices have limited resources that must be used optimally in order to meet the application user requirements (e.g. safety, usability, reliability, performance). This paper reports on a case study of a Chronic Obstructive Pulmonary Disease telemonitoring application delivered by the mobihealth system. This system relies on a commercial mobile device with multiple (wireless) Network Interfaces (NI). Our study focuses on how NI activation strategies affect the application end-to-end data delay (important in case of an emergency situation) and the energy consumption of the device (important for device sustainability while a patient is mobile). Our results show the trade-off between end-to-end delay and battery life-time achieved by various NI activation strategies, in combination with application-data flow adaptation for realtime and near real-time data transmission. For a given mobile device, our study shows an increase in battery life-time of 40- 90 %, traded against higher end-to-end data delay. The insights of our studies can be used for application-data flow adaptation aiming to increase battery life-time and device sustainability for mobile patients; which effectively increases the healthcare application usability.