Human Behavioral Data Research Articles

Visual Analytics of Smartphone-Sensed Human Behavior and Health.

Smartphone health sensing tools, which analyze passively gathered human behavior data, can provide clinicians with a longitudinal view of their patients' ailments in natural settings. In this Visualization Viewpoints article, we postulate that interactive visual analytics (IVA) can assist data scientists during the development of such tools by facilitating the discovery and correction of wrong or missing user-provided ground-truth health annotations. IVA can also assist clinicians in making sense of their patients' behaviors by providing additional contextual and semantic information. We review the current state-of-the-art, outline unique challenges, and illustrate our viewpoints using our work as well as those of other researchers. Finally, we articulate open challenges in this exciting and emerging field of research.

Data science: a game changer for science and innovation

This paper shows data science’s potential for disruptive innovation in science, industry, policy, and people’s lives. We present how data science impacts science and society at large in the coming years, including ethical problems in managing human behavior data and considering the quantitative expectations of data science economic impact. We introduce concepts such as open science and e-infrastructure as useful tools for supporting ethical data science and training new generations of data scientists. Finally, this work outlines SoBigData Research Infrastructure as an easy-to-access platform for executing complex data science processes. The services proposed by SoBigData are aimed at using data science to understand the complexity of our contemporary, globally interconnected society.

A neurocomputational theory of how rule-guided behaviors become automatic.

This article introduces a biologically detailed computational model of how rule-guided behaviors become automatic. The model assumes that initially, rule-guided behaviors are controlled by a distributed neural network centered in the prefrontal cortex, and that in addition to initiating behavior, this network also trains a faster and more direct network that includes projections from sensory association cortex directly to rule-sensitive neurons in the premotor cortex. After much practice, the direct network is sufficient to control the behavior, without prefrontal involvement. The model is implemented as a biologically detailed neural network constructed from spiking neurons and displaying a biologically plausible form of Hebbian learning. The model successfully accounts for single-unit recordings and human behavioral data that are problematic for other models of automaticity. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Temporal and cultural limits of privacy in smartphone app usage

Large-scale collection of human behavioural data by companies raises serious privacy concerns. We show that behaviour captured in the form of application usage data collected from smartphones is highly unique even in large datasets encompassing millions of individuals. This makes behaviour-based re-identification of users across datasets possible. We study 12 months of data from 3.5 million people from 33 countries and show that although four apps are enough to uniquely re-identify 91.2% of individuals using a simple strategy based on public information, there are considerable seasonal and cultural variations in re-identification rates. We find that people have more unique app-fingerprints during summer months making it easier to re-identify them. Further, we find significant variations in uniqueness across countries, and reveal that American users are the easiest to re-identify, while Finns have the least unique app-fingerprints. We show that differences across countries can largely be explained by two characteristics of the country specific app-ecosystems: the popularity distribution and the size of app-fingerprints. Our work highlights problems with current policies intended to protect user privacy and emphasizes that policies cannot directly be ported between countries. We anticipate this will nuance the discussion around re-identifiability in digital datasets and improve digital privacy.

Patterns of human exposure to early evening and outdoor biting mosquitoes and residual malaria transmission in Ethiopia

Ethiopia has shown a notable progress in reducing malaria burden over the past decade, mainly due to the scaleup of vector control interventions such as long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS). Based on the progress, the country has set goals to eliminate malaria by 2030. However, residual malaria transmission due to early evening and outdoor biting vectors could pose a challenge to malaria elimination efforts. This study assessed vector behavior, patterns of human exposure to vector bites and residual malaria transmission in southwestern Ethiopia. Anopheles mosquitoes were collected monthly from January to December 2018 using human landing catches (HLCs), human-baited double net traps, CDC light traps and pyrethrum spray catches. Human behavior data were collected using questionnaire to estimate the magnitude of human exposure to mosquito bites occurring indoors and outdoors at various times of the night. Enzyme-linked immunosorbent assay (ELISA) was used to determine mosquito blood meal sources and sporozoite infections. A total of 2,038 female Anopheles mosquitoes comprising Anopheles arabiensis (30.8%), An. pharoensis (40.5%), An. coustani (28.1%), An. squamosus (0.3%) and An. funestus group (0.2%) were collected. Anopheles arabiensis and An. pharoensis were 2.4 and 2.5 times more likely to seek hosts outdoors than indoors, respectively. However, 66% of human exposure to An. arabiensis and 39% of exposure to An. pharoensis bites occurred indoors for LLIN non-users. For LLIN users, 75% of residual exposure to An. arabiensis bites occurred outdoors while 23% occurred indoors before bed time. Likewise, 84% of residual exposure to An. pharoensis bites occurred outdoors while 15% occurred indoors before people retired to bed. Anopheles arabiensis and An. pharoensis were 4.1 and 4.8 times more likely to feed on bovine than humans, respectively. Based on the HLC, an estimated indoor and outdoor EIR of An. arabiensis was 6.2 and 1.4 infective bites/person/year, respectively, whereas An. pharoensis had an estimated outdoor EIR of 3.0 infective bites/person/year. In conclusion, An. arabiensis and An. pharoensis showed exophagic and zoophagic behavior. Human exposure to An. arabiensis bites occurred mostly indoors for LLIN non-users, while most of the exposure to both An. arabiensis and An. pharoensis bites occurred outdoors for LLIN users. Malaria transmission by An. arabiensis occurred both indoors and outdoors, whereas An. pharoensis contributed exclusively to outdoor transmission. Additional control tools targeting early-evening and outdoor biting malaria vectors are required to complement the current control interventions to control residual transmission and ultimately achieve malaria elimination.

Can an Instrument Validated to Assess Parent-Child Interactions in the Laboratory Setting Be Applied to Home-Based Observations?

Background: The Toothbrushing Observations Scale (TBOS) was developed in a laboratory setting to measure child and parent behaviors during toothbrushing. However, we required an instrument to assess home based behaviors. We assessed the feasibility of applying TBOS to observations of parents and their child (<3 years of age) in urban homes.Methods: Sample consisted of 36 families recruited from university and community pediatric dental/medical clinics and a Women, Infants, and Children center in Chicago as part of a pilot study for a larger clinical trial. The average age of children in our sample was 20.7 months. Most of the parent participants were mothers (90%), and 75% of the parents identified as Hispanic. Parent–child dyads were video-recorded during home-based toothbrushing activities and footage was reviewed by two independent TBOS coders.Results: The TBOS instrument consists of 12 parent and 18 child items. We were able to code five parent and ten child items.Conclusion: The feasibility of applying the TBOS measure to our study population was somewhat limited by factors related to home-based observations and the young age of children in our study. Instruments need to be validated across natural settings, such as the home, to increase the quality and accuracy of human behavioral data.

SS-ITS: secure scalable intelligent transportation systems

This paper introduces a secure and scalable intelligent transportation and human behavior system to accurately discover knowledge from urban traffic data. The data are secured using blockchain learning technology, where the scalability is ensured by a threaded GPU. In addition, different optimizations are provided to efficiently process data on the GPU. A reinforcement deep learning algorithm is also established to merge local knowledge discovered on each site into global knowledge. To demonstrate the applicability of the proposed framework, intensive experiments have been carried out on well-known intelligent transportation and human behavior data. Our results show that our proposed framework outperforms the baseline solutions for the outlier detection use case.

A meta-model for modular composition of tailored human digital twins in production

Multiple and diverse factory digital twins have been proposed in the literature. However, despite the recognized growing importance of workers in smart and autonomous industrial settings, such models still lack or oversimplify human representation. Human digital twins must include human monitoring and behavioural data and models based on psychophysical status, knowledge, skills, and personal needs to manage production systems that aim, at the same time, to achieve process performance and workers’ wellbeing. This paper proposes a meta-model based on data, events, and connectors that supports the modular composition of tailored human digital twins. This work also addresses an industrial application of the meta-model for preliminary validation.

A systematic literature review on intelligent automation: Aligning concepts from theory, practice, and future perspectives

With the recent developments in robotic process automation (RPA) and artificial intelligence (AI), academics and industrial practitioners are now pursuing robust and adaptive decision making (DM) in real-life engineering applications and automated business workflows and processes to accommodate context awareness, adaptation to environment and customisation. The emerging research via RPA, AI and soft computing offers sophisticated decision analysis methods, data-driven DM and scenario analysis with regard to the consideration of decision choices and provides benefits in numerous engineering applications. The emerging intelligent automation (IA) – the combination of RPA, AI and soft computing – can further transcend traditional DM to achieve unprecedented levels of operational efficiency, decision quality and system reliability. RPA allows an intelligent agent to eliminate operational errors and mimic manual routine decisions, including rule-based, well-structured and repetitive decisions involving enormous data, in a digital system, while AI has the cognitive capabilities to emulate the actions of human behaviour and process unstructured data via machine learning, natural language processing and image processing. Insights from IA drive new opportunities in providing automated DM processes, fault diagnosis, knowledge elicitation and solutions under complex decision environments with the presence of context-aware data, uncertainty and customer preferences. This sophisticated review attempts to deliver the relevant research directions and applications from the selected literature to the readers and address the key contributions of the selected literature, IA’s benefits, implementation considerations, challenges and potential IA applications to foster the relevant research development in the domain.

Why we need to account for human behavior and decision-making to effectively model the non-linear dynamics of livestock disease

Animal disease costs the livestock industries billions of dollars annually. These costs can be reduced using effective biosecurity. However, costs of biosecurity are steep and benefits must be weighed against the uncertain infection risks. Much effort has gone into determining efficacy of different biosecurity tactics and strategies. Unfortunately, the variability in human behavior and decision-making when confronted with risk information has largely been overlooked. Here we show that use of the human behavioral component is necessary to understand the patterns of infection incidence in livestock industries. Using an agent-based model developed with a foundation of supply chain and industry structural data, we integrate human behavioral data generated using experimental games that parameterizes communication strategies, learning, psychological discounting and categorization of human behavior along a risk aversion spectrum. The influence of risk communication strategies on human behavior can be tested with experimental gaming simulations and their impact on the system can be projected using agent-based models, delivering feedback to increase disease resiliency of production systems.

Quantifying Face-to-Face Communication among Multidisciplinary Medical Professions in a Convalescent Rehabilitation Ward by using a Name Tag-Type Information Communication Device

Context: Although quantitative analyses of human behavior data collected through wearable sensors have uncovered principles of complex human behaviors and organization performances, quantitative assessment has not been conducted on communication among the medical team, despite its importance. Aims: The aim of the study is to quantify face-to-face communication in a multiprofessional medical team using a sociophysical method and wearable sensors. Settings and Design: The study was conducted at a rehabilitation ward in Japan; this was a pilot study. Subjects and Methods: The amount of time spent on face-to-face communication among 36 multidisciplinary medical professional in a rehabilitation ward was recorded using name tag-type sensors for 2 weeks. The average communication time per week for each pair of participants was calculated; the network diagrams were drawn by using Cytoscape software. The unpaired t-test and Welch's F-test used together with the Games–Howell post hoc test were performed to compare the communication times between two groups and among three groups, respectively. Results: The following three groups emerged: ward staff (nurses and care workers), rehabilitation therapists, and physicians/medical social workers. The communication time between physicians and ward staff (mean, 8.0 min/week) did not differ from that between the physicians and rehabilitation therapists (5.5). The communication time was longer among rehabilitation therapists (61.7) than among the ward staff (45.7), both of which were longer than that between the ward staff and rehabilitation therapists (4.8). Conclusions: Quantitative measurement of face-to-face communication time using wearable sensors characterized the staff behavior in a multiprofessional medical practice in a rehabilitation ward, which may affect medical team performance, such as patient outcomes.

Innovation Trend of Edge Computing Technology Based on Patent Perspective

Edge computing is an important foundation for building 5G networks, but in my country, there are few applications or inventions based on edge computing. In order to improve the application of edge computing, this article innovatively designs a human behavior recognition system based on a patent perspective, which provides a reference for other researchers. This paper discusses and designs the software and hardware schemes and related communication methods of a new edge computing framework that combines edge devices and cloud computing centers. After processing the collected human behavior data, the behaviors of the corresponding monitoring objects are classified and modeled, and then the distributed computing of edge devices is used to modify these models. These systems are characterized by low energy consumption and fast response. The experimental results prove. The recognition efficiency of edge computing technology from the patent perspective has been greatly improved. Its recognition speed is more than 30% faster than other algorithm calculations, and the accuracy of recognition reaches 0.852, which is about 20% higher than traditional recognition. The authors show that edge computing technology based on a patent perspective can play an important role in our lives.

A computational model of language functions in flexible goal-directed behaviour

The function of language in high-order goal-directed human cognition is an important topic at the centre of current debates. Experimental evidence shows that inner speech, representing a self-directed form of language, empowers cognitive processes such as working memory, perception, categorization, and executive functions. Here we study the relations between inner speech and processes like feedback processing and cognitive flexibility. To this aim we propose a computational model that controls an artificial agent who uses inner speech to internally manipulate its representations. The agent is able to reproduce human behavioural data collected during the solution of the Wisconsin Card Sorting test, a neuropsychological test measuring cognitive flexibility, both in the basic condition and when a verbal shadowing protocol is used. The components of the model were systematically lesioned to clarify the specific impact of inner speech on the agent’s behaviour. The results indicate that inner speech improves the efficiency of internal representation manipulation. Specifically, it makes the representations linked to specific visual features more disentangled, thus improving the agent’s capacity to engage/disengage attention on stimulus features after positive/negative action outcomes. Overall, the model shows how inner speech could improve goal-directed internal manipulation of representations and enhance behavioural flexibility.

From likely to likable: The role of statistical typicality in human social assessment of faces

Humans readily form social impressions, such as attractiveness and trustworthiness, from a stranger’s facial features. Understanding the provenance of these impressions has clear scientific importance and societal implications. Motivated by the efficient coding hypothesis of brain representation, as well as Claude Shannon’s theoretical result that maximally efficient representational systems assign shorter codes to statistically more typical data (quantified as log likelihood), we suggest that social “liking” of faces increases with statistical typicality. Combining human behavioral data and computational modeling, we show that perceived attractiveness, trustworthiness, dominance, and valence of a face image linearly increase with its statistical typicality (log likelihood). We also show that statistical typicality can at least partially explain the role of symmetry in attractiveness perception. Additionally, by assuming that the brain focuses on a task-relevant subset of facial features and assessing log likelihood of a face using those features, our model can explain the “ugliness-in-averageness” effect found in social psychology, whereby otherwise attractive, intercategory faces diminish in attractiveness during a categorization task.

Deep learning for pedestrian collective behavior analysis in smart cities: A model of group trajectory outlier detection

This paper introduces a new model to identify collective abnormal human behaviors from large pedestrian data in smart cities. To accurately solve the problem, several algorithms have been proposed in this paper. These can be split into two categories. First, algorithms based on data mining and knowledge discovery, which study the different correlation among human behavioral data, and identify the collective abnormal human behavior from knowledge extracted. Secondly, algorithms exploring convolution deep neural networks, which learn different features of historical data to determine the collective abnormal human behaviors. Experiments on an actual human behaviors database have been carried out to demonstrate the usefulness of the proposed algorithms. The results show that the deep learning solution outperforms both data mining as well as the state-of-the-art solutions in terms of runtime and accuracy performance. In particular, for large datasets, the accuracy of the deep learning solution reaches 88%, however other solutions do not exceed 81%. Additionally, the runtime of the deep learning solution is below 50 seconds, whereas other solutions need more than 80 seconds for analyzing the same database.

Patterns of human exposure to malaria vectors in Zanzibar and implications for malaria elimination efforts

BackgroundZanzibar provides a good case study for malaria elimination. The islands have experienced a dramatic reduction in malaria burden since the introduction of effective vector control interventions and case management. Malaria prevalence has now been maintained below 1% for the past decade and the islands can feasibly aim for elimination.MethodsTo better understand factors that may contribute to remaining low-level malaria transmission in Zanzibar, layered human behavioural and entomological research was conducted between December 2016 and December 2017 in 135 randomly selected households across six administrative wards. The study included: (1) household surveys, (2) structured household observations of nighttime activity and sleeping patterns, and (3) paired indoor and outdoor mosquito collections. Entomological and human behavioural data were integrated to provide weighted estimates of exposure to vector bites, accounting for proportions of people indoors or outdoors, and protected by insecticide-treated nets (ITNs) each hour of the night.ResultsOverall, 92% of female Anopheles mosquitoes were caught in the rainy season compared to 8% in the dry season and 72% were caught outdoors compared to 28% indoors. For individual ITN users, ITNs prevented an estimated two-thirds (66%) of exposure to vector bites and nearly three quarters (73%) of residual exposure was estimated to occur outdoors. Based on observed levels of ITN use in the study sites, the population-wide mean personal protection provided by ITNs was 42%.Discussion/conclusionsThis study identified gaps in malaria prevention in Zanzibar with results directly applicable for improving ongoing programme activities. While overall biting risk was low, the most notable finding was that current levels of ITN use are estimated to prevent less than half of exposure to malaria vector bites. Variation in ITN use across sites and seasons suggests that additional gains could be made through targeted social and behaviour change interventions. However, even for ITN users, gaps in protection remain, with a majority of exposure to vector bites occurring outdoors before going to sleep. Supplemental interventions targeting outdoor exposure to malaria vectors, and groups that may be at increased risk of exposure to malaria vectors, should be explored.

Modeling the formation of social conventions from embodied real-time interactions

What is the role of real-time control and learning in the formation of social conventions? To answer this question, we propose a computational model that matches human behavioral data in a social decision-making game that was analyzed both in discrete-time and continuous-time setups. Furthermore, unlike previous approaches, our model takes into account the role of sensorimotor control loops in embodied decision-making scenarios. For this purpose, we introduce the Control-based Reinforcement Learning (CRL) model. CRL is grounded in the Distributed Adaptive Control (DAC) theory of mind and brain, where low-level sensorimotor control is modulated through perceptual and behavioral learning in a layered structure. CRL follows these principles by implementing a feedback control loop handling the agent's reactive behaviors (pre-wired reflexes), along with an Adaptive Layer that uses reinforcement learning to maximize long-term reward. We test our model in a multi-agent game-theoretic task in which coordination must be achieved to find an optimal solution. We show that CRL is able to reach human-level performance on standard game-theoretic metrics such as efficiency in acquiring rewards and fairness in reward distribution.

Human-Behavior and QoE-Aware Dynamic Channel Allocation for 5G Networks: A Latent Contextual Bandit Learning Approach

With the rapid advance of smart wireless technologies, a plethora of human behavioral data are generated in 5G networks, which is reported capable to improve network performance by leveraging intelligent channel resource allocation through big data analytics. However, what information can be extracted for the network mobility management, how to exploit the knowledge for resource allocation and to meet the user-centric quality of experience (QoE) are not well understood and fully explored. To address this problem, we propose an online learning algorithm for dynamic channel allocation based on contextual multi-armed bandit (CMAB) theory. Especially, we divide the stochastic human behavioral data into two categories: the user location and the QoE-driven context. Noticing that the distributions of CSI vary spatially, we define a set of user’s geographic locations that shares the same set of CSI distributions as a cluster , and the stochastic channel distributions vary across clusters. The problem is formulated as a novel latent SCB problem, where the proposed agnostic SCB algorithm could automatically find the underlying clusters and significantly improve the learning performance. We then extend our online learning algorithm into the practical multi-user random access scenario. We conduct experiments on a real dataset collected from China Mobile, which indicate that our algorithms outperform existing approaches tremendously and perform extremely well in large-scale and high-mobility networks.

0386 Identification of Sleep Complaints Using Social Media: Effect of the Daylight Savings Time to Standard Time Transition

Abstract Introduction Transitions to and from daylight savings time (DST) are natural experiments of circadian disruption and are associated with negative health consequences. Yet, the majority of the United States and several other countries still adopt these changes. Large observational studies focused on understanding the impact of DST transitions on sleep are difficult to conduct. Social media platforms, like Twitter, are powerful sources of human behavior data. We used machine learning to identify tweets reporting sleep complaints (TRSC) during the week of the standard time (ST)-DST transition. Next, we evaluated the circadian patterns of TRSC and compared their prevalence before and after the transition. Methods Using data publicly available via the Twitter API, we collected 500 tweets with evidence of sleep complaints, and manually annotated each tweet to validate true sleep complaints. Next, we calculated term frequency-inverse document frequency of each word in each tweet and trained a random forest to classify TRSC using a 3-fold cross-validation design. The trained model was then used to annotate a collection of tweets captured between Oct. 30, 2019-Nov. 6, 2019, overlapping with the DST-ST transition, which occurred on Nov. 3, 2019. Results Random forest demonstrated good performance in classifying TRSC (AUC[95%CI]=0.85[0.82-0.89]). This model was applied to 3,738,383 tweets collected around the DST-ST transition, and identified 11,044 TRSC. Posting of these tweets had a circadian pattern, with peak during nighttime. We found a higher frequency of TRSC after the DST-ST transition (0.33% vs. 0.27%, p&amp;lt;0.00001), corresponding to a ~20% increase in the odds of reporting sleep complaints (OR[95%CI]=1.21[1.16-1.25]). Conclusion Using machine learning and Twitter data, we identified tweets reporting sleep complaints, described their circadian patterns and demonstrated that the prevalence of these types of tweets is significantly increased after the transition from DST to ST. These results demonstrate the applicability of social media data mining for public health in sleep medicine. Support NIH (K01LM012870); AASM Foundation (194-SR-18)

Multimodal warning design for take-over request in conditionally automated driving

PurposeHumans are required to respond to a vehicle’s request to take-over anytime even when they are not responsible for monitoring driving environments in automated driving, e.g., a SAE level-3 vehicle. Thus, a safe and effective delivery of a take-over request from an automated vehicle to a human is critical for the successful commercialization of automated vehicles.MethodsIn the current study, a set of human-in-the-loop experiments was conducted to compare diverse warning combinations by applying visual, auditory, and haptic modalities under systematically classified take-over request scenarios in conditionally automated driving. Forty-one volunteers consisting of 16 females and 25 males participated in the study. Vehicle and human data on response to take-over request were collected in two take-over scenarios, i.e., a disabled vehicle on the road ahead and a highway exit.ResultsVisual-auditory-haptic modal combination showed the best performance in both human behavioral and physiological data and visual-auditory warning in vehicle data. Visual-auditory-haptic warning combination showed the best performance when considering all performance indices. Meanwhile, visual-only warning, which is considered as a basic modality in manual driving, performed the worst in the conditionally automated driving situation.ConclusionsThese findings imply that the warning design in automated vehicles must be clearly differentiated from that of conventional manual driving vehicles. Future work shall include a follow-up experiment to verify the study results and compare more diverse multimodal combinations.