Semantic Enrichment of Vital Sign Streams through Ontology-based

Context Modeling using Linked Data Approach

Sachiko Lim, Rahim Rahmani and Paul Johannesson

Department of Computer and Systems Science, Stockholm University, Kista, Sweden

Keywords: Internet of Things (IoT), Semantic Enrichment, Ontology, Linked Data, Patient Health Monitoring, Patient

Management, Vital Sign, Healthcare, Infectious Disease Outbreak.

Abstract: The Internet of Things (IoT) creates an ecosystem that connects people and objects through the internet. IoT-

enabled healthcare has revolutionized healthcare delivery by moving toward a more pervasive, patient-

centered, and preventive care model. In the ongoing COVID-19 pandemic, it has also shown a great potential

for effective remote patient health monitoring and management, which leads to preventing straining the

healthcare system. Nevertheless, due to the heterogeneity of data sources and technologies, IoT-enabled

healthcare systems often operate in vertical silos, hampering interoperability across different systems.

Consequently, such sensory data are rarely shared nor integrated, which can undermine the full potential of

IoT-enabled healthcare. Applying semantic technologies to IoT is a promising approach for fulfilling

heterogeneity, contextualization, and situation-awareness requirements for real-time healthcare solutions.

However, the enrichment of sensor streams has been under-explored in the existing literature. There is also a

need for an ontology that enables effective patient health monitoring and management during infectious

disease outbreaks. This study, therefore, aims to extend the existing ontology to allow patient health

monitoring for the prevention, early detection, and mitigation of patient deterioration. We evaluated the

extended ontology using competency questions and illustrated a proof-of-concept of ontology-based semantic

representation of vital sign streams.

1 INTRODUCTION

Healthcare has marked a significant paradigm shift

from a centralized, professional-focused, and reactive

model to a more pervasive, patient-centered, and

preventive care model (Epstein et al., 2010). The

Internet of Things (IoT) creates an ecosystem that

connects people and objects through the internet. By

revolutionizing healthcare service delivery, IoT-

enabled healthcare has a high potential to improve

population health and transform a healthcare model

toward a more comprehensive personalized care

model (Kelly et al., 2020). For example, it enables

preventive primary healthcare services to be more

accessible and available by enabling remote and real-

time monitoring of the patient's health status and daily

activities, leading to a more proactive prediction of

health issues (Kelly et al., 2020).

The current COVID-19 pandemic has posed

devastating effects on global health and economies.

Given that the incidence of emerging infectious

diseases has been increasing at an unprecedented rate

(Jones et al., 2008), it also has manifested the pressing

need for more resilient healthcare systems against

future emerging infectious diseases. IoT technology

has been exerting its power through its potential to

mitigate the impacts of COVID-19 on individuals and

health systems. Those innovative IoT have been used

for screening and early diagnosis of COVID-19, triage

of patients, epidemiological surveillance (Golinelli et

al., 2020), and contact tracing (Swayamsiddha &

Mohanty, 2020). Other utility includes maintaining

social distancing, remote and real-time monitoring of

confirmed/asymptomatic/suspected cases, ensuring

adherence to isolation/quarantine, and after-recovery

follow-ups to understand long-time sequelae and

possible re-infection (Nasajpour et al., 2020). By

providing such capabilities, IoT-enabled healthcare

systems can also help prevent overstretching

healthcare systems (Swayamsiddha & Mohanty, 2020).

Thus, the successful integration of IoT technology in

existing healthcare systems seems to be a key to

increase preparedness and resilience for future

pandemics.

292

Lim, S., Rahmani, R. and Johannesson, P.

Semantic Enrichment of Vital Sign Streams through Ontology-based Context Modeling using Linked Data Approach.

DOI: 10.5220/0010582202920299

In Proceedings of the 10th International Conference on Data Science, Technology and Applications (DATA 2021), pages 292-299

ISBN: 978-989-758-521-0

A dramatic increase in IoT usage has generated

a massive amount of heterogeneous sensory data.

Nevertheless, due to the heterogeneity of data sources

and technologies, IoT-enabled healthcare systems are

often operating in vertical silos, which hampers

interoperability across different systems (Kelly et al.,

2020). As a result, such sensory data are rarely shared

nor integrated. Moreover, a previous systematic

literature review identified contextualization and

situation-awareness as some of the challenges IoT

applications in healthcare have been facing (Lim &

Rahmani, 2020). An underlying reason for the issue

is that raw sensory data are mere numerical values

that are not necessarily easy to associate with

meaningful and understandable information unless

the context of data is provided (Ganz et al., 2016).

Applying semantic web technologies to represent

IoT data is a promising approach for fulfilling

heterogeneity, contextualization, and situation-

awareness requirements for real-time healthcare

solutions. Adding semantic description can transform

raw data into an unambiguous machine-interpretable

form. The semantically enriched data can further be

processed and interpreted by machines to generate

meaningful knowledge (Biswanath, 2017).

The need for and execution of IoT-enabled

healthcare services largely varies depending on the

context (Alirezaie et al., 2017). Ontologies are among

the most appropriate approaches to perform context

modeling (Strang & Linnhoff-Popien, 2004).

Ontology refers to "a formal naming and definition of

the types, properties, and relationships of the entities

that really or fundamentally exist in a particular

domain of discourse" (Biswanath, 2017). Ontologies

are machine-interpretable due to the formal and

explicit specification of conceptualizations, which

has capabilities of knowledge sharing, logic

inferencing, knowledge reuse, and knowledge

integration (Biswanath, 2017; Perera et al., 2014).

Linked Data is another critical pillar of semantic

technologies, which refers to "a set of best practices

for publishing and interlinking structured data on the

Web." Linked Data connect items across different

data sources in a single global data space (Heath &

Bizer, 2011). Using an ontology complements Linked

Data by supporting data integration, schema

alignment, reasoning, and inferencing over data

(Biswanath, 2017).

Relevant previous studies have mainly focused on

annotating cross-sectional/categorical data rather

than continuous data. Jabbar et al. proposed an IoT-

based Semantic Interoperability Model (IoT-SIM) to

improve semantic interoperability among

heterogeneous IoT devices in the healthcare domain

(Jabbar et al., 2017). In their study, physicians made

a diagnosis using IoT devices and prescribed

medicine accordingly. They then semantically

annotated the diagnosis and prescription results (i.e.,

cross-sectional data) in RDF. Carbonaro et al.

proposed an ontology-based cognitive computing

eHealth system to achieve semantic interoperability

among heterogeneous IoT fitness and wellness

applications (Carbonaro et al., 2018). However, they

did not describe any details about the steps to annotate

sensor data.

There is, thus, a lack of studies that have

performed the enrichment of the sensor data streams

with their spatial, temporal, semantic meaning (Pacha

et al., 2020). Furthermore, to our knowledge, there is

no existing ontology enabling continuous patient

health monitoring for more effective patient

management during infectious disease outbreaks.

Therefore, this study aims to extend the IoT-

Stream ontology (Elsaleh et al., 2020) in order to

enable patient health monitoring for the prevention,

early detection, and mitigation of patient deterioration.

Using the extended ontology, we performed

ontology-based context modeling of vital signs (i.e.,

mapping vital sign data to ontology concepts) to add

contextual information to raw sensor data (i.e.,

semantic enrichment). We formulated the following

research question: "How can ontology-based context

modeling be used for the semantic representation of

vital sign streams from heterogeneous data sources

for enabling patient health monitoring and

management?" To address the research question, we

performed the following four steps:

1) Extend the existing ontology, named the IoT

for patient health monitoring (IoT4PHM)

ontology, to enable patient health monitoring

and management.

2) Extract vital signs data from two different data

sources.

3) Build and apply a semantic model to

semantically enrich vital signs using the

Resource Description Framework (RDF).

4) Perform a semantic search using a SPARQL

query language.

2 RELATED WORK

2.1 Ontology-based Context Modeling

of Sensor Data

Semantic Sensor Network (SSN) is an OWL 2

ontology that is a de-facto standard ontology for

Semantic Enrichment of Vital Sign Streams through Ontology-based Context Modeling using Linked Data Approach

293

describing sensors, their accuracy and capabilities,

observations, and sensing methods (Compton et al.,

2012).

Nevertheless, context modeling using heavy-

weight ontologies can increase computational

complexity and processing time, especially when data

volume is high (Perera et al., 2014). Such heavy-

weight ontologies are thus not suitable for (near) real-

time IoT applications. Light-weight ontologies,

therefore, have been created to cope with those real-

time applications.

For example, the Sensor, Observation, Sample,

and Actuator (SOSA) ontology is a light-weight,

general-purpose ontology that provides a flexible but

coherent framework to represent the entities, relations,

and activities involved in sensing, sampling, and

actuation (Janowicz et al., 2019). IoT-Lite is another

light-weight semantic model for IoT (Bermudez-Edo

et al., 2017). The ontology instantiates the SSN

ontology to describe key IoT concepts, enabling

interoperability and discovery of sensory data among

heterogeneous IoT platforms. However, SOSA and

IoT-Lite focus on devices rather than IoT data

streams (Elsaleh et al., 2020). Thus, IoT-Stream, a

light-weight semantic model for annotating IoT

streaming data, was created by extending SOSA.

2.2 Semantic Enrichment of Sensor

Data Streams

Discovering and analyzing sensor data requires

spatial, temporal, and thematic information.

Nevertheless, sensor observation data are by nature

opaque. Thus, metadata is crucial for managing

sensor data. A semantic sensor web (SSW) enriches

sensor data by providing the meaning for the sensor

data. The semantic enrichment facilitates

interoperability and enables situational awareness

and advanced application from heterogeneous

sensors (Sheth et al., 2008).

As written in Section 1, few studies have focused

on annotating sensor data streams. However, there are

still several pioneering studies that aimed to address

the research gap. Pacha et al. proposed a novel

framework called SEmantic Annotation over

Summarized sensOr Data stReam (SEASOR),

enabling the real-time semantic annotations of

streaming sensor data (Pacha et al., 2020). Their

framework facilitates sensor data stream analytics

through summarization, semantic annotation, and

query processing (Pacha et al., 2020). Semantic

annotation is performed over the summarized sensor

data using the base ontology extended from the SSN

ontology.

Alirezaire et al. presented a system called E-

care@home, which can integrate measurements from

heterogeneous sensor sources used for ambient

assisted living using ontologies (Alirezaie et al.,

2017). By augmenting devices and their

measurements into semantic representation, the

system provides the semantic interpretation of events

and context awareness. To semantically represent

sensor data, they developed the SmartHome

ontology, which includes modules representing a

smart home environment's physical and conceptual

aspects. They proposed to use a network of

interlinked ontology modules because real-time

reasoning requires reducing the complexity of

semantic reasoning (Alirezaie et al., 2017).

3 METHODS

3.1 Data Preparation

3.1.1 Electronic Medical Record (EMR)

Vital Sign Dataset

We extracted patient vital signs from the early

prediction of sepsis from clinical data published on

the PhysioNet website (Reyna et al., 2020). The

dataset consists of hourly vital sign summaries,

laboratory values, and static descriptions of 60,000

ICU patients from two hospitals. It includes 40

clinical variables: 8 vital signs, 26 laboratory

variables, and six demographic variables (Reyna et

al., 2020). Of the 40 variables, this study utilized the

following seven variables: age, gender, heart rate,

oxygen saturation, temperature, systolic blood

pressure, and respiration rate.

In addition, we simulated sensor ID by random

number generation and the start and end time of the

IoT stream, assuming that the selected vital signs are

monitored simultaneously. Since the location data are

required for context-aware health systems, we

conveniently extracted the location data from the

epidemiological dataset of the COVID-19 outbreak

(Xu et al., 2020). We randomly selected ten patients

who developed sepsis within 10 hours after admission

to ICU and used their 10-hour vital sign observations

for the semantic enrichment.

3.1.2 Radar Vital Sign Dataset

In the EMR vital sign dataset, every vital sign is

provided in hourly summaries. To demonstrate the

semantic enrichment of raw sensor data, we extracted

ECG raw data from a publicly available dataset which

DATA 2021 - 10th International Conference on Data Science, Technology and Applications

294

consists of 24 h of synchronized data from radar and

a reference device (Schellenberger et al., 2020). The

dataset contains data including ECG, impedance

cardiogram, and non-invasive continuous blood

pressure collected from 30 healthy participants. We

extracted ECG signals, tfm_ecg1, which were

recorded at the sampling frequency of 2000 Hz.

First, we randomly selected 10 of the total 30

participants and assessed their ECG signals' length.

Since one participant had the shortest ECG length of

1200000 (corresponding 10 minutes recording), we

truncated other participants' ECG to this point so that

every participant has the same ECG length.

We then applied a sliding window for the real-

time processing of ECG. We set the size of the sliding

window to 150000 samples (i.e., 60 seconds) and the

step size of the sliding window to 120000 samples

(i.e., 60 seconds), which indicates the overlapped

interval between sliding windows is 30000 samples

(i.e., 15 seconds).

After setting the sliding window, step size, and

overlapped interval, we applied the Pan Tompkins

algorithm for each sliding window using the

MATLAB function, "Complete Pan Tompkins

Implementation ECG QRS detector" (Sedghamiz,

n.d.). The algorithm is most widely used to detect

QRS complex for detecting and monitoring various

cardiovascular diseases (Fariha et al., 2020). After

detecting R peaks in each sliding window, RR

intervals were determined to compute heartbeats.

Please note that optimizing R-peak detection and

the size of sliding window and step size is out of this

study's scope. Other preprocessing methods for

denoising ECG and detecting QRS complexes can

undoubtedly be used to achieve clinical relevance.

We prepared the datasets from both data sources

with MATLAB ver. R2020b.

3.2 Context Modeling of Vital Signs

using Linked Data

We chose ontology-based context modeling because

the approach was identified to be the most promising

asset for context modeling in ubiquitous computing

(Strang & Linnhoff-Popien, 2004). Figure 1 shows

the overview of the IoT4PHM ontology. We reused

the IoT-Stream ontology. The ontology focused on

modeling an IoT stream, stream observations

belonging to the IoT stream, and analysis used for and

events detected from the IoT stream. Those concepts

are captured in four classes: IoTStream,

StreamObservation, Analytics, and Event, depicted in

sky blue rectangles in Figure 1 (Elsaleh et al., 2020).

The IoTStream class is the central concept of the

ontology, representing an IoT data stream generated

by an IoT source. The StreamObservation class is

continuous stream observations belonging to the IoT

stream, observed by a sensor device captured as a data

point over a time instant or a subset of data points

over a defined time interval. The Analytic class

captures the data analytics that has been applied to

analyze the IoT data stream. The Event class abstracts

the event that has been detected by using an analytics

process to an IoT data stream (Elsaleh et al., 2020).

Moreover, the IoT-Stream ontology is linked with

six concepts from external ontologies (Figure 1):

qoi:Quality, iot-lite:Service, sosa:Sensor,

qu:QuanityKind, qu:Unit, and geo:Point. The

qoi:Quality is the top class to describe the quality of

IoT data sources. The class has a sub-class called

Timeliness which defines a metrics category to

represent which rate a data source provides data

within a defined time span or age. The iot-lite:Service

is an abstract of a service provided by an IoT device.

The sosa:Sensor is a device, agent (including

humans), or software (simulation) that generates an

IoT stream. The qu:QuanityKind represents a

quantity without any numerical value or unit, while

qu:Unit abstracts the concept of measurement unit.

The geo:Point represents the latitude, longitude, and

altitude of the location where an IoT stream

originates. For a complete description of the

ontology, see (Elsaleh et al., 2020).

We added three concepts (Figure 1) to support the

data integration and sharing, knowledge

representation and reasoning, and computer-assisted

data analysis to enable patient health monitoring and

management for the prevention, early detection, and

mitigation of patient deterioration.

The first concept, Patient, stores the patient-

related information such as ID, age, sex, and social

determinants of health, which can have a considerable

effect on health outcomes. For example, the

increasing evidence shows that social determinants of

health, including poverty, physical environment,

race, or ethnicity, impacts COVID-19 morbidity and

mortality profoundly and unevenly (Abrams &

Szefler, 2020). Thus, the concept is essential for

understanding the patient's needs to provide optimal

healthcare services.

The second concept, UnderlyingHealthCondition,

is added to understand the patient's underlying health

conditions because they can significantly increase the

risk of a worse disease prognosis. For example, a

modeling study shows that the population with

underlying health conditions such as chronic kidney

Semantic Enrichment of Vital Sign Streams through Ontology-based Context Modeling using Linked Data Approach

295

Figure 1: The overview of the IoT4PHM ontology.

disease, diabetes, cardiovascular disease, and chronic

respiratory disease are at increased risk of severe

COVID-19 and hospitalization (Clark et al., 2020).

Having the UnderlyingHealthCondition class enables

identifying high-risk groups, which is crucial for

performing triage and rolling out effective epidemic

management (e.g., identifying individuals who may

need to be shielded or vaccinated first).

The third class, PatientManagement, is added

because it is essential to enable a computer-assisted

analysis to recommend patient management measures

(e.g., perform semantic reasoning to compute early

warning score and recommend the corresponding

patient management measure). The class has three

subclasses: HealthEducation, EmergencyAlert, and

ReferralToHealthcare. The HealthEducation

represents the concept of general public health

practices recommended by, for example, a public

health agency to reduce the transmission of infectious

diseases. The EmergencyAlert class abstracts a

concept of clinical alert that can be critically

important for an individual (e.g., dispatch an

ambulance). The ReferralToHealthcare represents the

concept of referring the patient to healthcare.

We also added PhysiologicalParameter and

EnvironmentalParameter as subclasses of

QuantityKind, which refers to an "aspect common to

mutually comparable quantities" and represents the

essence of a quantity without any numerical value or

unit (e.g., humidity) (Elsaleh et al., 2020). In addition

to physiological parameters, monitoring

environmental parameters can also play an essential

role in transmitting some infectious diseases such as

water-associated and vector-borne infectious diseases

(Yang et al., 2012). The environmental parameters

are also crucial for assessing their possible effects on

emerging infectious diseases and supporting

decision-making to control the disease effectively, as

has been done by (Poirier et al., 2020) during the early

phase of the ongoing COVID-19 pandemic.

We implemented an extension of the IoT-Stream

ontology using Protégé ver. 5.5.0 (Musen & Team,

2015) and evaluated the IoT4PHM ontology using

competency questions (CQ) listed in Table 1.

We performed semantic enrichment by mapping

data to the ontology classes using the open-source

tool called Karma, which is a data integration tool that

allows the transformation of the data into Linked Data

by creating URIs for entities (Knoblock et al., 2012).

Finally, we demonstrate basic query search on the

enriched RDF using SPARQL Protocol and RDF

Query Language (SPARQL) to extract some patient

health information. We used a SPARQL processor

called ARQ, which is part of the Apache Jena

framework (The Apache Software Foundation, n.d.).

Table 1: Competency Questions.

CQ1

What t

pes of patient data are collected?

CQ2

Is there any information gathered that can be

associated with a worse disease pro

nosis?

CQ3

What are the possible types of quantity kind

to monitor patient health in the context of an

infectious disease outbreak?

CQ4

What are the main types of patient

management that can be utilized for patient

tria

e accordin

to the detected event?

DATA 2021 - 10th International Conference on Data Science, Technology and Applications

296

4 ILLUSTRATION OF IoT4PHM

ONTOLOGY AND PROOF OF

CONCEPT

We illustrated how the IoT4PHM ontology enables

the semantic representation of the vital sign streams

to monitor the patient's health condition. We present

a proof-of-concept of ontology-based semantic

modeling and semantic enrichment of vital sign data

from the two different data sources.

First, we built the heart rate dataset's semantic

model from the EMR vital sign dataset. We semi-

automatically performed the semantic enrichment by

mapping data to the ontology entities using the Karma

Data Integration Tool. In addition, to handle the

volume and velocity of streaming ECG data from the

Radar Vital Sign dataset and realize its real-time

semantic enrichment, we performed summarization

to enrich ECG streams, inspired by the previous study

(Pacha et al., 2020). To summarize the 10-minute

ECG streams, we applied a sliding window to

compute the average heart rate over a set of one-

minute periods. After the summarization step, the

same semantic model used to annotate vital signs

from the EMR vital sign dataset can also be applied

to automatically annotate the summarized ECG data

streams from the Radar Vital Sign dataset. Figure 2

shows the enrichment of the first sliding window

from the patient's ECG stream whose ID is

"GDN0021" (from the Radar Vital Sign dataset) in a

turtle format. Semantic IoT data can later be queried,

interpreted, and reasoned to generate new information

and knowledge (Zgheib et al., 2020).

After converting patient data into RDF using

semantic enrichment, we performed a basic SPARQL

query search to extract some patient's information.

Figure 3 shows the SPARQL query we run to

identify the female patients older than or at the age of

70, with a respiration rate greater than or equal to 25

breaths per minute, and the search results.

5 CONCLUSIONS

There is a lack of an ontology that enables IoT-based

patient health monitoring and management in the

context of infectious disease outbreaks. To answer the

research question, we extended the IoT-Stream

ontology to create the IoT4PHM ontology to enable

patient health monitoring and management for the

prevention, early detection, and mitigation of patient

deterioration during infectious disease outbreaks.

We evaluated the IoT4PHM ontology using four

CQs in Table 1. The answer to the CQ1 is that the

current version of the ontology can store the patient's

ID, age, gender, symptoms, and (if any) recent

contact with a wild animal. Such information is

essential to collect, especially at an early phase of a

newly emerging infectious disease outbreak, for

investigating possible high-risk groups and

symptoms and virus spillover from wildlife.

Figure 2: An excerpt of ontology-based semantic

enrichment for the first sliding window of the ECG stream

for patient "GDN0021".

Figure 3: An excerpt of the SPARQL query and results for

extracting the IDs of the patients with a respiration rate ≥

25.

Semantic Enrichment of Vital Sign Streams through Ontology-based Context Modeling using Linked Data Approach

297

The answer to the CQ2 is that the IoT4PHM

ontology has the UnderlyingHealthCondition class

that can capture risk factors associated with the

increased risk of developing complications. This

information is critical to identify high-risk groups

who need to be prioritized for the treatment and

public health interventions.

The answer to the CQ3 is that both physiological

and environmental parameters are included as

subclasses of the QuantiyKind class. Since

environmental factors play a crucial role in

transmitting some infectious diseases, they

significantly impact public health strategies.

Finally, the answer to the CQ4 is the patient

management can be classified as either

HealthEducation, ReferralToHealthcare, and

EmergencyAlert, depending on the severity of the

detected event (e.g., an early warning score). The

classification contributes to ensuring the limited

resources are effectively allocated to those who need

most and prevent overburdening healthcare systems.

Therefore, the IoT4PHM ontology successfully

addressed all the CQs and can potentially

be used for effective patient health monitoring and

management during infectious disease outbreaks.

In our future study, we will further extend the

ontology by adding relevant concepts to annotate the

aggregated individual patient data to obtain

population-level data.

Furthermore, to evaluate the capability of the

IoT4PHM ontology more rigorously and improve the

quality of the ontology, we will invite domain experts

to assess the ontology in terms of accuracy, clarity,

and completeness. We also plan to evaluate the

validity of our ontology through a series of real

annotation scenarios.

ACKNOWLEDGEMENTS

We sincerely appreciate Dr. Stefano Bonacina

at Health Informatics Centre, Department of learning,

informatics, management, and ethics, Karolinska

Institutet, for his constructive advice on software

usage and the ontology extension.

REFERENCES

Abrams, E. M., & Szefler, S. J. (2020). COVID-19 and the

impact of social determinants of health. The Lancet

Respiratory Medicine, 8(7), 659–661.

Alirezaie, M., Renoux, J., Köckemann, U., Kristoffersson,

A., Karlsson, L., Blomqvist, E., Tsiftes, N., Voigt, T.,

& Loutfi, A. (2017). An Ontology-based Context-aware

System for Smart Homes: E-care@home. Sensors

(Basel, Switzerland), 17(7). https://doi.org/10.3390/

s17071586

Bermudez-Edo, M., Elsaleh, T., Barnaghi, P., & Taylor, K.

(2017). IoT-Lite: a lightweight semantic model for the

internet of things and its use with dynamic semantics.

Personal and Ubiquitous Computing, 21(3), 475–487.

https://doi.org/10.1007/s00779-017-1010-8

Biswanath, D. (2017). Examining the interrelatedness

between ontologies and Linked Data. Library Hi Tech,

35(2), 312–331. https://doi.org/10.1108/LHT-10-2016-

0107

Carbonaro, A., Piccinini, F., & Reda, R. (2018). Integrating

Heterogeneous Data of Healthcare Devices to enable

Domain Data Management. Journal of E-Learning and

Knowledge Society, 14, 45–56. https://doi.org/

10.20368/1971-8829/1450

Clark, A., Jit, M., Warren-Gash, C., Guthrie, B., Wang, H.

H. X., Mercer, S. W., Sanderson, C., McKee, M.,

Troeger, C., Ong, K. L., Checchi, F., Perel, P., Joseph,

S., Gibbs, H. P., Banerjee, A., Eggo, R. M.,

Nightingale, E. S., O’Reilly, K., Jombart, T., … Jarvis,

C. I. (2020). Global, regional, and national estimates of

the population at increased risk of severe COVID-19

due to underlying health conditions in 2020: a

modelling study. The Lancet Global Health, 8(8),

e1003–e1017. https://doi.org/10.1016/S2214-109X(20)

30264-3

Compton, M., Barnaghi, P., Bermudez, L., García-Castro,

R., Corcho, O., Cox, S., Graybeal, J., Hauswirth, M.,

Henson, C., Herzog, A., Huang, V., Janowicz, K.,

Kelsey, W. D., Le Phuoc, D., Lefort, L., Leggieri, M.,

Neuhaus, H., Nikolov, A., Page, K., … Taylor, K.

(2012). The SSN ontology of the W3C semantic sensor

network incubator group. Journal of Web Semantics,

17, 25–32. https://doi.org/https://doi.org/10.1016/

j.websem.2012.05.003

Elsaleh, T., Enshaeifar, S., Rezvani, R., Acton, S. T.,

Janeiko, V., & Bermudez-Edo, M. (2020). IoT-Stream:

A Lightweight Ontology for Internet of Things Data

Streams and Its Use with Data Analytics and Event

Detection Services. Sensors (Basel, Switzerland),

20(4), 953. https://doi.org/10.3390/s20040953

Epstein, R. M., Fiscella, K., Lesser, C. S., & Stange, K. C.

(2010). Why the nation needs a policy push on patient-

centered health care. Health Affairs (Project Hope),

29(8), 1489–1495. https://doi.org/10.1377/hlthaff.

2009.0888

Fariha, M. A. Z., Ikeura, R., Hayakawa, S., & Tsutsumi, S.

(2020). Analysis of Pan-Tompkins Algorithm

Performance with Noisy ECG Signals. Journal of

Physics: Conference Series, 1532, 12022.

https://doi.org/10.1088/1742-6596/1532/1/012022

Ganz, F., Barnaghi, P., & Carrez, F. (2016). Automated

Semantic Knowledge Acquisition From Sensor Data.

IEEE Systems Journal, 10(3), 1214–1225.

https://doi.org/10.1109/JSYST.2014.2345843

Golinelli, D., Boetto, E., Carullo, G., Nuzzolese, A. G.,

Landini, M. P., & Fantini, M. P. (2020). Adoption of

DATA 2021 - 10th International Conference on Data Science, Technology and Applications

298

Digital Technologies in Health Care During the

COVID-19 Pandemic: Systematic Review of Early

Scientific Literature. Journal of Medical Internet

Research, 22(11), e22280. https://doi.org/10.2196/

22280

Heath, T., & Bizer, C. (2011). Linked data: Evolving the

web into a global data space. In Synthesis lectures on

the semantic web: theory and technology (Vol. 1, Issue

1). Morgan & Claypool Publishers.

Jabbar, S., Ullah, F., Khalid, S., Khan, M., & Han, K.

(2017). Semantic Interoperability in Heterogeneous IoT

Infrastructure for Healthcare. Wireless

Communications and Mobile Computing, 2017,

9731806. https://doi.org/10.1155/2017/9731806

Janowicz, K., Haller, A., Cox, S. J. D., Le Phuoc, D., &

Lefrançois, M. (2019). SOSA: A lightweight ontology

for sensors, observations, samples, and actuators.

Journal of Web Semantics, 56, 1–10. https://doi.org/

10.1016/j.websem.2018.06.003

Jones, K. E., Patel, N. G., Levy, M. A., Storeygard, A.,

Balk, D., Gittleman, J. L., & Daszak, P. (2008). Global

trends in emerging infectious diseases. Nature,

451(7181), 990–993. https://doi.org/10.1038/nature

06536

Kelly, J. T., Campbell, K. L., Gong, E., & Scuffham, P.

(2020). The Internet of Things: Impact and Implications

for Health Care Delivery. Journal of Medical Internet

Research, 22(11), e20135. https://doi.org/10.2196/

20135

Knoblock, C. A., Szekely, P., Ambite, J. L., Goel, A.,

Gupta, S., Lerman, K., Muslea, M., Taheriyan, M., &

Mallick, P. (2012). Semi-automatically Mapping

Structured Sources into the Semantic Web BT - The

Semantic Web: Research and Applications (E. Simperl,

P. Cimiano, A. Polleres, O. Corcho, & V. Presutti

(eds.); pp. 375–390). Springer Berlin Heidelberg.

Lim, S., & Rahmani, R. (2020). Toward Semantic IoT Load

Inference Attention Management for Facilitating

Healthcare and Public Health Collaboration: A Survey.

Procedia Computer Science, 177, 371–378.

https://doi.org/10.1016/j.procs.2020.10.050

Musen, M. A., & Team, P. (2015). The Protégé Project: A

Look Back and a Look Forward. AI Matters, 1(4), 4–

12. https://doi.org/10.1145/2757001.2757003

Nasajpour, M., Pouriyeh, S., Parizi, R. M., Dorodchi, M.,

Valero, M., & Arabnia, H. R. (2020). Internet of Things

for Current COVID-19 and Future Pandemics: an

Exploratory Study. Journal of Healthcare Informatics

Research, 4(4), 325–364. https://doi.org/10.1007/

s41666-020-00080-6

Pacha, S., Murugan, S. R., & Sethukarasi, R. (2020).

Semantic annotation of summarized sensor data stream

for effective query processing. The Journal of

Supercomputing, 76(6), 4017–4039. https://doi.org/

10.1007/s11227-017-2183-7

Perera, C., Zaslavsky, A., Christen, P., & Georgakopoulos,

D. (2014). Context Aware Computing for The Internet

of Things: A Survey. IEEE Communications Surveys &

Tutorials, 16(1), 414–454. https://doi.org/10.1109/

SURV.2013.042313.00197

Poirier, C., Luo, W., Majumder, M. S., Liu, D., Mandl, K.

D., Mooring, T. A., & Santillana, M. (2020). The role

of environmental factors on transmission rates of the

COVID-19 outbreak: an initial assessment in two

spatial scales. Scientific Reports, 10(1), 1–11.

Reyna, M. A., Josef, C. S., Jeter, R., Shashikumar, S. P.,

Westover, M. B., Nemati, S., Clifford, G. D., &

Sharma, A. (2020). Early Prediction of Sepsis From

Clinical Data: The PhysioNet/Computing in Cardiology

Challenge 2019. Critical Care Medicine, 48(2), 210–

217. https://doi.org/10.1097/CCM.0000000000004145

Schellenberger, S., Shi, K., Steigleder, T., Malessa, A.,

Michler, F., Hameyer, L., Neumann, N., Lurz, F.,

Weigel, R., Ostgathe, C., & Koelpin, A. (2020). A

dataset of clinically recorded radar vital signs with

synchronised reference sensor signals. Scientific Data,

7(1), 291. https://doi.org/10.1038/s41597-020-00629-5

Sedghamiz, H. (n.d.). Complete Pan Tompkins

Implementation ECG QRS detector. MATLAB Central

File Exchange.

Sheth, A., Henson, C., & Sahoo, S. (2008). Semantic Sensor

Web. Internet Computing, IEEE, 12, 78–83.

https://doi.org/10.1109/MIC.2008.87

Strang, T., & Linnhoff-Popien, C. (2004). A Context

Modeling Survey. In Proceedings of the Workshop on

Advanced Context Modeling, Reasoning and

Management as Part of UbiComp.

Swayamsiddha, S., & Mohanty, C. (2020). Application of

cognitive Internet of Medical Things for COVID-19

pandemic. Diabetes & Metabolic Syndrome: Clinical

Research & Reviews, 14(5), 911–915. https://doi.org/

10.1016/j.dsx.2020.06.014

The Apache Software Foundation. (n.d.). ARQ - A SPARQL

Processor for Jena (2.0). https://jena.apache.org/

documentation/query/index.html

Xu, B., Gutierrez, B., Mekaru, S., Sewalk, K., Goodwin, L.,

Loskill, A., Cohn, E. L., Hswen, Y., Hill, S. C., Cobo,

M. M., Zarebski, A. E., Li, S., Wu, C.-H., Hulland, E.,

Morgan, J. D., Wang, L., O’Brien, K., Scarpino, S. V.,

Brownstein, J. S., … Kraemer, M. U. G. (2020).

Epidemiological data from the COVID-19 outbreak,

real-time case information. Scientific Data, 7(1), 106.

https://doi.org/10.1038/s41597-020-0448-0

Yang, K., LeJeune, J., Alsdorf, D., Lu, B., Shum, C. K., &

Liang, S. (2012). Global distribution of outbreaks of

water-associated infectious diseases. PLoS Negl Trop

Dis, 6(2), e1483.

Zgheib, R., Kristiansen, S., Conchon, E., Plageman, T.,

Goebel, V., & Bastide, R. (2020). A scalable semantic

framework for IoT healthcare applications. Journal of

Ambient Intelligence and Humanized Computing.

https://doi.org/10.1007/s12652-020-02136-2

Semantic Enrichment of Vital Sign Streams through Ontology-based Context Modeling using Linked Data Approach

299