Journal of Information Technology Case and Application

Research

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/utca20

Network analytics: an introduction and illustrative

applications in health data science

Pankush Kalgotra & Ramesh Sharda

To cite this article: Pankush Kalgotra & Ramesh Sharda (2023) Network analytics:
an introduction and illustrative applications in health data science, Journal of
Information Technology Case and Application Research, 25:3, 305-315, DOI:
10.1080/15228053.2023.2187995

To link to this article: https://doi.org/10.1080/15228053.2023.2187995

Published online: 15 Jun 2023.

Submit your article to this journal

Article views: 192

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=utca20
JOURNAL OF INFORMATION TECHNOLOGY CASE AND APPLICATION RESEARCH
2023, VOL. 25, NO. 3, 305–315
https://doi.org/10.1080/15228053.2023.2187995

EDITORIAL PREFACE ARTICLE

Network analytics: an introduction and illustrative

applications in health data science

ABSTRACT
Analytics researchers are widely using network analysis as a part
of their methodology. In this review paper, we discuss different
network concepts while summarizing some studies conducted
using descriptive, predictive, and prescriptive analytics
approaches. These applications illustrate the value of incorpor­
ating network properties of a phenomenon in better under­
standing the problem, prediction, and optimization of an
outcome of interest, especially in the health domain.

Introduction
Network Analysis is a popular method for analyzing complex problems
involving interactions among features or observations. While network analysis
is not a new technique, it has recently gained momentum due to the avail­
ability of cheap computing as the algorithms to analyze large networks require
large processing power. In addition, its suitability for analyzing large datasets
involving underlying relationships or connectedness has made it one of the top
choices for analytics researchers.
A network comprises nodes connected through well-defined edges. One
major area that generates network-type data is social media, where relation­
ships are explicitly embedded. In other words, the nodes make decisions to
connect to other nodes. For instance, two friends on Facebook make
a connection in the Facebook network, which is an explicit edge. However,
there are other types of networks with implicit relationships that are defined
using some underlying exchanges derived using some computation. Examples
include product co-purchase network (Dhar et al., 2014), ingredient network
(Teng et al., 2012), comorbidity network (Hidalgo et al., 2009; Kalgotra et al.,
2017), text-based network (Celardo & Everett, 2020), brain parts network
(Kalgotra & Sharda, 2018) and others.
In this paper, our focus is on discussing a variety of data science research
emerging using network analysis. It is important to note that our focus is not
specifically on social networks. We refer the interested reader to a review paper
by Borgatti et al. (2009) in the Science journal in which the authors elaborated

© 2023 The Author(s). Published with license by Taylor & Francis Group, LLC.
306 EDITORIAL

on the history of social network analysis, the theories emerging from the social
network analysis, and the type of research questions studied in the past.
In this review paper, we start by discussing the representation of network
data and common outputs from network analysis. Then, different types of
network analytics research are discussed. In each type, we include some of our
published papers as examples in addition to other papers. Finally, we conclude
by describing the potential contributions of network analytics research.

Network data and analysis output

As noted, previously, there are two main elements in a network: nodes and
edges. To connect the nodes, the edges are required to be defined. If the
connections are binary, i.e., present or absent, the edges are unweighted. In
contrast, if the strength of the ties between the nodes varies, the edges are
weighted. Moreover, if there is a direction in a connection between one node
to the other, the edges are directed, whereas the edges are undirected if there is
no direction in the relationship between two nodes. Therefore, a clear defini­
tion of the nodes and edges is required in the network analysis. Consider an
example of 100 friends forming an online social network. These 100 friends
can form an unweighted, undirected network based on who knows whom. At
the same time, the same 100 friends can form another weighted directed
network based on the frequency/volume of conversations between specific
friends. The latter network could be quite different from the first network; that
simply describes who is connected to who. Thus, it entirely depends on the
definition of the nodes and edges.
Unlike the above networks, where the relationship between nodes is expli­
citly defined by the nodes or by a user, in an implicit network, the network
itself is derived from the underlying relationship between the elements. It
requires an additional step to organize the data in the form of nodes and edges.
For instance, in a co-occurrence network where edges are defined based on the
presence of certain nodes in the same transactions, the nodes and edges are
derived from a transactional dataset. The common metrics applied in the
network research to derive the edges include Jaccard’s index (Jaccard, 1901),
Salton Cosine index (Salton & McGill, 1983), pointwise mutual information
(Teng et al., 2012), relative risk, and correlation coefficient, among others. For
example, Kalgotra et al. (2017) used Salton Cosine Index to define an edge in
the comorbidity network based on the co-occurrence of diseases in the
patients. The set of patients with multiple diseases during the hospital visits
made the transactional set from where the edges between the nodes were
derived using a similarity index. Common representations of the edges in
a network include an adjacency matrix, an adjacency list, and an edge list. The
network dataset is typically stored in a graph database.
 JOURNAL OF INFORMATION TECHNOLOGY CASE AND APPLICATION RESEARCH 307

The most common outputs obtained from the network analysis are the node
centralities such as degree, closeness, betweenness, and eigenvector. Other
measures at the node level may include clustering coefficient and page rank,
among others. At the structural level (macro), the metrics such as distribution
of centralities, average path length, and density result in a typical network
analysis. The micro (node) and macro (structure) level measures are used to
understand the properties of the network. For instance, Watts and Strogatz
(1998) used clustering coefficient and path length to derive the small-world
phenomenon, whereas Barabási and Albert (1999) used the distribution of the
degree centrality to describe the scale-free property of the network.
In addition to the mathematical computations, the common output from
a network analysis is a visualization. Since it is difficult to make sense of the
visualization of a network consisting of a large number of nodes and edges,
visualization researchers have developed different layouts to construct a view
of the network. Common layouts include Fruchterman-Reingold
(Fruchterman & Reingold, 1991), Yifan Hu Multilevel (Hu, 2005), etc.
Although a layout simplifies the structure of the network by reorganizing
the nodes in a visual space, it does not provide precise information about
the network, especially when the number of nodes is large. A network visual is
of little value if it is not annotated properly or presented creatively. Therefore,
the onus is on the researchers to present the visual by annotating it or
creatively organizing it. A good example of the annotation of a network can
be found in Hidalgo et al. (2009). On the other hand, Figure 1 presents
a network of diseases developed by the authors of this paper. In this undirected
comorbidity network, the nodes are diseases. Two diseases are connected if
these co-occur in the patients. The strength of a connection is computed using
Salton Cosine Index. To provide meaning to the network, we organized the
network in the shape of a human body and placed the diseases at the corre­
sponding organ system. The size of a node is based on the degree centrality. It
is easy to interpret from the visual that mental disorders and heart disorders
have the highest degree, which would have been difficult to infer using
a typical network layout. The same network was presented in Kalgotra et al.
(2017) with the Fruchterman-Reingold layout.
The original method of presenting the network is a graph matrix, which is
a mathematical representation of the network. With the advent of graphical
user interfaces, the two-dimensional visualizations of large networks became
popular with software such as UCINET (Borgatti et al., 2002), Gephi (Bastian
et al., 2009), etc. We expect the next step in the visual network analysis to be
the analysis through virtual and augmented reality, which is more immersive
and will likely increase the adoption of the network method further across the
disciplines (See Figure 2). It seems to be the natural evolution of network
analysis. Some researchers and companies are already exploring this idea of
network analysis in virtual reality such as VRNetzer (Pirch et al., 2021).
 308
EDITORIAL

Figure 1. Comorbidity network.

 JOURNAL OF INFORMATION TECHNOLOGY CASE AND APPLICATION RESEARCH 309

Figure 2. Evolution of network output.

Types of network analytics research

Most analytics studies use the above-discussed network measures and outputs
in one or another forms. However, researchers have used these innovatively to
perform descriptive, predictive, and prescriptive analytics research. In this
section, we discuss some examples of each type of analytics.

Descriptive network analytics

In descriptive analytics, the broad purpose is to find interesting patterns in

a large historical sample (representing the population) and develop novel
questions or observations about what is being observed. One interesting
descriptive study was performed more than two decades ago by Albert et al.
(1999) in which the authors demonstrated the network of websites on the
World Wide Web (WWW). The purpose was to find the diameter of the
WWW network. The authors found that any two randomly selected web pages
were only 19 connections away from each other, indicating the small world
phenomenon in the WWW. The study further argued that with the 1000%
increase in the number of web pages, the average diameter will only increase
by 2 units. This is an interesting descriptive analytics study that found WWW
to be a small network despite billions of nodes in it. Such an analysis of the
current WWW can yield interesting findings.
Another application of descriptive analytics is our study on health dispa­
rities by race (Kalgotra, Sharda, & Croff, 2020), in which we identified the
comorbidity differences in seven different races through comorbidity net­
works developed from patient-level data. The seven races were White,
African American, Asian, Hispanic, Native American, Bi- or Multi-racial,
and Pacific Islander. By using more than 18 million patients’ medical records,
one comorbidity network was developed for each race based on the multiple
diseases co-occurring in the patients. To define an edge, Salton Cosine Index
was used. The comorbidity networks of races were compared to find unique
and common comorbidities across the races. For instance, the relationship
between infectious and parasitic disorders with respiratory, circulatory, and
310 EDITORIAL

genitourinary system disorders was stronger among African Americans than

others. On the other hand, the connection of mental disorders with respira­
tory, musculoskeletal system, and connective tissue disorders was more pre­
valent in Whites than in other races.
Likewise, we conducted a similar network analysis to find comorbidity
differences between men and women (Kalgotra et al., 2017). A comorbidity
network for men and another one for women was created. The women’s
comorbidity network was denser than the men’s network. Specifically, mental
disorders had a higher number of connections with other diseases in women’s
multimorbidity network than in the men’s network. On the other hand, the
connection of chronic heart disorders with other disorders was stronger in
men than women.
Another interesting example of descriptive analytics is an ingredients net­
work created by Teng et al. (2012) in which the authors used the co-occurrence
of multiple ingredients in the same recipe to create the network and identify
the clusters of ingredients to suggest new recipes. Similar other descriptive
analytics applications are available that utilize network analysis as a core
methodology to model interactions in large datasets from the business and
other scientific domains.

Predictive network analytics

Since the main criticism of network analysis is its descriptive nature as
discussed by Borgatti et al. (2009), many researchers have attempted to create
predictive analytics applications. More recently, we are seeing network analy­
tics techniques used in conjunction with other approaches, such as deep
learning or natural language processing.
There are two potential types of predictive analytics using the network
method. First, the attributes of the nodes or edges of a network are predicted.
In this case, the attributes are an internal part of the network. One such
example is a study by Dhar et al. (2014) in which the authors predicted the
sales of books in a book co-purchase network created using a recommendation
system on the Amazon website. In this example, books are the nodes and sales
for the books were predicted based on the position of the nodes in the network.
Another common type of research in the first category of predictive analytics is
the link prediction problem in which the likelihood of the formation of new
edges is estimated. The primary objective of link prediction studies is to model
the evolution of a network. Examples include Davazdahemami et al. (2023). Lü
and Zhou (2011) highlighted different methods and applications of link pre­
diction in their survey.
Second and more recent, other external outcomes are predicted using novel
features generated using network methodology as predictors. In this case,
network analysis is a crucial part of the bigger methodology. One such study
 JOURNAL OF INFORMATION TECHNOLOGY CASE AND APPLICATION RESEARCH 311

is by Kalgotra and Sharda (2021), in which network analysis has been used to
predict an outcome exogenous to the network. Specifically, the comorbidity
network was used to predict hospital length of stay (LOS). In this paper, the
authors used electronic health records of more than 24.7 million patients
across 662 US hospitals over 16 years (2000–2015). The authors used a two-
step approach to create the machine learning model – creating comorbidity
networks in the first step and then creating machine learning models for
predicting LOS in the second step.
First, an independent sample of about three million patients was used to
create comorbidity networks in which the diseases were the nodes, and two
diseases were connected if they appeared in a patient during the same hospital
visit. The comorbidity networks were used to create new variables for the
remaining patients who were not part of the network analysis. To understand
the new features generated using comorbidity networks, consider a patient
who visits the hospital with a complaint of a hypothetical disease A. The only
disease-related information available at the time of admission is disease A. In
our application case, the network was searched for disease A, and the top five
connected diseases were identified. These diseases were labeled as probable
diseases as these were likely to be diagnosed during the hospital stay. In
addition, a patient may have a history of diseases in the system, termed
historical diseases. Together, the probable and historical diseases were called
latent comorbidities. The new construct of latent comorbidities was then used
in modeling and predicting LOS at the time of admission. The predictive
models for LOS were created with patient demographics, the known diseases
at the time of admission, and latent comorbidities as the independent vari­
ables. The Long-Short-Term Memory (LSTM) models were created without
and with the latent comorbidities to compute the explanatory and predictive
power added by the proposed variables. In terms of variance explained, the
new construct added 3.6%, and in terms of mean absolute percent error
(MAPE), the latent variable improved the MAPE by 1.9%. Although the
numbers seem low, these are equivalent to the improvement in the forecast
by $882.8 million. Therefore, the gain is practically significant. See Kalgotra et
al. (2023) for another such study.
As evident from the examples above, networks can be used for predictive
modeling to gain additional predictive power. More such network-driven
methods are required to predict outcomes that are endogenous or exogenous
to the networks.

Prescriptive network analytics

The third type of analytics is prescriptive analytics in which the objec­
tive is to find the optimal solution. Network analysis is a popular
research area for prescriptive analytics. The majority of the prescriptive
312 EDITORIAL

analytics research applying network analysis involves efficient processing

and traversing of the network. Examples include creating prescriptive
models to identify cliques in the network, as discussed by Miao and
Balasundaram (2017), to traverse the network and identify the shortest
path between nodes (Selim & Zhan, 2016), etc. Several Operations
Research and Computer Science researchers are focusing on the topics
of efficiently computing different network metrics. See Balakrishnan
(2019) for such models. More prescriptive analytics applications in
different domains are required to be explored using network analytics.

Conclusions
Network science has been used as a theory to understand an emergent
phenomenon and as a methodology to model relationships. Although our
focus is not on theory-driven network analysis, it is worthwhile to mention
some important concepts and theories derived from network analysis. Some
of the well-known concepts include random graphs (Erdős & Rényi, 1959),
scale-free networks (Barabási & Albert, 1999), strength of weak ties
(Granovetter, 1973), power law distribution of WWW (Adamic &
Huberman, 2000), small world phenomenon (Watts & Strogatz, 1998),
Benford’s law in online social network (Golbeck, 2015), role of cliques
(Provan & Sebastian, 1998), information diffusion (Bakshy et al., 2012),
preferential attachment (Newman, 2001), network flow (Borgatti, 2005) and
community detection (Reichardt & Bornholdt, 2006), among others. In
addition, studies have been designed to understand the validity or the
structure of the network (Kalgotra, Sharda, & Luse, 2020).
In data-driven research, the purpose is to discover new theories and pat­
terns. Therefore, it is important to generalize the novel relationships between
concepts involving network analysis. In network analytics studies, the con­
textual and methodological contributions are more apparent. In the papers
with contextual contributions, novel networks are created. In other words,
a problem is studied through a novel network lens. On the other hand, in the
papers with methodological contributions, the network analysis is a crucial
part of a bigger methodological process and thus, contributes to the method of
the study. Subsequently, it is important to generalize the methodology so that
it can be applied in different settings and different problem domains.
In this paper, we attempted to review different types of analytics research
conducted using network analysis. In each type of analytics, several papers
across the disciplines were discussed. In addition, the relevant concepts and
references were listed throughout the paper. Therefore, our paper can be used
as a guide by researchers and educators interested in learning and applying
network methodology.
 JOURNAL OF INFORMATION TECHNOLOGY CASE AND APPLICATION RESEARCH 313

Notes on contributors
Pankush Kalgotra is an Assistant Professor of Business Analytics at Auburn University. He
earned his Ph.D. in Management Science and Information Systems from Oklahoma State
University. His research interests include healthcare analytics, network science, neuroimaging
in information systems, and the dark side of information technology. His recent papers have
appeared in journals such as Journal of Management Information Systems, Journal of Business
Research, Nature: Scientific Reports, International Journal of Medical Informatics, Computer
Methods and Programs in Biomedicine, Information Systems Frontiers, Decision Sciences
Journal of Innovative Education and others.
Ramesh Sharda is the Vice Dean for Research and Graduate Programs, Watson/ConocoPhillips
Chair and a Regents Professor of Management Science and Information Systems in the Spears
School of Business at Oklahoma State University. His research has been published in major
journals in management science and information systems including Management Science,
Information Systems Research, Journal of Management Information Systems, Operations
Research Nature: Scientific Reports, INFORMS Journal on Computing, and many others. He
has coauthored two textbooks: Analytics, Data Science, and Artificial Intelligence: Systems for
Decision Support, 11th edition, and Business Intelligence, Analytics, and Data Science: A
Managerial Perspective, 4th edition. He is a member of several editorial boards, served as the
Executive Director of Teradata University Network, and was inducted into the Oklahoma Higher
Education Hall of Fame. Dr. Sharda is a Fellow of INFORMS and AIS.

References
Adamic, L. A., & Huberman, B. A. (2000). Power-law distribution of the world wide web.
Science, 287(5461), 2115. https://doi.org/10.1126/science.287.5461.2115a
Albert, R., Jeong, H., & Barabási, A. L. (1999). Diameter of the world-wide web. Nature, 401
(6749), 130–131. https://doi.org/10.1038/43601
Bakshy, E., Rosenn, I., Marlow, C., & Adamic, L. (2012, April 16 - 20). The role of social
networks in information diffusion. In Proceedings of the 21st international conference on
World Wide Web, Lyon, France, (pp. 519–528).
Balakrishnan, V. K. (2019). Network optimization. Chapman and Hall/CRC.
Barabási, A. L., & Albert, R. (1999). Emergence of scaling in random networks. Science, 286
(5439), 509–512. https://doi.org/10.1126/science.286.5439.509
Bastian, M., Heymann, S., & Jacomy, M. (2009, March). Gephi: An open source software for
exploring and manipulating networks. Proceedings of the International AAAI Conference on
Web and Social Media, 3(1), 361–362. https://doi.org/10.1609/icwsm.v3i1.13937
Borgatti, S. P. (2005). Centrality and network flow. Social Networks, 27(1), 55–71. https://doi.
org/10.1016/j.socnet.2004.11.008
Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet for windows: Software for social
network analysis. Harvard, MA: Analytic Technologies, 6, 12–15. https://pages.uoregon.edu/
vburris/hc431/Ucinet_Guide.pdf
Borgatti, S. P., Mehra, A., Brass, D. J., & Labianca, G. (2009). Network analysis in the social
sciences. Science, 323(5916), 892–895. https://doi.org/10.1126/science.1165821
Celardo, L., & Everett, M. G. (2020). Network text analysis: A two-way classification approach.
International Journal of Information Management, 51, 102009. https://doi.org/10.1016/j.
ijinfomgt.2019.09.005
314 EDITORIAL

Davazdahemami, B., Kalgotra, P., Zolbanin, H. M., & Delen, D. (2023). A developer-oriented
recommender model for the app store: A predictive network analytics approach. Journal of
Business Research, 158, 113649. https://doi.org/10.1016/j.jbusres.2023.113649
Dhar, V., Geva, T., Oestreicher-Singer, G., & Sundararajan, A. (2014). Prediction in economic
networks. Information Systems Research, 25(2), 264–284. https://doi.org/10.1287/isre.2013.
0510
Erdős, P., & Rényi, A. (1959). On random graphs. I. Publicationes Mathematicate, 6(3–4),
290–297. https://doi.org/10.5486/PMD.1959.6.3-4.12
Fruchterman, T. M., & Reingold, E. M. (1991). Graph drawing by force‐directed placement.
Software: Practice & Experience, 21(11), 1129–1164. https://doi.org/10.1002/spe.4380211102
Golbeck, J. (2015). Benford’s law applies to online social networks. Plos One, 10(8), e0135169.
https://doi.org/10.1371/journal.pone.0135169
Granovetter, M. S. (1973). The strength of weak ties. The American Journal of Sociology, 78(6),
1360–1380. https://doi.org/10.1086/225469
Hidalgo, C. A., Blumm, N., Barabási, A. L., & Christakis, N. A. (2009). A dynamic network
approach for the study of human phenotypes. PLoS Computational Biology, 5(4), e1000353.
https://doi.org/10.1371/journal.pcbi.1000353
Hu, Y. (2005). Efficient, high-quality force-directed graph drawing. Mathematica Journal, 10
(1), 37–71.
Jaccard, P. (1901). Distribution de la flore alpine dans le bassin des Dranses et dans quelques
régions voisines. Bull Soc Vaudoise Sci Nat, 37, 241–272.
Kalgotra, P., & Sharda, R. (2018). BIARAM: A process for analyzing correlated brain regions
using association rule mining. Computer Methods and Programs in Biomedicine, 162,
99–108. https://doi.org/10.1016/j.cmpb.2018.05.001
Kalgotra, P., & Sharda, R. (2021). When will I get out of the hospital? Modeling length of stay
using comorbidity networks. Journal of Management Information Systems, 38(4),
1150–1184. https://doi.org/10.1080/07421222.2021.1990618
Kalgotra, P., Sharda, R., & Croff, J. M. (2017). Examining health disparities by gender:
A multimorbidity network analysis of electronic medical record. International Journal of
Medical Informatics, 108, 22–28. https://doi.org/10.1016/j.ijmedinf.2017.09.014
Kalgotra, P., Sharda, R., & Croff, J. M. (2020). Examining multimorbidity differences across
racial groups: A network analysis of electronic medical records. Scientific Reports, 10(1), 1–9.
https://doi.org/10.1038/s41598-020-70470-8
Kalgotra, P., Sharda, R., & Luse, A. (2020). Which similarity measure to use in network
analysis: Impact of sample size on phi correlation coefficient and Ochiai index.
International Journal of Information Management, 55, 102229. https://doi.org/10.1016/j.
ijinfomgt.2020.102229
Kalgotra, P., Sharda, R., & Parasa, S. (2023). Quantifying disease-interactions through co-
occurrence matrices to predict early onset colorectal cancer. Decision Support Systems,
113929. https://doi.org/10.1016/j.dss.2023.113929
Lü, L., & Zhou, T. (2011). Link prediction in complex networks: A survey. Physica A: Statistical
Mechanics and Its Applications, 390(6), 1150–1170. https://doi.org/10.1016/j.physa.2010.11.
027
Miao, Z., & Balasundaram, B. (2017). Approaches for finding cohesive subgroups in large‐
scale social networks via maximum k‐plex detection. Networks, 69(4), 388–407. https://doi.
org/10.1002/net.21745
Newman, M. E. (2001). Clustering and preferential attachment in growing networks. Physical
Review E, 64(2), 025102. https://doi.org/10.1103/PhysRevE.64.025102
 JOURNAL OF INFORMATION TECHNOLOGY CASE AND APPLICATION RESEARCH 315

Pirch, S., Müller, F., Iofinova, E., Pazmandi, J., Hütter, C. V., Chiettini, M., Menche, J. . . .
Menche, J. (2021). The VRNetzer platform enables interactive network analysis in virtual
reality. Nature Communications, 12(1), 1–14. https://doi.org/10.1038/s41467-021-22570-w
Provan, K. G., & Sebastian, J. G. (1998). Networks within networks: Service link overlap,
organizational cliques, and network effectiveness. Academy of Management Journal, 41(4),
453–463. https://doi.org/10.2307/257084
Reichardt, J., & Bornholdt, S. (2006). Statistical mechanics of community detection. Physical
Review E, 74(1), 016110. https://doi.org/10.1103/PhysRevE.74.016110
Salton, G., & McGill, M. J. (1983). Introduction to modern information retrieval. Mcgraw-Hill.
Selim, H., & Zhan, J. (2016). Towards shortest path identification on large networks. Journal of
Big Data, 3(1), 1–18. https://doi.org/10.1186/s40537-016-0042-7
Teng, C. Y., Lin, Y. R., & Adamic, L. A. (2012, June). Recipe recommendation using ingredient
networks. In Proceedings of the 4th Annual ACM Web Science Conference, Austin, TX, USA,
(pp. 298–307).
Watts, D. J., & Strogatz, S. H. (1998). Collective dynamics of ‘small-world’ networks. Nature,
393(6684), 440–442. https://doi.org/10.1038/30918

Pankush Kalgotra
Harbert College of Business, Auburn University, AL, USA,
[email protected] http://orcid.org/0000-0003-2684-0342
Ramesh Sharda
Oklahoma State University, Stillwater, OK, USA