ArticlePDF Available

The Role of Complex Systems in Predictive Analytics for E-Commerce Innovations in Business Management

Authors:

Abstract and Figures

This review explores the incorporation of complex systems theory into predictive analytics in the e-commerce sector, particularly emphasizing recent advancements in business management. By analyzing the intersection of these two domains, the review emphasizes the potential of complex systems models—including agent-based modeling and network theory—to improve the precision and efficacy of predictive analytics. It will provide a comprehensive overview of the applications of emergent predictive analytics techniques and tools, including real-time data analysis and machine learning, in inventory optimization, dynamic pricing, and personalization of customer experiences. In addition, this review will suggest future research directions to advance the discipline and address the technical, ethical, and practical challenges encountered during this integration phase.
Content may be subject to copyright.
Citation: Madanchian, M. The Role of
Complex Systems in Predictive
Analytics for E-Commerce
Innovations in Business Management.
Systems 2024,12, 415. https://
doi.org/10.3390/systems12100415
Academic Editor: Gandolfo Dominici
Received: 26 August 2024
Revised: 23 September 2024
Accepted: 2 October 2024
Published: 5 October 2024
Copyright: © 2024 by the author.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
systems
Review
The Role of Complex Systems in Predictive Analytics for
E-Commerce Innovations in Business Management
Mitra Madanchian
Department of Arts, Communications and Social Sciences, University Canada West, Vancouver, BC V6Z 0E5,
Canada; mitra.madanchian@gmail.com
Abstract: This review explores the incorporation of complex systems theory into predictive analytics
in the e-commerce sector, particularly emphasizing recent advancements in business management.
By analyzing the intersection of these two domains, the review emphasizes the potential of complex
systems models—including agent-based modeling and network theory—to improve the precision
and efficacy of predictive analytics. It will provide a comprehensive overview of the applications of
emergent predictive analytics techniques and tools, including real-time data analysis and machine
learning, in inventory optimization, dynamic pricing, and personalization of customer experiences.
In addition, this review will suggest future research directions to advance the discipline and address
the technical, ethical, and practical challenges encountered during this integration phase.
Keywords: complex systems; predictive analytics; business management; customer engagement;
operational efficiency; strategic decision-making
1. Introduction
Globally, the electronic commerce (e-commerce) sector is expanding at a never-before-
seen pace [
1
]. The reason for the expansion can be attributed to the belief held by businesses
that e-commerce is an essential tool derived from internet technology that allows them to
compete on a worldwide scale [
2
,
3
]. E-commerce also helps businesses plan strategically,
provide customer service, cut costs, increase productivity and efficiency in the workplace,
propel their growth and development, and open up new markets [
4
,
5
]. Zion Market
Research estimates that the global predictive analytics market was valued at USD 7.1 billion
in 2019 and would increase at a compound annual growth rate (CAGR) of 21% between
2020 and 2026, reaching USD 26.3 billion. Businesses use predictive analytics models
to analyze transactional data and identify risks. These models capture the relationships
between many variables to assess risk or opportunity and guide transaction decisions [6].
Several elements have fueled this expansion, including the rising acceptance of cell
phones, the ease of online buying, and the possibility of obtaining a larger spectrum of
products [
7
,
8
]. Companies depend more on data as e-commerce grows to guide their oper-
ations. E-commerce businesses can maximize their websites, personalize the purchasing
experience, and change their marketing plans using data analytics [
9
]. Analyzing consumer
behavior, tastes, and buying habits will enable companies to acquire insightful information
that will enable them to remain competitive in the fast-changing e-commerce scene [10].
Despite the value of e-commerce to businesses, the volume and variety of data gener-
ated by its activities have been growing at an ever-increasing rate due to the development
of web technologies and their applications [
11
]. However, depending more on data also
brings difficulties. While following different data protection laws, e-commerce firms must
ensure they gather and utilize data ethically and securely [
12
]. The sheer amount of data
produced by e-commerce transactions can be daunting. Hence, companies must have
strong data management systems [9].
Systems 2024,12, 415. https://doi.org/10.3390/systems12100415 https://www.mdpi.com/journal/systems
Systems 2024,12, 415 2 of 20
To drive choices and actions to the appropriate stakeholders, business analytics refers
to the wide use of data from varied sources, fact-based management, predictive and ex-
planatory models, and quantitative and statistical analysis [
13
,
14
]. To achieve this, business
analytics uses techniques from information systems, data science, machine learning, and
operational research [
15
]. Predictive analytics uses statistics, data mining, machine learning,
AI, and business modeling to analyze historical and present data, predicting future events
in the context of management and IT. Businesses that use predictive analytics can benefit
from big data. It may benefit proactive, forward-thinking organizations that can foresee
data-driven behavior. It has grown rapidly with big data systems [16].
Complex systems have many interconnected components that cause behaviors that
cannot be comprehended by looking at the individual components. Complex systems
exhibit emergent traits, which originate from the interactions of their components rather
than their independent features. For instance, researching individual species only partially
explains ecosystem behavior; rather, complex interactions between species form ecosystem
dynamics. This perspective supports the premise that “the whole is more than the sum
of its parts” by emphasizing relationships in understanding challenging situations [
17
].
Complex systems are non-linear, so small changes in one area can have large effects
elsewhere. This non-linearity often causes inconsistent behavior, making linear approaches
difficult to model and evaluate for complex systems. Complex systems have feedback loops,
flexibility, and self-organization. Self-organization is a system’s ability to organize itself
into a structured pattern or behavior without external guidance. Feedback loops increase
or lessen system changes. These qualities in biology, society, and technology highlight
complex interactions in complex systems [18].
Comprising ideas from several disciplines like physics, biology, sociology, and com-
puter science, studying complex systems is naturally multidisciplinary. This multidisci-
plinary approach lets scientists investigate how ideas of complexity might be implemented
in many fields. For example, while biology clarifies the dynamics of ecosystems and bio-
logical processes, in social sciences, complex systems theory can assist in understanding
collective behavior and social networks. Integrating several disciplines promotes a bet-
ter knowledge of complicated systems and their behaviors, resulting in creative ideas to
address challenging issues [19,20].
Their complex character calls for advanced analytical techniques and approaches,
including network theory, agent-based modeling, and systems dynamics, to properly
investigate and control them [
21
]. Modern enterprises abound in complex systems, en-
compassing many different fields. In manufacturing, sophisticated systems comprise
modern production lines with multiple interacting components [
22
]. Within the energy
sector, smart grids and renewable energy sources show intricate system dynamics [
23
].
People and products are moved across complicated systems with emergent characteristics
like traffic congestion in the transportation networks [
24
]. Financial markets are intricate
systems that combine interactions between investors, institutions, and world events [
25
].
Complex information systems with complex interactions abound in telecommunications
networks, social media platforms, and the Internet [
26
]. With personnel, departments,
and non-linear interactions across strategies, even businesses and firms can be seen as
complicated systems [27].
Although current research shows the transforming power of predictive analytics in
improving operational efficiency and decision-making in e-commerce, they sometimes need
to pay more attention to integrating these analytics inside more general complex systems
frameworks, including environmental, technological, and organizational aspects [
28
]. Par-
ticularly in terms of knowledge of how complicated systems might enable the integration
of analytics into corporate processes and enhance strategic decision-making, the systematic
assessment of the e-commerce literature shows the demand for studies bridging these
gaps [
29
,
30
]. Dealing with these gaps will offer insightful analysis of how best to maximize
e-commerce technologies and guarantee compliance with ethical data use.
Systems 2024,12, 415 3 of 20
This review investigates how complex systems concepts and methodology can im-
prove predictive analytics in the e-commerce industry and result in novel business man-
agement strategies. Traditional predictive models frequently fail when dealing with the
increasingly complex datasets that e-commerce platforms handle, which reflect dynamic
client behaviors, supply chain variability, and changing market conditions. Complex sys-
tems theory, characterized by adaptability, emergence, and non-linearity, can be integrated
into predictive analytics to help firms acquire a deeper understanding and create more
resilient strategies. The initial goal of this review’s structure is to lay a basic grasp of
complex systems and how e-commerce fits into them. After that, it discusses the important
e-commerce predictors while underlining the drawbacks of conventional methods. Predic-
tive analytics and complex systems are explored, highlighting novel ideas and practical
applications. This review concludes by discussing the broader implications for corporate
management, including the opportunities and problems that sophisticated forecasting
tools bring.
2. Complex Systems
A conceptual diagram of the many elements of complex systems in e-commerce, such
as supply chain networks, market competitiveness, and user behavior dynamics, is shown
in Figure 1, highlighting their interdependence and non-linearity. It has been shown that in
order to fully understand supply networks’ intrinsic complexities, they need to be seen as
complex systems. The CAS perspective emphasizes the value of flexibility and adaptation
in management strategies by enabling a more nuanced understanding of how supply chains
function under various circumstances. The unpredictable character of supply networks is
frequently overlooked by traditional approaches that rely on static models, particularly in
stormy contexts where quick changes can occur [31,32].
Systems2024,12,4154of21
dicultiesthate-commerceplatformsencounterarehighlightedbytheintegrationof
theseelementsintocoherentsystems,asdemonstratedbyIBM’sWebSphereimplementa-
tion[37].Theideaofa“synergyeld”helpsexplainhowe-commercesystemsareim-
pactedbybothinternalandexternaldisruptions.Thisapproachemphasizestherequire-
mentforadaptabilityinresponsetodiverseshocksandaidsinunderstandingthetrans-
formationmechanismsinsidethesesystems[38].E-commercegrowthisdrivenbyboth
internalandexternalcauses,accordingtoempiricalresearchfromDianping.com.Accord-
ingtoYu-lin[39],theintricacyofthesenetworks,whichincludebusinessandhardware
networks,demonstratesthedicultyofe-commerceactivities.
Figure1.Overviewofcomplexsystemsine-commerce.
Complexsystemsaredistinguishedbytheirinterconnectedness,thatis,byeachcom-
ponentinuencingandbeinginuencedbyothers.Interdependenceresultsfromthiscon-
nectivity,sochangesinoneareaofthesystemmighthaveknock-onconsequencesallover
thenetwork.Inane-commercesupplychain,forexample,asupplier-leveldisturbance
mayinuenceinventorylevels,pricingpolicies,andcustomerhappiness.Likewise,inan
e-commerceplatform,consumerbehaviorisimpactedbyseveralelements,including
productrecommendations,peerevaluations,andpricing,allrelatedincomplicatedways
[40].
Anotherimportantfeatureisemergence,inwhichthecombinedbehavioroftheele-
mentsinthesystemproducesfresh,usuallysurprisingresultsthatneedtobeclaried
fromtheindividualcomponentanalysis.Forinstance,theunexpectedriseindemandfor
someproductsduringapandemicoraviralsocialmediatrendcanresultfromthecom-
binedactionsofconsumersthatcannotbereadilyforecastusingconventionallinearmod-
els[41].Thestudyofemergencechallengesreductionistapproachesthataempttoeluci-
datecomplexphenomenabydividingthemintosimplercomponents,insteademphasiz-
ingthesignicanceofcomprehendingthedynamicsandrelationshipswithinthesystem.
Thisconcepthassubstantialimplicationsforeldssuchasengineeringandsystemsthink-
ing,wheretheidenticationandmanagementofemergentpropertiescanresultinim-
proveddesignanddecision-makingprocesses[42,43].Becausemodestchangesininitial
Figure 1. Overview of complex systems in e-commerce.
Global economic conditions, consumer tastes, and technological improvements are
some of the elements that impact market competitiveness in e-commerce. These variables’
non-linear interactions may result in difficult-to-predict emergent phenomena. For example,
Systems 2024,12, 415 4 of 20
the emergence of digital platforms has changed the dynamics of conventional markets and
created new competitive environments where companies need to constantly innovate to stay
in business. According to research studies, viewing these competitive dynamics through
the perspective of complex systems may help organizations better traverse obstacles and
take advantage of market possibilities [32,33].
Large volumes of data are produced by user–platform interactions, which can be ex-
amined to learn more about user preferences and purchase trends. Because of its non-linear
character, which causes modest changes in user experience to have a substantial impact
on customer behavior, these behavioral data are intrinsically complicated. Businesses can
create more successful marketing strategies that foresee user requirements and improve cus-
tomer involvement by utilizing complex systems theory. By matching offerings to customer
expectations, this strategy not only increases user pleasure but also boosts sales [33,34].
Intricate decision-making processes are exposed by the interaction of pricing, service
standards, and emission reduction initiatives in e-commerce supply chains. For example,
rapid price strategy adjustments can cause uncontrollably large profit variations, high-
lighting the fine balance needed to manage supply chain dynamics [
35
]. The difficulties
in adjusting to erratic consumer behavior and quick changes in the market are illustrated
with a case study of a high-tech bicycle manufacturer making its way into the American
market. To successfully negotiate the intricacies of a dynamic and competitive market, this
scenario requires agile product development and strategic project management [36].
An additional degree of complexity is introduced by the architecture of e-commerce
systems, which consists of databases, web servers, and security services. The technical
difficulties that e-commerce platforms encounter are highlighted by the integration of these
elements into coherent systems, as demonstrated by IBM’s WebSphere implementation [
37
].
The idea of a “synergy field” helps explain how e-commerce systems are impacted by
both internal and external disruptions. This approach emphasizes the requirement for
adaptability in response to diverse shocks and aids in understanding the transformation
mechanisms inside these systems [
38
]. E-commerce growth is driven by both internal
and external causes, according to empirical research from Dianping.com. According to
Yu-lin [
39
], the intricacy of these networks, which include business and hardware networks,
demonstrates the difficulty of e-commerce activities.
Complex systems are distinguished by their interconnectedness, that is, by each
component influencing and being influenced by others. Interdependence results from this
connectivity, so changes in one area of the system might have knock-on consequences all
over the network. In an e-commerce supply chain, for example, a supplier-level disturbance
may influence inventory levels, pricing policies, and customer happiness. Likewise, in an e-
commerce platform, consumer behavior is impacted by several elements, including product
recommendations, peer evaluations, and pricing, all related in complicated ways [40].
Another important feature is emergence, in which the combined behavior of the
elements in the system produces fresh, usually surprising results that need to be clarified
from the individual component analysis. For instance, the unexpected rise in demand
for some products during a pandemic or a viral social media trend can result from the
combined actions of consumers that cannot be readily forecast using conventional linear
models [
41
]. The study of emergence challenges reductionist approaches that attempt
to elucidate complex phenomena by dividing them into simpler components, instead
emphasizing the significance of comprehending the dynamics and relationships within the
system. This concept has substantial implications for fields such as engineering and systems
thinking, where the identification and management of emergent properties can result in
improved design and decision-making processes [
42
,
43
]. Because modest changes in
initial conditions can produce disproportionately huge effects—a phenomenon sometimes
referred to as the “butterfly effect”—this non-linearity makes forecasting difficult.
Complex systems exhibit adaptability and evolution over time. This is evident in
an e-commerce environment in how businesses constantly modify their customer service,
inventory control, and marketing plans in response to shifting consumer tastes and com-
Systems 2024,12, 415 5 of 20
petition demands. Direct implementation of this idea in predictive analytics is adaptive
algorithms, which learn and grow depending on fresh data, allowing e-commerce plat-
forms to offer tailored experiences that change with user behavior [
44
]. Businesses that
use data analytics, for example, are able to recognize new trends and modify their product
offerings accordingly. This flexibility builds resilience against market volatility in addition
to improving customer satisfaction [45,46].
Because of the many interactions among consumers, vendors, platforms, and outside
variables like market trends and economic situations, e-commerce settings are great models
of complicated systems. These interactions produce a dynamic environment where the
interaction of several elements produces results like profitability, customer happiness, and
sales. User behavior dynamics is one field where intricate systems are absolutely important.
On e-commerce sites, consumers do not act alone; recommendations, reviews, social media,
and even other consumer behavior shape their actions. For instance, a consumer’s choice
to buy a product might be shaped by a mix of targeted ads, past purchases, and peer
reviews—all elements of a system of effects [
47
]. In this sense, predictive analytics must
consider the non-linear interactions among these elements to accurately project customer
behavior [
27
,
48
]. Das and Jadhav [
49
] identified non-linear pricing in e-commerce and its
use in their study. This research examined post-digitization non-linear commodity pricing.
By building trust, non-linear pricing has become a choice for consumers in the digital
market. Researchers employed theoretical models and empirical data to identify a new
non-linear pricing approach and its effect on e-commerce market behavior. Online meal
orders commonly use e-wallets, although offline cash payment is preferable. Therefore, the
offer matters to consumers.
Supply chain networks also illustrate another point: commodities’ transit from sup-
pliers to consumers entails several linked nodes, including manufacturers, warehouses,
logistics providers, and stores. These networks are complex because of their interdepen-
dencies; a delay or disturbance at any point can spread across the system, influencing
inventory levels, delivery schedules, and customer satisfaction. Maintaining efficiency and
satisfying customer expectations depend on predictive models that foresee and adjust to
such disturbances [
50
]. The resilience against disruptions can be further enhanced through
collaboration with logistics partners who specialize in commodities transport. These part-
nerships facilitate the sharing of resources and expertise, which are essential for navigating
the complexities of modern supply chains [
51
,
52
]. As businesses continue to encounter
challenges in logistics and customer demands evolve, it will be essential to maintain an
agile and responsive supply chain in order to maintain a competitive advantage in the
marketplace [53,54].
Complex systems also help to shape market trends and e-commerce rivalry. A new
product, a pricing modification, or a change in consumer tastes can all set off feedback
loops that magnify some trends while negating others. For instance, introducing a novel
product may cause demand to rise, which would inspire rivals to release similar products,
generating a competitive feedback loop that stimulates more invention and market evo-
lution [
55
]. In complex systems, the quantity and quality of data greatly determine how
effective predictive analytics is. High-dimensional data produced by complex systems
mean that there are numerous variables to examine with their possible interactions and
dependencies. To forecast future behavior, an e-commerce platform might gather, for ex-
ample, user demographics, browsing behavior, purchase history, and social media activity.
However, the sheer volume and range of data provide difficulties for data processing,
storage, and analysis [56].
Data sparsity and noise define complicated systems quite a bit. Data sparsity in an
e-commerce environment refers to the difficulty of creating reliable predictive models re-
sulting from insufficient information for certain goods or consumer segments. Conversely,
noise is meaningless or random data that might mask important trends. For instance, sea-
sonal variations in sales data could add noise that affects demand forecasting attempts [
57
].
Many approaches to resolving these problems have been investigated in recent studies.
Systems 2024,12, 415 6 of 20
One method, for instance, uses machine learning techniques that can efficiently handle
sparse datasets and are robust against noise. Techniques like Gaussian process regression
can improve the accuracy of prediction models by reducing the impact of noise by smooth-
ing out inconsistencies in the data [
58
]. Enhancing robustness against sparsity and noise
has been demonstrated to be possible with hybrid models that integrate various machine
learning approaches [
59
]. Reducing overfitting and enhancing generalizability are two
benefits of using techniques like regularization and dimensionality reduction to expedite
data input into predictive models [58,60].
3. Predictive Analytics in E-Commerce
Even while predictive analytics is becoming more and more common, several obstacles
remain to overcome before it can be used effectively. As seen in Table 1, businesses that
employ a traditional approach to predictive analytics frequently encounter obstacles in
several crucial areas.
Table 1. Challenges and solutions in the adoption of predictive analytics tools.
Category Challenge Users Tools Issues Solutions
Expertise Deep expertise
needed
Data
scientists
Predictive
analytics solutions
Inaccessible to most
application teams
Hire dedicated data
scientists for usage
Adoption
Difficult to adopt
End users Traditional
analytics tools
Disrupts workflows,
hard to scale
Integrate with primary
business applications
Empowering
End Users
Fails to enable
action End users Predictive
analytics tools
Time-wasting,
interrupts workflow
Empower users to act
within regular applications
Predictive analytics enables companies to find trends and patterns using statistical
algorithms and machine learning approaches, which guide their decisions. Across many
industries, including finance, healthcare, marketing, and manufacturing, this method is
increasingly used to improve operational efficiency, lower risk, and maximize resource
allocation. Predictive analytics is necessary for companies trying to remain competitive in
a data-driven environment since its main goal is to offer actionable insights that might lead
to better results [
61
]. Usually involving numerous important processes, predictive analytics
is data collection, cleaning, analysis, model construction, and validation. Organizations
first compile pertinent information from many sources, including operational measures,
consumer contacts, and market trends. To guarantee consistency and correctness, these data
are next cleansed and pre-processed. Analyzers then use several modeling approaches—
including neural networks, decision trees, and regression analysis—to find relationships
inside the data. The last models are evaluated for predictive power and dependability
using real-world results [62].
The capacity of predictive analytics to improve risk management and strategic plan-
ning is among its main benefits. Forecasting possible future events helps companies to
solve problems and grab possibilities aggressively. Predictive analytics can help companies
forecast consumer behavior, maximize inventory levels, and instantly identify fraudulent
activity. Predictive analytics capacity is predicted to grow even more as the area develops
with developments in artificial intelligence and machine learning, therefore allowing ever
more exact and insightful forecasts [63].
Modern e-commerce relies heavily on predictive analytics, which helps companies
anticipate client wants, streamline processes, and increase overall profitability [
64
,
65
].
Data-driven decision-making is mostly driven by three important predictive elements in
e-commerce: pricing strategies, demand forecasts, inventory management, and customer
behavior and purchase patterns. Table 2outlines a variety of predictive analytics tools.
Systems 2024,12, 415 7 of 20
Table 2. Tools and methods for predictive analytics in e-commerce.
Predictive Analytics
Tool/Technique Description Real-World Applications in E-Commerce
References
Machine Learning
Algorithms
Algorithms that learn from data to
make predictions or decisions
without being explicitly programmed.
Used for customer segmentation, recommendation
systems, and inventory management. Companies
like Amazon use ML for personalized
recommendations based on user behavior and
preferences.
[66]
Deep Learning
Models (LSTM)
A type of neural network particularly
suited for sequence
prediction problems.
Applied in predicting customer behavior over time,
such as forecasting future purchases based on past
buying patterns. Netflix uses LSTM for content
recommendation based on viewing history.
[66,67]
Deep Learning
Models (RNN)
Recurrent Neural Networks are
designed to recognize patterns in
sequences of data.
Utilized for sentiment analysis of customer reviews
and feedback, helping firms like Walmart
understand consumer sentiment towards products.
[66]
Statistical Methods
(ARIMA)
Autoregressive Integrated Moving
Average is a statistical analysis model
that uses time series data to predict
future points.
Employed in demand forecasting to optimize
inventory levels during peak shopping seasons, such
as Black Friday sales by retailers like Walmart.
[68]
Support Vector
Machines (SVM)
A supervised learning model that
analyzes data for classification and
regression analysis.
Used for classifying customer segments and
predicting churn rates, which is crucial for
companies like Amazon to retain customers.
[67]
Random Forests
An ensemble learning method that
operates by constructing multiple
decision trees during training time
and outputting the mode of
their predictions.
Applied in fraud detection systems to identify
fraudulent transactions in e-commerce platforms,
enhancing security measures for companies
like PayPal.
[66,68]
3.1. Customer Behavior and Purchase Patterns
Success in e-commerce requires an understanding of consumer behavior. Based on
prior data, predictive models estimate future customer behavior (e.g., likelihood of making
a purchase, likelihood of churning, and response to marketing initiatives). Collaborative
filtering is a common technique in recommendation systems that makes product recommen-
dations based on historical purchasing behavior. This personalizes the shopping experience
and boosts conversion rates [69].
Research indicates that by utilizing copious amounts of clickstream data, both deep
learning techniques—like LSTM—and machine learning techniques—like Random Forest
and Gradient Boosting—can forecast consumer behavior with accuracy rates ranging from
72% to 75% [
70
]. To find complicated linkages within datasets, innovative data mining
techniques must be employed by effective predictive models to account for the complexity
of consumer behavior [
71
]. Distinct client categories can be identified with machine learning
techniques such as clustering, which enables more focused marketing campaigns and better
product offerings [72,73].
3.2. Demand Forecasting and Inventory Management
In e-commerce, precise demand forecasting is essential to inventory management.
Predictive analytics models use seasonality, trends, and previous sales data to project
future product demand. For this, methods like time series analysis and machine learning
models like LSTM (Long Short-Term Memory) networks and ARIMA (AutoRegressive
Integrated Moving Average) are frequently employed [
74
,
75
]. E-commerce businesses can
reduce the likelihood of stockouts and overstocks by using effective demand forecasts.
These situations are critical for preserving customer happiness and controlling operating
expenses. As an illustration, Walmart optimizes its inventory turnover and significantly
lowers stockouts using predictive analytics [76].
Systems 2024,12, 415 8 of 20
3.3. Pricing Strategies and Dynamic Pricing
E-commerce platforms use predictive models to identify pricing policies that optimize
sales and market share. Random Forest Classifier analyzes user interactions to predict
purchase likelihood, helping businesses tailor marketing strategies effectively [
77
]. Utilizing
Recurrent Neural Networks (RNNs), these models capture temporal dependencies in
user behavior, significantly improving prediction accuracy for purchasing patterns [
78
].
Predictive analytics helps in anticipating product demand, thus optimizing inventory
management and reducing costs associated with overstocking [
79
]. Advanced models like
QLBiGRU enhance forecasting accuracy, enabling e-commerce platforms to make informed
pricing decisions and operational plans [
80
]. While predictive models offer substantial
advantages, they also face challenges, such as data limitations and the need for continuous
model refinement to adapt to changing market dynamics [81].
The critical factors contributing to the success of e-commerce are summarized in
Table 3
, which emphasizes key aspects such as website quality, customer support, person-
alization, electronic word of mouth (EWOM), technological advancements, and organiza-
tional strategies.
Table 3. Key factors influencing e-commerce success.
Aspect Description Key Features Impact on E-Commerce
Success
References
Website Service
Quality
Critical factor
influencing
e-commerce success.
Ease of use, website design, and
functionality.
Enhances user experience,
increases customer satisfaction,
and boosts conversion rates.
[82,83]
Customer Support
System
Vital role in
e-commerce success.
Responsive customer service,
clear communication channels,
and efficient problem resolution.
Builds trust and loyalty and
improves customer retention. [83]
Personalization Key predictor of
e-commerce success.
Personalized product
recommendations, customized
content, and targeted marketing.
Increases customer
engagement and drives sales. [82,83]
Electronic Word of
Mouth (EWOM)
Significant impact on
e-commerce success.
Customer reviews, ratings, and
social media discussions.
Enhances brand reputation,
influences purchase decisions,
and drives customer
acquisition.
[83]
Technological
Factors
Crucial for e-commerce
success.
AI for customer service and
personalization, secure payment
systems, and data analytics for
decision-making.
Improves customer service,
enhances security, and
provides customer insights.
[7,84]
Organizational
Factors
Significant role,
especially for SMEs.
Innovation culture, investment
in R&D, and effective supply
chain management.
Keeps companies competitive
and enhances operational
efficiency.
[7,8,84]
4. Intersection of Complex Systems and Predictive Analytics
4.1. Complex Systems as Predictive Analytics Models
Predictive analytics greatly benefits from modeling e-commerce as a complex system,
especially when comprehending customer behavior and streamlining operational tactics.
Complex systems are naturally suited to encapsulating the dynamic nature of e-commerce
settings since they comprise multiple interacting components that display emergent be-
haviors. By implementing complex systems theory, businesses can better understand how
several factors, including market trends, consumer preferences, and technical improve-
ments, interact to influence outcomes. Using a holistic perspective can result in enhanced
decision-making processes and more accurate predictions, ultimately boosting customer
happiness and profitability [85,86].
Online transactions generate enormous volumes of data, which offer a wealth of
information for studying customer behavior. Businesses can use sophisticated analytical
approaches to find patterns and trends that guide product recommendations, marketing
Systems 2024,12, 415 9 of 20
tactics, and inventory control. Predictive models, for example, can project future sales by
analyzing past purchasing patterns, which helps businesses better customize their goods to
match client expectations. This data-driven strategy promotes a more tailored shopping
experience for customers while increasing operational efficiency [82,87].
E-commerce can be modeled as a complex system, which makes it easier to explore
different situations and how they affect business performance. Businesses can use agent-
based modeling and simulations to test various approaches and evaluate the results in a
safe setting. This capacity is especially useful in a changing market, where companies must
adjust quickly to new possibilities and obstacles. E-commerce companies can minimize the
risks associated with uncertainty by making well-informed decisions that align with their
strategic objectives by knowing the possible outcomes of different actions [88,89].
Complex systems modeling’s interdisciplinary approach fosters stakeholder coop-
eration, including data scientists, marketers, and business strategists. This cooperative
approach may result in creative fixes and a deeper comprehension of the e-commerce
environment. Establishing an environment that encourages multidisciplinary cooperation
will be essential to maximizing the potential of complex systems in predictive analytics, as
businesses depend more and more on technology and data to make choices. Organizations
can improve their resilience and adaptation to changing market conditions by incorporating
knowledge from other sectors [82,86,87].
Complex systems have extensive interactions between their components, resulting
in emergent phenomena that cannot be foreseen. Machine learning is needed for good
forecasting since linear models cannot capture these non-linear dynamics. Dahia and
Szabo’s [
90
] study showed that machine learning can predict emergent behaviors from
huge datasets without knowing variable relationships. By combining post-mortem and live
analysis, this method improves our comprehension of how interactions result in emergent
phenomena. Traditional modeling efforts are complicated by complex systems’ non-linear
dynamics, wherein modest changes can have huge effects [
91
]. Plant systems highlight the
need for sophisticated computational tools because emergent features result from several
interactions at different scales [
92
]. A multiscale approach is necessary to comprehend
complex systems because emergent characteristics change how they appear at different
organizational levels [
17
]. Recent developments in predictive analytics underline the need
for models to reflect the dynamics of intricate systems. Table 4lists several key methods.
Table 4. Modeling approaches in complex systems and predictive analytics.
Model Type Description Key Feature Applications Example References
Agent-Based
Modeling (ABM)
Simulates actions and
interactions of
autonomous agents to
study emergent
phenomena.
Emergent Behavior
Understanding
flocking behavior
in birds
Revealed hidden
interactions
overlooked by
traditional methods
[93]
Graph Neural
Networks
(GNNs)
Models complex systems
focusing on relationships
between entities and
non-linear interactions.
Structure Learning Social networks,
biological systems
Identifies non-linear
interactions among
agents
[93]
Deep Learning
Frameworks
Utilizes techniques like
RNNs and CNNs to
capture temporal and
spatial patterns in data.
Temporal and
Spatial Pattern
Recognition
Analysis of vast
data for future
predictions
Identifies underlying
structures and
predicts future states
[93,94]
Hybrid Models
Combines machine
learning with traditional
statistical methods to
improve predictive
accuracy and incorporate
domain knowledge.
Integration of
Machine Learning
and Statistical
Methods
Healthcare, supply
chain management
Enhanced risk
assessment and
decision-making
[94,95]
Systems 2024,12, 415 10 of 20
Numerous studies have explored integrating complex systems into predictive analytics
(Table 5).
Table 5. Recent research and applications in complex systems-based predictive analytics.
Research Area Focus Method Application Complexity
Element Reference
Predictive Learning
Analytics
Educational
settings
Machine
learning
Capturing student
interactions
Complex
interactions [96]
Big Data and Predictive
Analytics Large datasets Non-linear
models
Uncovering
hidden patterns Big data analysis [95]
Prescriptive Analytics Business
decision-making Optimization Outcomes
optimization
Emergent
behaviors [88]
Onfirmed. Real-Time
Supply Chain Risk
Mitigation
Supply chain
management
Machine
learning Risk assessment Complex
interactions [94]
Unraveling Hidden
Interactions Complex systems Deep learning Revealing hidden
interactions
AgentNet
framework [93]
4.2. Innovations in Predictive Analytics through Complex Systems
4.2.1. Use of Agent-Based Modeling and Simulations
Figure 2shows a flowchart or framework highlighting the steps involved in data
gathering, model adaptation, real-time analysis, and decision-making in an e-commerce
scenario. It incorporates concepts from complex systems theory into predictive analytics.
Physical data from e-commerce transactions with digital modeling allow firms to construct
responsive and adaptive systems that can predict market shifts and consumer preferences.
Complex systems include complex interdependencies and emergent behaviors, making
classic analytical methods unsuitable [97,98].
Systems2024,12,41510of21
HybridMod-
els
Combinesmachinelearningwithtra-
ditionalstatisticalmethodstoim-
provepredictiveaccuracyandincor-
poratedomainknowledge.
IntegrationofMa-
chineLearning
andStatistical
Methods
Healthcare,sup-
plychainman-
agement
Enhancedriskas-
sessmentanddeci-
sion-making
[94,95]
Numerousstudieshaveexploredintegratingcomplexsystemsintopredictiveana-
lytics(Table5).
Tab l e5.Recentresearchandapplicationsincomplexsystems-basedpredictiveanalytics.
ResearchAreaFocusMethodApplicationComplexityEle-
ment
Refer-
ence
PredictiveLearningAnalyticsEducationalseingsMachinelearn-
ing
Capturingstudentin-
teractions
Complexinterac-
tions[96]
BigDataandPredictiveAna-
lyticsLargedatasetsNon-linear
models
Uncoveringhidden
paernsBigdataanalysis[95]
PrescriptiveAnalyticsBusinessdecision-
makingOptimizationOutcomesoptimiza-
tion
Emergentbehav-
iors[88]
Onrmed.Real-TimeSupply
ChainRiskMitigation
Supplychainman-
agement
Machinelearn-
ingRiskassessmentComplexinterac-
tions[94]
UnravelingHiddenInterac-
tionsComplexsystemsDeeplearningRevealinghiddenin-
teractions
AgentNetframe-
work[93]
4.2.InnovationsinPredictiveAnalyticsthroughComplexSystems
4.2.1.UseofAgent-BasedModelingandSimulations
Figure2showsaowchartorframeworkhighlightingthestepsinvolvedindata
gathering,modeladaptation,real-timeanalysis,anddecision-makinginane-commerce
scenario.Itincorporatesconceptsfromcomplexsystemstheoryintopredictiveanalytics.
Physicaldatafrome-commercetransactionswithdigitalmodelingallowrmstocon-
structresponsiveandadaptivesystemsthatcanpredictmarketshiftsandconsumerpref-
erences.Complexsystemsincludecomplexinterdependenciesandemergentbehaviors,
makingclassicanalyticalmethodsunsuitable[97,98].
Figure2.Predictiveanalyticsframeworkintegratingcomplexsystems.
Complexsystemandagent-basedmodeling(ABM)integrationisdrivingmoreand
moreinnovationsinpredictiveanalytics.Bysimulatingtheinteractionsofautonomous
agentsindiversesituations,ABMenablesresearcherstogaininsightsintohowindividual
behaviorscanresultinemergentphenomenaatthesystemlevel.Thismodelingapproach
hasgainedpopularityfromsociologytoeconomicsbecauseitcancapturecomplicated
systemdynamicswellandmakeiteasiertoexplorescenariosthatstandardmodeling
toolscouldmiss.Tosimulatehumansystems,forexample,whereagentinteractions
mightresultinunanticipatedconsequencesessentialforcomprehendingcomplicatedso-
cialphenomena,Bonabeau[99]emphasizestheusefulnessofABM.
UsingmachinelearningtechniquesimprovestheapplicabilityofABMinpredictive
analytics.RecentresearchhasshownthatmachinelearningcanenhanceABMcalibration,
Figure 2. Predictive analytics framework integrating complex systems.
Complex system and agent-based modeling (ABM) integration is driving more and
more innovations in predictive analytics. By simulating the interactions of autonomous
agents in diverse situations, ABM enables researchers to gain insights into how individual
behaviors can result in emergent phenomena at the system level. This modeling approach
has gained popularity from sociology to economics because it can capture complicated
system dynamics well and make it easier to explore scenarios that standard modeling tools
could miss. To simulate human systems, for example, where agent interactions might result
in unanticipated consequences essential for comprehending complicated social phenomena,
Bonabeau [99] emphasizes the usefulness of ABM.
Using machine learning techniques improves the applicability of ABM in predictive
analytics. Recent research has shown that machine learning can enhance ABM calibration,
increasing their resemblance to real-world data. Researchers can close the gap between
simulation and reality by updating the behavior rules of agents based on empirical evidence
by employing learning-based methodologies. This is especially useful in domains like
biomedicine, where agent-based models can forecast treatment results based on patient-
specific data and simulate the dynamics of cancer [
100
]. The combination of ABM and
Systems 2024,12, 415 11 of 20
machine learning improves model predictiveness and offers a framework for ongoing
development as new data become available.
Complicated problems in mobility and transportation networks are being addressed
with ABM. An extensive analysis of ABM’s applications in mobility transitions shows how
well it models the spread of innovations like electric cars. To predict market dynamics
and adoption rates, these models can consider several variables, such as infrastructure
development and consumer behavior [
101
]. Policymakers and other stakeholders can
assess the possible effects of different actions by simulating various scenarios, making
well-informed decisions about sustainable mobility projects easier.
A major development in the subject is the growth of multi-level agent-based modeling
(MLABM). By enabling the depiction of interactions across several levels of organization,
from individual agents to broader social systems, MLABM expands on the principles of
traditional ABM. This method is especially helpful for comprehending complex adaptive
systems, where interactions take place at several sizes and impact the system’s behavior as
a whole. Because MLABM can avoid some of the drawbacks of traditional ABM, as noted
by Morvan [
102
], it is becoming increasingly popular as a potential direction for future
complex systems and predictive analytics research. Understanding complex systems and
predictive analytics across a range of areas will continue to grow due to the continuous
development and improvement of these modeling tools.
4.2.2. Network Analysis and Its Applications in Customer Segmentation and Marketing
Large datasets are arranged into useful insights through big data analysis (BDA),
which enhances market segmentation precision and makes targeted marketing tactics possi-
ble [
103
]. Businesses can boost customer loyalty and retention by using predictive analytics
to generate targeted marketing efforts based on understanding customer preferences and
behaviors [
104
]. Businesses can increase profitability by allocating resources and marketing
expenditures more efficiently by identifying high-value client segments [105].
Examining relationships and interactions among consumers helps network analysis
greatly assist in consumer segmentation. This approach helps businesses to find groups of
like-minded consumers and grasp their interactions. Social network research, for example,
can highlight consumer information and recommendation sharing, which is important for
creating focused marketing campaigns. Understanding these networks helps companies
create plans using social influence, enhancing their marketing campaigns’ success [
106
,
107
].
More predictive analytics capacities have been improved by including big data ana-
lytics in marketing strategies. Organizations can use advanced analytics to find insights
hitherto impossible, as they compile enormous volumes of data from many sources. Facts-
driven marketing, in which judgments are grounded in empirical facts instead of intuition,
has evolved from this change. Real-time consumer behavior analysis enables marketers
to quickly modify their plans, optimizing campaigns for maximum performance and
increased conversion rates [85,108].
The continuous developments in artificial intelligence and machine learning are chang-
ing predictive analytics. Increased complex consumer behavior modeling made possible by
these technologies lets one forecast future activities more accurately. Learning from fresh
data, algorithms keep improving, giving companies a dynamic instrument for targeted
marketing and client segmentation. This improves operational effectiveness and helps de-
velop a better awareness of consumer demands, promoting customer loyalty and business
expansion [108].
4.2.3. Adaptive Algorithms and Real-Time Learning Systems
Recently, predictive analytics has revolutionized many fields, especially using real-time
learning systems and adaptable algorithms. These developments use sophisticated systems
to improve decision-making procedures by real-time data analysis, covering enormous
volumes. Adaptive algorithms let computers learn from fresh input constantly since they
can change their behavior depending on it. This capacity is very valuable in educational
Systems 2024,12, 415 12 of 20
environments, where predictive analytics can predict student performance and customize
learning experiences to meet individual requirements. Studies have demonstrated, for
example, that machine and deep learning models can accurately forecast academic results,
therefore enabling teachers to provide timely interventions for at-risk pupils [96,109].
Predictive analytics also depends critically on real-time learning systems. These sys-
tems constantly update predictive models using continuous data streams, guaranteeing
that the insights are based on the most recent information. This method is vital in sectors
including supply chain management, where the capacity to change with the times helps
to reduce risks and improve operational agility. Research has shown, for instance, that
adaptive decision-making algorithms can use real-time data to maximize supply chain
operations, enhancing resilience and responsiveness to changes in the market [
110
]. Inte-
gration of such systems increases productivity and encourages a proactive approach to
solving problems.
Predictive analytics applied via adaptive algorithms affects many sectors, including
banking and healthcare, outside supply chains, and education. Predictive models can
examine patient data in healthcare to project future health outcomes and maximize therapy
programs, enhancing patient care. In finance, these models can also evaluate risk and
guide investment plans using market trend analysis and customer behavior. Driven by
adaptive algorithms, predictive analytics’s adaptability highlights its ability to transform
conventional processes in many fields, enabling more informed decision-making and
strategic planning [111].
4.3. Case Studies and Practical Applications
E-commerce platforms are progressively employing intricate predictive analytics sys-
tems to optimize decision-making and enhance customer engagement. These systems
employ sophisticated data mining and machine learning methodologies to forecast market
trends and analyze user behavior, resulting in personalized experiences and optimized
operations. Table 6compares various studies on predictive analytics in e-commerce, empha-
sizing their focal areas and impacts, including operational efficiency, customer experience
enhancement, market trend prediction, and customer engagement. Although challenges
such as data privacy and algorithmic bias remain critical considerations in its applica-
tion [
112
], these examples illustrate the transformative potential of predictive analytics
in e-commerce.
Table 6. Comparison of predictive analytics focus and impact on e-commerce studies.
Criteria Rajeshkumar and Rajakumari [81] Zhu [113] Jakkula [114]
Topic
Customer Insights and Engagement
Agricultural
E-Commerce
Inventory Management
and Sales Forecasting
Analyzes Consumer Activities
Predicts Market Trends
Improves Processing Time
Enhances Customer Engagement
Improves Predictive Accuracy
Operational Efficiency
Customer Satisfaction
These platforms can forecast consumer behavior, optimize inventory management, and
personalize marketing efforts using machine learning algorithms and big data analytics.
Recommender systems, sentiment analysis, and personalization are the main areas of
attention for integrating artificial intelligence (AI) in e-commerce. These areas are crucial
for enhancing consumer experiences and increasing sales [108,115].
Using machine learning models to anticipate sales transactions is one prominent
way that predictive analytics is being used in e-commerce. A case study by Morsi [
116
]
illustrated how a predictive analytics model could help decision-making processes by
Systems 2024,12, 415 13 of 20
efficiently analyzing historical sales data. Businesses can optimize stock levels and cut
expenses associated with surplus inventory by using this model to predict swings in
demand. E-commerce businesses can improve their agility and reactivity to market changes
by implementing such predictive technologies, which will eventually boost consumer
satisfaction and retention.
Big data analytics has shown to be useful in e-commerce marketing for comprehend-
ing customer preferences and habits. Recent research shows that e-commerce platforms
might use big data to drive innovation and competitiveness. Businesses can ensure that
their offerings meet consumer expectations by identifying patterns and adjusting their
marketing strategy by analyzing large amounts of customer contact data [
117
]. In addition
to improving marketing efficacy, this data-driven strategy offers insights about prospective
new product and service developments.
E-commerce platforms are investigating AI applications for supply chain management
and its application in sales forecasting and marketing optimization. Supply chain agility can
be increased overall, and real-time risk mitigation can be facilitated by predictive analytics.
Businesses can provide a smoother operating flow by proactively addressing potential
disruptions using machine learning algorithms to examine various risk factors [94].
Impact on Customer Experience, Logistics, and Business Strategies
Research suggests that consumers engage with businesses through various touch-
points, and their experiences at each stage can significantly impact their overall satisfaction
and loyalty. For instance, effective customer experience design can increase engagement
and retention, driving sales and profitability [118,119].
The integration of information technology (IT) systems, including Enterprise Resource
Planning (ERP) and Customer Relationship Management (CRM), has been demonstrated
to improve customer service processes in logistics. More efficient logistics operations result
from these technologies, which improve communication and coordination among supply
chain partners. The customer experience is improved by implementing IT in logistics, which
streamlines processes and enhances responsiveness to customer requirements. Research
has shown that organizations that capitalize on their IT capabilities experience substantial
operational efficiency enhancements, ultimately leading to enhanced consumer satisfaction
and loyalty [8,119].
The strategic implementation of customer experience initiatives can alter business
strategies. Innovative service offerings and enhanced brand loyalty are frequently the
result of companies prioritizing customer experience and implementing a customer-centric
approach to their business models. For example, businesses that actively engage with
consumers and solicit feedback can adjust their strategies in real time to align with chang-
ing customer expectations. This adaptability is crucial in the current fast-paced market
environment, where consumer preferences can change swiftly. Companies can develop
more personalized experiences that resonate with their target audience by aligning business
strategies with customer insights [120,121].
The influence of consumer experience is not limited to immediate sales; it also affects
the long-term sustainability of a business. Organizations that effectively manage consumer
relationships and provide exceptional experiences can distinguish themselves from their
competitors. This differentiation is becoming increasingly significant in saturated mar-
kets, where consumers have many options. Companies can cultivate loyalty, encourage
repeat purchases, and achieve sustainable growth by prioritizing customer experience as a
fundamental element of their logistics and business strategies [118,120].
5. Discussion
5.1. Implications for Business Management
The interconnected and dynamic nature of e-commerce ecosystems necessitates a more
nuanced understanding, which complex systems theory provides. This theory emphasizes
non-linearity, interdependencies, and emergent behaviors, making it a better fit for ana-
Systems 2024,12, 415 14 of 20
lyzing and managing the complexities of contemporary markets. By leveraging predictive
analytics grounded in complex systems, businesses can enhance their decision-making
processes, developing more resilient and adaptable strategies for the unpredictable nature
of market trends, consumer behaviors, and competitive dynamics [62,122].
Enhancing predictive insights can lead to better decision-making, a noteworthy conse-
quence of this change. Businesses can more effectively prepare for eventualities and predict
a larger variety of possible outcomes by modeling the interplay between multiple factors
that come with complex systems. Conventional models, for instance, in inventory manage-
ment, might forecast stock levels based on previous sales data. Nevertheless, companies
may account for variables like abrupt customer demand shifts, supply chain interruptions,
or even outside economic shocks by integrating complex systems into predictive analytics.
This results in more precise projections and better-prepared operations. This improved
predictive ability is essential in the fast-evolving world of e-commerce, where companies
have to manage shifting customer tastes and international supply chains [123125].
Predictive analytics may segment customers and adapt products and services based
on consumer behavior and preferences to improve customer happiness and loyalty [
126
].
Companies can predict demand, optimize supply chains, and minimize inventory costs
by studying past sales data and market trends [
126
,
127
]. Marketing methods that include
machine learning boost consumer engagement and happiness, emphasizing the necessity
of individualized marketing [
128
,
129
]. Advanced systems in predictive analytics help firms
analyze customer behavior and improve operational efficiency, making it a crucial tool for
innovation and competitiveness [130].
However, there are obstacles to this move toward sophisticated systems-based man-
agement techniques. These sophisticated analytical models demand a lot of computer
power and technical know-how to implement them. Companies need to invest in data
science skills and the IT infrastructure required to manage the intricate algorithmic cal-
culations and massive data processing involved [
131
]. Corporate managers and analysts
need to possess a wider range of skills due to the interdisciplinary nature of complex
systems, which draws from disciplines like biology, physics, and economics [
132
]. The
need for specialized knowledge and the difficulty of effectively managing interrelated
subsystems present challenges for organizations attempting to integrate these advanced
analytical models into their operations, which can compromise the overall effectiveness of
management strategies in a quickly changing business environment [133,134].
The application of complex systems in predictive analytics raises issues related to
data protection and ethics. Businesses must ensure they follow data protection laws and
protect customer privacy as they collect and analyze massive volumes of data to model
complex systems. Robust governance mechanisms are necessary to supervise the ethical
deployment of these technologies since there is a large danger of data misuse or unfore-
seen consequences from complex system interactions. Incorporating intricate systems
into predictive analytics has significant advantages for e-commerce business management;
nonetheless, it necessitates meticulously evaluating the associated technological, organi-
zational, and moral dilemmas. A matrix or infographic depicting the effects of complex
systems-based predictive analytics on several facets of corporate management can be found
in Figure 3.
5.2. Challenges and Considerations
Predictive analytics for e-commerce presents formidable technological and strategic
obstacles when implementing complex systems. The computational complexity of mod-
eling and analyzing complex systems is one of the main obstacles. Complex algorithms
that can manage high-dimensional data and non-linear interactions are frequently needed
for these systems. For this reason, conventional data processing tools and approaches
are frequently insufficient, calling for the creation of more sophisticated methods like
agent-based modeling and deep learning. Even though these techniques are strong, they
Systems 2024,12, 415 15 of 20
need significant computational capacity and knowledge, which may be prohibitive for
many businesses.
Systems2024,12,41515of21
protectcustomerprivacyastheycollectandanalyzemassivevolumesofdatatomodel
complexsystems.Robustgovernancemechanismsarenecessarytosupervisetheethical
deploymentofthesetechnologiessincethereisalargedangerofdatamisuseorunfore-
seenconsequencesfromcomplexsysteminteractions.Incorporatingintricatesystemsinto
predictiveanalyticshassignicantadvantagesfore-commercebusinessmanagement;
nonetheless,itnecessitatesmeticulouslyevaluatingtheassociatedtechnological,organi-
zational,andmoraldilemmas.Amatrixorinfographicdepictingtheeectsofcomplex
systems-basedpredictiveanalyticsonseveralfacetsofcorporatemanagementcanbe
foundinFigure3.
Figure3.Businessmanagementimplicationsofcomplexsystems-basedpredictiveanalytics.
5.2.ChallengesandConsiderations
Predictiveanalyticsfore-commercepresentsformidabletechnologicalandstrategic
obstacleswhenimplementingcomplexsystems.Thecomputationalcomplexityofmod-
elingandanalyzingcomplexsystemsisoneofthemainobstacles.Complexalgorithms
thatcanmanagehigh-dimensionaldataandnon-linearinteractionsarefrequentlyneeded
forthesesystems.Forthisreason,conventionaldataprocessingtoolsandapproachesare
frequentlyinsucient,callingforthecreationofmoresophisticatedmethodslikeagent-
basedmodelinganddeeplearning.Eventhoughthesetechniquesarestrong,theyneed
signicantcomputationalcapacityandknowledge,whichmaybeprohibitiveformany
businesses.
Integratingcomplicatedsystemswithcurrentbusinessprocessesalsopresentsastra-
tegicchallenge.Theadvantagesofsophisticatedpredictivemodelsmustbeweighed
againste-commercecompanies’implementationchallenges,suchasintegratingthem
withtheirexistingITinfrastructureandensuringtheycanfunctionatscale.Thisisespe-
ciallychallengingine-commercesituationsbecausemarketconditions,consumerbehav-
ior,andtechnologicaladvancementscanquicklychange,makingpredictivemodelsout-
dated.Asaresult,companiesmustuseresource-intensivebutexibleandadaptivemeth-
odsfordevelopinganddeployingmodels.
Predictiveanalyticsinvolvescomplicatedsystemdeployment,andethicalconsider-
ationsarecritical.Sensitiveclientdataarefrequentlyincludedintheenormousamounts
neededtoeectivelymodelcomplexsystems,causingprivacyanddatasecurityconcerns.
Theimplementationprocessisfurthercomplicatedbytheneedtoensurecompliancewith
dataprotectionlaws,suchastheGeneralDataProtectionRegulation(GDPR)[135].
Figure 3. Business management implications of complex systems-based predictive analytics.
Integrating complicated systems with current business processes also presents a strate-
gic challenge. The advantages of sophisticated predictive models must be weighed against
e-commerce companies’ implementation challenges, such as integrating them with their
existing IT infrastructure and ensuring they can function at scale. This is especially chal-
lenging in e-commerce situations because market conditions, consumer behavior, and
technological advancements can quickly change, making predictive models outdated. As
a result, companies must use resource-intensive but flexible and adaptive methods for
developing and deploying models.
Predictive analytics involves complicated system deployment, and ethical considera-
tions are critical. Sensitive client data are frequently included in the enormous amounts
needed to effectively model complex systems, causing privacy and data security concerns.
The implementation process is further complicated by the need to ensure compliance with
data protection laws, such as the General Data Protection Regulation (GDPR) [
135
]. Deci-
sions made by opaque algorithms may be hard to understand or defend, eroding public
confidence in automated systems. As a result, companies need to carefully manage these
moral dilemmas to preserve client confidence and fully utilize sophisticated predictive
analytics technologies.
6. Conclusions
Combining sophisticated systems theory with predictive analytics for e-commerce
offers a more nuanced and all-encompassing method of grasping the dynamic and linked
character of digital markets. Businesses can greatly improve their decision-making pro-
cedures by including ideas of non-linearity, interdependencies, and emergent behaviors.
By modeling erratic market trends, customer behaviors, and supply chain interruptions
more precisely, the use of complex systems theory helps companies to create more robust
strategies and operational efficiencies.
Still, effectively using these sophisticated analytics tools calls for overcoming signif-
icant obstacles. Significant challenges arise from the computational needs of modeling
complicated systems, the requirement for great technical knowledge, and the connection
with current IT infrastructure. Maintaining consumer confidence depends on ethical issues
about data privacy and the openness of predictive algorithms given serious attention.
Systems 2024,12, 415 16 of 20
This review’s theoretical contribution is in closing the gap between the emergent
behaviors seen in complicated systems and conventional predictive analytics models. This
connection gives companies more actionable data and better-informed strategic decisions,
helping them to grasp their operations and market situations. Managerial consequences in-
clude the need for companies to invest in the required technology and expertise to properly
use complicated systems, therefore guaranteeing adaptation in a fast-changing environment.
Future studies can concentrate on improving predictive analytics tools using adap-
tive systems reacting dynamically to market changes and real-time learning algorithms.
Investigating how sophisticated systems theory might be used to develop e-commerce
sectors like cross-border trade and worldwide supply chain management is another topic
of interest. Businesses striving to stay ahead of the curve and keep competitiveness in a
digital environment growingly complicated will depend on these developments.
Funding: This research received no external funding.
Conflicts of Interest: The author declares no conflicts of interest.
References
1.
Roy, H. E-Commerce-Is-Growing-at-An-Unprecedented-Rate-Allover-the-Globe. The Next Scoop-An Internet Marketing Blog.
2019. Available online: https://thenextscoop.com/e-commerce-is-growing-atan-unprecedented-rate-all-over-the-globe (ac-
cessed on 1 July 2024).
2.
Agwu, E.M.; Murray, P.J. Empirical study of barriers to electronic commerce adoption by Small and Medium scale businesses in
Nigeria. Int. J. Innov. Digit. Econ. 2015,6, 1. [CrossRef]
3.
Tolstoy, D.; Nordman, E.R.; Vu, U. The indirect effect of online marketing capabilities on the international performance of
e-commerce SMEs. Int. Bus. Rev. 2022,31, 101946. [CrossRef]
4. Laudon, K.C.; Laudon, J.P. Management Information Systems: Managing the Digital Firm; Pearson Educación: London, UK, 2004.
5.
Taherdoost, H. E-Business Models and Strategies. In E-Business Essentials: Building a Successful Online Enterprise; Springer:
Berlin/Heidelberg, Germany, 2023; pp. 25–50.
6.
Henrys, K. Role of Predictive Analytics in Business. 2021. Available online: https://ssrn.com/abstract=3829621 (accessed on 1
July 2024).
7.
Hendricks, S.; Mwapwele, S.D. A systematic literature review on the factors influencing e-commerce adoption in developing
countries. Data Inf. Manag. 2024,8, 100045. [CrossRef]
8.
Zennaro, I.; Finco, S.; Calzavara, M.; Persona, A. Implementing E-commerce from logistic perspective: Literature review and
methodological framework. Sustainability 2022,14, 911. [CrossRef]
9.
Faccia, A.; Le Roux, C.L.; Pandey, V. Innovation and E-commerce models, the technology catalysts for sustainable development:
The Emirate of Dubai case study. Sustainability 2023,15, 3419. [CrossRef]
10. Rybaczewska, M.; Sparks, L. Ageing consumers and e-commerce activities. Ageing Soc. 2022,42, 1879–1898. [CrossRef]
11.
Beath, C.; Becerra-Fernandez, I.; Ross, J.; Short, J. Finding value in the information explosion. MIT Sloan Manag. Rev. 2012,53, 18.
12.
Yue, H. Research on e-commerce data standard system in the era of digital economy from the perspective of organizational
psychology. Front. Psychol. 2022,13, 900698. [CrossRef]
13.
Soltanpoor, R.; Sellis, T. Prescriptive analytics for big data. In Proceedings of the Databases Theory and Applications: 27th
Australasian Database Conference, ADC 2016, Sydney, NSW, Australia, 28–29 September 2016; Proceedings 27. Springer:
Berlin/Heidelberg, Germany, 2016; pp. 245–256.
14. Davenport, T.H. Competing on analytics. Harv. Bus. Rev. 2006,84, 98.
15.
Mortenson, M.J.; Doherty, N.F.; Robinson, S. Operational research from Taylorism to Terabytes: A research agenda for the analytics
age. Eur. J. Oper. Res. 2015,241, 583–595. [CrossRef]
16. Kumar, V.; Garg, M. Predictive analytics: A review of trends and techniques. Int. J. Comput. Appl. 2018,182, 31–37. [CrossRef]
17. San Miguel, M. Frontiers in complex systems. Front. Complex Syst. 2023,1, 1080801. [CrossRef]
18. Nicolis, G.; Rouvas-Nicolis, C. Complex systems. Scholarpedia 2007,2, 1473. [CrossRef]
19.
Freeborn, L.; Andringa, S.; Lunansky, G.; Rispens, J. Network analysis for modeling complex systems in SLA research. Stud.
Second. Lang. Acquis. 2023,45, 526–557. [CrossRef]
20.
McGill, E.; Marks, D.; Er, V.; Penney, T.; Petticrew, M.; Egan, M. Qualitative process evaluation from a complex systems perspective:
A systematic review and framework for public health evaluators. PLoS Med. 2020,17, e1003368. [CrossRef]
21. Page, S. Diversity and complexity. In Diversity and Complexity; Princeton University Press: Princeton, NJ, USA, 2010.
22.
Jemai, H.; Badri, A.; Fredj, N.B. Towards better understanding of the complex industrial systems: Case of production systems. Int.
Rev. Appl. Sci. Eng. 2023,14, 383–391. [CrossRef]
Systems 2024,12, 415 17 of 20
23.
Baugh Littlejohns, L.; Near, E.; McKee, G.; Rasali, D.; Naiman, D.; Faulkner, G. A scoping review of complex systems methods
used in population physical activity research: Do they align with attributes of a whole system approach? Health Res. Policy Syst.
2023,21, 18. [CrossRef]
24.
Siegenfeld, A.F.; Bar-Yam, Y. An introduction to complex systems science and its applications. Complexity 2020,2020, 6105872.
[CrossRef]
25.
Gao, J.; Xu, B. Complex systems, emergence, and multiscale analysis: A tutorial and brief survey. Appl. Sci. 2021,11, 5736.
[CrossRef]
26.
McCarthy, I.P.; Rakotobe-Joel, T.; Frizelle, G. Complex systems theory: Implications and promises for manufacturing organisations.
Int. J. Manuf. Technol. Manag. 2000,2, 559–579. [CrossRef]
27. Funtowicz, S.; Ravetz, J.R. Emergent complex systems. Futures 1994,26, 568–582. [CrossRef]
28.
Solano, M.C.; Cruz, J.C. Integrating Analytics in Enterprise Systems: A Systematic Literature Review of Impacts and Innovations.
Adm. Sci. 2024,14, 138. [CrossRef]
29. Rachmadewi, I.P.; Purnomo, H. The Role of e-Commerce: A Systematic Literature Review. iJIM 2022,16, 119.
30.
Bilgic, E.; Duan, Y. E-commerce and business analytics: A literature review. In Proceedings of the Digital Economy. Emerging
Technologies and Business Innovation: 4th International Conference, ICDEc 2019, Beirut, Lebanon, 15–18 April 2019; Proceedings
4. Springer International Publishing: Berlin/Heidelberg, Germany, 2019; pp. 173–182.
31.
Marchi, J.J.; Erdmann, R.H.; Rodriguez, C.M.T. Understanding supply networks from complex adaptive systems. BAR-Braz. Adm.
Rev. 2014,11, 441–454. [CrossRef]
32.
Abbasi, M.; Varga, L. Steering supply chains from a complex systems perspective. Eur. J. Manag. Stud. 2022,27, 5–38. [CrossRef]
33.
Luévano, A.N.; Barrientos, A.H.; Valverde, N.K. A Hybrid Modelling Framework for E-Commerce Supply Chain Simulation:
Complex Adaptive Systems Perspective. Informatica 2023,47. [CrossRef]
34.
Cheng, C.-Y.; Chen, T.-L.; Chen, Y.-Y. An analysis of the structural complexity of supply chain networks. Appl. Math. Model. 2014,
38, 2328–2344. [CrossRef]
35.
Xi, X.; Zhang, Y. Complexity analysis of pricing, service level, and emission reduction effort in an e-commerce supply chain under
different power structures. Int. J. Bifurc. Chaos 2022,32, 2250023. [CrossRef]
36.
Baumann, T.; Kaiser, M. Mastering successful parallel areas of complexity—A structured PM approach! Bicycle E-commerce in
USA—A fast growing high tech company with an agile product portfolio entering the most complex market! In Proceedings of
the 2017 12th International Scientific and Technical Conference on Computer Sciences and Information Technologies (CSIT), Lviv,
Ukraine, 5–8 September 2017; pp. 179–186.
37. Jhingran, A. Anatomy of a real e-commerce system. ACM SIGMOD Rec. 2000,29, 571–572. [CrossRef]
38.
Lai, D.H.; Hou, X.H.; Guo, W. Dynamic Complexity of E-Commerce Systems Base on Synergy Field. Adv. Mater. Res. 2012,403,
383–388. [CrossRef]
39.
Yu-lin, X.; Ting-Jie, L.; Jia-Yin, Q.; Xia, C. E-commerce Operation under Complex Network Based on Analytic Hierarchy
Process#br#—An Empirical Study of Dianping com. J. Beijing Univ. Posts Telecommun. (Soc. Sci. Ed.) 2014,16, 34.
40.
Simon, H.A. The Sciences of the Artificial, Reissue of the Third Edition with a New Introduction by John Laird; MIT Press: Cambridge,
MA, USA, 2019.
41. Holland, J.H. Studying complex adaptive systems. J. Syst. Sci. Complex. 2006,19, 1–8. [CrossRef]
42.
Aziz-Alaoui, M.; Bertelle, C. From System Complexity to Emergent Properties; Springer Science & Business Media: Berlin/Heidelberg,
Germany, 2009.
43.
Johnson, C.W. What are emergent properties and how do they affect the engineering of complex systems? Reliab. Eng. Syst. Saf.
2006,91, 1475–1481. [CrossRef]
44. Mitchell, M. Complexity: A Guided Tour; Oxford University Press: Oxford, UK, 2009.
45. Merelli, E.; Paoletti, N.; Tesei, L. Adaptability checking in complex systems. Sci. Comput. Program. 2016,115, 23–46. [CrossRef]
46. Liu, J.; Tong, T.W.; Sinfield, J.V. Toward a resilient complex adaptive system view of business models. Long Range Plan. 2021,54,
102030. [CrossRef]
47.
Bakshy, E.; Rosenn, I.; Marlow, C.; Adamic, L. The role of social networks in information diffusion. In Proceedings of the 21st
International Conference on World Wide Web, Lyon, France, 16–20 April 2012; pp. 519–528.
48.
Moeyersoms, J. The Value of Customer Behavior Data in Predictive Modeling; Universiteit Antwerpen (Belgium): Antwerp, Belgium,
2016.
49. Das, D.; Jadhav, V.S. Non-linear pricing in E-commerce: An exploration. J. Econ. Stud. 2022,49, 1453–1475. [CrossRef]
50.
Choi, T.Y.; Dooley, K.J.; Rungtusanatham, M. Supply networks and complex adaptive systems: Control versus emergence. J. Oper.
Manag. 2001,19, 351–366. [CrossRef]
51.
Hosoda, J.; Maher, S.J.; Shinano, Y.; Villumsen, J.C. Location, transshipment and routing: An adaptive transportation network
integrating long-haul and local vehicle routing. EURO J. Transp. Logist. 2022,11, 100091. [CrossRef]
52.
Karimi, B.; Bashiri, M. Designing a Multi-commodity multimodal splittable supply chain network by logistic hubs for intelligent
manufacturing. Procedia Manuf. 2018,17, 1058–1064. [CrossRef]
53.
Nagurney, A. Supply Chains and Transportation Networks. In Handbook of Regional Science, Fischer, M.M., Nijkamp, P., Eds.;
Springer: Berlin/Heidelberg, Germany, 2019; pp. 1–23.
Systems 2024,12, 415 18 of 20
54.
Wang, L. Efficient Green Supply Chain Management for Transit Manufacturing Enterprises Integrating AHP, K-NN, and MILP in
the Context of Sustainable Development. IEEE Access 2024,12, 36680–36701. [CrossRef]
55. Anderson, P. Perspective: Complexity theory and organization science. Organ. Sci. 1999,10, 216–232. [CrossRef]
56. Kelleher, J.D.; Tierney, B. Data Science; MIT Press: Cambridge, MA, USA, 2018.
57.
Mantegna, R.N.; Stanley, H.E. Introduction to Econophysics: Correlations and Complexity in Finance; Cambridge University Press:
Cambridge, UK, 1999.
58.
Poulinakis, K.; Drikakis, D.; Kokkinakis, I.W.; Spottswood, S.M. Machine-learning methods on noisy and sparse data. Mathematics
2023,11, 236. [CrossRef]
59.
França, T.; Braga, A.M.B.; Ayala, H.V.H. Feature engineering to cope with noisy data in sparse identification. Expert Syst. Appl.
2022,188, 115995. [CrossRef]
60.
Abdullah, F.; Wu, Z.; Christofides, P.D. Handling noisy data in sparse model identification using subsampling and co-teaching.
Comput. Chem. Eng. 2022,157, 107628. [CrossRef]
61.
Rustagi, M.; Goel, N. Predictive Analytics: A study of its Advantages and Applications. IARS’Int. Res. J. 2022,12, 60–63.
[CrossRef]
62. Lainjo, B. Enhancing program management with predictive analytics algorithms (paas). Algorithms Models Appl. 2021, 103.
63.
Hlosta, M.; Herodotou, C.; Papathoma, T.; Gillespie, A.; Bergamin, P. Predictive learning analytics in online education: A deeper
understanding through explaining algorithmic errors. Comput. Educ. Artif. Intell. 2022,3, 100108. [CrossRef]
64.
Lv, H. Intelligent e-commerce Framework for Consumer Behavior Analysis Using Big Data Analytics. Adv. Data Sci. Adapt. Anal.
2022,14, 2250007. [CrossRef]
65.
Bediako, G. The Application of Big Data Analytics in Improving eCommerce Processes. The Retail Sector User Experience. 2023.
Available online: https://www.theseus.fi/handle/10024/812302 (accessed on 1 July 2024).
66.
Sarker, I.H. Machine learning: Algorithms, real-world applications and research directions. SN Comput. Sci. 2021,2, 160.
[CrossRef]
67.
Rajula, H.S.R.; Verlato, G.; Manchia, M.; Antonucci, N.; Fanos, V. Comparison of conventional statistical methods with machine
learning in medicine: Diagnosis, drug development, and treatment. Medicina 2020,56, 455. [CrossRef]
68.
Rainio, O.; Teuho, J.; Klén, R. Evaluation metrics and statistical tests for machine learning. Sci. Rep. 2024,14, 6086. [CrossRef]
[PubMed]
69. Aggarwal, C.C. Recommender Systems; Springer: Berlin/Heidelberg, Germany, 2016; Volume 1.
70.
Liu, D.; Huang, H.; Zhang, H.; Luo, X.; Fan, Z. Enhancing customer behavior prediction in e-commerce: A comparative analysis
of machine learning and deep learning models. Appl. Comput. Eng. 2024,55, 190–204. [CrossRef]
71.
Al Ayub Ahmed, A.; Rajesh, S.; Lohana, S.; Ray, S.; Maroor, J.P.; Naved, M. Using Machine Learning and Data Mining to Evaluate
Modern Financial Management Techniques. In Proceedings of the Second International Conference in Mechanical and Energy
Technology: ICMET 2021, Greater Noida, India, 28–29 October 2022; pp. 249–257.
72.
Lee, I.; Shin, Y.J. Machine learning for enterprises: Applications, algorithm selection, and challenges. Bus. Horiz. 2020,63, 157–170.
[CrossRef]
73.
Hicham, N.; Karim, S. Analysis of unsupervised machine learning techniques for an efficient customer segmentation using
clustering ensemble and spectral clustering. Int. J. Adv. Comput. Sci. Appl. 2022,13, 122–130. [CrossRef]
74. Hyndman, R. Forecasting: Principles and Practice; OTexts: Melbourne, Australia, 2018.
75.
Siami-Namini, S.; Tavakoli, N.; Namin, A.S. A comparison of ARIMA and LSTM in forecasting time series. In Proceedings of the
2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA), Orlando, FL, USA, 17–20 December
2018; pp. 1394–1401.
76.
Choi, T.M.; Wallace, S.W.; Wang, Y. Big data analytics in operations management. Prod. Oper. Manag. 2018,27, 1868–1883.
[CrossRef]
77.
Chaudhuri, N.; Gupta, G.; Vamsi, V.; Bose, I. On the platform but will they buy? Predicting customers’ purchase behavior using
deep learning. Decis. Support Syst. 2021,149, 113622. [CrossRef]
78.
Liu, Z. Prediction Model of E-commerce Users’ Purchase Behavior Based on Deep Learning. Front. Bus. Econ. Manag. 2024,15,
147–149. [CrossRef]
79.
Manoharan, G.; Sharma, A.; Vani, V.D.; Raj, V.H.; Jain, R.; Nijhawan, G. Predictive Analytics for Inventory Management in E-
commerce Using Machine Learning Algorithms. In Proceedings of the 2024 International Conference on Advances in Computing,
Communication and Applied Informatics (ACCAI), Chennai, India, 9–10 May 2024; pp. 1–5.
80.
Zhao, G.; Tian, R.; Wang, J. Hyperparameter Adaptive Neural Network Model for E-Commerce Sales Prediction. In FSDM; IOS
Press: Amsterdam, The Netherlands, 2023; pp. 589–599.
81.
Rajeshkumar, C.; Rajakumari, K. Predictive Analytics in E-Commerce Leveraging Data Mining for Customer Insights. In
Proceedings of the 2023 7th International Conference on Electronics, Communication and Aerospace Technology (ICECA),
Coimbatore, India, 22–24 November 2023; pp. 783–787.
82.
Costa, P.; Rodrigues, H. The ever-changing business of e-commerce-net benefits while designing a new platform for small
companies. Rev. Manag. Sci. 2023,18, 2507–2545. [CrossRef]
83.
Sharma, H.; Aggarwal, A.G. Finding determinants of e-commerce success: A PLS-SEM approach. J. Adv. Manag. Res. 2019,16,
453–471. [CrossRef]
Systems 2024,12, 415 19 of 20
84.
Salah, O.H.; Ayyash, M.M. E-commerce adoption by SMEs and its effect on marketing performance: An extended of TOE
framework with ai integration, innovation culture, and customer tech-savviness. J. Open Innov. Technol. Mark. Complex. 2024,10,
100183. [CrossRef]
85.
Le, T.M.; Liaw, S.-Y. Effects of pros and cons of applying big data analytics to consumers’ responses in an e-commerce context.
Sustainability 2017,9, 798. [CrossRef]
86. Pardi, A.-L.; Burina, E. Democratising complex system modelling. Sci. Rep. 2024,14, 10590. [CrossRef] [PubMed]
87.
Sayama, H.; Pestov, I.; Schmidt, J.; Bush, B.J.; Wong, C.; Yamanoi, J.; Gross, T. Modeling complex systems with adaptive networks.
Comput. Math. Appl. 2013,65, 1645–1664. [CrossRef]
88.
Lepenioti, K.; Bousdekis, A.; Apostolou, D.; Mentzas, G. Prescriptive analytics: Literature review and research challenges. Int. J.
Inf. Manag. 2020,50, 57–70. [CrossRef]
89.
Kraus, S.; Jones, P.; Kailer, N.; Weinmann, A.; Chaparro-Banegas, N.; Roig-Tierno, N. Digital transformation: An overview of the
current state of the art of research. Sage Open 2021,11, 21582440211047576. [CrossRef]
90.
Dahia, S.S.; Szabo, C. Detecting Emergent Behavior in Complex Systems: A Machine Learning Approach. In Proceedings of the
Proceedings of the 38th ACM SIGSIM Conference on Principles of Advanced Discrete Simulation, Atlanta, GA, USA, 24–26 June
2024; pp. 81–87.
91.
Solis, I.M.; Van Rossem, M.; Madeleine, T.; Podoliak, N.; Orlova, T.; D’Alessandro, G.; Brodzki, J.; Kaczmarek, M. Learning Complex
Systems Dynamics from Vector Fields over Discrete Measure Spaces; Research Square: Durham, NC, USA, 2023.
92. Tomkins, M. Towards modelling emergence in plant systems. Quant. Plant Biol. 2023,4, e6. [CrossRef]
93. Ha, S.; Jeong, H. Unraveling hidden interactions in complex systems with deep learning. Sci. Rep. 2021,11, 12804. [CrossRef]
94.
Aljohani, A. Predictive analytics and machine learning for real-time supply chain risk mitigation and agility. Sustainability 2023,
15, 15088. [CrossRef]
95.
Jamarani, A.; Haddadi, S.; Sarvizadeh, R.; Haghi Kashani, M.; Akbari, M.; Moradi, S. Big data and predictive analytics: A
sytematic review of applications. Artif. Intell. Rev. 2024,57, 176. [CrossRef]
96.
Sghir, N.; Adadi, A.; Lahmer, M. Recent advances in Predictive Learning Analytics: A decade systematic review (2012–2022).
Educ. Inf. Technol. 2023,28, 8299–8333. [CrossRef]
97.
Huo, H.; Guo, W.; Yang, R.; Liu, X.; Xue, J.; Peng, Q.; Deng, Y.; Sun, X.; Lv, C. Data-Driven Strategies for Complex System
Forecasts: The Role of Textual Big Data and State-Space Transformers in Decision Support. Systems 2024,12, 171. [CrossRef]
98.
Chakrabarti, A.S.; Bakar, K.S.; Chakraborti, A. Data Science for Complex Systems; Cambridge University Press: Cambridge, UK,
2023.
99.
Bonabeau, E. Agent-based modeling: Methods and techniques for simulating human systems. Proc. Natl. Acad. Sci. USA 2002,99,
7280–7287. [CrossRef] [PubMed]
100.
Cogno, N.; Axenie, C.; Bauer, R.; Vavourakis, V. Agent-based modeling in cancer biomedicine: Applications and tools for
calibration and validation. Cancer Biol. Ther. 2024,25, 2344600. [CrossRef] [PubMed]
101.
Mehdizadeh, M.; Nordfjaern, T.; Klöckner, C.A. A systematic review of the agent-based modelling/simulation paradigm in
mobility transition. Technol. Forecast. Soc. Change 2022,184, 122011. [CrossRef]
102. Morvan, G. Multi-level agent-based modeling-a literature survey. arXiv 2012, arXiv:1205.0561.
103.
Özemre, M.; Kabadurmus, O. A big data analytics based methodology for strategic decision making. J. Enterp. Inf. Manag. 2020,
33, 1467–1490. [CrossRef]
104.
Amajuoyi, C.P.; Nwobodo, L.K.; Adegbola, A.E. Utilizing predictive analytics to boost customer loyalty and drive business
expansion. GSC Adv. Res. Rev. 2024,19, 191–202. [CrossRef]
105.
Abdul-Yekeen, A.M.; Kolawole, M.A.; Iyanda, B.; Abdul-Yekeen, H.A. Leveraging Predictive Analytics to Optimize SME
Marketing Strategies in the US. J. Knowl. Learn. Sci. Technol. 2024,3, 73–102. [CrossRef]
106.
Alves Gomes, M.; Meisen, T. A review on customer segmentation methods for personalized customer targeting in e-commerce
use cases. Inf. Syst. e-Bus. Manag. 2023,21, 527–570. [CrossRef]
107.
Rosário, A.T.; Dias, J.C. How has data-driven marketing evolved: Challenges and opportunities with emerging technologies. Int.
J. Inf. Manag. Data Insights 2023,3, 100203. [CrossRef]
108.
Haleem, A.; Javaid, M.; Qadri, M.A.; Singh, R.P.; Suman, R. Artificial intelligence (AI) applications for marketing: A literature-
based study. Int. J. Intell. Netw. 2022,3, 119–132. [CrossRef]
109.
Gligorea, I.; Cioca, M.; Oancea, R.; Gorski, A.-T.; Gorski, H.; Tudorache, P. Adaptive learning using artificial intelligence in
e-learning: A literature review. Educ. Sci. 2023,13, 1216. [CrossRef]
110.
Xu, Y.; Liu, X.; Cao, X.; Huang, C.; Liu, E.; Qian, S.; Liu, X.; Wu, Y.; Dong, F.; Qiu, C.-W. Artificial intelligence: A powerful
paradigm for scientific research. Innovation 2021,2, 100179. [CrossRef] [PubMed]
111.
Sarker, I.H. AI-based modeling: Techniques, applications and research issues towards automation, intelligent and smart systems.
SN Comput. Sci. 2022,3, 158. [CrossRef] [PubMed]
112.
Malleswari, T.N.; Ushasukhanya, S.; Karthikeyan, M.; Cherian, A.K.; Vaidhehi, M. Role of Predictive Analytics for Enhanced
Decision Making in Business Applications. In Harnessing AI and Digital Twin Technologies in Businesses; IGI Global: Hershey, PA,
USA, 2024; pp. 313–326.
113.
Zhu, H. User Behavior Analysis and Market Forecasting of Agricultural E-commerce Platforms. Appl. Math. Nonlinear Sci. 2024,9,
1–15. [CrossRef]
Systems 2024,12, 415 20 of 20
114.
Jakkula, A.R. Predictive Analytics in E-Commerce: Maximizing Business Outcomes. J. Mark. Supply Chain. Manag. 2023,2, 1–3.
[CrossRef]
115.
Bawack, R.E.; Wamba, S.F.; Carillo, K.D.A.; Akter, S. Artificial intelligence in E-Commerce: A bibliometric study and literature
review. Electron. Mark. 2022,32, 297–338. [CrossRef]
116.
Morsi, S. A Predictive analytics model for e-commerce sales transactions to support decision making: A case study. Int. J. Comput.
Inf. Technol. (2279-0764) 2020,9, 17–24. [CrossRef]
117.
Hicham, E.Y.; Abdelilah, Z.; Imane, N.; Abdelghaffar, I. The Use of Big Data Analytics in E-commerce Marketing: A Case Study of
Morocco. In The International Workshop on Big Data and Business Intelligence; Springer Nature Switzerland: Cham, Switzerland,
2024; pp. 430–440.
118.
Lemon, K.N.; Verhoef, P.C. Understanding customer experience throughout the customer journey. J. Mark. 2016,80, 69–96.
[CrossRef]
119.
Mesjasz-Lech, A. Effects of IT use in improving customer service logistic processes. Procedia Comput. Sci. 2015,65, 961–970.
[CrossRef]
120.
Ahmed, B.; Zada, S.; Zhang, L.; Sidiki, S.N.; Contreras-Barraza, N.; Vega-Muñoz, A.; Salazar-Sepúlveda, G. The impact of
customer experience and customer engagement on behavioral intentions: Does competitive choices matters? Front. Psychol. 2022,
13, 864841. [CrossRef]
121.
Johnson, E.; Karley, J. Impact of Service Quality on Customer Satisfaction. 2018. Available online: https://www.semanticscholar.
org/paper/Impact-of-service-quality-on-customer-satisfaction-Johnson-Karley/79a99957bcc846f8cbe344bb3361a418b7bd1146
(accessed on 25 August 2024).
122.
Sarker, I.H. Data science and analytics: An overview from data-driven smart computing, decision-making and applications
perspective. SN Comput. Sci. 2021,2, 377. [CrossRef]
123.
Lăzăroiu, G.; Neguri¸tă, O.; Grecu, I.; Grecu, G.; Mitran, P.C. Consumers’ decision-making process on social commerce platforms:
Online trust, perceived risk, and purchase intentions. Front. Psychol. 2020,11, 890. [CrossRef]
124.
Gadekar, R.; Sarkar, B.; Gadekar, A. Key performance indicator based dynamic decision-making framework for sustainable
Industry 4.0 implementation risks evaluation: Reference to the Indian manufacturing industries. Ann. Oper. Res. 2022,318,
189–249. [CrossRef]
125.
Enescu, F.M.; Bizon, N.; Onu, A.; Răboacă, M.S.; Thounthong, P.; Mazare, A.G.;
S
,
erban, G. Implementing blockchain technology
in irrigation systems that integrate photovoltaic energy generation systems. Sustainability 2020,12, 1540. [CrossRef]
126.
Aldoseri, A.; Al-Khalifa, K.N.; Hamouda, A.M. AI-Powered Innovation in Digital Transformation: Key Pillars and Industry
Impact. Sustainability 2024,16, 1790. [CrossRef]
127. Batko, K.; ´
Sl˛ezak, A. The use of Big Data Analytics in healthcare. J. Big Data 2022,9, 3. [CrossRef]
128.
De Mauro, A.; Sestino, A.; Bacconi, A. Machine learning and artificial intelligence use in marketing: A general taxonomy. Ital. J.
Mark. 2022,2022, 439–457. [CrossRef]
129.
Davenport, T.; Guha, A.; Grewal, D.; Bressgott, T. How artificial intelligence will change the future of marketing. J. Acad. Mark.
Sci. 2020,48, 24–42. [CrossRef]
130.
Dwivedi, Y.K.; Ismagilova, E.; Hughes, D.L.; Carlson, J.; Filieri, R.; Jacobson, J.; Jain, V.; Karjaluoto, H.; Kefi, H.; Krishen, A.S.
Setting the future of digital and social media marketing research: Perspectives and research propositions. Int. J. Inf. Manag. 2021,
59, 102168. [CrossRef]
131.
Zubair, A.; Erdo˘gdu, A. Modeling challenges to implement HR analytics in IT sector using ISM. J. Life Econ. 2024,11, 49–60.
[CrossRef]
132.
Aggrawal, N.; Pandey, A. Enhancing People Management Through Strategic Analytics in Indian Corporate Landscape. In Using
Strategy Analytics for Business Value Creation and Competitive Advantage; IGI Global: Hershey, PA, USA, 2024; pp. 116–155.
133.
Pérez, V.L. Levels of sophistication of information management systems and their upgrading. Inf. Manag. 1984,7, 29–38.
[CrossRef]
134.
Hallioui, A.; Herrou, B.; Santos, R.S.; Katina, P.F.; Egbue, O. Systems-based approach to contemporary business management: An
enabler of business sustainability in a context of industry 4.0, circular economy, competitiveness and diverse stakeholders. J.
Clean. Prod. 2022,373, 133819. [CrossRef]
135.
Zwitter, A.; Gstrein, O.J. Big data, privacy and COVID-19–learning from humanitarian