Integration of Data Science in Institutional Management Decision

Support System

Scăunașu Monica-Teodora and Mocanu Mariana Ionela

National University for Science and Technology POLITECHINICA Bucharest,

Computer Science Department, Bucharest, Romania

Keywords: Data Science, Decision Support Systems, Institutional Management, Predictive Analytics, Group

Decision- Making.

Abstract: This article explores the integration of data science into Decision Support Systems (DSS) as a transformative

framework for institutional management. Using advanced analytics such as Random Forest classifiers,

ARIMA models, and optimization algorithms, the research demonstrates how organizations can transition

from static decision-making frameworks to adaptive, data-driven systems. Case studies, including IT risk

management and group decision-making frameworks, illustrate the practical application and benefits of these

methodologies. The study compares the proposed DSS with traditional systems, underscoring the

advancements in predictive analytics, resource optimization, and collaborative decision-making. By aligning

predictive insights with institutional priorities, the proposed framework fosters operational efficiency,

strategic foresight, and inclusivity, setting a new standard for modern management practices.

1 INTRODUCTION

The digital age has transformed organizational

management, raising challenges to traditional

decision-making processes. Data science is now part

of institutional management, changing conventional

frameworks by leveraging vast data to reshape

decision-making paradigms (Davenport & Patil,

2012; Brynjolfsson & McAfee, 2017). Regardless of

sector, organizations can harness this data to gain

insights that enhance both strategic and tactical

decisions (Provost & Fawcett, 2013; Mayer-

Schönberger & Cukier, 2013).

Modern Decision Support Systems (DSS),

facilitative tools into critical assets for navigating

global markets and internal complexities. By

integrating machine learning and advanced

algorithms, these systems analyze data in real time,

transforming it into actionable intelligence. This shift

represents a move towards evidence-based

management, where data-driven insights replace

intuition or experience (Kroeber, 1952; Dicționar de

filosofie, 1978). Additionally, data science fosters

agility, equipping organizations with predictive

capabilities to remain proactive rather than merely

reactive in their strategies (Hoecke, 2002; Craiovean,

2020).

This paper examines methodologies for

embedding data science into DSS, detailing

algorithms and models across sectors. Expanded case

studies and comparisons with systematic literature

reviews underscore the unique contributions of this

research (Keyton, 2005).

2 DECISION MAKING SYSTEMS

Institutional management ensures organizations

achieve their objectives through structured

coordination and effective decision-making.

Traditionally, decision-making relied on manual

processes and delayed insights, leading to

inefficiencies across various sectors (Popa, 2022;

Psihologia personalității, 2024). The integration of

data science has transformed this landscape, enabling

proactive strategies, risk mitigation, and operational

stability (Griffin, 2016).

Unlike routine decisions, management decisions

require selecting from multiple alternatives, directly

influencing both operations and organizational

structures. Their impact underscores the necessity of

820

Monica-Teodora, S. and Ionela, M. M.

Integration of Data Science in Institutional Management Decision Support System.

DOI: 10.5220/0013352400003929

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 27th International Conference on Enterpr ise Information Systems (ICEIS 2025) - Volume 1, pages 820-829

ISBN: 978-989-758-749-8; ISSN: 2184-4992

data-driven approaches in enhancing institutional

performance and employee experiences (Goncalves &

Campos, 2018). As organizations increasingly depend

on technology, decision-making processes have

become more intertwined with information systems

and advanced analytics. Without Decision Support

Systems (DSS), decision quality relies heavily on

managerial expertise and hierarchical structures,

emphasizing the need for continuous training and

improved data accessibility (Berman & Bell, 2017).

Data science enhances decision-making by

offering structured analytical frameworks. Statistical

models such as regression analysis and time-series

forecasting help organizations detect patterns and

optimize resource allocation. Machine learning

techniques, including decision trees and neural

networks, refine predictive capabilities, while

optimization methods like linear programming

improve efficiency and strategic planning.

These advancements have reshaped key industries.

In healthcare, predictive models analyze patient data

to identify high-risk individuals, enabling targeted

interventions and improving patient outcomes. In

retail, machine learning enhances inventory

management by forecasting demand, preventing stock

shortages, and optimizing pricing strategies through

real-time data analysis (Smith, 2018).

The integration of data science into DSS marks a

significant shift in institutional management. By

analyzing large datasets and incorporating internal

performance metrics with external market indicators,

DSS generates actionable insights that drive

efficiency. These systems predict outcomes, optimize

resources, and refine service delivery. In hospitals, for

instance, DSS can forecast patient inflows, allowing

better allocation of staff and resources.

A key strength of data-driven DSS is real-time

analysis, essential for dynamic environments such as

finance and supply chain management. Organizations

can quickly adapt to demand fluctuations and

operational disruptions, ensuring informed, timely

decision-making. By leveraging machine learning,

optimization algorithms, and real-time analytics, DSS

improves decision-making, enhances institutional

performance, and secures a competitive advantage

across industries.

3 ORGANIZATION

MANAGEMENT

Management involves planning, decision adaptation,

organizing, leading, and controlling to optimize

human, financial, material, and informational

resources in pursuit of organizational goals

(Goncalves & Campos, 2018). Goncalves (2018)

considers management both an art and a science,

while Griffin (2016) highlights four fundamental

functions: planning, decision-making, organizing,

leading,

and controlling. These functions are

influenced by external factors such as market

conditions and globalization, alongside internal

elements like education, societal behaviors, and

technology. The most significant internal influence

remains organizational culture (Popa, 2022;

Cîmpeanu & Pîrju, 2010).

Planning sets objectives and strategies, decision-

making selects optimal courses of action, organizing

coordinates resources, leading motivates

collaboration, and controlling ensures compliance

with standards and goals (Griffin, 2016; Goncalves &

Campos, 2018). Data science strengthens

management by enhancing quality processes through

predictive analytics, monitoring tools, and continuous

feedback, reinforcing adaptability and efficiency. It

refines clarity through benchmarking, ensures

consistent implementation via automation, and

fosters long-term improvement through data-driven

feedback mechanisms (Bughin et al., 2017).

Gonçalves and Campos (2022) propose a four-

stage model for managing organizational culture:

strategic analysis, planning and diagnosis, action, and

validation. Data science enhances each phase,

analyzing employee feedback and performance data

to detect trends, forecast cultural impacts, and

simulate interventions before implementation (Popa,

2022; Cîmpeanu & Pîrju, 2010; Hudrea, 2015).

During execution, real-time monitoring ensures

adoption and engagement, while natural language

processing extracts insights from employee

responses, allowing timely adjustments (Hudrea,

2015). Validation employs statistical analysis and

machine learning to measure success, linking cultural

shifts to productivity and satisfaction metrics while

continuously refining strategies.

By integrating data science at all levels, cultural

management becomes dynamic, precise, and

adaptable, ensuring sustainable and effective

transformations (Gonçalves & Campos, 2022).

To contextualize the theoretical framework and

enhance comprehension of the study, the

organizational chart (Figure 1) represents a model of

how a DSS operates within an institutional structure.

It depicts an organization managing IT-related

projects, highlighting decision-making flows and the

integration of DSS into institutional management to

Integration of Data Science in Institutional Management Decision Support System

821

Figure 1: Organizational Chart.

support collaboration across departments (Griffin,

2016).

The hierarchical structure begins with the CEO,

who oversees the organization's strategic direction.

Directly reporting to the CEO are the Chief

Technology Officer (CTO), Chief Financial Officer

(CFO), and Chief Operating Officer (COO), each

responsible for key operational domains (Goncalves

& Campos, 2018; Bughin et al., 2017). Below them,

functional departments manage core activities: the IT

Security and Risk Analysis Team, led by the IT

Security Manager, focuses on identifying and

mitigating IT risks (Burdus & Popa, 2018); the

Software Development and Application

Development Team, managed by the Software

Development Manager, oversees technical project

execution; and the Operations and Finance Teams

ensure logistical efficiency and resource allocation

(Griffin, 2016; Goncalves & Campos, 2018). Within

this framework, specialized roles such as developers,

risk analysts, and accountants collaborate to achieve

organizational objectives, streamlining decision-

making and resource distribution (Griffin, 2016;

Bughin et al., 2017).

DSS acts as a central intelligence tool, integrating

data from all departments to provide actionable

insights that enhance decision-making and

coordination. The IT Security Team relies on DSS to

evaluate risks, prioritize mitigation strategies, and

allocate resources effectively. Using machine

learning models like Random Forest, the system

classifies risks based on probability and impact,

ensuring critical issues receive immediate attention

(Burdus & Popa, 2018). The CFO and Finance Team

use DSS to monitor budgets and optimize

expenditures. ARIMA-based forecasting predicts

cost trends and operational risks, enabling proactive

financial planning that aligns investments with

institutional priorities (Bughin et al., 2017).

Operational teams utilize DSS to address

logistical challenges and enhance efficiency through

real-time data processing and predictive

recommendations. Meanwhile, DSS aids executive

leadership, including the CTO and CEO, in aligning

IT strategies with long-term organizational goals by

providing a holistic view of institutional performance

and areas requiring intervention (Bughin et al., 2017).

Beyond individual departments, DSS fosters

cross-functional collaboration by serving as a shared

analytical platform. Visual representations, including

scatter plots and bar charts, allow stakeholders to

grasp risk levels, project statuses, and resource

distributions at a glance. This transparency ensures

alignment across departments and strengthens data-

driven decision-making throughout the organization.

4 DATA SCIENCE IN

INSTITUTIONAL

MANAGEMENT

Figure 2: Distribution of risk levels.

The distribution of risk levels is presented as a bar

chart, categorizing risks into low, moderate, and high

levels based on their calculated scores. This

visualization reveals the overall risk landscape, with

the majority of risks falling into the moderate and

high categories. High risks, such as “API

Vulnerabilities” and “Denial-of-Service Attacks,”

represent critical issues that require immediate

attention (Burdus & Popa, 2018). Moderate risks,

such as “Network Latency” and “Access Control

Issues,” demand careful monitoring to prevent

escalation. Low risks, while less urgent, still require

attention to avoid potential long-term complications.

By summarizing risks in this manner, the chart

enables decision-makers to prioritize their efforts

efficiently (Burdus & Popa, 2018; Bughin et al.,

2017).

ICEIS 2025 - 27th International Conference on Enterprise Information Systems

822

Figure 3: IT-specific risks.

The scatter plot of IT-specific risks delves deeper

into individual risks, plotting their probability against

impact. Each point on the plot represents a specific

risk, with color coding indicating its severity. High-

risk items, such as “System Scalability” and

“Phishing Attacks,” occupy the upper-right quadrant,

signaling their high probability and impact. In

contrast, low-risk items, like “Logging Failures,” are

positioned in the lower-left quadrant. This

visualization serves as a powerful tool during

collaborative meetings, allowing teams to assess the

relative severity of risks and devise appropriate

mitigation strategies (Bughin et al., 2017).

Figure 4: ARIMA forecast for database downtime.

Finally, the ARIMA forecast for database

downtime provides a forward-looking perspective.

By analyzing historical data, the ARIMA model

predicts how the risk score for database downtime

will evolve over time. The forecasted trend shows a

steady increase, underscoring the need for preemptive

action (Bughin et al., 2017; ). This prediction offers

invaluable foresight, enabling organizations to

allocate resources effectively and prevent costly

disruptions (Berman & Bell, 2017; McKinsey Global

Institute, 2018).

The risk management framework presented in this

study is not merely a theoretical construct; it is

designed to be seamlessly integrated into real-world

operations. Its applications span strategic planning,

operational management, and collaborative decision-

making, making it a valuable tool for organizations

aiming to improve their risk management practices

(Bughin et al., 2017;).

At the strategic level, senior leaders can use the

framework's predictive tools to guide investment

decisions. For example, the ARIMA forecast for

database downtime could prompt leaders to prioritize

infrastructure upgrades, while the Random Forest

model might highlight the need for enhanced security

measures. These insights ensure that limited

resources are allocated where they are needed most

(McKinsey Global Institute, 2018).

On an operational level, IT teams benefit from the

framework's ability to monitor risks dynamically. By

addressing high-priority risks, such as unpatched

software vulnerabilities, before they escalate, teams

can prevent significant disruptions and improve

service reliability. The scatter plot and bar chart

provide immediate clarity on where attention should

be focused, enabling teams to respond proactively

(Burdus & Popa, 2018; McKinsey Global Institute,

2018).

In collaborative settings, the framework's visual

tools foster a shared understanding of risks among

stakeholders. By presenting complex data in

accessible formats, the framework facilitates

productive discussions and ensures that decisions are

based on a comprehensive assessment of the risk

landscape (McKinsey Global Institute, 2018).

The integration of Random Forest and ARIMA

into this risk management framework highlights the

transformative potential of data-driven

methodologies. By combining machine learning,

time-series analysis, and intuitive visualizations, the

framework moves beyond traditional approaches to

risk management, offering actionable insights and

predictive capabilities (Bughin et al., 2017;

McKinsey Global Institute, 2018).

As organizations continue to face evolving

challenges, the scalability of this framework ensures

its relevance. By incorporating real-time data

streams, scenario analysis, and continuous

monitoring, it can adapt to new risks and changing

conditions. This adaptability positions the framework

as a cornerstone for resilient and efficient institutional

management, making it a vital area for further

research and development (Bughin et al., 2017;

McKinsey Global Institute, 2018).

Integration of Data Science in Institutional Management Decision Support System

823

Figure 5: DSS Recommendation.

The integration of the Decision Support System

(DSS) into the risk management framework

significantly enhances its practicality and relevance

for real-world applications. The DSS utilizes

predictive insights derived from machine learning

and time-series forecasting to generate actionable

recommendations, thereby enabling organizations to

address risks in a proactive and structured manner.

The Decision Support System (DSS) layer

dynamically interprets outputs from the framework,

including ARIMA forecasts for risk trends and

criticality assessments from the Random Forest

model. For instance, the ARIMA model forecasted an

upward trend in the "Database Downtime" risk score,

surpassing a predefined threshold. In response, the

DSS recommended infrastructure upgrades and

enhanced load balancing to mitigate the risk before it

could lead to operational disruptions.

Similarly, the Random Forest model classified

risks as critical or non-critical using probability and

impact scores. It prioritized high-risk items requiring

immediate attention, and the DSS proposed resource

reallocation strategies to address these critical risks

efficiently. This approach is especially valuable in

resource-constrained environments, enabling

judicious allocation to areas of greatest need.

The DSS outputs are presented in an accessible

format, including visual recommendations saved as

images. This bridges the gap between technical

analysis and strategic decision-making, offering clear

and actionable insights. By simplifying complex

datasets into concise recommendations, the DSS

empowers IT teams and senior management to align

their efforts and prioritize risk responses effectively.

By integrating the DSS, the framework evolves

from an analytical tool into a comprehensive risk

management system. It not only identifies and

predicts risks but also provides actionable guidance

for mitigation, enhancing both operational efficiency

and organizational resilience. This makes the DSS an

indispensable asset for modern institutional

management.

5 INTEGRATION OF

PREDICTIVE ANALYTICS

AND MACHINE LEARNING IN

DECISION SUPPORT SYSTEMS

FOR INSTITUTIONAL

MANAGEMENT: A PROPOSED

SOLUTION

Institutional management is a dynamic and

collaborative process that requires effective decision-

making to address complex challenges such as

resource allocation, risk management, and operational

efficiency. The integration of predictive analytics and

machine learning into Decision Support Systems

(DSS) is a transformative approach that enhances

decision-making processes by combining data-

driven insights with group collaboration. This chapter

explores

how

these

tools

can

implemented

within institutional structures, using the theoretical

organizational chart and practical examples of group

decisions as a guiding framework.

In institutions where decisions impact multiple

levels of operation, group decision-making is crucial.

By incorporating diverse perspectives, group

decisions ensure that all relevant factors are

considered, leading to more robust and

comprehensive outcomes. The decision-making

process involves collaboration across various

departments, such as IT, finance, operations, and

development, to address issues such as database

downtime and resource allocation. For instance, key

actions include upgrading database servers,

improving load balancing, and allocating additional

resources to mitigate critical risks. Each department

contributes its expertise, helping to prioritize and

resolve the most pressing challenges in a timely and

efficient manner.

As Griffin (2016) emphasizes, effective

management involves planning, organizing, leading,

and controlling, all of which benefit from

collaborative inputs. The organizational chart

discussed earlier provides a structural foundation for

these processes, illustrating how different roles and

departments interact to support group decision-

making. For instance, the Chief Technology Officer

(CTO) oversees technical feasibility, while the Chief

Financial Officer (CFO) evaluates budgetary

implications. The Risk Analysis Team, led by the IT

Security Manager, collaborates with the Application

Development Team and Operations Team to address

high-priority risks. This interconnected structure

ICEIS 2025 - 27th International Conference on Enterprise Information Systems

824

enables decisions to be both inclusive and data-

driven.

Predictive analytics and machine learning form

the backbone of the proposed DSS framework. These

tools provide actionable insights that inform group

decisions and enhance institutional resilience. Two

key algorithms are central to this framework: Random

Forest for risk classification and ARIMA for trend

forecasting (Burdus & Popa, 2018).

Random Forest is a machine learning algorithm

that classifies risks based on their probability and

impact. By analyzing historical data, it identifies

critical risks that require immediate attention. For

example, in the dataset analyzed, risks such as

"Database Downtime" and "API Vulnerabilities"

were flagged as critical due to their high probability

and impact scores (McKinsey Global Institute, 2018).

This classification provides a starting point for group

discussions, enabling stakeholders to prioritize their

efforts effectively.

ARIMA (AutoRegressive Integrated Moving

Average) is used to forecast trends and predict future

challenges. By analyzing historical patterns, ARIMA

models provide a forward-looking perspective,

enabling proactive planning. In the example of

"Database Downtime," the ARIMA model predicted

a steady increase in risk scores, prompting

discussions about resource allocation and mitigation

strategies. This foresight allows institutions to act

preemptively, avoiding disruptions and optimizing

operations.

The DSS framework not only generates

predictive insights but also integrates inputs from

various departments to support group decisions. For

example, when addressing a critical risk such as

database downtime, the DSS aggregates inputs from:

• IT Security: Focus on addressing high-risk

vulnerabilities (McKinsey Global Institute,

2018).

• Finance: Allocate budget for server upgrades

and load balancing.

• Operations: Ensure resource availability for

technical improvements.

• Development: Prioritize fixes for API

vulnerabilities before implementing new

features.

These inputs are synthesized alongside predictive

analytics outputs to produce comprehensive

recommendations. This collaborative process ensures

that decisions are informed by both data and

expertise, fostering alignment among stakeholders.

The outputs of the DSS are presented in accessible

formats that facilitate group discussions and

consensus-building. For instance, a scatter plot

mapping the probability and impact of various risks

allows stakeholders to assess the relative severity of

issues. High-risk items, such as "Phishing Attacks" or

"Denial-of-Service (DoS) Attacks," are visually

distinguished, enabling teams to focus on the most

pressing challenges. Similarly, time-series graphs

generated by ARIMA models provide insights into

anticipated trends, such as increasing workloads or

resource constraints.

The DSS output shown in the accompanying

image highlights recommendations based on group

inputs and predictive insights. The visualization

includes:

i. Forecasts for critical risks, such as database

downtime.

ii. Recommended actions, such as upgrading

servers or reallocating resources.

iii. Contributions from key departments, illustrating

the collaborative nature of the decision-making

process.

This combination of visual analytics and group

contributions fosters transparency and ensures that

decisions are both inclusive and actionable.

The proposed DSS framework is designed to be

seamlessly integrated into institutional operations,

supporting strategic planning, operational

management, and collaborative decision-making. At

the strategic level, the DSS enables senior leaders to

align decisions with organizational goals. For

example, if ARIMA predicts escalating risks for

database downtime, leaders can prioritize

investments in infrastructure upgrades. At the

operational level, the DSS helps teams monitor risks

dynamically, allowing them to address high-priority

issues proactively.

In group settings, the DSS serves as a mediator,

consolidating data and facilitating discussions. For

instance, during a meeting to address critical risks, the

DSS provides a comprehensive overview of the

organization's risk landscape, enabling stakeholders

to evaluate trade-offs and agree on the best course of

action. This collaborative approach aligns with the

principles outlined in current research on the

integration of advanced tools and diverse inputs in

decision-making processes.

The DSS output presented in this study

demonstrates the power of integrating predictive

analytics and group decision-making into

institutional management. By combining advanced

algorithms with collaborative tools, the DSS provides

organizations with a robust framework for addressing

complex challenges. The recommendations generated

Integration of Data Science in Institutional Management Decision Support System

825

by the system are not only data-driven but also

enriched by the expertise of various departments,

ensuring that decisions are comprehensive and

aligned with organizational priorities.

The visualization of DSS outputs enhances

communication and transparency, enabling

stakeholders to understand the rationale behind

recommendations and contribute meaningfully to

discussions. This fosters a culture of inclusivity and

accountability, where all participants feel empowered

to influence decisions.

The integration of predictive analytics and

machine learning into DSS represents a

transformative approach to institutional management.

By enabling group decision- making, the proposed

framework fosters collaboration, transparency, and

efficiency across organizations. The organizational

chart serves as a practical reference, illustrating how

DSS can facilitate collaboration among diverse roles

and departments. This approach not only addresses

current challenges but also lays the groundwork for

future innovation, demonstrating the potential of DSS

to drive success in complex organizational

environments.

6 METHODOLOGY

This study adopts a theoretical and applied research

approach to explore the integration of data science

into Decision Support Systems (DSS) for institutional

management. It is based on an extensive literature

review, conceptual analysis, and practical insights

derived from IT project management experience.

While no direct empirical data was used, the research

relies on academic sources, industry reports, and real-

world applications to support the proposed DSS

framework.

The methodology consists of three key

components. First, a literature review and

comparative analysis identify challenges in

institutional decision-making and evaluate traditional

DSS versus data-driven approaches. This comparison

highlights how predictive analytics, machine

learning, and optimization techniques enhance

decision processes.

Second, a DSS framework is developed,

integrating machine learning models such as Random

Forest for risk classification, ARIMA for trend

forecasting, and optimization techniques like Linear

Programming. A hypothetical institutional model is

introduced to demonstrate how DSS supports

decision-making across departments. To clarify its

functionality, a flowchart illustrates how DSS

processes data, assesses risks, and generates

actionable insights.

Finally, the framework is conceptually validated

through comparisons with existing methodologies.

The study assesses its scalability and adaptability,

discussing its potential application in various

institutional contexts. This structured approach

ensures that predictive analytics and real-time

optimization align with organizational needs,

fostering more effective and adaptive management

strategies.

7 COMPARATIVE ANALYSIS:

PROPOSED DATA SCIENCE

INTEGRATION VS. EXISTING

DSS METHODOLOGIES

Institutional management requires robust frameworks

to handle organizational complexities. Decision

Support Systems (DSS) facilitate data-driven

decision-making, aligning actions with institutional

goals. However, traditional DSS struggle with

adaptability, managing large-scale data, and

supporting collaborative decision-making. Integrating

data science techniques overcomes these limitations

through predictive analytics, optimization, and

improved decision-making.

The process begins with Risk Assessment, where

potential risks are identified. This step is connected to

Data Analysis & Forecasting, which analyzes

historical data to predict future risks and outcomes. By

identifying patterns and trends, the system pinpoints

High-Risk Areas that require immediate attention.

Once high-risk areas are identified, Group

Decision Making is engaged, bringing together

various departments, such as IT, finance, and

operations, to discuss, assess, and prioritize these

risks. This collaborative approach ensures that all

relevant factors are considered before moving

forward.

Following this, the system generates DSS

Recommendations that offer actionable insights.

These recommendations focus on necessary actions

like Database Upgrades to improve performance and

prevent failures, and adjustments in Resource

Allocation to ensure the organization can respond

effectively to identified risks.

The next step involves determining Resource

Allocation Needs. Here, the decision is made

regarding the optimal distribution of resources to

address the risks identified and mitigate potential

disruptions.

ICEIS 2025 - 27th International Conference on Enterprise Information Systems

826

Finally, the process concludes with a focus on

Security Improvements. This includes implementing

measures to strengthen security and ensure that the

organization’s critical infrastructure is protected

against potential threats.

This study compares traditional DSS with data

science-driven DSS, evaluating technical structure,

operational impact, and scalability. Traditional DSS

rely on static rule-based architectures and database

management systems (DBMS). Rule-based DSS

follow predefined rules, ensuring consistency but

lacking flexibility. Adjusting them requires extensive

reprogramming, limiting responsiveness. DBMS-

based DSS effectively manage structured data but

lack predictive analytics and struggle with

unstructured data. Gonçalves and Campos (2018)

highlight that DBMS-centric DSS are poorly suited

for machine learning, reducing their ability to

generate insights. Similarly, expert systems based on

if-then rules require extensive knowledge

engineering, making them unsuitable for dynamic

environments.

The proposed DSS integrates machine learning

algorithms, such as Random Forest for risk

classification and ARIMA for trend forecasting.

Random Forest enhances decision-making by

aggregating multiple decision trees, improving

accuracy and adaptability. This model analyzes

complex datasets, identifies hidden correlations, and

prioritizes high-risk events like database downtime or

API vulnerabilities. Unlike static rule-based DSS, it

dynamically adjusts to evolving risks, enhancing

resource allocation (Burdus & Popa, 2018).

ARIMA enables proactive decision-making by

forecasting trends. Unlike traditional DSS, which rely

solely on historical data, ARIMA predicts risks based

on trend analysis, allowing timely interventions such

as server upgrades or load balancing strategies.

To optimize resource allocation, the DSS

integrates Linear Programming and Genetic

Algorithms, automating personnel scheduling, budget

distribution, and workload management. Unlike

traditional DSS, which rely on heuristic approaches,

these models minimize inefficiencies and maximize

efficiency through dynamic optimization.

A key innovation of the proposed DSS is its

emphasis on group decision-making. By aggregating

inputs from IT, finance, operations, and development,

it ensures decisions are comprehensive, transparent,

and strategically aligned. Unlike traditional DSS,

which operate in silos, this framework synthesizes

predictive insights and stakeholder contributions into

actionable recommendations.

AI-powered DSS, such as IBM Watson Health,

have transformed medical decision-making but lack

collaborative decision-making for multidisciplinary

teams. The proposed DSS addresses this gap by

integrating stakeholder inputs for holistic resource

management and risk assessment. Similarly,

traditional manufacturing DSS, like Siemens

SIMATIC IT, rely on static rule-based models. By

contrast, the proposed DSS integrates predictive

analytics and optimization techniques, making it

more adaptable in dynamic environments.

Unlike traditional DSS that require manual

updates, the proposed framework continuously learns

from new data, improving scalability and

performance. Using text mining and natural language

processing (NLP), it processes structured and

unstructured data, making it adaptable to evolving

institutional needs. By shifting from reactive to

proactive decision-making, the DSS enhances

resource utilization, risk mitigation, and scalability.

This framework represents a major advancement

in DSS by integrating predictive analytics,

collaborative decision-making, and optimization. It

overcomes the limitations of rule-based models,

DBMS, and expert systems, offering a dynamic, AI-

powered approach. With machine learning and data-

driven methodologies, it remains adaptive and

efficient, fostering institutional transparency and

strategic alignment.

By leveraging data science, the proposed DSS

empowers institutions to navigate complexity with

confidence, ensuring resilience and long-term success

in an ever-changing environment.

8 CONCLUSIONS

This study highlights the transformative role of data

science in institutional management by

demonstrating how predictive analytics, machine

learning, and collaborative frameworks improve

decision-making. The proposed Decision Support

System (DSS) enhances resource allocation,

operational efficiency, and strategic planning,

enabling institutions to transition from static,

intuition-based decision-making to dynamic, data-

driven approaches.

The study demonstrates how Random Forest

algorithms classify risks and prioritize interventions,

ensuring that critical issues such as system

vulnerabilities and operational disruptions are

addressed proactively. Additionally, ARIMA

forecasting allows organizations to anticipate

challenges, offering insights that support

Integration of Data Science in Institutional Management Decision Support System

827

infrastructure scaling and risk mitigation strategies. A

key contribution of this research is its emphasis on

collaborative decision-making, integrating insights

from multiple departments to improve transparency,

coordination, and cross-functional alignment.

As institutions face increasingly complex

environments, real-time adaptability, crisis response,

and ethical governance become essential for decision-

making systems. One key area for future research is

the development of adaptive DSS capable of

processing real-time data and responding instantly to

dynamic institutional conditions. Investigating how

reinforcement learning and streaming analytics can

enhance DSS would enable institutions to react

swiftly to risks such as cybersecurity threats, financial

instability, or healthcare crises.

Further research should also focus on DSS for

institutional risk management and crisis response.

The integration of machine learning-based risk

classification, early-warning systems, and scenario

simulations could improve the ability of institutions

to predict and mitigate risks before they escalate. By

leveraging stochastic models and Bayesian inference,

DSS could offer more comprehensive risk

assessments, improving preparedness and crisis

response.

The enhancement of collaborative decision-

making is another crucial area. Future studies could

explore how Natural Language Processing (NLP) and

AI-driven discussion platforms facilitate more

effective communication among decision-makers.

Additionally, the integration of sentiment analysis

and gamification-based decision simulations could

foster more engaged and participatory decision-

making processes.

Ethical considerations and data governance must

also be prioritized. Research into Explainable AI

(XAI) could enhance DSS transparency, ensuring that

machine learning-driven decisions remain

interpretable, fair, and accountable. Institutions must

also develop data governance frameworks to ensure

compliance with privacy laws and ethical standards,

particularly in sectors where automated decision-

making impacts financial, legal, or public-sector

operations.

This study provides a foundation for data science-

driven DSS, but further research is needed to make

these systems more adaptive, resilient in risk

management, collaborative, and ethically

responsible. By advancing these areas, institutions

can develop next-generation DSS that ensure long-

term strategic alignment, agility, and accountability

in an increasingly complex data-driven world.

REFERENCES

Davenport, T. H., & Patil, D. J. (2012). Data scientist: The

sexiest job of the 21st century. In Harvard Business

Review. Harvard Business Publishing. Retrieved from

https://hbr.org/2012/10/data-scientist-the-sexiest-job-

of-the-21st-century

Provost, F., & Fawcett, T. (2013). Data science for

business: What you need to know about data mining

and data-analytic thinking. O'Reilly Media.

Brynjolfsson, E., & McAfee, A. (2017). Machine, platform,

crowd: Harnessing our digital future. W. W. Norton &

Company.

Mayer-Schönberger, V., & Cukier, K. (2013). Big data: A

revolution that will transform how we live, work, and

think. Houghton Mifflin Harcourt.

Kroeber, A. L. (1952). Culture: A critical review of

concepts and definitions. Vintage Books.

Dicționar de filosofie. (1978). Dicționar de filosofie. Ed.

Politică, București.

Hoecke, M. V. (2002). Law as communication. Oxford

University Press.

Craiovean, I. (2020). Tratat de teoria generală a dreptului.

Universul Juridic.

Keyton, J. (2005). Communication and organizational

culture: A key to understanding work experiences. Sage

Publications.

Popa, Ș. C. (2022). Cultura organizațională. Editura Pro

Universitaria, București.

Psihologia personalității. (2024, January 28). Psihologia

personalității. Retrieved from

https://ro.wikipedia.org/wiki/Psihologia_personalit%C

4%83%C8%9Bii

Isa, M. F., Ugheoke, S. O., & Noor, W. S. (n.d.). The

influence of organizational culture on employee’s

performance: Evidence from Oman. In Journal of

Entrepreneurship and Business, 4(2), 1-12. Retrieved

from https://doi.org/10.17687/JEB.0402.01

Hudrea, A. (2015). Cultura organizațională în România: O

analiză a cercetărilor în domeniu. In Revista

Transilvană de Științe Administrative, 2(37), 120-131.

Hofstede, G. (1980). Culture's consequences: International

differences in work-related values. Sage Publications.

Cîmpeanu, M.-A., & Pîrju, I. (2010). The specificity of the

organizational culture in European management. In

European Integration - Realities and Perspectives

(EIRP Proceedings), 290-295.

Griffin, R. (2016). Fundamentals of management (8th ed.).

Cengage Learning.

Goncalves, V., & Campos, C. (2018). The human change

management body of knowledge: Best practice and

advances in program management series. CRC Press.

Burdus, E., & Popa, I. (2018). Metodologii manageriale

(2nd ed.). Pro Universitaria.

Bughin, J., Seong, J., Manyika, J., Chui, M., & Joshi, R.

(2017). Artificial intelligence: The next digital frontier?

McKinsey Global Institute.

Olariu, A. A. (2023). Decizie, Proces Decizional și

Performanță. Pro Universitaria.

ICEIS 2025 - 27th International Conference on Enterprise Information Systems

828

Celi, L. A., Mark, R. G., Stone, D. J., & Montgomery, R.

A. (2016). Big data in the intensive care unit:

Improving diagnoses, prognoses, and outcomes. In

American Journal of Respiratory and Critical Care

Medicine, 194(2), 161-172.

McKinsey Global Institute. (2018). Unlocking the potential

of AI in public sector management. McKinsey &

Company. Retrieved from https://www.mckinsey.com

Integration of Data Science in Institutional Management Decision Support System

829