Explainable Assessment Model for Digital Transformation Maturity

Jihen Hlel

, Nesrine Ben Yahia

and Narj

es Bellamine Ben Saoud

RIADI Laboratory, National School of Computer Sciences, Manouba, Tunisia

Keywords:

Digital Transformation, Maturity, Assessment, Machine Learning, Explainability, Organizations.

Abstract:

Digital transformation has become a critical factor for organizational success in the modern business land-

scape. However, effectively and automatically assessing the maturity of this transformation remains a signiﬁ-

cant challenge. In this paper, we address the need for a uniﬁed and explainable digital maturity model to guide

organizations in their transformation journey. Our primary research questions focus on the development of a

core digital maturity model, the automatic validation of its effectiveness, and its explainability. To this end,

we propose a core model composed of seven key dimensions (Technology, Strategy, Skills, Culture, Organi-

zation, Data, and Leadership) derived from an extensive literature review. Each dimension is assessed across

ﬁve maturity levels (Basic, Discovery, Developed, Integrated, and Leadership). We then validate the proposed

model by leveraging machine learning techniques to assess its applicability within organizations. Finally, we

introduce an ensemble learning approach that combines unsupervised and supervised learning methods to en-

hance the explainability of the proposed digital maturity model. This approach aims not only to assess but also

to elucidate the impact of different dimensions on digital maturity.

1 INTRODUCTION

In today’s business world, digital transformation (DT)

has become a key focus in both information systems

research and business practice, with 84% of global

companies considering it as critical to their survival

in the next ﬁve years (Van Veldhoven and Vanthienen,

2022). Going digital is becoming a necessity as a

study in 2011 done by MIT Center for Digital Busi-

ness and Capgemini Consulting emphasized compa-

nies face common pressures from customers, com-

petitors, and employees to initiate their DT (McAffee

et al., 2011). According to the McKinsey research re-

port in 2018, the current success rate for DT in enter-

prises is only 30 % (McKinsey and company, 2018).

This is due to the fact that DT is a complex system

engineering, which is affected by the interaction of

many factors to jointly promote the success of such

transformation. Thus, while navigating the complexi-

ties associated with the DT, managers ﬁnd themselves

overwhelmed by the range of possible dimensions to

consider (Kiron, 2016).

Signiﬁcant progress has been made in understand-

ing DT, with growing research on its driving factors.

https://orcid.org/0000-0003-2753-7070

https://orcid.org/0000-0003-4788-4475

https://orcid.org/0000-0002-8071-0189

Still, a structured perspective is essential to effectively

guide DT efforts (Neff, 2014). As a result, organiza-

tions require tailored guidance to navigate their DT

journeys and assess their current level of digital ma-

turity (DM). Hence, the need for an effective man-

agement of the stages of DT requires that digital ma-

turity models (DMMs) to be put into practice (Thord-

sen et al., 2020). DM provides an accurate projection

of a company’s DT progress; it assesses the impact of

such transformation, making it essential to measure

the company’s current position. It evolves with the

ever-changing digital landscape (Akdil, 2017). There-

fore, ﬁrms must continuously assess their maturity to

adapt effectively in this dynamic environment. Such

assessments are crucial and depend on models that

provide reference frameworks, incorporating evalua-

tion criteria and indicators. In this context, the main

research questions explored in this paper are:

(1) How to build a core DMM for an efﬁcient and rel-

evant DM assessment?

(2) How to automatically validate the core model?

(3) How can the core model be interpreted and made

explainable?

The structure of the paper is as follows: ﬁrst, we con-

duct a comprehensive literature review to explore the

current state of knowledge, including key concepts

and DMMs in the ﬁeld. Second, we identify exist-

310

Hlel, J., Ben Yahia, N., Ben Saoud and N. B.

Explainable Assessment Model for Digital Transformation Maturity.

DOI: 10.5220/0013459300003967

In Proceedings of the 14th International Conference on Data Science, Technology and Applications (DATA 2025), pages 310-317

ISBN: 978-989-758-758-0; ISSN: 2184-285X

ing research gaps related to DT and DM. Next, we

present our core model for DM assessment. In the fol-

lowing section, we use machine learning (ML) tech-

niques to validate the proposed model. Finally, we

combine supervised and unsupervised learning meth-

ods to enhance its explainability.

2 BACKGROUND AND

RESEARCH GAP

DM refers to the scenario in which an organization

has successfully undergone a transformation (Akdil,

2017). It is not a static concept due to the digital

landscape change (Akdil, 2017). Thus, an organiza-

tion needs to assess it over time. In this context, ma-

turity models (MMs) are conceived as “frameworks

that evaluate the maturity of an organization through

the deﬁnition of a set of structured levels” (Battista

and Schiraldi, 2013). By examining the aspects of dif-

ferent MMs, common characteristics were extracted.

They consist of (a) maturity level, (b) descriptor for

each maturity level, (c) a generic description of each

level, (d) dimensions, (e) elements linked to corre-

sponding dimensions, and (f) a description of each el-

ement for each level of maturity (Fraser et al., 2002).

Although numerous MMs have been proposed,

several factors contribute to their limitations in effec-

tively assess an organization’s DT (Akdil, 2017). For

instance, many studies focus on speciﬁc regions and

sectors and rely on diverse but potentially insufﬁcient

sample sizes. Additionally, the complexity of validat-

ing these models often necessitates third-party assis-

tance, such as consultants, which increases costs and

extends the time required for evaluations. In fact, val-

idation of these models is frequently based on litera-

ture reviews or expert interviews, with minimal em-

pirical evidence provided to substantiate their effec-

tiveness. Moreover, a critical limitation of existing

models is their lack of explainability. While they may

provide an assessment of DM, they do not articulate

the reasoning behind their conclusions. This opac-

ity makes it difﬁcult for organizations to understand

why certain results are obtained or to derive action-

able insights from the evaluation. Without explain-

ability, organizations face challenges in building trust

in the assessment process and aligning the ﬁndings

with strategic decision-making. In this paper, we ad-

dress these issues to develop an effective and action-

able model that can better support organizations in

navigating their DT.

3 THE PROPOSED CORE

MATURITY MODEL

The proposed MM (Figure 1) serves as a core model

with the minimum and sufﬁcient dimensions for a rel-

evant and efﬁcient DM assessment. This study ﬁrst

undertakes a deep and up-to-date literature review, to

develop a core holistic DMM that uniﬁes the previ-

ous ones and covers several aspects of an organiza-

tion’s DT. The model is also generic built from multi-

ple models from different sectors.

Figure 1: The proposed core DT maturity model.

3.1 The Dimensions of Proposed

Maturity Model

As a basis for creating the core model, a review of the

current literature on MMs was conducted. It involved

a comparative analysis of several established mod-

els to assess the key dimensions crucial for achiev-

ing successful DT. The review led to the selection of

seven core dimensions including technology, strategy,

skills, leadership, culture, the organization, and data.

Table 1 summarizes some of the existing models.

Table 1: Existing digital maturity models.

Dimension Reference

Technology (Foundation, 2023),(Tubis, 2023),(Khourshed et al., 2023),

(Al-Ali and Marks, 2022), (van Tonder et al., 2024),(Kalender

and

Zilka, 2024)

Strategy (Foundation, 2023), (Tubis, 2023), (Khourshed et al.,

2023),(Al-Ali and Marks, 2022), (van Tonder et al.,

2024),(Kalender and

Zilka, 2024)

Skill (Spaltini et al., 2022),(Khourshed et al., 2023), (Al-Ali and

Marks, 2022), (van Tonder et al., 2024)

Culture (Tubis, 2023),(

Avila Boh

orquez and Gil Herrera, 2022), (van

Tonder et al., 2024),(Kalender and

Zilka, 2024)

Organization (Foundation, 2023), (Khourshed et al., 2023),(van Tonder

et al., 2024)

Data (Foundation, 2023),(Tubis, 2023),(

Avila Boh

orquez and

Gil Herrera, 2022),(Khourshed et al., 2023)

Leadership (

Avila Boh

orquez and Gil Herrera, 2022), (Khourshed

et al., 2023),(Al-Ali and Marks, 2022), (van Tonder et al.,

2024),(Kalender and

Zilka, 2024)

Explainable Assessment Model for Digital Transformation Maturity

311

3.1.1 Digital Technologies

The emergence of digital technologies mentioned

with the popular SMACIT have triggered DT as mate-

rial antecedents (Vial, 2019). The literature describes

digital technologies as inherent disrupters of the DT

wave (Vial, 2019). They shape it due to their speciﬁc

characteristics, referred to as digital properties. DT

starts with the adoption of digital technologies, then

evolving into an implicit holistic reshape of an orga-

nization, or deliberate pursuit of value creation (Vial,

2019).

3.1.2 Strategy

DT is depicted as a phenomenon that demands a

rapid organizational response. Although the con-

cept of strategy is often invoked to explain these re-

sponses, some researchers supported the traditional

view which refers to the IT strategy as a subordinated

functional-level strategy that must be aligned with the

ﬁrm’s business strategy. Others argue that strategic

responses require two novel concepts in line with the

DT: digital business strategy, which reﬂects a fusion

between IT and business strategy, and DT strategy

which is not part of any other strategy (Jimmy Bu-

mann, 2019).

3.1.3 Skills

Prior research shows that human factors can signiﬁ-

cantly impact DT capacity (Kwon, 2017). Employee

skills positively moderate the relationship between or-

ganizational capabilities and the success of DT. DT

requires employees to depend more heavily on their

analytical skills to solve increasingly complex busi-

ness problems (Dremel, 2017). The survey done by

MIT Center for Digital Business and Capgemini Con-

sulting in 2011, reveals that some IT departments

have established special units to track emerging tech-

nology skills and innovation centers to go with the

digital disruption impact.

3.1.4 Leadership

In line with the DT, organizational leaders must en-

sure that their organizations develop a digital mind-

set to be capable of responding to the disruptions as-

sociated with the use of digital capabilities (Haffke,

2017). To that end, the literature highlights the cre-

ation of new leadership roles (Horlacher, 2016) as the

chief digital ofﬁcer (CDO). The role of the CDO is

to implement digital business strategy into a series of

concrete actions.

3.1.5 Culture

Most of the ﬁrms that have initiated the DT often ex-

perience failures due to inert organizational cultures

that resist change (Hartl, 2017). Yet, a suitable or-

ganizational culture is a key requirement for the suc-

cessful transformation of businesses.

3.1.6 Organization

(Berghaus, 2017) considers partnerships and ecosys-

tems an important element of this dimension. Hence,

organizations must embrace a collaborative and

partnership-driven approach by actively emerging and

fusing organizational and IS strategy together to pur-

sue respectful relationships with various stakehold-

ers. While initially seen as competitors, partner-

ships should leverage each other’s strengths to meet

increasing customer needs. According to (Udovita,

2020), this dimension also encompasses the organi-

zation’s agility, which refers to its ability to respond

quickly to changes. Here, organizations should move

away from traditional hierarchies and embrace leaner

and ﬂatter organizational structures (Vial, 2019).

3.1.7 Data

Data has a decisive role in the DT journey. The

broad literature outlines that even strategic decision-

making will be based on data-driven insights (Haffke,

2017). Furthermore, ﬁrms are engaging in analytics

and combining with integrated data to gain a strategic

advantage over competitors. As a consequence, orga-

nizations are compelled to enhance their proﬁciency

in harnessing and leveraging data. Moreover, they can

maximize the advantages of technologies by gather-

ing data and using the derived insights to anticipate

customer behavior.

3.2 Maturity Levels of Our Model

For the assessment of DM related to each dimen-

sion, we deﬁne a ﬁve-level maturity scale: 1—ba-

sic, level 2—discovery, level 3—developed, level

4—integrated, and level 5—leadership inspired from

(Tubis, 2023). A detailed description of the assess-

ment levels is presented in Table 4 in the appendix.

4 DATA-DRIVEN VALIDATION

OF THE PROPOSED MM

In order to validate the proposed model, we rely on

data-driven approaches. Thus, empirical evidence

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

312

will be provided based on real data collected through

an assessment survey. The process from data collec-

tion to the model validation is presented in Figure 2.

Figure 2: Model validation approach.

4.1 Data Acquisition and Interpretation

Our MM is structured around seven dimensions each

is mapped to the criterion at each level according to

the degree of maturity. Individual areas and sub-areas

are assessed on ﬁve levels. The general characteris-

tics of individual levels are presented in Table 2. This

mapping is then used to produce a questionnaire ﬁlled

out by employees at the management level, which

is later carried out to assess organizational maturity

level and collect the necessary data. After collecting

216 survey responses, the data was compiled into a

structured tabular dataset. Our proposed scoring sys-

tem uses Python-deﬁned functions to compute DM

scores for each dimension—stored as new features in

the dataset—and calculates the overall maturity level

as the average of these dimension scores based on or-

ganizational responses. Our ﬁnal dataset features and

their deﬁnition are shown in Table 2.

4.2 Principal Component Analysis

In our study, we rely on principal component analy-

sis (PCA) to extract important insights from the col-

lected data, to do so it is crucial to, ﬁrst, determine

how many principal components (PC) to select. The

elbow test, is used to identify the best number of com-

ponents (Abdi and Williams, 2010). It suggests using

7 PCs, which explain nearly 80% of the total variance.

In Figure 4 we plot the loading of each feature on

every PC. PC1, dominated by features like Integration

of Digital Skills (0.701), Training Sessions (0.416),

and Mode of Analytics (0.270), represents organiza-

tional DM. PC2 captures a trade-off between skills in-

tegration (0.546) and factors like Culture of Digitiza-

tion (-0.491) and AI Integration (-0.391), highlighting

a tension between skill-building (0.546) and techno-

Table 2: Dataset description.

Features Deﬁnition Domains in the survey

Implementation of DT Plans Yes, No

Budget for DT Yes, No

Approach to Digitization

Strategy

Business development is driven by the digiti-

zation strategy, implementing and optimizing

its practices; Business development is driven

by the digitization strategy

Openness to New Ideas High, Low

Continuous Improvement

Strategy

Yes, No

Culture of Digitization Encourages advanced solutions, rewarding in-

novators; Promotes human-machine collabo-

ration with transparent risk management

ICT Devices Usage Yes, No

AI Integration M2M (machine-to-machine) deployed;

Cloud, IoT and AI; Cloud and IoT

Data Collection Strategies Yes, No

Mode of Analytics Descriptive; Descriptive, Predictive and Pre-

scriptive; Descriptive and Predictive

Real Time Analytics Yes, No

Digital Literacy Programs Yes, No

Training Sessions Planned, Random, Systematic

Integration of Digital Skills Enhances operational efﬁciency and collab-

oration; Continuously improves knowledge

sharing and skill development

Awareness of DT Yes, No

Presence of CDO Yes, No

Digital Mindset Strongly Deﬁned, Fully Embedded, Clearly

Deﬁned

Managing Key Partners Yes, No

Managing Partnerships Established partnerships contribute to opera-

tions; Fully integrated partnerships with real-

time data exchange; Personalized partnerships

with real-time data for decision-making

Overall Digital Maturity 1, 2, 3, 4, 5

Figure 3: PCA loadings.

logical or cultural factors (-0.491). PC3 is inﬂuenced

by Culture of Digitization (0.744), contrasting it with

AI-driven strategies (-0.384), while PC4 emphasizes

AI Integration (0.682). The later components, like

PC5 to PC7, represent more speciﬁc patterns, such

as partnerships, analytical capabilities, and strategic

alignment (Digital Mindset).

Explainable Assessment Model for Digital Transformation Maturity

313

4.3 Machine and Deep Learning

Algorithms for Model Validation

In our study, we explore machine and deep learn-

ing (DL) techniques for the validation of our model.

The idea is to evaluate the effectiveness of the model

in predicting actual outcomes. High prediction re-

sults suggest that the core model is valid and accu-

rate for DM assessment. Since our dataset is struc-

tured, recommended ML algorithms were explored

such as, K nearest neighbor (KNN), support vector

machine (SVM), decision tree, ensemble of decision

trees (random forest (RF) and extreme gradient boost-

ing (XGBoost)), and multiple layer perceptron (MLP)

(Yahia et al., 2021). In tabular data, relationships be-

tween features are often intricate and interdependent,

requiring the model to capture both local and global

patterns. Convolutional Neural Networks (CNNs)

(Mziou-Sallami et al., 2023) are suited for this task.

To effectively train them a large amount of data is re-

quired. Therefore, we used a data augmentation tech-

nique. This process expanded our dataset from 226 to

1000 records. Thus, we aim to validate the proposed

model not only on a small dataset but also on a larger

dataset to ensure its robustness and generalizability.

4.4 Results and Discussion

As mentioned above, we have applied KNN, SVM,

decision tree, RF, XGBoost, and MLP for small real

data, and CNN for augmented data. Table 3 shows the

obtained results.

Table 3: Performance evaluation results.

Algorithm Acc. Prec. Rec. F1

KNN 0.80 0.83 0.78 0.78

SVM 0.95 0.93 0.97 0.95

Decision Tree 0.66 0.62 0.65 0.62

RF 0.79 0.81 0.81 0.81

XGBoost 0.78 0.76 0.87 0.78

MLP 0.83 0.81 0.84 0.82

CNN 0.73 0.75 0.73 0.72

Combined, these metrics imply that the model ac-

curately captures underlying patterns in the data and

generalizes well. This underscores that our proposed

MM is accurate and can efﬁciently assess the DM.

In our study, we identiﬁed the minimal set of di-

mensions from the literature necessary for evaluat-

ing maturity and validated their sufﬁciency using ML

and DL algorithms. By systematically excluding fea-

tures representing speciﬁc dimensions, we observed a

signiﬁcant drop in model accuracy meaning that the

evaluation of the digital maturity is no more optimal,

demonstrating that this minimal set is both essential

and optimal for assessing digital maturity without re-

quiring additional dimensions thus the evaluation of

the digital maturity is no more optimal.

5 EXPLAINABILITY OF THE

PROPOSED MATURITY

MODEL

Existing studies provide insights into the coverage of

dimensions within MMs, but they fail to address how

to explain their results. To bridge this gap, we sug-

gest using a decision tree algorithm to identify rules

derived from the outputs of a K-Means clustering al-

gorithm. By doing so, our approach (illustrated in

Figure 5) provides clear explanations for the clusters,

maintaining the interpretability of results.

Figure 4: Cluster-based-classiﬁcation for maturity model

explainability.

Clustering offer a powerful tool for grouping or-

ganizations based on their maturity characteristics

(Wani, 2024). Ensuring that derived decisions can be

clearly understood is a fundamental requirement aim-

ing at making clustering results transparent and mean-

ingful. Thus, we enhanced our dataset with cluster la-

bels generated by k-means and subsequently applied

a decision tree classiﬁer. By combining k-means’

capability to detect patterns with the decision tree’s

explainability, we intend to make the assessment in-

sightful. This approach was applied on both a small

and augmented dataset, underscoring that our primary

focus is not the validation of the data itself but the

evaluation of the pipeline and the core MM.

5.1 A Cluster-Based-Classiﬁcation

Within Small Dataset

K-means require a predeﬁned number of clusters (k),

therefore, the elbow method is used. It involves calcu-

lating the WCSS for a range of k values and plotting

WCSS against k. The optimal k is identiﬁed at the

elbow point, where the rate of WCSS decrease levels

off (Wani, 2024). Here, We got an ”elbow” at k=3.

Then, having applied k-means, we assigned the

resulting cluster labels to the data points, creating a

newly labeled dataset. Below are the rules derived

from the hybrid of the k-means and the decision tree

algorithm. Each rule highlights the conditions under

which a particular class (cluster) is predicted.

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

314

Rules extracted from the decision tree within the

small dataset

1. Rule 1: If Training Sessions ≤ 2.50 and Mode of Analytics ≤ 1.50, then:

(a) If Training Sessions ≤ 1.50, then:

i. If Training Sessions ≤ 0.50: Class = 2.

ii. If Training Sessions > 0.50, then:

A. If Managing Partnerships ≤ 2.00: Class = 2.

B. If Managing Partnerships > 2.00: Class = 0.

(b) If Training Sessions > 1.50, then:

i. If AI Integration ≤ 1.50, then:

A. If Digital Mindset ≤ 1.50: Class = 2.

B. If Digital Mindset > 1.50: Class = 0.

ii. If AI Integration > 1.50, then:

A. If Awareness of DT ≤ 0.50, then:

B. If Approach to Digitization Strategy ≤ 1.50: Class = 2.

C. If Approach to Digitization Strategy > 1.50: Class = 0.

D. If Awareness of DT > 0.50: Class = 0.

2. Rule 2: If Training Sessions ≤ 2.50 and Mode of Analytics > 1.50, then:

(a) If Approach to Digitization Strategy ≤ 0.50, then:

i. If Training Sessions ≤ 1.50, then:

A. If AI Integration ≤ 3.50: Class = 2.

B. If AI Integration > 3.50: Class = 0.

ii. If Training Sessions > 1.50: Class = 0.

(b) If Approach to Digitization Strategy > 0.50, then:

i. If AI Integration ≤ 1.50, then:

A. If Culture of Digitization ≤ 0.50: Class = 2.

B. If Culture of Digitization > 0.50: Class = 0.

ii. If AI Integration > 1.50: Class = 0.

3. Rule 3: If Training Sessions > 2.50: Class = 1.

5.2 A Cluster-Based-Classiﬁcation

Within Augmented Dataset

Even with augmented data, the elbow test remains ro-

bust, with k=3 .

Rules extracted from the decision tree within the

augmented dataset

1. Rule 1: If Training Sessions ≤ 2.50 and Mode of Analytics ≤ 1.50, then:

(a) If Training Sessions ≤ 1.50, then:

i. If Digital Literacy Programs ≤ 0.50: Class = 1.

ii. If Digital Literacy Programs > 0.50, then:

A. If Managing Partnerships ≤ 1.50: Class = 1.

B. If Managing Partnerships > 1.50, then:

C. If Data Collection Strategies ≤ 0.50: Class = 1.

D. If Data Collection Strategies > 0.50: Class = 2.

(b) If Training Sessions > 1.50, then:

i. If AI Integration ≤ 2.50, then:

A. If Managing Partnerships ≤ 2.50: Class = 1.

B. If Managing Partnerships > 2.50: Class = 2.

ii. If AI Integration > 2.50, then:

A. If Approach to Digitization Strategy ≤ 1.50, then:

B. If Awareness of DT ≤ 0.50: Class = 1.

C. If Awareness of DT > 0.50: Class = 2.

D. If Approach to Digitization Strategy > 1.50: Class = 2.

2. Rule 2: If Training Sessions ≤ 2.50 and Mode of Analytics > 1.50, then:

(a) If Implementation of DT Plans ≤ 0.50, then:

i. If Training Sessions ≤ 1.50, then:

A. If AI Integration ≤ 3.50, then:

B. If Managing Partnerships ≤ 2.50: Class = 1.

C. If Managing Partnerships > 2.50: Class = 2.

D. If AI Integration > 3.50, then:

E. If Digital Literacy Programs ≤ 0.50: Class = 2.

F. If Digital

Literacy Programs > 0.50: Class = 1.

ii. If Training Sessions > 1.50: Class = 2.

(b) If Implementation of DT Plans > 0.50, then:

i. If ICT Devices Usage ≤ 0.50, then:

A. If Continuous Improvement Strategy ≤ 0.50: Class = 1.

B. If Continuous Improvement Strategy > 0.50: Class = 2.

ii. If ICT Devices Usage > 0.50, then:

A. If Managing Key Partners ≤ 0.50, then:

B. If Mode of Analytics ≤ 2.50: Class = 1.

C. If Mode of Analytics > 2.50: Class = 2.

D. If Managing Key Partners > 0.50: Class = 2.

3. Rule 3: If Training Sessions > 2.50: Class = 0.

5.3 Findings Interpretation

The interpretation of the two generated rule bases re-

veal key insights into the DM. On the one hand, the

rules exposes the importance of core features such as

”training sessions”, ”mode of analytics”, and ”AI in-

tegration” in determining classes/ clusters which val-

idates the results obtained previously from the PCA.

They, also, offer practical implications by guiding or-

ganizations to address gaps in areas like ”data col-

lection” and ”partnership management” and empha-

sizing actionable steps that can enhance DM. On the

other hand, ﬁndings emphasize the scalability of our

MM, proving its adaptability to varying contexts. The

consistency of interpretable rules across original and

augmented datasets validates the framework’s robust-

ness, illustrating its capability to generate meaningful

assessments regardless of the dataset.

6 CONCLUSIONS

This paper presents a uniﬁed and explainable digital

maturity model. By focusing on seven key dimen-

sions, the model evaluates digital maturity across ﬁve

levels. A data-driven validation approach based on

machine learning has been used to validate the model.

Then, an ensemble learning approach combining un-

supervised and supervised methods is proposed to en-

hance the model’s effectiveness and explainability. As

organizations evolve, the dynamic nature of digital

maturity must be considered. Future work should fo-

cus on expanding the model to track organizational

changes over time, providing a continuous feedback

to optimize digital transformation efforts.

Explainable Assessment Model for Digital Transformation Maturity

315

REFERENCES

Abdi, H. and Williams, L. J. (2010). Principal component

analysis. Wiley interdisciplinary reviews: computa-

tional statistics, pages 433–459.

Akdil, K.Y.;Ustundag, A. C. E. (2017). Maturity and readi-

ness model for industry 4.0 strategy. Springer Series

in Advanced Manufacturing.

Al-Ali, M. and Marks, A. (2022). A digital maturity model

for the education enterprise. Perspectives: Policy and

Practice in Higher Education.

Avila Boh

orquez, J. H. and Gil Herrera, R. J. (2022). Pro-

posal and validation of an industry 4.0 maturity model

for smes.

Battista, C. and Schiraldi, M. M. (2013). The logistic matu-

rity model: Application to a fashion company. Int. J.

Eng. Bus. Manag, 5.

Berghaus, S., B. A. . K. B. (2017). Digital maturity & trans-

formation report. St.Gallen.

Dremel, C., W. J. H. M. W. J.-C. B. W. (2017). How audi

ag established big data analytics in its digital transfor-

mation. MIS Quart.Execut.

Foundation, I. (2023). Industry 4.0 readiness on-

line self-check for businesses. https://www.

industrie40-readiness.de/?lang=en (accessed on 10

January 2025).

Fraser, P., Moultrie, J., and Gregory, M. (2002). The use of

maturity models/grids as a tool in assessing product

development capability. In IEEE international engi-

neering management conference, pages 244–249.

Haffke, I., K. B. . B. A. (2017). The transformative role of

bimodal it in an era of digital business. proceedings

of the 50th Hawaii international conference on system

sciences.

Hartl, E., . H. T. (2017). The role of cultural values for

digital transformation: Insights from a delphi study.

Americas Conference on Information Systems.

Horlacher, A., H. T. (2016). What does a chief digital ofﬁcer

do? managerial tasks and roles of a new c-level posi-

tion in the context of digital transformation. In: Sys-

tem Sciences 49th Hawaii International Conference.

Jimmy Bumann, M. K. P. (November 2019). Action ﬁelds

of digital transformation - a review and comparative

analysis of digital transformation maturity models and

frameworks. In book: Digitalisierung und andere In-

novationsformen im Management.

Kalender, Z. T. and

Zilka, M. (2024). A comparative analy-

sis of digital maturity models to determine future steps

in the way of digital transformation. Procedia Com-

puter Science, pages 903–912.

Khourshed, N. F., Elbarky, S. S., and Elgamal, S. (2023).

Investigating the readiness factors for industry 4.0

implementation for manufacturing industry in egypt.

Sustainability, page 9641.

Kiron, D.; Kane, G. P. D. P. A. B. N. (2016).

Does it pay to be a multinational? a large-

sample, cross-national replication assessing the multi-

nationality–performance relationship. MIT Sloan

Manag.Rev, 58.

Kwon, E.H.; Park, M. (2017). Critical factors on ﬁrm’s dig-

ital transformation capacity: Empirical evidence from

korea. Int. J. Appl. Eng.

McAffee, A., Ferraris, P., Bonnet, D., Calm

ejane, C., and

Westerman, G. (2011). Digital transformation: A

roadmap for billion-dollar organizations. MIT Sloan

Management Review.

McKinsey and company (2018). Unlocking success in dig-

ital transformations. McKinsey and company.

Mziou-Sallami, M., Khalsi, R., Smati, I., Mhiri, S., and

Ghorbel, F. (2023). Deepgcss: a robust and explain-

able contour classiﬁer providing generalized curvature

scale space features. Neural Computing and Applica-

tions, 35(24):17689–17700.

Neff, A.A., H. F. H. T. U. F. B. W. v. B. J. (2014). Devel-

oping a maturity model for service systems in heavy

equipment manufacturing enterprises. Inf. Manag,

page 895–911.

Spaltini, M., Acerbi, F., Pinzone, M., Gusmeroli, S., and

Taisch, M. (2022). Deﬁning the roadmap towards in-

dustry 4.0: the 6ps maturity model for manufacturing

smes. Procedia CIRP, pages 631–636.

Thordsen, T., Murawski, M., and Bick, M. (2020). How to

measure digitalization? a critical evaluation of dig-

ital maturity models. In Responsible Design, Im-

plementation and Use of Information and Communi-

cation Technology: 19th IFIP WG 6.11 Conference

on e-Business, e-Services, and e-Society, I3E 2020,

Skukuza, South Africa, April 6–8, 2020, Proceedings,

Part I 19, pages 358–369.

Tubis, A. A. (2023). Digital maturity assessment model for

the organizational and process dimensions. Sustain-

ability, page 15122.

Udovita, P. (2020). Conceptual review on dimensions of

digital transformation in modern era. International

Journal of Scientiﬁc and Research Publications, pages

520–529.

van Tonder, C., Bossink, B., Schachtebeck, C., and

Nieuwenhuizen, C. (2024). Key dimensions that mea-

sure the digital maturity levels of small and medium-

sized enterprises (smes). Journal of technology man-

agement & innovation, pages 110–130.

Van Veldhoven, Z. and Vanthienen, J. (2022). Digital

transformation as an interaction-driven perspective

between business, society, and technology. Electronic

markets, pages 629–644.

Vial, G. (2019). Understanding digital transformation: A

review and a research agenda. Journal of Strategic

Information Systems, pages 118–144.

Wani, A. A. (2024). Comprehensive analysis of clustering

algorithms: exploring limitations and innovative solu-

tions. PeerJ Computer Science.

Yahia, N. B., Hlel, J., and Colomo-Palacios, R. (2021).

From big data to deep data to support people analytics

for employee attrition prediction. Ieee Access, pages

60447–60458.

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

316

APPENDIX

Table 4: Digital maturity assessment levels.

Basic Discovery Developed Integrated Leadership

Strategy

The organization

has prepared

assumptions for the

DT implementation

or has implemented

an initial plan

with speciﬁc

highlights.

The organization

has an implemented

strategy for digitization

and assesses its

effeciency through

analyses.

Managers and consultants

notify the readiness for

starting digital changes.

The organization has a

devoted budget for

DT. Employees at all

levels of the organization

are engaged in the

DT processes.

Business development is

drived by the digitization

strategy.

The organization conduted

systematic risk assessment

concerning DT.

The organization

implements practices of

digitization strategies

and optimizes

them.

Culture

Change management is

supported by employees’

openness and the

assisstance

of managers in

implementing new ideas

and innovations

The organization has

adopted a continuous

improvement strategy and a

change management system.

Employees are kept

informed about the risks

and changes associated

with digitization.

Advanced solutions

are fostered in the

organization, and their

owners are awarded.

Employees beneﬁt from

a support in risk

management related to

digitization.

The organization

worked with a

culture open to

digitization and use

of new technologies.

Employees agree with

the coorporaton with

active human-machine .

The organization applies

optimal practices to

cultivate a culture of

collaborative

human-machine

interaction,

arising from transparent

risk management

regarding digitization.

Technology

The organization uses

information and

communication

technology for

horizontal and

vertical integration

in the internal

value chain. It also

uses mobile devices

for communication

among employees.

The organization

uses sensors

for data collection and

operation monitoring,

in addition to cloud

computing to save

and share data.

The organization

exploit the Internet

of Things for device

connection and data

transfer among them.

The organization

uses autonomous

devices to help

in decision making,

and AI to optimize

processes.

A machine-to-machine

communication

system (M2M) is deployed

to allow devices to

interact autonomously.

Data

The organization

collected data

periodically without

a clear strategy.

Basic standards are

developed for data

collection and

share at least

manually and

insights are realised

sporadically.

The organization

implemented a data

management strategy

supported by advanced

tools (data life cycle

management and

data quality).

Basic descriptive

analytics are used.

A clear data strategy

has been established.

The organization has

automated data

collection and

distribution,

including the automatic

generation of reports

sent to relevant

managers.

Descriptive and

diagnostic analytics

are used.

Data strategy is well

-established and aligns

closely with the

strategic goals of

the organization.

The organization has

deployed a data

integration platform,

to guarantee real-time

data access.

Predictive analytics are

integrated into business

processes

The organization

completely integrates

data as a core strategic

asset into culture

and operations.

The organization

use real-time and

automated transfer

of data between the

existing systems.

Prescriptive analytics

are used to provide

actionable

recommendations for

business outcomes

enhancement.

Skills

The organization

provides its employees

with training sessions

to enhance their

digital skills

(e.g., assistance

for novel digital

solutions, data

analytics,...).

Employees imrpove

their data analytical

and analyses skills.

A plan for acquiring

and developing digital

skills for employees

and managers has also

been in place.

The organization

has implemented a

systematic knowledge

management and

employee development

strategy using analytical

tools for its

implementation.

The required skills

related to DT and an

cross-disciplinary

mindset are ubiquitous

which covers the

whole levels

of management.

The organization

implements the best

practices of knowledge

management and

employee development.

Leadership

There is little

awareness or

understanding

of DT within

leadership.

There is no assigned

leader in charge of

riding digital initiatives.

Few leaders are

starting to understand

the necessity for a digital

mindset.Initial steps are

being taken to clarify

the responsibilities

of a CDO.

A digital mindset

is clearly deﬁned and

articulated within the

leadership team.

A CDO is formally

appointed with a

clear mandate to

drive and

coordinate DT

efforts.

Leadership shows a

strong digital mindset,

using data-driven

decision making and

a culture of

innovation and

continuous improvement.

The CDO role

is well-established

and integrated

into the executive

leadership team.

The digital mindset

is fully embedded within

the leadership and

organizational culture.

The CDO role evolves

into a central

strategic function,

driving DT.

Organization

Collaboration with

partners is ad hoc

and unstructured.

There is minimal

inetgration with

business partners

and information

exchange is limited.

Organization is rigid

to change and its

structure is hierarchical

and siloed.

There is basic processes

for managing

relationships

with key partners.

Some processes have

been adjusted to allow

for quiker changes.

Inititives to improve

croos-functional

communication and

collaboration.

Key business partners

are informationally

integrated with some

of the processes

carried out as part of

the organization.

The organizational

structure include clear

roles and

responsabilities.

The cooperation of the

organization with

business partners is

individualized and

managed based on

analyses and infor

-mation integration, as

well as data available

in real-time.

There is a systematic

approach to respond

to changes and the

organizational structure

is ﬂexible and supports

dynamic reconﬁguraton.

Partnerships are fully

integraed, there is a

continuous xchange of

real-time data.

The organization exhibits

peak agility, with optimized

processes for responsiveness

and innovation optimized

The organizational structure

is highly ﬂuid and adaptive

Explainable Assessment Model for Digital Transformation Maturity

317