Evaluating the Perceived Quality and Functionality of DEMO Models’

Representations in the Health Domain

David Aveiro

1,2,3 a

, V

ıtor Freitas

1,3 b

, Dulce Pacheco

1,2,4 c

and Duarte Pinto

1 d

ARDITI - Regional Agency for the Development of Research, Technology and Innovation, 9020-105 Funchal, Portugal

NOVA-LINCS, Universidade NOVA de Lisboa, Campus da Caparica, 2829-516 Caparica, Portugal

Faculty of Exact Sciences and Engineering, University of Madeira, Caminho da Penteada 9020-105 Funchal, Portugal

School of Technology and Management, University of Madeira, Caminho da Penteada 9020-105 Funchal, Portugal

Keywords:

Enterprise Models, Workﬂow, Business Process Modelling, DEMO.

Abstract:

Demo’s (Design and Engineering Methodology for Organizations) Way of Modelling encompasses a collec-

tion of interconnected models and diagrams designed to depict an organization’s structure and operations in

a cohesive and platform-independent manner. Nevertheless, there has been a contention that the syntax and

semantics of DEMO models are overly intricate and cluttered, posing challenges for laypeople in terms of

interpretation. Our research team has been working on improvements to the DEMO Modelling language for

Enterprise Ontology. Previous work had shown challenges in using standard DEMO notations for model com-

munication and validation, prompting the development of new representations. This study evaluates these

representations through quality and functionality testing using a health domain case and health professionals

with domain knowledge. The results of the conducted tests reveal signiﬁcant differences in perceived quality

and functionality between the new and traditional DEMO representations. These ﬁndings indicate a strong

preference for the new representations over traditional ones. This study underscores the importance of fo-

cusing on users in enhancing the effectiveness of modelling languages like DEMO, particularly in complex

domains such as healthcare. The results suggest that these new representations have the potential to improve

the perceived quality and functionality of DEMO models in various practical applications, including health-

related information systems.

1 INTRODUCTION

DEMO (Design and Engineering Methodology for

Organizations) consists of a method and language

standard based on the theories of Enterprise Ontol-

ogy. As the essence of enterprise engineering lies

in the systematic analysis and design of an organi-

zation’s business processes, enabling improvements

in efﬁciency and effectiveness, our research team has

been working on improvements to the DEMO Mod-

elling language representations. The latest develop-

ments are published in (Pinto et al., 2021) and (Gou-

veia et al., 2021), reporting the results of a large-scale

modelling project on the legal domain, where practice

has shown that the standard DEMO notations were

https://orcid.org/0000-0001-6453-3648

https://orcid.org/0009-0002-0667-5749

https://orcid.org/0000-0002-3983-434X

https://orcid.org/0000-0002-8451-5727

making model communication and validation a dif-

ﬁcult process. The new notations were, afterwards,

formally evaluated, in the same domain, in studies

published in (Pacheco et al., 2022b) and (Pacheco

et al., 2022a), expressing that these offer greater ac-

cessibility and a more straightforward understanding,

whether for professionals engaged in the represented

processes or those possessing expertise in DEMO.

The research contributions of this paper are pro-

viding a new assessment of the perceived quality and

functionality of the newly proposed enterprise engi-

neering’s DEMO model representations, now on the

health domain, through the analysis of NexusBRaNT,

an information system to support cognitive rehabilita-

tion.

Quality and functionality testing is widely recog-

nized as the most effective method for identifying the

genuine problems that can impact user performance

and preference (Wang and Caldwell, 2002). To as-

sess the perceived quality/functionality of the newly

Aveiro, D., Freitas, V., Pacheco, D. and Pinto, D.

Evaluating the Perceived Quality and Functionality of DEMO Models’ Representations in the Health Domain.

DOI: 10.5220/0012260400003598

In Proceedings of the 15th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2023) - Volume 2: KEOD, pages 331-338

ISBN: 978-989-758-671-2; ISSN: 2184-3228

331

proposed representations for DEMO Process and Fact

Models, we recruited a sample of health professionals

in the ﬁeld of psychology, with domain knowledge of

the modelled processes and modelled system.

Section 2 presents our research context, ﬁrst the

theoretical basis of DEMO, a summary of the afore-

mentioned NexusBRaNT scenario, and the DEMO

models addressed in our study. Section 3 presents

the study context with participants’ characterization,

method and procedures followed, and the main contri-

butions of this paper, namely the evaluation and com-

parison of the Perceived Quality and Functionality of

the traditional and newer diagrammatical representa-

tions of DEMO’s Process and Fact Models. Conclu-

sions and future work are found in section 4. Owing

to constraints on page space within the paper, images

had to be resized, but higher resolution versions can

still be accessed

2 RESEARCH CONTEXT

In this section, we will provide a comprehensive

overview of DEMO’s theories, models, and represen-

tations to ensure that readers are well-informed. Sub-

sequently, we will delve into an examination of the

information system utilized in this study, speciﬁcally

within the health domain. Following that, we will in-

troduce the representations of the DEMO models that

are subject to evaluation in this research.

2.1 DEMO Theories, Models and

Representations

The Design and Engineering Methodology for Orga-

nizations (DEMO) is a methodology centred on the

PSI theory of Enterprise Ontology to provide a set

of models and diagrams for representing an organi-

zation, which are interconnected and allow a compre-

hensive speciﬁcation of an enterprise in a neutral way

(Dietz and Mulder, 2020a).

Enterprise ontology aids in creating a shared

language within an organization, reducing misun-

derstandings and fostering effective communication.

Thus, every organization’s operations are a network

of transactions, according to the PSI theory (Dietz

and Mulder, 2020c). Each transaction represents a

path within the complete transaction pattern, which

is a universal pattern in all organizations (Dietz and

Mulder, 2020b). According to Dietz and Mulder (Di-

etz and Mulder, 2020c), a business process consists

of a sequence of procedural steps, which are in turn

http://bit.ly/keod-2023-paper-192

steps within transactions of speciﬁc transaction types

that are integral to a business process type within the

organizational structure of an enterprise.

A collection of models and diagrams are used in

DEMO’s way of modelling to depict an organization.

The Cooperation Model (CM), Action Model (AM),

Process Model (PM), and Fact Model (FM) are the

referenced aspect models. They are related to one

another and provide platform-independent represen-

tations of coherent information. As this paper is fo-

cused on the updated representations of DEMO’s PM

and FM, we will not delve into the other two aspect

models.

The Process Model outlines the business pro-

cesses that occur as a result of actions performed by

actors within the organization. The PM includes pro-

cess step types and applicable existence laws for both

internal and external transactions. It also reveals the

process step types and occurrence laws, including oc-

currence quantities, for different transaction types.

The Fact Model represents the organizational

products of the organization. It includes entity types,

value types, property types, and attribute types rele-

vant to the organization, along with the applicable ex-

istence laws. Additionally, it captures event types and

occurrence laws related to transitions in the organiza-

tion (Dietz and Mulder, 2020a).

2.2 The Information System to Support

Cognitive Rehabilitation

NexusBRaNT is an online platform to access the

BRaNT

(Belief Revision applied to Neurorehabili-

tation Therapy) project’s back ofﬁce. The BRaNT

project’s objective is to create technological tools that

support cognitive rehabilitation in domestic settings

through the assistance of artiﬁcial intelligence, while

also offering solutions to enhance the resilience of

healthcare systems.

This platform, then, speciﬁcally caters to health

professionals involved in cognitive rehabilitation,

including psychologists, neuropsychologists, thera-

pists, and others. It encompasses several key fea-

tures, including patient management, neuropsycho-

logical assessment management, and cognitive train-

ing management.

For the modelling of the NexusBRaNT system,

and following the notation proposed in (Gouveia

et al., 2021), in total, 224 fact types were speciﬁed,

with 31 concept types encompassing 193 attribute

types. A total of 20 neuropsychological instruments

https://www.arditi.pt/en/projetos-ﬁnalizados/brant-

project.html

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

332

are available for registration, collectively comprising

144 attributes. Each instrument has a varying number

of attributes, ranging from 2 to 17. These numbers

from the speciﬁed models highlight the complexity of

information requirements within NexusBRaNT. Con-

sequently, we considered it a suitable candidate for an

evaluation of the perceived quality and functionality

of the new DEMO Model representations.

2.3 Evaluated DEMO Model

Representations

In this subsection, we will initiate a comprehensive

examination of the disparities between the recently in-

troduced Process Model (Version A) and the conven-

tional DEMO PSD (Version B). Following this anal-

ysis, we will proceed to expand upon the distinctions

found between the newly proposed Fact Model (Ver-

sion C) and the established DEMO OFD (Version D).

It is essential to note that all the models discussed

herein are situated within the scope of the previously

mentioned NexusBRaNT case.

2.3.1 Process Model

In (Pinto et al., 2021), in the context of a large-

scale modelling project in a municipality, a novel ap-

proach to represent the PM was introduced (Version

A), which combines some elements of the standard

DEMO PM with both the CM and the AM, thus ex-

tending the conventional notation (Version B) pre-

sented in (Dietz and Mulder, 2020a). The main goal

was to provide a more agile and comprehensive so-

lution to specify and present the essence of organiza-

tional reality in a more visually appealing and con-

cise manner, making it easier for both modellers and

stakeholders to understand.

Figure 1 provides an explanation of each symbol

used in Version A.

Figure 2 (Version A) represents the Process Model

for NexusBRaNT’s Neuropsychological Assessments

and Training Programs Process, while Figure 3 shows

the traditional DEMO notation in (Dietz and Mulder,

2020a) (Version B). It is worth noting that the terms

“transaction” and “task” are used interchangeably due

Figure 1: Process Model Notation (Version A).

to quality and functionality concerns, as reported in

(Pinto et al., 2021).

Figure 2: Process Model (Version A) for NexusBRaNT’s

Neuropsychological Assessments and Training Programs

Process.

(a) DEMO traditional PSD (Version B) Part 1.

(b) DEMO traditional PSD (Version B) Part 2.

Figure 3: DEMO traditional PSD (Version B) for Nexus-

BRaNT’s Neuropsychological Assessments and Training

Programs Process.

The Neuropsychological Assessments and Train-

ing Programs Process starts with a healthcare pro-

fessional creating a new Neuropsychological As-

sessment Session for a patient, which includes one

or more neuropsychological assessment instruments.

These instruments, that can be of twenty different

kinds (resumed in the diagram with “...” to make it

Evaluating the Perceived Quality and Functionality of DEMO Models’ Representations in the Health Domain

333

legible), evaluate the patient’s current cognitive capa-

bilities and can be paused or resumed during the ses-

sion. Once all tests in the session are concluded, a

Cognitive Proﬁle Calculation based on Neuropsycho-

logical Assessment Instruments can take place. The

result of this calculation is then used by the health-

care professional as the baseline for the creation of

a Neuropsychological Training Program for that pa-

tient. The program must then start within the fol-

lowing 15 days, typically consisting of multiple Neu-

ropsychological Training Program Sessions (that are

a different process of their own, as they include multi-

ple tasks in each session depending on their speciﬁci-

ties) but can also be a single session. After complet-

ing all scheduled training sessions, another Cognitive

Proﬁle Calculation based on the Neuropsychological

Training Program is performed to assess the patient’s

progress compared to the initial assessment, conclud-

ing the process.

This representation (Pinto et al., 2021) solved sev-

eral issues regarding current DEMO CM and PM

representations. The Coordination Structure Dia-

gram and Process Structure Diagram of the latest

DEMO version (Dietz and Mulder, 2020a) were con-

sidered, both by stakeholders and modellers alike, to

be complicated to grasp and with extensive line clut-

ter (Pacheco et al., 2022b; Pacheco et al., 2022a). The

way we represent the PM - Figure 2 - is semantically

richer by: (a) presenting task names much closer to

day-to-day operations; (b) clearly separating the con-

cerns of process composition, task causation, and task

waiting; and (c) the connectors representing the com-

position perspective with diamonds, the causal links

by connectors with arrows, and the waiting links by

the connectors with double-crossed lines. Regard-

ing links between tasks, in DEMO’s notation dashed

meant optional and non-dashed mandatory. The use

of numbers at the end of connectors to represent that

essential concept is harder/slower to interpret than the

line expressing the concept, so their use was limited

to reﬂect cardinalities higher than 1. The proposed

notation in (Pinto et al., 2021) offered a new layer of

depth in the comprehension of the modelled process,

improving the readability of the limits of the scope

with the use of speciﬁc symbols to represent other re-

lated processes.

To address the gaps in the process information that

could overburden the model, a Transaction Descrip-

tion Table (TDT), that allows for the addition of rel-

evant text-intensive data such as descriptions, associ-

ated rules, conditions for transactions to take place,

time constraints, and other related elements was also

introduced in (Pinto et al., 2021), as can be seen in

Figure 4. For instance, let’s consider the “Cognitive

Training Program Execution” task. Its origin task is

“Cognitive Training Program Creation”. Upon com-

pletion, this task can trigger the “Cognitive Training

Program Session Execution” task and the “Cognitive

Proﬁle Calculation based on the Cognitive Training

Program” task. Additionally, speciﬁc conditions and

rules that must be taken into consideration for the task

to proceed are speciﬁed, along with any applicable

time constraints. In the example, only after all pro-

grammed sessions have ﬁnished, can the Cognitive

Proﬁle Calculation be executed.

By utilizing this table, we gain a comprehensive

understanding of each task’s characteristics and rela-

tionships within the system. It serves as a valuable

resource for analysing and organizing the various ele-

ments involved in task execution, facilitating not only

the implementation of more reﬁned system speciﬁca-

tions, but also the understanding and validation, by

an organization’s collaborators, of important details

of process ﬂow and execution.

2.3.2 Fact Model

In (Gouveia et al., 2021), a novel approach to rep-

resenting the FM was introduced (Version C), whose

resulting model on the context of NexusBRaNT can

be seen in Figure 5, which proposes a new way of

capturing and organizing facts in a more ﬂexible and

extensible manner compared to the traditional DEMO

FM (Version D), represented in Figure 6 in the same

context, presented by (Dietz and Mulder, 2020a). The

main goal of Version C is to provide a fact-oriented

and declarative solution that focuses on capturing the

essence of the system’s structure and behaviour. It is

important to note that Version C emphasizes the sepa-

ration of concerns between structure and behaviour,

while Version D integrates both aspects within the

FM.

The new representation proposed in (Gouveia

et al., 2021) addresses several issues identiﬁed with

the traditional DEMO FM representation. It simpli-

ﬁes the diagram by reducing line clutter and provides

a more intuitive and semantically rich depiction of

facts and their relationships. Version C of the Fact

Model introduces symbols and notations that improve

the comprehension of the modelled system, such as

the use of clear entity and fact type names that are

closer to the day-to-day operations.

In Version C, relationships between concepts are

depicted using arrows, representing an association

where an attribute in one concept refers to instances

of another concept (Gouveia et al., 2021). This allows

for a clear representation of the connections and de-

pendencies between concepts. To explicitly represent

dependency laws, a dark-ﬁlled circle is utilized. This

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

334

Figure 4: Excerpt of the Transaction Description Table (Version A).

Figure 5: Fact Model (Version C).

symbol indicates that the existence of an instance of

one concept is dependent on the existence of an in-

stance of the concept connected to it (Gouveia et al.,

2021). For example, a Cognitive Training Program

cannot exist without referencing an existing Patient

instance.

In addition to the improved notation, Version C of

the Fact Model separates the concerns of structure and

behaviour, focusing solely on capturing the facts and

their relationships, rather than explicitly modelling

state transitions or processes. Once the main con-

cepts and relationships are speciﬁed, the next step is

to identify the relevant attributes associated with each

concept. The Concept Attribute Diagram (CAD), of

which a section can be seen in Figure 7, proposed in

(Gouveia et al., 2021) is used for this purpose.

To provide additional context and details about the

modelled facts, a Fact Description Table (FDT), de-

picted in Figure 8, is introduced in (Gouveia et al.,

2021). The FDT allows for the inclusion of text-

intensive data such as fact descriptions, associated

Figure 6: DEMO traditional OFD (Version D).

Figure 7: Excerpt of the Concept Attribute Diagram (Ver-

sion C).

rules, conditions, and other relevant elements. This

enables a comprehensive understanding of the facts

and their properties.

Evaluating the Perceived Quality and Functionality of DEMO Models’ Representations in the Health Domain

335

Figure 8: Excerpt of the Fact Description Table (Version C).

In this depiction of the FDT, we provide an in-

depth analysis of each concept along with its corre-

sponding attributes. The table encompasses a com-

prehensive overview of all attributes associated with

each concept, including their scope, source, concept,

name, value type, referenced concept/values, descrip-

tion, and the tasks responsible for creating and mod-

ifying the concept. By organizing this information in

a structured manner, we gain valuable insights into

the characteristics and relationships of each concept

within the system. The table serves as a valuable re-

source for understanding the various attributes associ-

ated with each concept, facilitating effective concep-

tual analysis.

3 VALIDATION

Within the ﬁeld of information systems, many ap-

proaches to quality have been proposed (Krogstie,

2012), but it is still a problematic notion. Process

models may be difﬁcult to comprehend due to the

(un)formality of the modelling language, the com-

plexity or size of the model, or the effort needed

to deduce its important properties (Krogstie, 2016).

Our study hypothesized that subjects would evaluate

Versions A and C as having higher overall perceived

Quality (H1) and would also perceive the diagrams on

Versions A and C as more functional (H2).

In this section we will present the study’s partic-

ipants, method, procedures and results of the study

to validate the Perceived Quality and Functionality

of the DEMO Models, comparing the current nota-

tion with the newly proposed ones (Pinto et al., 2021;

Gouveia et al., 2021).

3.1 Participants

To assess the perceived functionality of the newly pro-

posed representations for DEMO Process and Fact

Models, we recruited a sample of health professionals

in the ﬁeld of psychology, with domain knowledge of

the modelled processes and implemented system (N

= 10, nine females and one male, Mdn age = 34, age

range = 28-58 years). All participants have a Human

and Social Sciences background. Namely, among

the participants, two are currently pursuing a Bach-

elor’s Degree in Psychology, one is currently pursu-

ing a Master’s Degree in Clinical, Health, and Well-

Being Psychology, and the remaining seven are health

professionals in the ﬁeld of psychology (Scholar lev-

els: Bachelors degree N = 2; Masters degree N = 7;

and Doctoral degree N = 1). Notably, two of these

health professionals are directly associated with the

NexusBRaNT project, adding valuable expertise and

insights to the study.

3.2 Method and Procedures

To evaluate the perceived quality and functionality

of the new and old versions of DEMO Models’ dia-

grams, a short questionnaire was designed, based on

previous work on the Quality of representations and

Functionality evaluations in the context of the SE-

QUAL framework (Krogstie, 2016). The question-

naire included questions related to functionality, such

as participants’ agreement on the diagrams’ function-

ality and their preference for one version over the

other. Two dimensions from the SEQUAL frame-

work, Empirical Quality (EQ) and Social Pragmati-

cal Quality (SPQ), were also included to assess the

perceived quality of the diagrams. The questionnaire

consisted of 10 items on a six-point scale ranging

from 1 = strongly disagree to 6 = strongly agree,

and one question related to Functionality (“is it func-

tional?”). Two questions were negatively phrased and

reversed before the statistical analyses.

Participants were instructed to evaluate both Ver-

sions A and B of the Process Model diagrams and

Versions C and D of the Fact Model diagrams. The

survey included three questions where subjects were

forced to pick which version (A or B; and C or D)

they considered as more functional (e.g., which ver-

sion is “more suitable to support the execution of your

tasks”).

Functional suitability, as deﬁned in software qual-

ity models, pertains to how well a system fulﬁlls func-

tions in line with both explicit and implicit require-

ments (ISO, 2011). It can be broken down into three

aspects: functional completeness, functional correct-

ness, and functional appropriateness. (ISO, 2011).

Functional completeness evaluates whether the set of

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

336

functions covers all speciﬁed tasks and user objec-

tives, ensuring no crucial functionality is missing.

Functional correctness assesses how accurately the

system produces results with the required precision,

verifying that it meets expectations without errors.

Functional appropriateness measures how well the

provided functions support users in achieving their in-

tended tasks and objectives effectively.

The experiment began with a brieﬁng about the

study and users providing informed consent. The

brieﬁng covered key DEMO concepts. The new and

old representations of the PM, and then the FM, di-

agrams were then presented to the participants, fol-

lowed by a questionnaire to assess their perceived

quality and functionality. The scale EQ revealed good

internal consistency (5 −items, N = 10, α = .72). The

scale SPQ also reached a good internal consistency

(5 − items, N = 10, α = .76). In the next section, par-

ticipants were given a scenario in which they had

to decide where to ﬁnd information when uncertain

about forwarding a speciﬁc process. They could

choose between Version A/C of the diagram, Ver-

sion B/D, or consulting the technical manual. Partic-

ipants assessed the likelihood of this scenario occur-

ring on a six-point scale ranging from 1 (deﬁnitely

not) to 6 (deﬁnitely yes). Additionally, the ques-

tionnaire included demographic questions about age,

gender, scholar level, and background. Participants

were also asked to self-report their knowledge lev-

els in three areas: Neuropsychological Assessments

and Training Programs Process, NexusBRaNT sys-

tem, and DEMO, using a scale from 1 (null) to 6

(very good). Comments and suggestions for improve-

ment were collected by the researchers. Statistical

data analyses were performed using computer soft-

ware (IBM SPSS Statistics, version 27 for MacOS X).

3.3 Results

Wilcoxon tests were employed to compare the per-

ceived Quality and Functionality of the diagrams. The

ﬁndings indicated that the Version A of the Process

Diagram and Transaction Description Table was per-

ceived to have a signiﬁcantly higher level of EQ (Mdn

= 5.1, SD = .56) compared to Version B (Mdn = 2.7,

SD = .51), with a z-value of -2.83 (p = .005), with a

large effect size (r = -.89) (Field, 2013). Regarding

SPQ, Version A is also perceived as higher (Version

A: Mdn = 5.46, SD = .42; Version B: Mdn = 2.72, SD

= .65), with a z-value of -2.81 (p = .005), also with a

large effect size (r = -.89) (Field, 2013).

When comparing the CRD, CAD and FDT (Ver-

sion C) to Version D, the ﬁrst is perceived as having

a higher level of EQ (Mdn = 4.7, SD = 1.1), when

compared to the current DEMO representation (Mdn

= 2.34, SD = .61), z = -2.67, p = .008, having a large

effect size (r = -.84) (Field, 2013). On variable SPQ,

Version C also obtained higher results (Version C:

Mdn = 4.88, SD = 1.24; Version D: Mdn = 2.6, SD

= .71), with a z-value of -2.71 (p = .007), also with a

large effect size (r = -.86) (Field, 2013). These results

provide full support for our hypothesis (H1), suggest-

ing that Versions A and C are perceived to possess

superior quality.

Previous research has suggested that some indi-

viduals prefer acquiring new information through for-

mal models, while others ﬁnd a combination of for-

mal and informal statements to be more comprehen-

sive (Krogstie, 2016). In our study, we asked par-

ticipants to pick, between Version A and Version B:

(a) which one they consider to be easier to under-

stand the sequence of tasks; (b) which one is easier

to visualize and understand the tasks related to their

professional activity; and, ﬁnally, (c) which one do

they consider to be more suitable to support the daily

execution of their tasks. All ten participants picked

Version A, indicating that they perceived Version A

as having higher quality, functionality, and overall at-

tractiveness. We asked analogous questions for com-

paring Versions C and D. Again, all participants chose

the newer representation (Version C). These ﬁndings

offer empirical support for our hypothesis (H2), indi-

cating that Versions A and C are perceived to exhibit

superior functionality.

When prompted to select the probability of con-

sulting the Version A of the diagrams, Version B, or

the Manual, in the case of doubt, the majority of the

participants stated that they prefer Version A (Mdn =

5.3, SD = .82), rather than Version B (Mdn = 2.9, SD =

.99) or the Manual (Mdn = 3.7, SD = 1.83). We asked

the same question but to compare Version C and D.

All participants have also shown a clear preference

for Version C (Version C: Mdn = 4.9, SD = 1.1; Ver-

sion D: Mdn = 2.8, SD = .79; Manual: Mdn = 3.6, SD

= 1.78).

The results clearly demonstrate a strong prefer-

ence for the newer representations over the conven-

tional ones, even when compared to the manual,

which participants are more familiar with. The new

representations of DEMO’s PM and FM present all

the essential information visually pleasing and con-

cisely, ensuring ease of comprehension for both mod-

ellers and stakeholders. Incorporating symbols and

notations that enhance the understanding of the mod-

elled system is a clear advantage of the new represen-

tations.

Evaluating the Perceived Quality and Functionality of DEMO Models’ Representations in the Health Domain

337

4 CONCLUSIONS AND FUTURE

WORK

In this paper, we have presented our research efforts

in enhancing the DEMO modelling language. Speciﬁ-

cally, we have focused on evaluating the functionality

of new DEMO model types using the case of Nexus-

BRaNT.

The evaluation of the perceived quality and func-

tionality of the new and conventional versions of

DEMO’s PM and FM diagrams was conducted

through a questionnaire-based survey. The results

indicated that the newer versions of the diagrams

are perceived to have signiﬁcantly higher quality and

functionality compared to the traditional versions, in

terms of comprehensibility, visualization, and suit-

ability for supporting participants’ daily tasks. The

results also indicate that the enhanced DEMO models

offer a promising approach to simplify and accelerate

the modelling of software solutions, as the new repre-

sentations are evaluated as cognitively more effective

than DEMO’s current representations.

The revised DEMO PM and FM diagrams have

also been integrated into the low-code platform

DISME (Dynamic Information System Modeller and

Executer) (Andrade et al., 2018; Aveiro and Freitas,

2023; Aveiro et al., 2023). New pilot projects aimed

at system implementation are on the horizon, promis-

ing additional instances in diverse domains to aug-

ment the existing dataset and bolster the ﬁndings of

this study.

However, there are still several areas that require

further attention and future work. While our quality

and functionality evaluation provided positive results,

it is essential to conduct additional evaluations with a

larger and more diverse user group to ensure the gen-

eralizability of the ﬁndings.

As the main limitation, we identify the small sam-

ple size. Therefore, results must be interpreted with

caution and generalizability of the ﬁndings may be

limited.

In conclusion, our research’s primary contribu-

tions encompass an enhanced understanding of how

users perceive the new DEMO models’ representa-

tions concerning their Quality and Functionality, with

a particular emphasis on the Process and Fact Models,

more accessible and inclusive to all stakeholders en-

gaged in an organization’s daily operations. With on-

going efforts, we envision the newly proposed DEMO

models becoming an invaluable tool in modelling ef-

ﬁcient information systems in various domains.

REFERENCES

(2011). Iso / iec 25010 : 2011 systems and software engi-

neering — systems and software quality requirements

and evaluation ( square ) — system and software qual-

ity models.

Andrade, M., Aveiro, D., and Pinto, D. (2018). Demo based

dynamic information system modeller and executer.

Aveiro, D. and Freitas, V. (2023). A new action meta-

model and grammar for a demo based low-code plat-

form rules processing engine. In Advances in Enter-

prise Engineering XVI, pages 33–52, Cham. Springer

Nature Switzerland.

Aveiro, D., Freitas, V., Cunha, E., Quintal, F., and Almeida,

Y. (2023). Traditional vs. low-code development:

comparing needed effort and system complexity in the

nexusbrant experiment. In 2023 IEEE 25th Confer-

ence on Business Informatics (CBI), pages 1–10.

Dietz, J. L. G. and Mulder, H. B. F. (2020a). The DEMO

Methodology, pages 261–299. Springer International

Publishing, Cham.

Dietz, J. L. G. and Mulder, H. B. F. (2020b). The Enterprise

Engineering Theories, pages 23–48. Springer Interna-

tional Publishing, Cham.

Dietz, J. L. G. and Mulder, H. B. F. (2020c). The PSI

Theory: Understanding the Operation of Organisa-

tions, pages 119–157. Springer International Publish-

ing, Cham.

Field, A. (2013). Discovering statistics using IBM SPSS

statistics. SAGE Publications, London, England, 4

edition.

Gouveia, B., Aveiro, D., Pacheco, D., Pinto, D., and Gou-

veia, D. (2021). Fact Model in DEMO - Urban Law

Case and Proposal of Representation Improvements,

pages 173–190.

Krogstie, J. (2012). Quality of business process models.

In The Practice of Enterprise Modeling, pages 76–90,

Berlin, Heidelberg. Springer Berlin Heidelberg.

Krogstie, J. (2016). SEQUAL Specialized for Business Pro-

cess Models, pages 103–138. Springer International

Publishing, Cham.

Pacheco, D., Aveiro, D., Gouveia, B., and Pinto, D. (2022a).

Evaluation of the perceived quality and functional-

ity of fact model diagrams in demo. In Advances in

Enterprise Engineering XV, pages 114–128, Cham.

Springer International Publishing.

Pacheco, D., Aveiro, D., Pinto, D., and Gouveia, B.

(2022b). Towards the x-theory: An evaluation of the

perceived quality and functionality of demo’s process

model. In Advances in Enterprise Engineering XV,

pages 129–148, Cham. Springer International Pub-

lishing.

Pinto, D., Aveiro, D., Pacheco, D., Gouveia, B., and Gou-

veia, D. (2021). Validation of DEMO’s Conciseness

Quality and Proposal of Improvements to the Process

Model, pages 133–152.

Wang, E. and Caldwell, B. (2002). An empirical study of

usability testing: Heuristic evaluation vs. user testing.

Proceedings of the Human Factors and Ergonomics

Society Annual Meeting, pages 774–778.

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

338