Product Line Engineering in Smart Governance Systems

Salvador Mu

noz-Hermoso

1 a

, Miguel A. Olivero

2 b

, Francisco J. Dom

ınguez-Mayo

2 c

and David Benavides

2 d

INPRO, Diputaci

on Provincial de Sevilla, Men

endez Pelayo 32, Seville, Spain

I3US, Universidad de Sevilla, Reina Mercedes s/n, Seville, Spain

Keywords:

Conﬁguration, Feature Models, Smart Governance, Software Product Lines, Software Reuse.

Abstract:

Smart governance systems are used to develop smart and sustainable cities. Small municipalities are users

of these systems. However, creating individual and custom systems for each municipality is costly and often

unfeasible due to the limited resources of local governments. Developing a modular and conﬁgurable system

could help reduce costs, enabling municipalities to tailor solutions to their speciﬁc needs without requiring

custom developments. Software Product Line Engineering (SPLE) can make this possible by fostering soft-

ware reuse, and creating a set of adaptable systems that share common features. Nonetheless, applying SPLE

in the smart governance domain remains challenging and, so far, there is no applications of SPLE in this area

in the existing literature. We propose an Smart Governance Product Line (SGPL), based on a multi-level

conﬁguration architecture for the customization of solutions in the smart governance domain. Based on the

SGPL, we also present a prototype conﬁgurator for customizable governance systems that has been used in a

case involving three municipalities with different needs. The tool was materialized with conﬁgurations of the

existing Decidim governance system. The prototype demonstrated its usefulness in deploying in an easy and

automatized way adapted conﬁgurations to the municipalities’ needs. Furthermore, the case study suggests

that this approach could evolve into a general framework to support different software systems and compo-

nents for providing a comprehensive smart governance platform tailored to institutional speciﬁcations.

1 INTRODUCTION

Smart governance boosts the quality of public ser-

vices and provide smart management (of territories

and societies), through a collaborative government,

open to all the stakeholders. Introducing smart gover-

nance requires the collaboration of different parties.

However, the variety of stakeholders with different

and sometimes conﬂicting interests introduces by a

complex challenge (Tran et al., 2019). The joint use

of ’smart’ and ’governance’ aims at maximizing pos-

itive results by means of intensive use of Informa-

tion and Communication Technology (ICT) (Nastjuk

et al., 2022). Materializing smart governance is com-

plex, as it involves multiple systems and services that

are required to interoperate. A smart governance, is

an ecosystem that presents a great variability due to

the different needs of multiple stakeholders, as well

https://orcid.org/0000-0002-8130-7869

https://orcid.org/0000-0002-6627-3699

https://orcid.org/0000-0003-3502-8858

https://orcid.org/0000-0002-8449-3273

as different software systems and services. Addi-

tionally, governance processes and public services in

small municipalities may not be the same to those in

big cities, or those in a regional or state level (Fajar

and Shoﬁ, 2017; Ojo et al., 2007). Smart governance

is key in the 2030 Agenda to achieve sustainable cities

(Ependi et al., 2022), facilitating the implementation

and conﬁguration of these systems is a relevant issue.

Software Product Line Engineering (SPLE) is

used to manage software variability by focusing on

the systematic production and reuse of shared as-

sets across related products, thus favoring variability

and reusability (Achour et al., 2011; Felfernig et al.,

2024). A Feature Model (FM) represents all possi-

ble conﬁgurations of a Software Product Line (SPL)

in a compact way (Felfernig et al., 2024; Pohl et al.,

2005). To the best of our knowledge, SPLE has not

been applied in the smart governance context so far.

In this paper, we propose a general Smart Gov-

ernance Product Line (SGPL) approach that con-

siders the main features and needs of smart gover-

nance. The SGPL would allow for the generation of

262

Muñoz-Hermoso, S., Olivero, M. A., Domínguez-Mayo, F. J. and Benavides, D.

Product Line Engineering in Smart Governance Systems.

DOI: 10.5220/0013522200003964

In Proceedings of the 20th International Conference on Software Technologies (ICSOFT 2025), pages 262-272

ISBN: 978-989-758-757-3; ISSN: 2184-2833

smart governance platforms adapted to each individ-

ual need by reusing commonalities. The SGPL ap-

proach is supported by a conﬁguration tool named

InGoverkno (conﬁgurator for INtelligent GOVERn-

ments KNOwledge-based platforms). Such a tool al-

lows the conﬁguration of desired features for a smart

governance system. The conﬁgurator is based on an

FM that collects the common features of smart gover-

nance. The FM allows to study the variability of this

SGPL, through the conﬁguration of the different alter-

natives. Once the desired smart governance has been

chosen, the conﬁgurator deploys a customized smart

governance platform using the features available in

the chosen platform (i.e., Decidim

). This conﬁgu-

rator was applied to the needs of three different-sized

local governments suggesting that the case study re-

sults would facilitate the deployment of smart gover-

nance solutions.

Our approach contributes to this area by (1) ex-

ploring the variability of smart governance systems,

(2) deﬁning a FM for smart governance, and (3) devel-

oping a conﬁguration tool that enables the automatic

deployment of customized systems.

This paper is organized into different sections:

Section 2 reviews background and related work. Sec-

tion 3 presents the SGPL. An FM summarizing the

features and constraints is proposed in Section 4.

Then, Section 5 outlines the application of our conﬁg-

uration tool, which is used in a case study in Section

6 involving three municipalities. Finally, Sections 7

and 8, discuss and summarize the ﬁndings, and pro-

pose future work.

2 BACKGROUND AND RELATED

WORK

Electronic governance (e-governance), uses ICT to

enhance public services and encourages citizen par-

ticipation in the decision-making (DM) processes

(Grigalashvili, 2022). Thus, the e-governance and

its DM processes are the basis for implementing par-

ticipatory processes such as public consultations or

participatory budgets (Bherer et al., 2016; Boudjelida

et al., 2016).

Smart governance implies e-governance in which

smart technologies (e.g., artiﬁcial intelligence [AI])

are jointly applied with civil society collaboration for

the co-creation of quality public services and pro-

mote optimal DM (Parycek and Viale, 2017; Ruijer

et al., 2023). Therefore, electronic collaboration (e-

collaboration), deﬁned as ’working together as a team

https://decidim.org/

in a particular situation to solve problems using tech-

nologies’ (Abdullah et al., 2019), plays a key role in

smart governance implementation. Since smart gov-

ernance is a recent and multifaceted ﬁeld, and still re-

mains complex to handle, it is rare to ﬁnd any tech-

nical progress (Nastjuk et al., 2022), and most ex-

isting applications are more focused on the ﬁeld of

democratic or collaborative e-governance rather than

on smart governance.

Among the existing e-governance systems, De-

cidim and Consul

highlight for their wide range of

functionalities and for covering all stages of the pub-

lic policy cycle (Deseriis, 2023). Furthermore, it was

reinforced by an overview of websites of 20 existing

applications. Some of these features are also useful

in the smart governance context. We focus on De-

cidim platform as it is open source and has ﬂexibility

and diversity of components. Furthermore, its design

is focused on public institutions being used in rele-

vant municipalities such as the city council of New

York, Helsinki and Barcelona, as well as the Euro-

pean Union, among other institutions. In addition, it

considers participation and collaboration in a broader

way. So that its participation feature could be applied

in any organization, and be specialized in speciﬁc par-

ticipatory processes for institutions. Like the vast

majority of platforms, it lacks features that enhance

knowledge and collective DM by applying group DM

and AI techniques, that they could move from ’Demo-

cratic Governance’ to ’Smart Governance’.

The conﬁgurability and variability management of

information systems is important. SPLE together with

the FM modeling method, enable software products

and services to be adapted to the needs of the orga-

nization and its stakeholders (Felfernig et al., 2024;

Pohl et al., 2005).

The e-collaboration ﬁeld has an inherent complex-

ity that is increased since its requirements and needs

vary according to the domain of application and the

type of organization (Munkvold and Zigurs, 2005).

Therefore, regarding the e-governance ﬁeld (which

involves e-collaboration), it also supports a great va-

riety of requirements related to the governance pro-

cesses, the institutions and its stakeholders (Fajar and

Shoﬁ, 2017; Ojo et al., 2007). Thus, considering the

high variability of those systems, a multi-level con-

ﬁguration approach is also suitable to manage their

conﬁgurability (Reiser and Weber, 2007). However,

we have not found any approach that speciﬁcally ad-

dresses the problem of conﬁgurability in smart gover-

nance or in the e-governance context. As evidenced

in the literature review and highlighted in (Cledou

and Barbosa, 2017), there are few studies that address

https://consuldemocracy.org/

Product Line Engineering in Smart Governance Systems

263

variability and SPLE in the general electronic govern-

ment (e-government) domain, so this is an area that

needs to be further explored.

We have identiﬁed some studies related to the e-

governance ﬁeld. Even though they do not focus on

improving conﬁgurability, some studies promote the

development and adaptation of these e-governance

systems to different needs by means of SPLE (Achour

et al., 2011; Cledou and Barbosa, 2017; Fajar and

Shoﬁ, 2017). An FM is proposed in (Debnath et al.,

2008) using a broader vision and considering the fea-

tures of e-government systems. The FM also con-

siders establishing a division by front-ofﬁce or back-

ofﬁce software, and another by applications typology,

such as Government to Government (G2G) or Gov-

ernment to Citizen (C2C). In (Fajar and Shoﬁ, 2017)

the authors further distinguish products for central or

local governments. It is appropriate as the latter of-

fers public services related to city government, quite

different from those offered by the state government.

SPLE is also applied in some particular use cases such

as the one proposed in (Lima et al., 2014) for content

management systems (CMS).

3 SMART GOVERNANCE

PRODUCT LINE

So far, no single system fully supports smart gov-

ernance services because of their complexity and

the diversity of data required from the organization.

Therefore, the services to be developed will need to

interoperate with the organization’s general and e-

government legacy systems, presenting a high vari-

ability and complexity (Fajar and Shoﬁ, 2017).

The SGPL involves a vast variability. Managing

an SPL and its customization in a context of large

variability and multiple features is challenging if us-

ing a ﬂat conﬁgurability (Clark et al., 2017). How-

ever, a multi-level approach favors conﬁgurability and

reduces the complexity of managing the variability

(Clark et al., 2017; Czarnecki et al., 2005; Reiser and

Weber, 2007).

3.1 Multi-Level Conﬁguration

Architecture

A multi-level approach enables division and organizes

the conﬁguration into different levels. Each one of

these levels is related to a different area such as sys-

tem parts or requirements groups, facilitating the con-

ﬁguration, customization, and reuse of the services

(Clark et al., 2017; Czarnecki et al., 2005).

SGPL considers a fourth-layered conﬁguration as

shown in the Figure 1: Level 1 (Process), collects the

variability of the business processes for each individ-

ual governance collaborative event. Level 2 (Model),

is proposed as an abstraction layer to assist in imple-

menting different smart governance models. These

ﬁrst levels manage the variability of the business logic

of collaborative processes. To divide the complexity

we opted for model two abstraction level since collab-

oration is a main pillar of smart governance (Nastjuk

et al., 2022; Purba and Arman, 2022). Thus, the spe-

ciﬁc smart governance domain is considered a spe-

cialization of the general e-collaboration domain. The

left side represents the conﬁguration of the general e-

collaboration domain, and the right side is the spe-

cialized (’white triangles’) conﬁguration of the smart

governance domain.

Level 3 (Platform), represents the set of features

characterizing the deployed customized smart gover-

nance system. Finally, level 4 (Organization) is used

to orchestrate those particular features of the system

for the speciﬁc organization. Whereas these ﬁrst two

are common to any organization, the latter two refer

to the software system that supports the smart gover-

nance instance for each individual process and spe-

ciﬁc organization.

Figure 1: Multi-level conﬁguration architecture.

The multi-level approach contributes to a more

precise and easier customization, as several organiza-

tions can be customized at once. In addition, the con-

ﬁgurations are inherited through the levels from top

to bottom. The whole system could be updated when

acting on levels 1 and 2 of the conﬁguration. Just by

modifying level 3, it can be modiﬁed those features

that globally affect the behavior of the whole system.

ICSOFT 2025 - 20th International Conference on Software Technologies

264

Finally, it is possible to customize any particular or-

ganization by only modifying the differential features

at level 4.

Level 1: Process. The participatory processes are

subject to citizen participation and collaborative gov-

ernance regulations. Since there are different regu-

lations and procedures depending on the country and

territorial scope of the institution, a way to conﬁgure

the participatory processes is needed. Therefore, level

1 deﬁnes and characterizes these process models, and

allows their adaptation to the needs of each organi-

zation. (e.g., a certain process might be modeled for

collaborative agenda setting, whereas another might

be modeled for participatory budgeting).

In the E-Collaboration Processes Conﬁguration

the common e-collaboration features of different

types of collaborative processes are set (e.g., whether

the process includes DM, which stakeholders can par-

ticipate, how the ﬁnal decision is made, the duration

of the process, etc.). The Smart Governance Pro-

cesses Conﬁguration allows specialized process ty-

pology for the smart governance domain (e.g., citizen

consultations or surveys) and conﬁguring their spe-

ciﬁc features. These process conﬁgurations are mate-

rialized in the next conﬁguration level 2.

Level 2: Model. Models are used as a structure

or reference, to be reused or adapted. The models of

e-collaboration and smart governance refer to a set of

principles, structures, processes or mechanisms that

guide how decisions are taken. Those processes are

already considered in Level 1. However, the same

process may vary according to the kind of organi-

zation, its scope, or its goals. The idea of having a

E-collaboration Models Conﬁguration and a Smart

Governance Models Conﬁguration is to help insti-

tutions and organizations deﬁne their structure and

DM behavior. Consequently, these models would al-

low reusability and adaptability of already existing

processes (relations represented by ’black arrows’).

These two levels enhance dynamic variability, as they

enable a system in production to meet new needs by

conﬁguring new types of collaborative processes or

new governance models.

Level 3: Platform. Previous levels character-

ize the smart governance models with their associ-

ated processes in software platforms and might be

seen as a back-ofﬁce conﬁguration. This Level 3

instances speciﬁc pre-conﬁgured smart governance

models with their process (’black arrow’) that can be

adapted to the speciﬁc smart governance platform to

be customized and deployed. This Smart Governance

Platform Conﬁguration outlines the software partic-

ularities to be used by speciﬁc organizations, allow-

ing the customization of a product from the SGPL

that supports the chosen customized smart gover-

nance services (i.e., a set of smart governance models,

and its associated processes). Thus, a unique software

platform can offer smart governance services to a set

of organizations. This conﬁguration level also fosters

the identiﬁcation and use of common and reusable as-

sets (i.e., components, tools, libraries and other arti-

facts) to be later employed through application engi-

neering to generate the customized platforms.

Level 4: Organization. Previous levels detail

the conﬁgurability of those features that can be of-

fered, whereas this level allows the customization

of those chosen features. Taking account the multi-

organization approach of level 3, in level 4 each or-

ganization can tailor a deployed platform according

to their requirements. Those requirements consider

general and speciﬁc features of the governance model

and its processes. For instance, some municipalities

may want their participatory processes to be binding,

whereas others may only want them to be consulta-

tive. It also applies to speciﬁc characteristics of the or-

ganization, and its participatory processes, such as the

duration, participation requirements or restrictions,

among others. It will also be the level where the user

interface of the platform can be adapted to the needs

of the organization. Consequently, this customized

Smart Governance Platform Organization Conﬁgura-

tion, constrained from the general platform conﬁgura-

tion (’black arrow’), makes up the complete platform-

speciﬁc conﬁguration for any particular organization.

3.2 Features

Below we brieﬂy describe, organized by groups, the

most relevant features desirable in the smart gov-

ernance systems and services. They are based on

the principles, drivers, key aspects, regulations, and

functionalities that we have identiﬁed in our research

about e-governance and smart governance. Further-

more, they have been endorsed and adjusted based on

the analysis of the features of existing e-governance

platforms and the feedback received from real users

on e-governance needs according to the case study

carried out (Section 6).

Core Features. They are the main features to sup-

port the core functionality and business logic of smart

governance managed at levels 1 (Process) and 3 (Plat-

form) of the conﬁguration. Since collaborative gover-

nance and shared DM with civil society are key for

smart governance (Ependi et al., 2022; Nastjuk et al.,

2022; Purba and Arman, 2022; Ruijer et al., 2023),

these features include the logic necessary for stake-

holders to interact with the system on a given topic

Product Line Engineering in Smart Governance Systems

265

by providing and sharing information and knowledge

to solve a problem or make a decision. Furthermore,

the evaluation of the results of decisions, projects,

and services by stakeholders will provide feedback

that facilitates appropriate changes to improve results,

transparency, and accountability (Hasan and Rizvi,

2021; Parycek and Viale, 2017; Valle-Cruz et al.,

2020). They are the basis of the collaborative pro-

cesses implementation.

About Techniques, Technologies and Methods.

Intelligent and intensive use of data and ICT are key

in smart governance. Thus, data analysis (DA) fos-

tered with knowledge management (KM) capabili-

ties allows obtaining relevant information for achiev-

ing effective evidence-based policies and intelligent

DM (Hong and Lee, 2023; Parycek and Viale, 2017).

DM techniques can improve individual and group

decisions, as well as favor negotiation and consen-

sus building (Tran et al., 2019). AI techniques can

be applied to DA, KM and DM, improving the co-

production of quality public services (Hong and Lee,

2023). Techniques such as machine learning and

deep learning allow extracting useful information and

knowledge to improve DM (Hong and Lee, 2023;

Hasan and Rizvi, 2021). Thus, this features group,

managed by level 3 of conﬁguration, favors collec-

tive knowledge management from the different data

sources and stakeholders.

About Guarantees. Since legal framework and eth-

ical principles are key aspects (Parycek and Viale,

2017; Pereira et al., 2018), these platform’s features

(level 3) enable rules related to norms, policies, pro-

cedures and ethical values of the organization, so

that actions and decisions are within the legal and

ethical frameworks that affect the organization and

the people related to it. To accomplish principles

of transparency and accountability of e-governance

(Abu-Shanab, 2015), features are contemplated to

support real-time information on the collaboration

process and DM; as well as on the criteria used to

choose certain options (explainability) and their re-

sults (Dwivedi et al., 2021).

About the e-Collaboration Model. These features

are related to the e-collaboration model used in smart

governance and addressed in conﬁguration level 2.

They aim to favor agile and continuous interaction

between and among the stages of the collaborative

processes providing value incrementally and itera-

tively (Parycek and Viale, 2017; Valle-Cruz et al.,

2020). Since a knowledgeable citizen will inﬂu-

ence more effective DM processes (Tiwari et al.,

2023), its qualiﬁcation may be reﬂected in the weight

of collaborative outcomes and DM. A comprehen-

sive and multi-level approach across the policy cy-

cle and decision levels maximizes the beneﬁts ob-

tained (Hong and Lee, 2023; Tiwari et al., 2023;

Valle-Cruz et al., 2020; Wirtz and M

uller, 2023).

Furthermore, multi-organization support enables col-

laboration of several organizations to solve common

problems, fostering intergovernmental collaboration

(Hasan and Rizvi, 2021). A data-driven model to-

gether with smart assistance, through AI and DM

features (above mentioned), fosters performing au-

tomated actions and evidence-based DM (Ju et al.,

2018; Parycek and Viale, 2017). All previous bene-

ﬁts can be enhanced with networked processes to ob-

tain greater knowledge, through aggregation relation-

ships to create more complex processes composed of

others, afﬁnity relations between similar processes, as

well as information or chronological dependencies.

About the e-Collaboration Process. Features must

be considered (level 1 of conﬁguration) to regulate

the functionality of collaborative processes, the adap-

tation to the speciﬁc typology of participatory pro-

cesses, and the different phases of the policy cy-

cle (i.e., policy strategy, agenda setting, policy for-

mulation, policy implementation and policy assess-

ment (Hong and Lee, 2023; Valle-Cruz et al., 2020)).

Thus, these features allow characterize the most com-

mon types of participatory processes contemplated

in the different regulations on citizen participation

(Bouzguenda et al., 2019; Parycek and Viale, 2017;

Yusuf et al., 2019). They also allow enabling the DM

stage in collaborative processes, and the assessment

of their outcomes. Other speciﬁc features are contem-

plated for the ﬁnest adjustment of processes working.

4 FEATURE MODEL

A Feature Model (FM) is a structured representation

of those features that characterize a system. These

features deﬁne the functionalities and key properties,

which are classiﬁed as mandatory, optional or alterna-

tives. Deﬁning the FM for the SGPL helps in visual-

izing how the features are orchestrated and which of

them are common for the products of this SGPL.

The FM that characterizes the SGPL approach

considers the multi-level conﬁguration architecture

integrating their levels to obtain a family of cus-

tomized products down to the speciﬁc platform of the

organization. Although this is a single FM, for greater

clarity we present the subtrees related to each level, in

a descending order respecting the natural hierarchical

ICSOFT 2025 - 20th International Conference on Software Technologies

266

order of the FM. The features outlined in the subsec-

tion 3.2 are speciﬁed and represented below.

4.1 Feature Model - Level 3 & 4

Figure 2 represents the general features of the smart

governance platform, showing collapsed groups of

features related to the other conﬁguration levels 1 and

2. The FM initial root represents the Smart Gov-

ernance Platform, which has E-Collaboration Com-

mon Services and Smart Governance Services. Fur-

thermore, on the one side, a Multi-Tenant architec-

ture can be chosen. It deﬁnes whether the platform is

going to be used by different organizations (and cus-

tomized for each one) or an independent platform will

be considered for each organization. On the other

hand, the group Graphical User Interface (GUI) is

optional deﬁning if a GUI is required or if an existing

external interface is used. If the system GUI is se-

lected (collapsed in the ﬁgure), then the organization

might choose between a Web Platform, a Mobile App,

or both; for increased interoperability and accessibil-

ity from any device. On these levels, the features of

the ﬁnal software platform are conﬁgured in a general

way and in speciﬁc organizations that use the plat-

form.

Regarding E-Collaboration Common Services,

Collaboration and Decision-Making are considered

core features. Both are mandatory in any conﬁgura-

tion, because these general services are necessary to

accomplish any process related to smart governance.

Nevertheless, these must be customized to adapt to

the organization’s needs, through their features and

subgroups. Assessment functionality (shown col-

lapsed) can be selected to accomplish Results Assess-

ment of processes and decisions, or Stakeholders As-

sessment. In smart governance, these services will en-

able public policy evaluation citizenship processes.

Technical Capabilities (some of which are re-

quired by other models and processes), enhances ad-

dressing complex collaboration and DM problems.

This is achieved by including technical capabilities

for data analysis (DA Support), knowledge manage-

ment from relevant data (KM Support), and individ-

ual and group knowledge-based DM (DM Support).

AI Support is required to support these capabilities.

To ensure accountability for the organization

and its stakeholders, the Guarantees features Trans-

parency, Ethics Control, and Legal Control (shown

collapsed), play crucial roles by activating mecha-

nisms that provide full real-time information, promote

regulatory compliance within the domain and the or-

ganization, and uphold ethical standards. These func-

tionalities require the activation of the KM Support

feature, which allows the system to manage the eth-

ical and legal rules modeled and perform inference

tasks effectively.

For domain-speciﬁc Smart Governance Services,

it is mandatory to establish if one or various E-

Collaboration Model (the root for Level 2). Its car-

dinality in a range 1..n, increases the variability, al-

lowing the deﬁnition of a set of models, and for each

of them, different types of processes. That is, dif-

ferent services depending on the models and types of

e-collaboration processes that they implement. Fur-

thermore, Systems Integration is also are mandatory.

Within this group, the feature Organizational Services

(shown collapsed) is mandatory, as the system must

interoperate with other existing information systems

in the organization (Nastjuk et al., 2022; Pereira et al.,

2018) such as E-Census for controlling user access, or

secure E-Voting services.

To address comprehensive smart governance

throughout the entire public policy cycle (Valle-Cruz

et al., 2020), speciﬁc services for collaboration in

urban processes and projects are desirable. This is

achieved by activating the Urban Process Services

and Urban Project Services, with the latter being es-

sential for collaboration within the former. Urban

processes typically progress to the implementation

phase, where ideas are developed into proposed so-

lutions

4.2 Feature Model - Level 2

Level 2 represents the e-collaboration and smart gov-

ernance models. In Figure 3, the variability of these

models is represented. Except for the selection of

at least a E-Collaboration Process to be included in

the model, all features are optional. They enhance

the capabilities of the core model established in E-

Collaboration Common Services from other levels.

The Project-oriented feature enables e-

collaboration at project, and project phase level

(required if Urban Project Services are activated).

Networked-Processes allows the creation of more

complex collaborative processes based on simpler

ones, or to relate processes with each other, forming

a network to interoperate between them.

The Multi-Level feature enables collaboration

across the organization’s DM levels (strategic, tac-

tical, and operational). Together with the Multi-

Organization feature, it supports complex, cross-

organizational processes to address shared problems,

requiring the activation of the Multi-Tenant feature.

An urban process generally advances projects during

the implementation phase, transitioning ideas into proposed

solutions (Tran et al., 2019).

Product Line Engineering in Smart Governance Systems

267

Figure 2: General FM for smart governance (level 3 & 4).

Figure 3: FM for e-collaboration model (level 2).

The Agile feature introduces feedback between

the different phases. Together with the previous one

(Multi-Level), they facilitate dynamic collaboration

throughout the public policy cycle.

The Data-Driven feature requires the activation

of DA Support to obtain relevant data. Enhanced by

the inﬂuence of the expertise and qualiﬁcation of the

collaborators (Qualiﬁed feature), in the DM process,

it allows knowledge to be promoted in the ﬁnal re-

sults of the decisions. Both features contribute to a

citizen-centric government. Finally, Smart Assistance

is envisaged, to support informed and effective DM,

through the activation of (AI Support) and DM tech-

niques (DM Support).

4.3 Feature Model - Level 1

Level 1 is used to represent the processes. Note that

only the feature group Process Speciﬁc is mandatory,

but not the process type, so that if no type is selected,

the process will have a default operation for a general

collaborative process.

Thus, the remaining features are optional: Deci-

sion Stage enable DM by the collaborators (DM sup-

port would be required) that can be Mandatory, Bind-

ing for the organization or a Weighted decision (re-

quired if Qualiﬁed is selected). Furthermore, several

decision types are considered: Simple choice, Multi-

ple choices, or Multiple Ordered.

Regarding the decisions, optionally a Assessment

Stage can be enabled to evaluate them by collabora-

tors, and consequently being necessary for the gen-

eral Assessment service activation. For effective in-

dividual and collective DM, it is necessary the DM

Support feature activation. Notiﬁcation feature fos-

ters transparency in DM, and can be reinforced with

Debate and Meeting features, above all in deliberative

processes.

About process types, according to the e-

collaboration process group exposed in Subsection

3.2, the following participatory and collaborative pro-

cesses are contemplated in the smart governance do-

main: Urban Process and Urban Project, Propos-

als (policies, normative or other nature), Deliberation

about public affairs, Instrument Making such as poli-

cies, normative or participatory budgets, collaborative

Solution Implementation, and several types of Con-

sultations: surveys, forums, or referendums. Further-

more, through a Assessment Process of outcomes or

decisions can be evaluated by citizenship, improving

accountability.

Regarding mandatory Process Speciﬁc features, it

is required to indicate the Creation Mode of the pro-

cesses: by organization (coordinators), collaborator,

smart agent, or event (e.g., a date established in the

agenda or the ending of a related process). Access

mode to processes must be set: open (to anyone inter-

ested), restricted or by invitation, by census, or col-

laborator type/role. Finally, some optional features

can be selected by characterizing the participation

ICSOFT 2025 - 20th International Conference on Software Technologies

268

mode: through delegation (ad hoc in a certain pro-

cess) or representation (permanent) to another collab-

orator, and if anonymous participation is established

(e.g., for matters protected by data protection laws).

Figure 4 represents the variability of e-

collaboration processes through features regarding

the collaborative process group described in the

previous subsection. To ease readability, Process

Speciﬁc sub-tree has been illustrated in Figure 5.

5 CONFIGURATION TOOL

To assess the usability of the conﬁguration solution,

a software tool has been built. In this section, we de-

scribe this tool used in an industrial environment. The

ﬁrst approach for the SGPL and its FM has been sup-

ported by a conﬁgurator (called InGoverkno) that al-

lows feature selection from an existing e-governance

platform (i.e., Decidim software). Then, the tool con-

ﬁgures and deploys an instance of such a platform

considering the customization without the need of

coding.

The conﬁgurator has been designed based on prior

general FM, and regarding coverage features. Previ-

ously to the conﬁgurator implementation, a reduced

FM has been made, removing those features that are

not present in Decidim. This means that any updates

to the FM would currently require manual software

updates. To this end, both FMs have been speciﬁed

in Universal Variability Language (UVL) to facilitate

their processing and sharing (Benavides et al., 2024).

They have been uploaded to UVLHub, a repository

of FMs in UVL format (Romero-Organvidez et al.,

2024)

The conﬁgurator starts from this reduced FM,

which is mapped to the speciﬁc implementation of

the Decidim platform, to generate the common arti-

facts for subsequent conﬁguration and deployment.

Through a web user interface (UI) an user may se-

lect features and common artifacts. Those features

come from the deﬁned FM according to: (1) the e-

governance software selected (the Decidim in this ﬁrst

approach), and (2) a subset of processes that this soft-

ware may manage. Then, the tool would use the val-

ues of this form to generate the necessary artifacts and

resources and obtain the SPL-derived product, i.e., an

instance of the Decidim platform. The deployed plat-

form contains the necessary and customized services

according to the general FM as described in the previ-

ous sections. We note that the conﬁgurator also acts at

SGPL Datasets are available at http://uvlhub.io/doi/10.

5281/zenodo.12697539 and http://uvlhub.io/doi/10.5281/

zenodo.15123479

level 4, allowing customizing features for the speciﬁc

organization such as the organization name or institu-

tional image

Regarding its implementation, the tool has been

developed as a web application using the Ruby pro-

gramming language mainly in addition to JavaScript.

To manage the conﬁguration and deployment of De-

cidim instances, container technology has been used

through the Docker tool to facilitate compatibility,

scalability and efﬁciency. The source code of the con-

ﬁgurator is available in the GitHub platform

6 CASE STUDY

To assess the usefulness of our conﬁgurability and

conﬁgurator proposals, the conﬁgurator was applied

to three municipalities as a case study. To this end,

we previously designed a survey to get information

about e-governance needs and participative processes

demanded. The survey was launched to experts and

practitioners in e-governance and related areas such

as citizen participation from city councils of different

sizes (from the Seville province in Spain). Since the

variability in the ﬁeld of e-governance depends along

with other variables, on the size of the city’s popula-

tion, this was the main criterion of choice.

The analysis of the survey responses gave us in-

sight into the e-governance features desired by these

municipalities. Then we selected three with different

needs and population sizes (one with less than 10,000,

another with between 15,000 and 20,000, and another

with more than 40,000 inhabitants). This diversity

would broaden the variability and make the case study

closer to a real environment. So, the smallest mu-

nicipality demanded citizen (political) proposals, sur-

veys, meetings and debates. The medium-sized one

was interested in general proposals, surveys and de-

bates. Lastly, the largest municipality demanded more

complex participatory processes such as participatory

budgeting and citizen forums. Finally, once the con-

ﬁgurator was deployed in a software platform, we

applied it to these municipalities, using the features

identiﬁed through the survey.

6.1 Results

The results supported the survey responses regarding

existing variability and conﬁrmed the demand for this

type of conﬁguration and automation solutions.

A video of this process can be seen at https://doi.org/

10.5281/zenodo.15119490

https://github.com/diverso-lab/ingoverkno

Product Line Engineering in Smart Governance Systems

269

Figure 4: FM for e-collaboration process (level 1).

Figure 5: FM for speciﬁc features of e-collaboration pro-

cess (level 1).

The design and development processes of the con-

ﬁgurator, revealed that the user interface (UI), the

logic of the conﬁguration process, and part of the

code, are common across any platform within the im-

plementation domain. This commonality improves

usability, ensures a similar conﬁguration process re-

gardless of the platform, and promotes the reuse of

code and resources.

In relation to the deployment and conﬁguration

process, the main insight drawn has been that opera-

tional instances of the e-governance platform tailored

to users’ needs can be easily obtained in an automated

manner. One of the reasons is that the conﬁguration

was carried out in the smart governance domain rather

than in the application domain, and the concepts are

therefore closer to the end user. Thus, the speciﬁc

details of the Decidim platform and its conﬁguration

are hidden. This means that users need less technical

support.

7 DISCUSSION

Concerning other related proposals, ours focuses on

the speciﬁc problem of conﬁgurability from a gen-

eral perspective by providing several complementary

methods and techniques integrated into the solution:

SPLE, multi-level conﬁguration architecture and FM.

In addition, we have provided an SGPL-based conﬁg-

uration tool to conﬁgure and deploy an instance of the

Decidim platform from the FM.

The results obtained from the deployment and

conﬁguration of Decidim in three real and different

local councils show that the SGPL, together with the

FM, contributes to the general area of e-collaboration

and in particular of smart governance. It facilitates

the deployment and conﬁguration of these systems,

also favoring their reusability, and adaptability with

respect to the particular and varying needs of the dif-

ferent stakeholders and organizations.

Nevertheless, due to the complexity and recent-

ness of smart governance, we recognize some chal-

lenges and limitations must be taken into account.

Thus, although different studies on smart governance

identify the same or similar features, there is no to-

tal consensus on this and other features could be con-

sidered. Furthermore, there are no SPLs or FMs that

can be taken as a reference. In addition, a system-

atic study of the features of the already existing so-

lutions has not been conducted. The vast majority of

solutions are limited to e-governance (democratic or

participatory) and do not consider smart governance.

Therefore we have opted for a domain analysis. Con-

sequently, a complementary work could be applica-

tion engineering to extract common features and in-

corporate them into the domain.

8 CONCLUSIONS AND FUTURE

WORK

The proposed SGPL addresses a gap in both the

literature and industry by offering a general smart

governance product line. Its goal is to serve as a

guide for practitioners, simplifying the implementa-

tion and conﬁguration process within a single soft-

ware ecosystem. Leveraging reuse and conﬁguration

provides signiﬁcant dynamic variability in services

for e-collaboration, particularly in the smart gover-

nance domain. This approach is adaptable to the di-

verse needs of various stakeholders and types of insti-

tutions.

To achieve this, the SGPL is built upon appropri-

ate methods and techniques, such as a multi-level con-

ﬁguration architecture. At the ﬁrst level, it manages

the variability of collaborative processes; at the sec-

ond level, it handles the different governance models.

At the third and fourth levels, the overall smart gover-

nance platform is conﬁgured and tailored to meet the

ICSOFT 2025 - 20th International Conference on Software Technologies

270

speciﬁc needs of the institution utilizing the platform.

To represent the SGPL, the FM provided includes

general e-collaboration features and those that are key

to smart governance along with their variability, mak-

ing it easy to obtain a customized line of services for

different organizations and stakeholders.

We envision, as a future work, an evolution to-

wards a general conﬁguration and deployment frame-

work to address the automation of the conﬁguration

and deployment of the various existing (or newly de-

veloped) smart and governance platforms, compo-

nents and tools. The conﬁgurator would choose the

most suitable common and speciﬁc implementation

artifacts to deploy a comprehensive smart governance

platform, adapted to the particularities of the institu-

tions and their stakeholders. For this purpose, as a

next step, the SGPL-based conﬁgurator could be en-

hanced considering several existing platforms such as

Decidim, Consul or other similar ones. This general

conﬁgurator would be applied to a wider set of mu-

nicipalities to validate its usefulness.

Finally, the availability of SGPL-based conﬁgu-

rators would allow us to deﬁne quantitative perfor-

mance metrics (e.g., related to time spent on con-

ﬁguration processes, their complexity or accuracy).

Those metrics would be used to empirically assess

the efﬁciency and effectiveness of our solution com-

pared to other approaches and proposals for managing

conﬁgurability. So that we could assess the improve-

ments that SGPL provides along with its methods and

techniques.

ACKNOWLEDGEMENTS

This work was supported by the VII Own Plan

research aid from the University of Seville,

FEDER/Ministry of Science, Innovation and

Universities/Junta de Andaluc

ıa/State Research

Agency/CDTI with the following grants: Data-pl

(PID2022-138486OB-I00), TASOVA PLUS re-

search network (RED2022-134337-T) and AquaIA

(GOPG-SE-23-0011)

REFERENCES

Abdullah, M., Almalki, M., and Blahmer, H. (2019). E-

collaboration knowledge management. International

Journal of Advanced Trends in Computer Science and

Engineering, 8(1):259–265.

Abu-Shanab, E. A. (2015). Reengineering the open gov-

ernment concept: An empirical support for a pro-

posed model. Government Information Quarterly,

32(4):453–463.

Achour, I., Labed, L., Helali, R., and Ghazela, H. B.

(2011). A Service Oriented Product Line Architecture

for E-Government. In The 2011 International Con-

ference on e-Learning, e-Business, Enterprise Infor-

mation Systems, and e-Government, EEE 2011, pages

186 – 192, Athens. The Steering Committee of The

World Congress in Computer Science, Computer En-

gineering and Applied Computing (WorldComp).

Benavides, D., Sundermann, C., Feichtinger, K., Galindo,

J. A., Rabiser, R., and Th

um, T. (2024). Uvl: Feature

Modelling with the Universal Variability Language.

Journal of Systems and Software.

Bherer, L., Dufour, P., and Montambeault, F. (2016). The

participatory democracy turn: an introduction. Jour-

nal of Civil Society, 12(3):225–230.

Boudjelida, A., Mellouli, S., and Lee, J. (2016). Electronic

citizens participation: Systematic review. In Pro-

ceedings of the 9th International Conference on The-

ory and Practice of Electronic Governance (ICEGOV

’15-16), volume 01-03-Marc, pages 31–39, Montev-

ideo, Uruguay. Association for Computing Machinery

(ACM).

Bouzguenda, I., Alalouch, C., and Fava, N. (2019). To-

wards smart sustainable cities: A review of the role

digital citizen participation could play in advancing

social sustainability.

Clark, T., Frank, U., Reinhartz-Berger, I., and Sturm, A.

(2017). A multi-level approach for supporting conﬁg-

urations: A new perspective on software product line

engineering. In CEUR Workshop Proceedings, pages

170–178, Graz, Austria. CEUR-WS.

Cledou, G. and Barbosa, L. S. (2017). Modeling families

of public licensing services: A case study. In Pro-

ceedings - 2017 IEEE/ACM 5th International FME

Workshop on Formal Methods in Software Engineer-

ing, FormaliSE 2017, pages 37–43, Buenos Aires, Ar-

gentina. IEEE.

Czarnecki, K., Helsen, S., and Eisenecker, U. (2005).

Staged conﬁguration through specialization and mul-

tilevel conﬁguration of feature models. Software pro-

cess: improvement and practice, 10(2):143–169.

Debnath, N., Felice, L., Montejano, G., and Riesco, D.

(2008). A feature model of E-government systems in-

tegrated with formal speciﬁcations. In Proceedings -

5th International Conference on Information Technol-

ogy: New Generations, ITNG 2008, pages 27–31, Las

Vegas, Nevada.

Deseriis, M. (2023). Reducing the Burden of Decision

in Digital Democracy Applications: A Comparative

Analysis of Six Decision-making Software. Science,

technology, & human values, 48(2):401–427.

Dwivedi, Y. K., Hughes, L., Ismagilova, E., and Aarts, G.

(2021). Artiﬁcial Intelligence (AI): Multidisciplinary

perspectives on emerging challenges, opportunities,

and agenda for research, practice and policy. Interna-

tional Journal of Information Management, 57:1–47.

Ependi, U., Rochim, A. F., and Wibowo, A. (2022). Smart

City Assessment for Sustainable City Development on

Smart Governance: A Systematic Literature Review.

In 2022 International Conference on Decision Aid

Product Line Engineering in Smart Governance Systems

271

Sciences and Applications, DASA 2022, pages 1088–

1097, Chiang Rai, Thailand. IEEE.

Fajar, A. N. and Shoﬁ, I. M. (2017). Reduced software

complexity for E-Government applications with ZEF

framework. Telkomnika (Telecommunication Comput-

ing Electronics and Control), 15(1):415–420.

Felfernig, A., Falkner, A., and Benavides, D. (2024). Fea-

ture models: AI-driven design, analysis and applica-

tions. SpringerBriefs in computer science (Springer).

Springer, Cham, Switzerland.

Grigalashvili, V. (2022). E-government and E-governance:

Various or Multifarious Concepts. International Jour-

nal of Scientiﬁc and Management Research, 5(1):183–

196.

Hasan, I. and Rizvi, S. (2021). Knowledge Management

Framework for Sustainability and Resilience in Next-

Gen e-Governance. In Data Engineering for Smart

Systems, volume 238 of Lecture Notes in Networks

and Systems, pages 191–203. Springer Singapore Pte.

Limited, Singapore.

Hong, S. G. and Lee, D. (2023). Development of a citizen

participation public service innovation model based

on smart governance. Service Business, 17:669–694.

Ju, J., Liu, L., and Feng, Y. (2018). Citizen-centered big

data analysis-driven governance intelligence frame-

work for smart cities. Telecommunications Policy,

42(10):881–896.

Lima, V. M. A. D., Marcacini, R. M., Lima, M. H. P.,

Cagnin, M. I., and Turine, M. A. S. (2014). A genera-

tion environment for front-end layer in e-government

content management systems. In Proceedings - 9th

Latin American Web Congress, LA-WEB 2014, pages

119–123, Minas Gerais, Brazil. IEEE.

Munkvold, B. E. and Zigurs, I. (2005). Integration

of e-collaboration technologies: Research opportu-

nities and challenges. International Journal of e-

Collaboration (IJeC), 1(2):1–24.

Nastjuk, I., Trang, S., and Papageorgiou, E. I. (2022). Smart

cities and smart governance models for future cities:

Current research and future directions. Electronic

Markets, 32(4):1917–1924.

Ojo, A., Janowski, T., and Estevez, E. (2007). A Composite

Domain Framework for Developing Electronic Pub-

lic Services. Technical report, United Nations (UN),

Tokyo, Japan.

Parycek, P. and Viale, G. (2017). Drivers of smart gov-

ernance: Towards to evidence-based policy-making.

In ACM International Conference Proceeding Series,

DG.O 2017, volume Part F1282, pages 564–565, New

York, USA. Proceedings of DG.O.

Pereira, G. V., Parycek, P., Falco, E., and Kleinhans,

R. (2018). Smart governance in the context of

smart cities: A literature review. Information Polity,

23(2):143–162.

Pohl, K., B

ockle, G., and Van Der Linden, F. (2005). Soft-

ware product line engineering: Foundations, princi-

ples, and techniques. Springer Berlin / Heidelberg,

Berlin, Heidelberg.

Purba, F. N. and Arman, A. A. (2022). A Systematic Liter-

ature Review of Smart Governance. In 2022 Interna-

tional Conference on Information Technology Systems

and Innovation, ICITSI 2022 - Proceedings, pages

70–75, Bandung, Indonesia. Institute of Electrical and

Electronics Engineers Inc.

Reiser, M. O. and Weber, M. (2007). Multi-level fea-

ture trees: A pragmatic approach to managing highly

complex product families. Requirements Engineering,

12(2):57–75.

Romero-Organvidez, D., Galindo, J. A., Sundermann, C.,

Horcas, J.-M., and Benavides, D. (2024). UVLHub:

A feature model data repository using UVL and open

science principles. The Journal of systems and soft-

ware, 216:112150.

Ruijer, E., Van Twist, A., Haaker, T., Tartarin, T., Schu-

urman, N., Melenhorst, M., and Meijer, A. (2023).

Smart Governance Toolbox: A Systematic Literature

Review. Smart Cities, 6(2):878–896.

Tiwari, A. A., Gupta, S., Zamani, E. D., Mittal, N., and

Agarwal, R. (2023). An Overarching Conceptual

Framework for ICT-enabled Responsive Governance.

Information Systems Frontiers, 1:3.

Tran, T. H. G., Dupont, L., and Camargo, M. (2019). Ap-

plication of Decision-Making Methods in Smart City

Projects: A Systematic Literature Review. Smart

Cities, 2(3):433–452.

Valle-Cruz, D., Criado, J. I., Sandoval-Almaz

an, R., and

Ruvalcaba-Gomez, E. A. (2020). Assessing the pub-

lic policy-cycle framework in the age of artiﬁcial in-

telligence: From agenda-setting to policy evaluation.

Government Information Quarterly, 37(4):101509.

Wirtz, B. W. and M

uller, W. M. (2023). An Integrative Col-

laborative Ecosystem for Smart Cities - A Framework

for Organizational Governance. International journal

of public administration, 46(7):499–518.

Yusuf, M., Alamsyah, N., Syarif, M., Muntasa, A., and

Muzakki, H. (2019). A novel framework of e-

participation for smart cities. Bulletin of Social In-

formatics Theory and Application, 3(2).

ICSOFT 2025 - 20th International Conference on Software Technologies

272