Challenges in Requirements Engineering and Its Solutions: A Systematic

Review

avio da Cruz Mello and Lisandra Manzoni Fontoura

Programa de P

os-Graduac¸

ao em Ci

encia da Computac¸

ao, Federal University of Santa Maria (UFSM), Brazil

Keywords:

Requirements Engineering (RE), Software Engineering (SE), Challenges, Solutions, Systematic Literature

Review (SLR).

Abstract:

Software development projects are susceptible to many adversities throughout their life cycle that can be orig-

inated, among several reasons, of a low-quality speciﬁcation and management of requirements. To ensure the

Requirements Engineering activities are conducted correctly, researchers study and apply various techniques

to predict and avoid the negative impacts that may occur in projects. The main goal of this research is to

identify which techniques have been used to solve problems related to requirements management in software

projects. We retrieved and reviewed studies published across various scientiﬁc databases to answer research

questions that were previously deﬁned. From this work, it was possible to obtain a better understanding of the

most common problems in the Requirements Engineering ﬁeld, as well as some techniques that currently exist

to solve them. We also identiﬁed that Artiﬁcial Intelligence has been widely explored to improve the activities

of the ﬁeld.

1 INTRODUCTION

Historically, it is usual for projects focused on soft-

ware development to face several problems through-

out their life cycle. Even after years of advances in

project management practices and new methodolo-

gies, about 31% of projects tend to be discontinued,

and 52% extrapolate ﬁnancial resources, have deliv-

eries after the agreed deadlines, and/or do not satisfy

the needs that were promised to the customer, accord-

ing to the 2019 CHAOS report (Sommerville, 2011)

(Standish Group International, 2019).

Most of these problems can be originated from the

way requirements are managed. Some of the essen-

tial criteria for the success of a project are the correct

deﬁnition and speciﬁcation of requirements in its ini-

tial stages, as well as good communication between

stakeholders to ensure that everyone has a clear un-

derstanding and is satisﬁed with the requirements that

have been speciﬁed (Iriarte and Bayona, 2020).

However, each project has different characteris-

tics, and other types of unpredictable obstacles can

generate negative impacts. To minimize possible

risks, it is necessary to understand the requirements

engineering process and activities and apply appro-

priate techniques.

Thereby, the present study aims to list the main

challenges existing today in the Requirements Engi-

neering ﬁeld, from a study in literature, and explore

current approaches to solve them. From this analysis,

we seek to identify solutions that have not yet been

explored by researchers for possible future projects.

We organized the paper as follows. In Section 2,

we discuss some related research found in the litera-

ture. In Section 3, we describe the methodology and

procedures for carrying out this study. We analyze

and discuss the initial results obtained in Section 4.

Finally, we present the conclusion and future works

in Section 5.

2 RELATED WORK

Numerous researchers have described systematic re-

views in the Requirements Engineering literature,

mapping some of the main problems found in stud-

ies in the ﬁeld.

Alam et al. (2017) performed a systematic review

based on the guidelines suggested by Kitchenham et

al. (2007), which consists of six basic activities: plan-

ning, identiﬁcation of keywords, inclusion and exclu-

sion, the deﬁnition of research questions, analysis and

review, and description of results. The researchers an-

alyzed studies related to Requirements Engineering in

Mello, O. and Fontoura, L.

Challenges in Requirements Engineering and Its Solutions: A Systematic Review.

DOI: 10.5220/0011079000003179

In Proceedings of the 24th International Conference on Enterprise Information Systems (ICEIS 2022) - Volume 2, pages 70-77

ISBN: 978-989-758-569-2; ISSN: 2184-4992

agile methodologies, seeking to answer what were the

practices adopted in agile projects and what were the

challenges and limitations in this same context.

Sch

on et al. (2017) followed the same guide-

lines as the previous study for their literature review.

Studies related to Requirements Engineering that also

addressed user and stakeholder involvement in agile

methodologies were selected. The researchers inves-

tigated what approaches exist to continuously involve

the stakeholder during the requirements engineering

process, what methodologies are used to present the

user perspective to the stakeholder, and how require-

ments management occurs.

Finally, Elghariani and Kama (2016), like the pre-

vious researchers, also followed the guidelines of

Kitcheham et al. (2007) to carry out a systematic re-

view of the literature related to practices and problems

found in the Requirements Engineering ﬁeld. Eighty

studies were analyzed and twenty were selected. The

researchers identiﬁed sixteen fundamental practices

and six common problems of the ﬁeld.

This project focuses on carrying out a systematic

review of the literature to identify the main challenges

of the Requirements Engineering ﬁeld and also iden-

tifying existing solutions for such problems. In ad-

dition, it is also expected to determine which chal-

lenges still do not have an adequate solution, so fu-

ture new research is developed to further improve the

requirements speciﬁcation and management activities

and, this way, avoid negative impacts on a project.

3 PROCEDURES FOR THE

REVIEW

To carry out the systematic literature review, we fol-

lowed the guidelines and the process proposed by

(Kitchenham and Charters, 2007). This process con-

sists of three main phases, which are illustrated in Fig-

ure 1. Each step has recommended procedures and

activities that must be followed so that the research

can produce adequate and meaningful results.

3.1 Planning

The initial planning phase involves the preparation ac-

tivities for carrying out the review. In planning, we

deﬁne the goal we want to achieve with the research

and the steps to reach such a goal. The steps are:

Deﬁne the Need for the Review. The need for this

research originated from the problems usually faced

by professionals in Requirements Engineering activ-

ities. In most cases, software development projects

Figure 1: Systematic Literature Review Process proposed

by Kitchenham et al.

end up being severely affected by these problems,

having impacts of great magnitude on previously esti-

mated deadlines and costs. Thus, this study is focused

on understanding what are the most frequent difﬁcul-

ties in this area and also what solutions currently exist

to prevent, or at least reduce, these difﬁculties.

Deﬁne Research Questions. Research questions help

the researcher focus his study, answering speciﬁc

questions deﬁned by him to achieve his goal. For the

present work, we deﬁned three research questions to

be answered.

• Q

: What are the most frequently faced chal-

lenges in Requirements Engineering currently?

• Q

: What approaches are proposed to prevent or

reduce such difﬁculties?

• Q

: What opportunities exist for future research

according to the current scenario?

To answer research question Q

, we analyzed

works published between 2015 and 2021 reporting

problems experienced in Requirements Engineering

during the development of software projects. After

mapping these problems, we identiﬁed in the litera-

ture techniques and approaches that currently exist to

Challenges in Requirements Engineering and Its Solutions: A Systematic Review

solve them to answer the research question Q

Finally,

we answered question Q

based on the two previous

research questions. We analyze which areas are open,

or can still be explored, to improve activities of the

ﬁeld and reduce risks.

3.2 Conducting

From the deﬁnition of the review protocol, which es-

tablishes methods and procedures that should be taken

when selecting and analyzing studies, it is possible to

start the execution of the activities that were planned

in the previous phase. Conducting the systematic lit-

erature review is composed of the following steps:

Deﬁne Search Terms. The studies were retrieved

from the IEEE Xplore, ACM Digital Library, and Sci-

enceDirect scientiﬁc databases. To ﬁlter the most rel-

evant publications to the present study, we deﬁned

search terms that take into account what was needed

to answer the research questions. The result was the

following search string:

(”requirements engineering” OR ”requirements

management” OR ”requirements elicitation” OR ”re-

quirements analysis” OR ”requirements speciﬁca-

tion” OR ”requirements validation” OR ”software

requirements”) AND (”challenges” OR ”issues”

OR” obstacles” OR ”difﬁculties” OR ”burden” OR

”problems” OR ”complications” OR ”hardship”)

Therefore, the string limits the search to publi-

cations that address speciﬁc problems related to the

main activities of Requirements Engineering, as well

as the ﬁeld as a whole.

Deﬁne Inclusion and Exclusion Criteria. Inclusion

and exclusion criteria ﬁlter publications that will be

relevant during the review process. This search used

the following criteria:

Inclusion Criteria.

• Studies published between 2015 and 2021.

• The text must be available in full.

• The studies are peer-reviewed publications.

• The studies are relevant to the deﬁned search

terms.

• The studies describe the challenges related to Re-

quirements Engineering experienced during the

development of a software project.

Exclusion Criteria.

• Studies that are not focused on or do not discuss

requirements engineering topics.

• Studies that describe neither challenges nor solu-

tions experienced in Requirements Engineering.

• Publications like Keynotes, White Papers, or

works consisting only of the abstract.

• Publications that are not in English.

• Studies that do not meet any inclusion criteria that

have been deﬁned.

• Duplicate studies.

Study Selection. The ﬁrst selection was performed

by reading the title and abstract of the publications

to understand if they were aligned with the proposed

theme. The ﬁrst selection resulted in 152 results.

After the initial selection, a more detailed analysis

was carried out in two stages, in which the publica-

tions were read and the inclusion and exclusion crite-

ria were applied. 36 of the 152 publications selected

initially were considered acceptable for review.

The number of publications retrieved from the ex-

ecution of the search string in the three deﬁned bases

and the number of ﬁltered studies in each stage can be

seen in Table 1.

Table 1: Selected studies.

Base 1st Sel. 2nd Sel. Final Sel.

IEEE Xplore 32 14 10

ACM Library 49 24 11

ScienceDirect 71 35 15

Total 152 73 36

4 INITIAL RESULTS AND

DISCUSSION

This section focuses on the exposure of the initial re-

sults obtained from the systematic review of the liter-

ature we performed. To answer the deﬁned research

questions, we identiﬁed in the selected studies the

main challenges faced in Requirements Engineering

activities and then classiﬁed each problem into related

categories. In the end, we calculated the frequency of

each challenge based on its occurrences in studies.

The mapping of the most frequent challenges is

available in Table 2. It helps to answer the ﬁrst re-

search question Q

. It is possible to see that the

lack of communication between the stakeholders is

the most cited challenge in Requirements Engineering

among the selected publications, occurring in about

56% of the cases.

Also with considerably high percentages are am-

biguous, incomplete, inconsistent, or incorrect re-

quirements with 33%, 31%, 22%, and 19% of oc-

currence, respectively. Another frequent problem, the

ICEIS 2022 - 24th International Conference on Enterprise Information Systems

Table 2: The main Requirements Engineering challenges

according to the analysis of selected publications.

Challenge Frequency %

(Total: 36)

Lack of communication 20 56

and/or between stakeholders

Ambiguous requirements 12 33

Incomplete requirements 11 31

Documentation 10 28

Requirements traceability 9 25

Inconsistent requirements 8 22

Insufﬁcient knowledge about 8 22

the domain

Requirements priorization 8 22

Incorrect requirements 7 19

Volatile requirements 7 19

Technical difﬁculties 5 14

maintenance of documentation, appears next, cited in

28% of the studies as a challenge.

In addition to those already mentioned, other ad-

versities tend to generate complications to the project,

such as insufﬁcient knowledge about the application

domain (22%), prioritization of requirements (22%),

and changing requirements (19%). A less frequent

challenge is general technical difﬁculties, which gen-

erally involve a lack of techniques and tools to man-

age requirements, with only 14% of occurrences.

For each problem identiﬁed, we searched the liter-

ature for some of the most current and effective tech-

niques that exist to try to solve them and thus answer

the research question Q

: Lack of Communication and/or Involvment

between the Stakeholders

Although communication and engagement depend on

stakeholder availability, there are some traditional

methodologies to improve team communication as

the MUST method, Joint Application Design (JAD),

User-Led Requirements Construction (ULRC), and

Soft Systems Methodology (SSM). These are focused

on deﬁning best practices for requirements elicitation

(Coughlan and Macredie, 2002).

Currently, many publications propose innovative

ways to make this communication happen effectively

in projects of different characteristics. In global soft-

ware development projects, for example, the lack

of communication between the parties involved is a

common problem, as the team and the client are dis-

tributed in many places around the planet. Thus,

Nadeem and Lee (2019) proposed a framework based

on Case-Based Reasoning (CBR) for knowledge stor-

age and elicitation techniques to improve the dialogue

between stakeholders and thus deﬁne best practices

for requirements gathering in the context of global

software development.

In a similar study, Shahzad et al. (2021) dis-

cuss the necessary changes for requirements elicita-

tion tasks in the context of the global COVID-19 pan-

demic and propose the use of blockchain-based tech-

nologies to improve various challenges related to the

Requirements Engineering ﬁeld, including commu-

nication between stakeholders, conﬂicts in decision-

making, negotiations, among others.

: Ambiguous Requirements

The detection of ambiguous requirements is currently

highly explored by researchers, who mainly use in-

telligent algorithms to correct already speciﬁed docu-

ments. Among the speciﬁc ﬁelds of Artiﬁcial Intelli-

gence, Machine Learning is perhaps one of the most

frequent in works related to requirements ambiguities

reduction.

Osman and Zaharin (2018) proposed a hybrid Text

Mining and Machine Learning approach to detect

and classify ambiguities in a dataset extracted from

Malaysian speciﬁcation documents. From the ex-

traction of information in the mining stage, the sys-

tem progressively learns to detect ambiguous require-

ments.

Also focused on Machine Learning, Sharma et

al. (2016) proposed an approach for detecting harm-

ful pronomial anaphor ambiguities in requirements

speciﬁcation documents. The researchers used clas-

siﬁcation algorithms to classify harmful and non-

harmful ambiguities. The chosen classiﬁer was able

to correctly detect 95% of harmful ambiguous re-

quirements.

: Incomplete Requirements

As well as ambiguous requirements, incomplete re-

quirements are another common challenge in require-

ments engineering and can be caused by the precar-

ious deﬁnition of requirements in the speciﬁcation

stage. It is often difﬁcult to identify requirements in-

completeness before functionality is developed, espe-

cially when documents are not periodically reviewed.

Thus, it is interesting to use tools that help the profes-

sional in this task.

DeVries and Cheng (2017) proposed the Ares-C

approach for detecting incomplete requirements, us-

ing symbolic analysis and evolutionary computation

to analyze hierarchical requirements models. This

approach was applied to a real system, and the re-

searchers were able to notice that Ares-EC was able

to automatically detect incomplete requirements and

generate completeness counter-examples.

Kalinowski et al. (2016) conducted a question-

naire in 88 small, medium, and large-sized Brazilian

and Austrian organizations to deﬁne the most com-

Challenges in Requirements Engineering and Its Solutions: A Systematic Review

mon causes for incomplete or hidden requirements in

a project. Focusing on these causes, the researchers

discussed mitigation practices and actions based on

the responses obtained from the respondents.

: Documentation

Having documentation that is not accessible to all par-

ties involved, that is not managed properly, or that

simply does not exist are major risk factors for a

project. It is necessary to have good practices deﬁned

to document and manage the documents to ensure the

software product quality.

Salvador and Santos (2016) presented the Do-

MaR application in their study, which is focused

on preventing problems related to requirements man-

agement and documentation by mapping these prob-

lems through questionnaires carried out with ex-

perts. Based on the experts feedback, their application

showed potential to solve the challenges frequently

mentioned in the questionnaires answered.

Behutiye et al. (2017) interviewed profession-

als from four different companies and presented

their ﬁndings regarding the documentation practices

adopted by organizations. Based on their analysis, the

researchers proposed guidelines for documenting re-

quirements in agile software development, in which

documents are generally not given priority.

: Requirements Traceability

Traceability is the ability to follow a requirement

from its origin and speciﬁcation through its devel-

opment and implementation. Ensuring traceability

throughout the software development cycle is a com-

plex task, but it helps maintain and control project re-

quirements.

To propose a solution to the traceability prob-

lem, Haidrar et al. (2018) developed a language for

requirements speciﬁcation named ReqDL. The lan-

guage has speciﬁc operators that reveal the explicit

and implicit links between requirements and artifacts.

From the use of ReqDL expressions, the traceability

task becomes less complex and more understandable.

Garcia and Paiva (2016) presented a tool to help

the user keep the traceability information of the re-

quirements updated, without errors. By using the de-

veloped tool, the researchers demonstrated that cre-

ating a link between requirements and implementa-

tion artifacts was more effective in maintenance than

traceability matrices, which are standard documents

to show the relationship between artifacts in Require-

ments Engineering.

: Inconsistent Requirements

Consistency between requirements is yet another cru-

cial aspect to maintain the quality of software product

development. The inconsistency goes unnoticed most

of the time and can cause damages to the project, such

as rework and wasted effort, generating an increase in

time and resources needed for a project to be com-

pleted.

To address requirements inconsistencies in the

context of global software development, Gull et

al. (2021) introduced the BOMO framework based

on blockchain technology and Model-Driven Soft-

ware Engineering (MDSE) methodology. This union

makes it possible to manage inconsistent require-

ments effectively and easily. The results of the case

study applying the framework proved to be positive,

helping the global software company to deal with the

inconsistency.

Mezghani et al. (2018) used the unsupervised Ma-

chine Learning algorithm, k-means, to identify re-

dundancy and inconsistency in requirements. The k-

means algorithm groups the data around centroids, ac-

cording to a deﬁned value of groups named ”k”. The

algorithm was validated using a real industrial base

containing inconsistent data and using a number de-

ﬁned by a Requirements Engineer who assisted in the

research as the value of ”k”. The experiments gen-

erated positive initial results in detecting inconsistent

and redundant requirements.

: Insufﬁcient Knowledge about the Application

Domain

An RE specialist who does not know the application

domain tends to fail to perform requirements speciﬁ-

cation and management activities correctly. Under-

standing the domain of an application to be devel-

oped is a task that demands time and dedication from

the professional involved in Requirement Engineering

activities and, for this reason, applying means of ob-

taining this knowledge in a simple and fast way in the

project can bring beneﬁcial results.

Li et al. (2020) proposed an automated way of ex-

tracting domain knowledge from requirements docu-

ments to help professionals. The tool represents the

documents in natural language as a vector using the

Doc2Vec algorithm and then applies clustering algo-

rithms to create the initial cluster feature tree. Ex-

tracted results containing the most important words

and phrases are returned to the user for analysis.

: Requirements Priorization

Prioritizing requirements is a task that plays a big role

during the implementation phase. Prioritizing fea-

tures is a highly complex process and takes into ac-

count aspects such as the team’s ability to develop at

the moment and the customer’s urgency for the re-

quirement.

There are currently several ways to perform re-

ICEIS 2022 - 24th International Conference on Enterprise Information Systems

quirements prioritization, some known to produce

better results. However, there is no consensus on

which is the best technique to prioritize require-

ments, since it all depends on the characteristics of the

project, the team, and the product. Studies proposing

new ways to prioritize focusing on usability, scalabil-

ity, and quality assurance are often published.

In this perspective, Mkpojiogu and Hashim (2017)

proposed a quality-based approach to requirements

prioritization, using the Kano model. Requirements

are prioritized according to attributes that take into ac-

count the stakeholders’ point of view on quality. The

result showed that the model generated consistent re-

sults and that it can serve as a good option for require-

ments prioritization.

Yaseen et al. (2020) suggested a Spanning Tree-

based approach for prioritizing requirements from the

developer’s perspective. The approach was validated

using a set of requirements from the ERP ODOO soft-

ware, which were assigned to four developers to have

dependencies. The researchers then performed the

time estimates for the prioritized and non-prioritized

requirements to assess the impact on the total project

estimate time. A signiﬁcant difference was noticed

between the time estimates of prioritized and non-

prioritized requirements, which demonstrated the im-

portance of prioritizing requirements.

: Incorrect Requirements

Incorrect requirements are generally wrong deﬁni-

tions of functionalities required by the customer, or

deﬁnitions that conﬂict with previously speciﬁcation

documents about the product. It can happen due to

an error of the specialist when preparing the require-

ments speciﬁcation document or due to a misunder-

standing about functionalities. An incorrect require-

ment can lead to a product that does not meet the cus-

tomer’s needs and requires reworking.

To improve the quality of the requirements anal-

ysis step, Nguyen et al. (2014) developed the GUI-

TAR tool to detect incorrect, incomplete, and incon-

sistent artifacts. The GUITAR approach is based on

domain ontologies and semantics for analysis of re-

quirements. The core of this methodology is the rep-

resentation of activities, which are composed of both

an action and an object (”create review”, for exam-

ple). These activities are related to one another, thus

it is easier to identify incompatibilities between arti-

facts or missing artifacts from these relationships.

: Volatile Requirements

Changes are inevitable within any project. It is im-

possible to deﬁne all that is expected from software at

the beginning of development, so the customer must

make frequent change requests to the team (Som-

merville and Kotonya, 1998). These requests may in-

volve changing an existing requirement in the project

or incorporating new requirements that were not pre-

viously planned.

Thus, to minimize negative impacts throughout

its life cycle, the project must have a well-deﬁned

Change Management Process, which does not present

conﬂicts to its practices.

Ali et al. (2018) proposed a framework for man-

aging requirements changes based on the Case-Based

Reasoning technique in the context of global software

development. With the application of the CBR-based

framework in the cloud, the authors noticed that the

communication and coordination of the global team

during change management, which was previously

challenging, became more effective. All services re-

quired became available for the users at all times on

a single platform without time and space restrictions,

when previously the team used tools on different plat-

forms with different logins, not taking cultural differ-

ences into account.

Naz et al. (2013) deﬁned and described a model

that integrates requirements change management with

the Case-Based Reasoning technique. For evaluation,

the researchers presented their model to experts in the

ﬁeld and asked about the experience gained once the

model was implemented, thus comparing the perfor-

mance differences before and after its use. The results

show that the framework helped to reduce the cost and

time impacts of a change, as well as to resolve re-

quirements conﬂicts and increase customer satisfac-

tion.

Regarding the C

challenge, which refers to tech-

nical difﬁculties, there are no speciﬁc solutions to be

used. Such challenge refers to general problems re-

lated to the organization’s practices, such as the lack

of clearly deﬁned tools and processes for managing

requirements.

Looking at the current scenario, as deﬁned in the

research question Q

, it is clear that researchers are

exploring innovative ways to solve classic require-

ments engineering problems. It is possible to no-

tice that the scientiﬁc community is looking for so-

lutions in other areas, such as Artiﬁcial Intelligence,

which was present in most of the publications stud-

ied, as seen in Figure 2. Many studies also suggested

best practices for Requirements Engineering activi-

ties (28%), while some proposed approaches based on

Blockchain technology (11%). Other types of tech-

niques were not as frequent.

According to (Harman, 2012), the relationship be-

tween AI and Software Engineering areas tends to

generate beneﬁcial results. Software Engineering is a

ﬁeld highly focused on knowledge, but uncertainty is

Challenges in Requirements Engineering and Its Solutions: A Systematic Review

Figure 2: Most frequent approaches in the studied solutions.

its main obstacle. From the techniques and algorithms

provided by Artiﬁcial Intelligence, such uncertainty

can be reduced.

Analyzing the number of studies containing the

keywords ”Artiﬁcial Intelligence” and ”Software En-

gineering” published between 2010 and 2021 in the

IEEE Xplore, ACM Digital Library, and ScienceDi-

rect databases, one can also see a growing trend in the

intersection of these two areas. As illustrated in Fig-

ure 3, in 2010 about 900 publications addressed both

topics, while in 2021 this number exceeded the 2500

mark. Analyzing the graph, the number of publica-

tions mentioning AI and Software Engineering today

is almost three times greater than it was in 2010.

Figure 3: Number of publications addressing Software En-

gineering and Artiﬁcial Intelligence, per year.

Based on this information, it is possible to note

that Artiﬁcial Intelligence is a branch of science that

has much to offer to the Software Engineering ﬁeld.

As it is a vast and constantly growing area, there are

still many open possibilities of also applying intelli-

gent techniques in Requirements Engineering.

5 CONCLUSION

Requirements Engineering is a subarea of Software

Engineering that encompasses activities such as anal-

ysis, elicitation, documentation, and requirements

management and is crucial for the success of a

project. Projects that do not have good techniques and

practices deﬁned to perform such activities tend to

face problems throughout their life cycle and conse-

quently extrapolate initially estimated deadlines and

resources.

We carried out a systematic literature review to

deﬁne the most frequent problems faced in Require-

ments Engineering today. Among the challenges

identiﬁed, the ones that had the highest occurrence

were: lack of communication between stakehold-

ers, ambiguous, incomplete, and inconsistent require-

ments, problems related to documentation, require-

ments traceability, insufﬁcient knowledge about the

domain, and requirements prioritization.

In addition, for each challenge encountered, we

studied the current scenario concerning published so-

lution proposals. We understand that, currently, Re-

quirements Engineering still presents numerous chal-

lenges, and researchers are continually proposing in-

novative solutions to make activities in this ﬁeld less

problematic, even intersecting RE concepts with other

areas of science.

In future work, we plan to use the knowledge ac-

quired by the elaboration of this work and apply them

in the development of a solution proposal for the most

common problems of Requirements Engineering.

THREATS TO VALIDITY

The fact that only a small set of studies were selected

for this systematic literature review poses a poten-

tial threat to its validity. Many studies also don’t go

in depth about problems faced, which could possibly

lead to the misinterpretation of such problems.

ACKNOWLEDGEMENTS

We thank the Brazilian Army Strategic Program AS-

TROS for the ﬁnancial support through the SIS-

ASTROS GMF (898347/2020) projects.

REFERENCES

Alam, S., Nazir, S., and Asim, S. (2017). Impact and chal-

lenges of requirement engineering in agile methodolo-

ICEIS 2022 - 24th International Conference on Enterprise Information Systems

gies: A systematic review. International Journal of

Advanced Computer Science and Applications, 8.

Ali, S., Iqbal, N., and Hafeez, Y. (2018). Towards require-

ment change management for global software devel-

opment using case base reasoning. Mehran University

Research Journal of Engineering and Technology, 37.

Behutiye, W., Karhap

a, P., Costal, D., Oivo, M., and

Franch, X. (2017). Non-functional requirements doc-

umentation in agile software development: Chal-

lenges and solution proposal. pages 515–522.

Coughlan, J. and Macredie, R. (2002). Effective commu-

nication in requirements elicitation: A comparison of

methodologies. Requirements Engineering, 7:47–60.

DeVries, B. and Cheng, B. (2017). Automatic detection of

incomplete requirements using symbolic analysis and

evolutionary computation. In SSBSE.

Elghariani, K. and Kama, N. (2016). Review on agile re-

quirements engineering challenges. pages 507–512.

Garcia, J. E. and Paiva, A. C. (2016). A requirements-to-

implementation mapping tool for requirements trace-

ability. J. Softw., 11:193–200.

Gull, N., Rashid, M., Azam, F., Rasheed, Y., and

Waseem Anwar, M. (2021). A block-chain oriented

model driven framework for handling inconsistent re-

quirements in global software development. In 2021

10th International Conference on Software and Com-

puter Applications, ICSCA 2021, page 105–111, New

York, NY, USA. Association for Computing Machin-

ery.

Haidrar, S., Anwar, A., Bruel, J., and Roudi

es, O. (2018).

A domain-speciﬁc language to manage requirements

traceability. J. Softw., 13:460–480.

Harman, M. (2012). The role of artiﬁcial intelligence in

software engineering. In RAISE ’12: Proceedings of

the First International Workshop on Realizing AI Syn-

ergies in Software Engineering, pages 1–6, Zurich,

Switzerland. IEEE Press.

Iriarte, C. and Bayona, S. (2020). It projects success factors:

a literature review. International Journal of Informa-

tion Systems and Project Management, 8:49–78.

Kalinowski, M., Felderer, M., Conte, T., Sp

ınola, R., Prik-

ladnicki, R., Winkler, D., M

endez Fern

andez, D., and

Wagner, S. (2016). Preventing incomplete/hidden re-

quirements: Reﬂections on survey data from austria

and brazil.

Kitchenham, B. and Charters, S. (2007). Guidelines for per-

forming systematic literature reviews in software en-

gineering. 2.

Li, Y., Schulze, S., Scherrebeck, H. H., and Fogdal, T. S.

(2020). Automated extraction of domain knowledge

in practice: The case of feature extraction from re-

quirements at danfoss. In Proceedings of the 24th

ACM Conference on Systems and Software Product

Line: Volume A - Volume A, SPLC ’20, New York,

NY, USA. Association for Computing Machinery.

Mezghani, M., Kang, J., and Sedes, F. (2018). Industrial

requirements classiﬁcation for redundancy and incon-

sistency detection in semios. pages 297–303.

Mkpojiogu, E. O. C. and Hashim, N. (2017). Quality based

prioritization: An approach for prioritizing software

requirements. Journal of Telecommunication, Elec-

tronic and Computer Engineering, 9:17–21.

Nadeem, M. and Lee, S. U.-J. (2019). Requirement elicita-

tion framework for global software development. In-

dian journal of science and technology, 12:1–6.

Naz, H., Motla, Y., Asghar, S., Abbas, M., and Khatoon,

A. (2013). Effective usage of ai technique for require-

ment change management practices. pages 121–125.

Nguyen, H., Grundy, J., and Almosy, M. (2014). Guitar: An

ontology-based automated requirements analysis tool.

Osman, M. H. and Zaharin, M. F. (2018). Ambiguous

software requirement speciﬁcation detection: An au-

tomated approach. In Proceedings of the 5th Inter-

national Workshop on Requirements Engineering and

Testing, RET ’18, page 33–40, New York, NY, USA.

Association for Computing Machinery.

Salvador, M. E. R. F. L. and dos Santos, L. B. R. (2016).

Domar: An approach to prevent problems related to

requirements documentation and management.

Sch

on, E.-M., Thomaschewski, J., and Cuaresma, M. J. E.

(2017). Agile requirements engineering: A systematic

literature review. Comput. Stand. Interfaces, 49:79–

91.

Shahzad, B., Javed, I., Shaikh, A., Sulaiman, A., Abro, A.,

and Ali Memon, M. (2021). Reliable requirements

engineering practices for covid-19 using blockchain.

Sustainability, 13(12).

Sharma, R., Sharma, N., and Biswas, K. K. (2016). Ma-

chine learning for detecting pronominal anaphora am-

biguity in nl requirements. 2016 4th Intl Conf on

Applied Computing and Information Technology/3rd

Intl Conf on Computational Science/Intelligence and

Applied Informatics/1st Intl Conf on Big Data, Cloud

Computing, Data Science & Engineering (ACIT-CSII-

BCD), pages 177–182.

Sommerville, I. (2011). Software Engineering. Pearson

Prentice Hall, S

ao Paulo, SP, Brasil, 9th edition.

Sommerville, I. and Kotonya, G. (1998). Requirements En-

gineering - Processes and Techniques. Wiley, Hobo-

ken, NJ, United States, 1st edition.

Standish Group International (2019). The standish group

report chaos.

Yaseen, M., Mustapha, A., and Ibrahim, N. (2020). Pri-

oritization of software functional requirements: Span-

ning tree based approach. International Journal of Ad-

vanced Computer Science and Applications, 10.

Challenges in Requirements Engineering and Its Solutions: A Systematic Review