A SURVEY ON AUTOMATED USER INTERFACE TESTING FOR MOBILE APPLICATIONS

Abstract

Nowadays, smartphones play a remarkable role in our lives. Testing mobile applications is significant to guarantee their quality. Automated testing is applied to minimize the cost and the interval of time instead of manual testing. There are different testing levels which are unit testing, integration testing, system testing and acceptance testing. Automated mobile application testing type methodologies are categorized into white-box testing, black-box testing and grey-box testing. Besides, there are several testing types such as functional testing and non-functional testing. Most of the existing studies focus on user interface testing which is type of functional testing. In this paper, testing approaches for user interface testing through different existing studies from 2013 to 2021 have been surveyed. Those approaches are classified into model-based testing, model learning testing, search-based testing, random-based testing, and record & replay testing. Several essential issues related to those approach such as the optimization and redundancy for generation of test suites have been mentioned. Finally, challenges in automated mobile applications user interface testing have been discussed.

Full text

IJICIS, Vol.22, No.2, 126-136
DOI: 10.21608/ijicis.2022.98138.1124
International Journal of Intelligent
Computing and Information Sciences
A SURVEY ON AUTOMATED USER INTERFACE TESTING FOR MOBILE
APPLICATIONS
Amira Samir *
Department of Information
Systems, Faculty of Computer
and
Information Sciences,
Ain Shams University
Cairo 11566, Egypt
amira_samir@cis.asu.edu.eg
https://orcid.org/0000-0003-4641-
0585
Huda Amin Maghawry
Department of Information
Systems, Faculty of Computer
and Information Sciences,
Ain Shams University
Cairo 11566, Egypt
huda_amin@cis.asu.edu.eg
https://orcid.org/0000-0001-
5550-5717
Nagwa Badr
Department of Information
Systems, Faculty of Computer and
Information Sciences,
Ain Shams University
Cairo 11566, Egypt
nagwabadr@cis.asu.edu.eg
https://orcid.org/0000-0002-5382-
1385
Received 2021-09-26; Revised 2022-04-27; Accepted 2022-04-28
Abstract: Nowadays, smartphones play a remarkable role in our lives. Testing mobile applications is
significant to guarantee their quality. Automated testing is applied to minimize the cost and the interval
of time instead of manual testing. There are different testing levels which are unit testing, integration
testing, system testing and acceptance testing. Automated mobile application testing type methodologies
are categorized into white-box testing, black-box testing and grey-box testing. Besides, there are
several testing types such as functional testing and non-functional testing. Most of the existing studies
focus on user interface testing which is type of functional testing. In this paper, testing approaches for
user interface testing through different existing studies from 2013 to 2021 have been surveyed. Those
approaches are classified into model-based testing, model learning testing, search-based testing,
random-based testing, and record & replay testing. Several essential issues related to those approach
such as the optimization and redundancy for generation of test suites have been mentioned. Finally,
challenges in automated mobile applications user interface testing have been discussed.
Keywords: mobile application testing, user interface testing, functional testing, system testing, black-
box testing
 Corresponding author: Amira Samir
Department of Information Systems, Faculty of Computer and Information Sciences, Ain Shams University Cairo 11566,
Egypt amira_samir@cis.asu.edu.egmail addresses: -E 0585-4641-0003-https://orcid.org/0000ORCID:
https://ijicis.journals.ekb.eg/

127 Amira Samir et al.
1. Introduction
Today, technology has a great influence on the world. It has helped in the prediction and detection of
severe diseases as lung cancer [1, 2] and the diagnosis of Corona virus disease 2019 (COVID-19) [3]. It
has also helped in connecting the world through smartphones that have played a vital role in our daily
life. Their applications are used in different fields to make life easier for users. For example, mobile
payment is one of the quickest ways of electronic payment [4]. Another example is mobile learning that
helps in the effectiveness of education through simplifying the connection between teachers and
students [5]. Testing is essential when constructing a mobile application to guarantee its quality. It
requires too much time to be performed manually. That is the reason of the presence of automated
testing for mobile applications. It minimizes the cost and time interval of testing [6]. Mobile application
testing could be performed using testing environments which are emulators and real devices. Emulator
is a virtual machine simulating real device [7]. The same testing approach could use both emulator and
device. The studies at mobile application testing have focused on several points: Optimization of the
generated test suites to increase the performance to testing [8–10], redundant exploration of the
Application Under Test (AUT) [11–14], generation of user interface events only without support to
system events [12, 15].
In this paper, the automated user interface testing approaches for mobile applications have been
surveyed through different existing studies from 2013 to 2021. In the next section, the testing types
have been presented, they are divided into functional testing and non-functional testing. Then, existing
studies of the mobile testing levels have been surveyed, specifically at unit testing and system testing. In
section 4, the three testing type methodologies are presented, which are white-box testing, black-box
testing and grey-box testing. Then, the challenges in automated mobile applications User Interface (UI)
testing have been presented. Finally, the conclusion of this survey is presented.
2. Testing types
There are two types of testing. They are functional testing and non-functional testing. Functional
testing is concerned with the application’s functionalities. For mobile application testing, it is concerned
with testing application’s UI, service functionalities and Application Programming Interface (API) [7,
16]. UI testing is a type of functional testing. It executes events through application’s UI for testing the
application’s behavior [16]. Most of the studies focus on UI testing as there are lots of libraries
supporting UI mobile application testing, this could help researchers working at this point [17]. Service
is a functionality that could work in the application’s background. Service testing is concerned with
appropriate management of its life cycle [18]. API is a set of software functionalities that could run by
another application. API testing ensures that there is no error could happen as result of this integration
[19].

A SURVEY ON AUTOMATED USER INTERFACE TESTING FOR MOBILE APPLICATIONS 128
Non-functional testing is concerned with testing the application’s non-functional requirements. There
are several testing types of non-functional testing for mobile applications as performance testing, stress
testing, security testing, compatibility testing, usability testing and accessibility testing. Performance
testing is concerned with checking the application’s performance behavior when it is used by several
users. It checks its response, time it could take for executing, the usage of mobile resources such as
memory, network, and power [16, 18]. Stress testing is concerned with checking the application’s
performance behavior when it is used by a huge number of users or at high usage of mobile resources. It
checks when the application could crash and its capability for restoration [18]. Security testing tests the
application’s capability of rejecting access of unauthorized users and protecting the user’s data. It
assures that the application is invulnerable to any attacks could happen. It also detects the weak points
of the application that could result into security breach in order to be solved [18]. Compatibility testing
finds out the mobile devices that could support the AUT. There are several configuration profiles at the
market, so it is important to check the devices that would fail to run the AUT. Emulators have decreased
the cost of compatibility testing [18]. Usability testing assures that the AUT is easy to be used, that the
user will not face any difficulties while dealing with AUT. The AUT should respond quickly to user’s
actions. However, this could be hard when there are more than one activity running at the same time
[18]. It checks that AUT’s UI design is usable to prevent any confusions for users [16, 20].
Accessibility testing is a type of usability testing. It tests the application’s behavior with disabled users.
There are some features at mobile operating systems that could help, as text-to-speech in android [18].
3. Testing Levels
Testing levels are classified into unit testing, integration testing, system testing and acceptance testing
[16, 21]. Unit testing is concerned with verifying the smallest component of AUT. It focuses on finding
out any error at the application’s code [16, 21]. Integration testing targets revealing the bugs that happen
after combining different application’s units [18]. It could be performed using different ways such as
integrating independent components at the beginning as database then combining other system
functionalities which is bottom-up integration. Top-down integration is the opposite of the bottom-up
integration as the components are tested and integrated to the system dependent components. For many
applications those two ways are mixed and applied together [18, 22]. System testing is concerned with
checking that the application has completed all its requirements after integration [18]. Acceptance
testing checks that the application has reached the user’s trust by completing all his requirements. It also
checks that it does not have any problems using real data by the user. This testing is performed by the
testers and users [21, 22].
Most of the studies focus on unit testing and system testing. The following subsections review the
approaches proposed for each testing level of them.

129 Amira Samir et al.
3.1 Unit testing
Unit testing is usually done by the application’s developers. It is also called component testing. It
could assure the application’s behavior by testing the response to different types of events. Moreover, it
could be used for checking that the life cycle events are properly managed [18], as at [23, 24] and the
user’s events as at [25].
3.1.1 Model-based / Model learning testing
Several studies have focused on checking that the management of life cycle events. At MobiGUITAR
[23] some errors have been found as result of the inappropriate life cycle management of AUT’s
activities. It is a model-based / model learning approach that has enhanced its previous model that
worked on desktop applications. It could not work with mobile applications as it has no state and did not
consider security. The application’s UI state has been modelled through a reverse engineering approach.
Model-based testing approaches create test cases based on the extracted model of AUTs [26]. Model
learning testing approaches learn the model of AUT while testing [27]. At [24] a model-based approach
has been proposed for generating application’s life cycle test cases. The problem of ensuring the quality
of mobile application’s data has been addressed through different states of application’s activity life
cycle. A model of application’s activity life cycle has been constructed using source code, parsed, and
analyzed. Then, a graph has been created for each activity. Moreover, automated test cases have been
generated and executed using the constructed model. The status of each system resource has been
checked that it is either acquired or released in the suitable life cycle method, to detect any failure could
happen at system resources.
3.1.2 Record & replay testing
Record & replay testing is an approach for recording the interactions by testers with application’s UI
components to create a script. This script could be automatically replayed later for the testing process
[28]. ACRT [25] is a record & replay testing approach. It is implemented for minimizing the testing
effort. It captures the user’s events and input. Then, a script is created by obtaining the coordinates of
the event at the screen or the code of the pressed key. The script is replayed by Robotium [29] which is
an open source automated testing framework. Assertions are used to find out if there are any errors at
the application’s UI components.
3.2 System Testing
System testing is concerned with finding any errors that could happen when executing the complete
application. It has three steps, first the application runs on an emulator. Then, the environment changes
gradually towards real device. Finally, it runs on a real device [18]. Many applications have applied
system testing, as with search-based testing [8–10, 14, 30], model-based testing [11, 12], model learning
testing [15] and random-based testing [13, 31–33].

A SURVEY ON AUTOMATED USER INTERFACE TESTING FOR MOBILE APPLICATIONS 130
3.2.1 Search-based testing
Search-based testing approaches apply metaheuristic approaches for generating optimized test cases
by evaluating some solutions with a fitness function [34]. EvoDroid [10] has applied an approach for
solving the defects resulting from using genetic algorithm to find the best individuals affecting the
performance of the search. The application’s behavior has been analyzed and modelled for detecting the
source code dependencies. Then, search approach has been applied on the detected independent code
parts to find the appropriate crossing over between genes and going in depth at the code to increase the
coverage. AGRippin [8] is a search-based approach for handling the inefficiency of generation of test
suites by other approaches. Crossover has been applied to generate test cases (chromosomes). They
have been ensured that they could be executed by applying some heuristic criteria. A mutation approach
has been applied and ranked test cases based on fitness. Genetic algorithm has been merged with hill
climbing algorithm. Sapienz [9] applies a genetic algorithm with randomness for optimization of test
sequences and increasing coverage. If the input is an Android Application Package (APK), then it is
instrumented, to find out the coverage. Monkey++ [14] generates control flow graph for AUT and
traverses through it by using Depth First Search (DFS) strategy. The generated test cases achieved more
coverage results with less execution time than Monkey. ADAPTDROID [30] performs adaptation for
test cases of AUTs with common methods. It extracts their semantics though their APKs. Then, it
matches between the events of AUTs through the implemented evolutionary algorithm. Finally, it
produces effective related test cases with more similarity in semantics between the AUTs compared to
other approaches.
3.2.2 Model-based testing
Different model-based approaches for UI testing were proposed in literature, they have focused on
maximizing coverage and generating relevant events. However, the exploration strategy could be a
challenging issue. Stoat [12] is a model-based approach for UI testing. It has addressed several
problems, as the exhaustive derivation of tests from the models to validate an application, the
redundancy of randomly generated tests that results from the previous point, there are no models for
some applications and the generation of UI events only. Dynamic analysis for exploration and static
analysis have been combined for building a stochastic model. A guided search for finding the best
models has been applied. Generation of system events by using intents is also supported. CrawlDroid
[11] is model-based approach for the redundant exploration of an application that could get to the same
state multiple of times. A feedback exploration approach has been applied and scores have been given
to actions based on their ability to get to a new state in the application. That leads to an increasing in the
coverage and discovering more failures.
3.2.3 Model learning testing
As the limitations of previous approaches were the susceptibility of errors from manual testing,
requirement of source code of application and the generation of UI events only GATS [15] has been
proposed. It learns the model of AUT using finite-state machine. The first state is the application’s

131 Amira Samir et al.
installation. An unexplored transition is selected, then the model is updated. The target is to find more
bugs in the least time. It generates a report when a crash happens. It supports system and UI events.
3.2.4 Random-based testing
Random-based testing approaches depend on the randomness of selecting the events. Several studies
have focused on generating relevant and minimizing redundancy compared to monkey. Monkey [31] is
a testing tool at Android Software Development Kit (SDK). It generates pseudo-random events.
Monkey testing was used in the beginning for testing Macintosh (Mac) programs in 1983 [35].
Dynodroid [32] generates relevant events for testing an application. Three approaches have been
implemented for selecting an event which are frequency, biased random and uniform random.
Frequency selects the event with the minimum number of selections. Uniform random selects a random
event. Biased random keeps a history of events’ number of selections. Dynodroid’s performance is
better than monkey. At [33] an approach has been presented for handling the excessive time and effort
required for generation of test cases. It is based on statistical analysis from mining users’ usage. First,
the executed events have been recorded to get the usage logs. Then, a behavioral model from those logs
has been generated. An event will be randomly selected based on probabilistic calculations.
MonkeyImprover [13] enhances AUT through refactoring the GUI without affecting the functionalities
of AUT. It extracts GUI elements of AUT that a user could interact with. Then, it generates a weight for
each functionality based on its complexity. Then, GUI components would be resized based on their
weight. As the size increases, the chance of interacting with monkey increases.
4. Testing Type Methodologies
There are three testing type methodologies. They are white-box testing, black-box testing and grey-
box testing.
White-box testing is also called structure-based methodology. It considers the internal structure of an
application, as the testing process is based on the presence of the application’s source code [21]. There
are several white-box testing approaches as AGRippin [8], EvoDroid [10], MonkeyImprover [13],
Monkey++ [14], ACRT [25] and [24]. Their input is source code of android application.
Black-box testing is also called specification-based methodology, as it based on the application’s
requirements. It does not consider the presence of the source code of application. However, it considers
interfaces [21]. There are several black-box testing approaches as at CrawlDroid [11], Stoat [12], GATS
[15], MobiGUITAR [23], Monkey [31], Dynodroid [32], ADAPTDROID [30] and [33]. Their input is
the application’s APK.
Grey-box testing is a combination of both black-box testing and white-box testing. Testing is based on
the application’s structure and requirements [21]. At Sapienz [9], where the input is only application’s
APK, it is extracted to get the source code for instrumentation. However, the approach could also be
performed if the source code is present.A summary for existing studies is presented at Table 1. They are
sorted according to the year of publication ascendingly.

A SURVEY ON AUTOMATED USER INTERFACE TESTING FOR MOBILE APPLICATIONS 132
Table 1 - Summary of UI testing existing studies
Paper
Name
Year
Testing Type
Methodology
Testing
Level
Testing
Approach
Testing
environment
Advantages
Disadvantages
[31]
Monkey
1983
Black-box
System
Random-
based
Mobile device
and emulator
Fast execution time
Irrelevant events
[32]
Dynodroid
2013
Black-box
System
Random-
based
Android 2.3.5
emulator
Relevant events
Slow execution
time
[23]
MobiGUITAR
2014
Black-box
Unit
Model-
based /
Model
learning
Not specified
Enhanced model of
desktop applications
Simple exploration
strategy
[10]
EvoDroid
2014
White-box
System
Search-
based
Emulators in
parallel
Fast execution time with
increased coverage
Slow execution
time with complex
applications
[25]
ACRT
2014
White-box
Unit
Record &
replay
Mobile device
Decreased time and effort
Not providing
coverage
[8]
AGRippin
2015
White-box
System
Search-
based
Android 2.3.3
emulator
Increased effectiveness
than applying hill
climbing solely
Same coverage at
successive
iterations
[9]
Sapienz
2016
Grey-box
System
Search-
based
Android 4.4
emulator and
mobile device
Optimization of test
sequences
Tested applications
may not cover all
existing fields
[12]
Stoat
2017
Black-box
System
Model-
based
Android 4.4.2
emulator
Relevant events
Incomplete
exploration of UI
[33]
Random GUI Testing
of Android
Application Using
Behavioral Model
2017
Black-box
System
Random-
based
Android Nexus
5 emulator
Mining users’ usage and
increasing coverage
Complex events not
handled
[11]
CrawlDroid
2018
Black-box
System
Model-
based
Android 4.4
emulator and
mobile device
Increasing coverage and
discovering more failures
Tested applications
may not cover all
existing fields
[15]
GATS
2019
Black-box
System
Model
learning
Android 5.1
emulator
Supporting system and UI
events
Slow execution
time
[24]
A Model for
Generating
Automated Lifecycle
Tests
2020
White-box
Unit
Model-
based
Android 11
mobile device
Detecting failures in
system resources
Failed resources are
released manually
[13]
MonkeyImprover
2020
White-box
System
Random-
based
Not specified
Minimizing redundant
events generated by
Monkey
Needs more
evaluation (one
AUT only)
[14]
Monkey++
2021
White-box
System
Search-
based
Not specified
Relevant events,
increase coverage and fast
execution time
Tested applications
may not cover all
existing fields
[30]
ADAPTDROID
2021
Black-box
System
Search-
based
Android
emulators
Semantic test cases
High cost of
computation

133 Amira Samir et al.
5. Challenges in Automated Mobile Applications UI Testing
Several challenges in automated mobile applications testing approaches through UI testing have been
surveyed.
 Selecting the AUTs covering all existing fields is a challenging issue. As the approach’s
performance could change between different applications [9, 11]. Sapienz [9] and CrawlDroid
[11] could need to select more AUTs in different fields to ensure their effectiveness .
 The effectiveness UI testing at model-based approaches could be affected by applying simple
exploration strategies such as using breadth-first search approach in MobiGUITAR [23] or the
inability to extract the whole behaviors leading into incomplete exploration of UI of an
application as in Stoat [12].
 The slow execution time at some approaches as GATS [15] and Dynodroid [32] compared to
some existing tools as monkey. However, They have achieved higher performance results than
monkey [31], as monkey generates irrelevant events to the application.
 Most of the studies as [33] focus on simple events such as to click, long-click, text and scroll.
Since handling complex input events such as dragging is a challenging issue.
 Complex AUTs with many conditions could have lower performance than other AUTs when
applying the same testing approach. By applying EvoDroid [10], complex applications got
slower execution time.
 The need of providing information about testing metrics as code coverage when validating the
performance of an approach to ensure that it has covered as much as possible of AUT’s code.
ACRT [25] does not provide information about code coverage. Maximizing the coverage is a
challenging issue for any approach as for EvoDroid [10]. For search-based approaches as
AGRippin [8], the ratio of crossover and mutation could lead into that some test suites could
have the same coverage at successive iterations. This could lead to decreasing the coverage.
 Detecting errors in system resources when using an application could be challenging issue, as
there are multiple resources that could be assessed. This approach [24] has focused on 3
resources only which are drive, camera and location.
 Extraction of GUI elements of AUT that a user could interact with could be a challenging issue.
As there are different formats of implementation for event handlers either by using particular
annotations or by declaring their names in layout files or by using specific functions as in
MonkeyImprover [13]. Moreover, common callback functions names could lead to inability of
detection of control and its related callback function as in Monkey++ [14].

A SURVEY ON AUTOMATED USER INTERFACE TESTING FOR MOBILE APPLICATIONS 134
6. Conclusion
Every day, the importance of smartphones increases. Testing mobile applications has become
essential to assure their quality. In this paper, a survey about automated UI testing for mobile
applications has been presented. Mobile testing types, testing type methodologies and testing levels
have been discussed. Existing studies have focused on the several issues such as discovering more
failures in AUTs, generating relevant events with minimized redundancy, the non-generation of
system events when testing AUT and optimization of test suites. Different existing UI mobile
application testing studies from 2013 to 2021 have been presented. Those studies have been
classified according to their year, testing type methodology, testing level, testing approach, testing
environment, their advantages, and disadvantages. Testing approaches that have been presented in
these studies are random-based testing, model-based testing, model learning testing, search-based
testing, and record & replay testing. Most of the existing mobile testing studies are about UI testing,
which is a type of functional testing, due to the presence of libraries supporting it. Moreover, the
challenges that could happen at automated mobile applications UI testing have been discussed. As
future direction, researchers can start by tackling those challenges.
References
1. Numan N., Abuelenin S., Rashad M.: Prediction of Lung Cancer Using Artificial Neural
Network Int. J. Intell. Comput. Inf. Sci., 16, pp. 1–19 (2016)
2. Aouf M.: Application of K-MTh Algorithm for Accurate Lunge Cancer Detection Int. J. Intell.
Comput. Inf. Sci., 19, pp. 1–20 (2019)
3. Ukaoha K., Ademiluyi O., Ndunagu J., Daodu S., Osang F.: Adaptive Neuro Fuzzy Inference
System for Diagnosing Coronavirus Disease 2019 (COVID-19) Int. J. Intell. Comput. Inf. Sci.,
20, pp. 1–31 (2020)
4. Nasr M., Farrag M., Nasr M.: E-Payment Systems Risks, Opportunities, and Challenges for
Improved Results in E-Business Int. J. Intell. Comput. Inf. Sci., 20, pp. 1–20 (2020)
5. Amasha M.: Using Actionscript 3.00 To Develop an Android Application for Mathematics
Course Int. J. Intell. Comput. Inf. Sci., 16, pp. 67–79 (2016)
6. Kaur A.: Review of Mobile Applications Testing with Automated Techniques Int. J. Adv. Res.
Comput. Commun. Eng., 4, pp. 503–507 (2015)
7. Gao J., Bai X., Tsai W.-T., Uehara T.: Mobile Application Testing: A Tutorial Computer (Long.
Beach. Calif)., 47, pp. 46–55 (2014)
8. Amalfitano D., Amatucci N., Fasolino A.R., Tramontana P.: AGRippin: A novel search based
testing technique for android applications 3rd Int. Work. Softw. Dev. Lifecycle Mobile,
DeMobile 2015 - Proc., pp. 5–12 (2015)
9. Mao K., Harman M., Jia Y.: Sapienz: multi-objective automated testing for Android applications
Proc. 25th Int. Symp. Softw. Test. Anal. - ISSTA 2016, pp. 94–105 (2016)
10. Mahmood R., Mirzaei N., Malek S.: Evodroid: Segmented evolutionary testing of android apps
Proceedings of the 22nd ACM SIGSOFT International Symposium on Foundations of Software
Engineering. pp. 599–609 (2014)

135 Amira Samir et al.
11. Cao Y., Wu G., Chen W., Wei J.: Crawldroid: Effective model-based gui testing of android apps
Proceedings of the Tenth Asia-Pacific Symposium on Internetware. pp. 1–6 (2018)
12. Su T., Meng G., Chen Y., Wu K., Yang W., Yao Y., Pu G., Liu Y., Su Z.: Guided, Stochastic
Model-Based GUI Testing of Android Apps Proceedings of the 2017 11th Joint Meeting on
Foundations of Software Engineering. pp. 245–256. Association for Computing Machinery, New
York, NY, USA (2017)
13. Paydar S.: Automated GUI Layout Refactoring to Improve Monkey Testing of Android
Applications Proc. RTEST 2020 - 3rd CSI/CPSSI Int. Symp. Real-Time Embed. Syst. Technol.,
(2020)
14. Doyle J., Saber T., Arcaini P., Ventresque A.: Improving mobile user interface testing with
model driven monkey search Proc. - 2021 IEEE 14th Int. Conf. Softw. Testing, Verif. Valid.
Work. ICSTW 2021, pp. 138–145 (2021)
15. Chen T., Song T., He S., Liang A.: A GUI-based automated test system for android applications
Adv. Intell. Syst. Comput., 760, pp. 517–524 (2019)
16. Mostefaoui G.K., Tariq F.: Mobile apps engineering: design, development, security, and testing,
CRC Press, (2018)
17. Tramontana P., Amalfitano D., Amatucci N., Fasolino A.R.: Automated functional testing of
mobile applications: a systematic mapping study Softw. Qual. J., (2018)
18. Amalfitano D., Fasolino A.R., Tramontana P., Robbins B.: Testing android mobile applications:
Challenges, strategies, and approaches Advances in Computers. vol. 89. pp. 1–52. Elsevier
(2013)
19. Jorgensen A., Whittaker J.A.: An api testing method Proceedings of the International Conference
on Software Testing Analysis \& Review (STAREAST 2000) (2000)
20. Bruegge B., Dutoit A.H.: Object-Oriented Software Engineering. Using UML, Patterns, and Java,
(2009)
21. Homès B.: Fundamentals of Software Testing, (2013)
22. Sommerville I.: Software Engineering, Pearson Education, (2006)
23. Amalfitano D., Fasolino A.R., Tramontana P., Ta B.D., Memon A.M.: MobiGUITAR:
Automated Model-Based Testing of Mobile Apps IEEE Softw., 32, pp. 53–59 (2014)
24. Motan M., Zein S.: Android App Testing: A Model for Generating Automated Lifecycle Tests
4th Int. Symp. Multidiscip. Stud. Innov. Technol. ISMSIT 2020 - Proc., (2020)
25. Liu C.H., Lu C.Y., Cheng S.J., Chang K.Y., Hsiao Y.C., Chu W.M.: Capture-replay testing for
android applications Proc. - 2014 Int. Symp. Comput. Consum. Control. IS3C 2014, pp. 1129–
1132 (2014)
26. Dias Neto A.C., Subramanyan R., Vieira M., Travassos G.H.: A survey on model-based testing
approaches: A systematic review Proc. - 1st ACM Int. Work. Empir. Assess. Softw. Eng. Lang.
Technol. WEASELTech 2007, Held with 22nd IEEE/ACM Int. Conf. Autom. Softw. Eng., ASE
2007, pp. 31–36 (2007)
27. Choi W., Necula G., Sen K.: Guided gui testing of android apps with minimal restart and
approximate learning Acm Sigplan Not., 48, pp. 623–640 (2013)
28. Börjesson E., Feldt R.: Automated system testing using visual GUI testing tools: A comparative
study in industry Proc. - IEEE 5th Int. Conf. Softw. Testing, Verif. Validation, ICST 2012, pp.
350–359 (2012)
29. GitHub - RobotiumTech/robotium: Android UI Testing,
https://github.com/robotiumtech/robotium
30. Mariani L., Pezze M., Terragni V., Zuddas D.: An Evolutionary Approach to Adapt Tests across
Mobile Apps Proc. - 2021 IEEE/ACM Int. Conf. Autom. Softw. Test, AST 2021, pp. 70–79

A SURVEY ON AUTOMATED USER INTERFACE TESTING FOR MOBILE APPLICATIONS 136
(2021)
31. UI Application Exerciser Monkey - Android Developers,
https://developer.android.com/studio/test/monkey
32. Machiry A., Tahiliani R., Naik M.: Dynodroid : An Input Generation System for Android Apps
Proc. 9th Jt. Meet. Found. Softw. Eng. ACM,pp. 224–234., (2013)
33. Muangsiri W., Takada S.: Random GUI Testing of Android Application Using Behavioral Model
Int. J. Softw. Eng. Knowl. Eng., 27, pp. 266–271 (2017)
34. Afzal W., Torkar R., Feldt R.: A systematic review of search-based testing for non-functional
system properties Inf. Softw. Technol., 51, pp. 957–976 (2009)
35. What is Monkey Testing_ Features, Types With Examples - Testbytes,
https://www.testbytes.net/blog/monkey-testing/