A Newly Proposed Ambidextrous Software Testing Model Based On Conventional Black Box Testing Strategy Using The Applications of Gaussian Distribution

International Research Journal of Innovations in Engineering and Technology (IRJIET)
ISSN (online): 2581-3048

Volume 5, Issue 8, pp 94-101, August-2021
https://doi.org/10.47001/IRJIET/2021.508016
A Newly Proposed Ambidextrous Software Testing Model

Based on Conventional Black Box Testing Strategy Using
the Applications of Gaussian Distribution
1
Shivankur Thapliyal, 2Renu Bahuguna
1
Assistant Professor, Computer Science and Engineering, Doon Institute of Engineering and Technology, Rishikesh, Uttarakhand,
India
2
Head of Department, Computer Science and Engineering, Doon Institute of Engineering and Technology, Rishikesh,
Uttarakhand, India
Abstract - To Tests Software Systems with all of its aspects I. INTRODUCTION

and areas are really very typical and challenging tasks for
today’s Software Engineering issues. Bug free and fully Software Testing are one of the most challenging issues in
tested software systems are very typical to build or today’s Software Engineering Concepts [1]. To Test software
manufacture and one of the most challenging tasks for with all its functionalities and its major tasks orientation
recent software industries issues in today’s information specifications are really a very typical and challenging issues
age, because many software systems crash out or failure for today’s software development industries. In the Real
due to the reason of bug presence, or partial testing environment scenarios, most of the software crises [2] and
approach or not fully tested the software systems and software failures [3] are to done through some bugs which
missing the examine of the behavior of the software reported on the software systems and strictly related to the
systems with all of its input values, but it’s a very complex functionality of software systems. So the need of the hour is
and typical tasks to tests any software systems with all of that we have some strong potential software testing
its possible input tests values. So the need of the hour is we methodology [4] or mechanisms which strictly navigate or to
have some potential and robust software system testing discover bugs or errors and software failures [6]
model which are responsible and also capable to tests any misconceptions in a very easily manner, because the main
software system product with all of its functionalities with reason which is responsible for behind the software failures
including all areas and investigate or to examine the are that we have not an appropriate or adequate mechanisms,
behavior of software system for all of its possible input which tests software at all aspects corresponding to it’s all
tests values with also achieve greater reliability. Some branching conditions and looping or jumping statements, and
Black Box testing strategies such as Equivalent also a one major reason behind that, we doesn’t tests software
Partitioning and Boundary Value Analysis (BVA), which for all of its input values or not to draw all test cases related to
are responsible to provide range of input vales among all input values. In this model we proposed a new software
multiple values, but we don’t have any strict mechanisms testing model which tests values for all of its input cases and
to filter or to select some input tests values among the set also defines the major importance of its input values. In the
of multiple input tests, because the major reason behind previous developed software testing methodology to draw or
the software crisis and software failures are that we don’t design test cases, we generally adopt some traditional
check software systems with all of its input values or mechanisms such as equivalent partitioning [7,11,14] and
missing to notify the behavior of software systems for some Boundary Value Analysis (BVA) [8,9,11,14] to design tests
particular input, Here we proposed a robust software cases for all of its inputs, but In this previously developed
testing model, which adopts the mechanisms of Gaussian traditional approach based model also contains many
Distribution for selection of input data and work at the misconceptions that it’s doesn’t tests software for all of its
module level of software systems and this model also input cases and very slow to perform, but In this newly
applies with each module of software systems. proposed software testing model we adopts or to use Statistical
Distribution [10]. There are many distributions are available
Keywords: Software Testing model, Black Box testing for mapping any value with its probability density function
strategies, Gaussian based Software testing model, Software [10] with its importance, which based upon the nature of the
testing strategies. variable, which is discrete and continuous in nature. Here with
the use of probability density function [10] we find out the
importance of each values to look through the probabilistic
© 2021-2017 IRJIET All Rights Reserved www.irjiet.com 94


value and their occurring patterns, so for using this Equivalent Partitioning are the strategy of Black Box testing
methodology, we tests software with all of its aspects with all technique, where the number of inputs are to be divided into
of its input values, so we examine or tests software in a very some similar classes. In equivalence portioning, equivalence
efficiently manner. Here we proposed or adopted a modular classes are evaluated for given input conditions. Whenever
designing approach [11] of software system architecture any input is given, then type of input condition is checked,
[12,13], which work with all of its module in a very similar then for this input conditions, Equivalence class represents or
manner, but the working criteria of all modules are must be describes set of valid or invalid states.
different. To tests a software with all of its aspects such as:
code coverage and path coverage or branching and looping are Guidelines for Equivalence Partitioning:
really a very challenging and typical concepts. But In this
 If we gives the range condition as an input then one valid
newly proposed model we don’t require or examine any
and two invalid equivalences classes are to be defined.
WHITE BOX TESTING [14] methodology, here we adopt
 If a specific values are gives as an input, then also one
some BLACK BOX [14] testing methodology which works in
valid and two invalid inputs are to be defined.
a very fast efficient manner, but here we only checked the
 If we gives some member of a set as an input, then one
functionality of the software with all of its aspect of the
valid and one invalid equivalence classes are to be
software system in a new manner, we don’t check any non-
defined.
functional requirements such as the code coverage, branch
 If we gives Boolean number as an input, then also one
coverage, looping and some jumping statements of the
valid and one invalid equivalences classes are to be
software systems. So that’s why we will design all of its test
defined.
cases with all of its input values using this newly proposed
methodology for software systems architecture. Because
Normal Distribution [10] or Gaussian distribution [10] are bell
shaped curve, which describe or to define all of its values with
probability density functions. To pick up any values we find
out the importance factor of each values using the Gaussian
Probability Density function [10]. We will design test cases
for all inputs, which contains a low probabilistic value or those
Fig. a: Example of Equivalent Partitioning
which contains a high probabilistic value. We also design test
cases for invalid inputs. We will check any software systems Boundary Value Analysis (BVA): In the concept of
with all of its functional area using this model in a very Boundary Value Analysis (BVA) we test any software
efficient and reliable manner. Its Computational Cost are very systems at the boundaries such as if we take a range of input
low and gives an accurate result and have capability to easily values such as 0-100, then we check software systems at -1
detect some error prone areas inside the software system and 101 for invalid cases and 0 and 100 for valid cases
architecture. Now the working mechanisms of this model are corresponding to range, because a software systems provides
detailed define in the upcoming section of this research paper. some abnormal results at the boundaries of the range of the
input values. Boundary Value Analysis (BVA) are testing
II. SOME PREVIOUSLY KNOWN METHODOLOGY
strategy of Black Box Testing where any software product
Equivalence Partitioning: In this methodology we have to be tested at extremely ends because a software can
partitioned some input values into some range of values, In give some abnormal and unsatisfactory results at the boundary
detailed we can say that if we have multiple inputs varieties of the input partitioning. These extremely ends like Start-End,
then In the mechanisms of Equivalent Partitioning we Upper-Lower, Maximum-Minimum, these are called the
boundary values of any input partition.
partitioned or range the values of inputs which gives similar
types of output or which behaviors are same. In general we
The basic idea in normal boundary value testing is to select
can say that Equivalence partitioning or equivalence class input variable values at their:
partitioning (ECP) is a software testing technique or strategy
that divides or partitioned the input data of a software unit into  Minimum
partitions of equivalent or similar data from which test cases  Just above the minimum
can be derived. In this mechanisms of equivalent partitioning a  A nominal value
very reliable and accurate test cases have to generate such as  Just below the maximum
test cases are easy to generate for all input values using the  Maximum
mechanisms of Equivalent Partitioning. The concepts of
Equivalent Partitioning come from the Equivalence Relation.


Gaussian Distribution behind the reason is that Gaussian

Distribution defines the probability of each number which
existing in the range of values and it’s a continuous
probabilistic distribution where each numbers we assign a
unique probability for their occurrence, so with the use of
Fig. b: Diagrammatic representation of Boundary Value Analysis probability, we better tests each values which placed in our
test case and better observe or explore the behavior of
In Boundary Value Analysis, Equivalence Partitioning software system on each value. Gaussian Distribution also
play’s a good role. Boundary Value Analysis comes after the called the Normal Distribution or Laplacian Distribution.
Equivalent Partitioning. Because in the Gaussian Distribution values which are near far
the means contains a high probability relatively to those which
III. A NEWLY PROPOSED MODEL FOR TESTING
are far away from mean. Generally Gaussian Distribution
The diagrammatic representation of this Software Testing works on the theorem of Central Tendency, Here the middle
model is representing in fig c: line represents the mean of the sample size or ranges, and
other values which puts between the ranges are equally
This software testing model applies on each modules of the distributed both sides of the means, that’s why it’s contains a
while software product, generally at the broad level this model bath tub curve. It’s a continuous probabilistic distribution, so
comes under the category of Black Box Testing, because these using the mechanisms of probability we better tests software
model checks the functionalities of the software systems rather systems with each values and select or filter values for tests
the code coverage and branch coverage or some looping or through their respective probabilities and it’s gives a crisp
jumping. Because to test overall functionalities of the software clearance about values that, which values have to select for
with tracing or investigate and examine each path or branching testing. The detailed about Gaussian Distribution are as
coverage are really a very typical tasks. So In this Paper we follows:
modeled some Black Box testing strategies for conduct a
software testing in a very easy and reliable manner. Now we Gaussian Distribution [16] are also known as the Normal
describe these models by using some phases or steps: Distribution [16] or the Laplace Distribution [16] . It is a
continuous probability distribution which based on some
First phase: In the first phase we categorized data into probabilistic approach and works on real valued numbered
various categories which based on their types, behaviors and variables. In this distribution we have a sample size of various
their tasks to perform. We categorized data into three types random numbers and we find out the occurring probability of
which is: each numbers using some mathematical and statistical
techniques. Gaussian Distribution [16] contains some
 Numeric Data probability density function. The formula of this probability
 Textual Data density function are as follows:
 File Data
1 1 𝑥 −𝜇 2
Further Numeric data also subdivided into some types such 𝑓 𝑥 = 𝜎 𝑒 −2( 𝜎
)
2𝜋
as the floating point numbers, or some integers numbers,
similarly textual data further subdivided into some categories The parameter 𝜇 is the mean or expectation of the
and file data are further subdivided into image, audio, video, distribution (also it’s median and mode), while the parameter
documents or some pdf files. After categorized data into some 𝜎 is it’s standard deviation. The Variance of the distribution is
categories, the next step is applying some interval arithmetic 𝜎 2 . A random variable with a Gaussian distribution is said to
and initializes the range of values, which will be the tests by be normally distributed, and is called a normal deviate.
the software module. But this type of range are easy to
developed with Numeric Data, but data like textual or file type Normal distributions play’s a much significant role
it’s very complex to define the range especially textual and in statistics and are frequently used in the physics and
image or some multimedia data, so here for textual and file mathematical modelling to represent real-valued random
data we assign some random numeric identity, this process are variables whose distributions are not known. It’s works on
called the Randomization after Randomization of textual data Central Limit Theorem. It states that, under some conditions,
and some file data, we applying the interval arithmetic to the average of various samples (observations) of a random
define the range. variable with finite mean and variance is itself a random
variable—whose distribution converges to a normal
Second Phase: In the second phase we are applying the distribution as the number of samples increases. Therefore,
Gaussian Distribution on that ranges of data. Here we applies


physical quantities that are expected to be the sum of many

independent processes, such as measurement errors, extremely
have distributions that are nearly normal.
Moreover, Gaussian distributions [16] contains some
unique properties and features that are valuable in analytic
studies and mathematical modelling. For instance, any linear
combination of a fixed collection of normal deviates is a
normal deviate. Many results and methods, such
as propagation of uncertainty and least squares parameter
fitting, can be derived analytically in explicit form when the
relevant variables are normally distributed.
A normal distribution [16] is sometimes informally called
a bell curve. However, many other distributions are bell-
Fig. e: Commulative Distribution Function of Normal Distribution
shaped (such as the Cauchy, Student's t,
and logistic distributions). Table 1: Significant expressions of Gaussian Distributions
Normal distribution, which is frequently known as the

Notation 𝑁(𝜇 , 𝜎 2 )
Gaussian distribution [16] or Laplace distribution [16], is
1 1 𝑥−𝜇 2
a probability distribution that is symmetric about the mean, 𝑓 𝑥 = 𝜎 𝑒 −2( 𝜎
)
PDF 2𝜋
showing that data near about the mean are more frequent in
occurrence than data far from the mean. In graph form, normal 1 𝑥−𝜇
distribution will appear as a bell curve. CDF [1 + erf⁡
( 𝜎 )
2 2
𝜎
Quantile 𝜇+ 2 𝑒𝑟𝑓 −1 (2𝑝 − 1)
Mean 𝜇
Median 𝜇
Mode 𝜇
Variance 𝜎2
𝜎
MAD 2 𝜋
1 1
Entropy log 2𝜋𝜎2 +
2 2
MGF exp⁡ (𝜇𝑡 + 𝜎 2 𝑡 2 /2)
CF (𝑖𝜇𝑡 − 𝜎 2 𝑡 2 /2)
exp⁡
Fig. c: Graph of Normal Distribution
1/𝜎 2 0
𝐼 𝜇 ,𝜎 =
0 2/𝜎 2
Fisher 1/𝜎 2 0
𝐼 𝜇 , 𝜎2 =
Information 0 1/(2𝜎 4 )
Kullback-
1 𝜎0 2 𝜇 1 −𝜇 0 2 𝜎1
Leibler { + − 1 + 2𝐼𝑛 }
2 𝜎1 𝜎12 𝜎0
divergence
Fig. d: Probability Density Function of Normal Distribution


Gaussian or Normal Distribution, because this distribution

contains the continuous probabilistic distribution, that’s why
it’s distribute each numbers which is within the range based
on their probability. So these are the three phases of this
model. This model equally applies on each of the modules of
the software product, because In the previous models of Black
Box testing are Equivalence Partitioning and Boundary Value
Analysis (BVA), but in this model we also make some range
of input values, but we don’t have any mechanisms that we
decided which input gives best results or which gives worst
result before tests any software systems, but In this newly
proposed model, we easily decided that which inputs
occurrence pattern frequently or not, and which inputs occur
rarely based on the probability of that number, so that’s why
we easily take judgments for occurring patterns of the inputs.
IV. FUTURISTIC SCOPE AND IT’S APPLICATIONS
Software testing is one of the most challenging issues for

today’s information age and software industry issues.
Software Testing contains a very typical and complex
procedure to draw out an exact outcomes corresponding to
particular input in the form of test cases and test suits. Because
a most significant activity, which is Software Quality
Assurance are totally based on the designing or modelling of
the test cases, but in today’s software engineering scenario it’s
very hard to draw or to find out test cases for each of its
values, which placed on the interval set. So how we draw out
potential mechanisms to assured that this system works
perfectly or adequately and also contains reliability features
without reporting or to examine some bugs or errors are to be
placed inside the code coverage of software modules are major
Fig. f: A newly proposed Gaussian distribution based Software testing issues for today’s Software Engineering issues. Black Box
model testing are the most significant testing methodology, which
tests software with whole aspects which based on their
Third Phase: In the third phase we designed test cases for functionalities and task oriented or task perform features of
each module and write the actual results on test case. Now software systems, but again a one major questions have to
after designing the test cases of each module, at the end we arise that can we test software as a whole with all of its input
compare the resultant data with expected data and find out the values, unfortunately the answer of this question are No,
error rate or success percentage through comparing and after because when we making some interval set of the values then
do the detailed analysis we validate the test data. We also it’s very hard to tests each value which placed inside the
design test cases for invalid inputs or those inputs which occur interval, so here we elicit value based on the Random manner,
rarely, because to decide that which inputs occur frequently or
that’s why to judge or to examine or notify behavior of the
rarely are clearly decided through the probability of that
some elements are have to be missing. But In this newly
number. This model gives much accuracy and clearance about proposed model, we use some statistical techniques and
the selection of input tests data. Because to tests or validate distribution theory with the interval arithmetic.
any software systems with all path coverage and branching
coverage are really a very typical and complex tasks, so with Here we adopt the Gaussian or Normal distribution,
the use of Gaussian Distribution we easy this tasks and also because this distribution are based on some probability
take concern the accuracy and clearance of tests data inputs. features and contains continuous distributions approach, that’s
There are many distributions are available such as Binomial why we elicit values based on their probability. Using the
Distribution, chi-square Distribution , Discrete unit probability features of values we judge that which value
distribution, Cauchy Distribution, but here we elicit the occurring utmost during task performing and we also judge or


examine those elements behavior which contains lower Box testing strategies such as Equivalent Partitioning and
probability, because Gaussian Distribution works on the Boundary Value Analysis (BVA), are also defining the range
concepts of the theorem of Central of Tendency, here a middle of the values, but it’s not give a crisp clearance and reliability
line represents the means and left or right side along the means about the selection of the inputs, that In other words we have
the elements are equally distributed based on their to say that some previously developed Black box testing
probabilities. Using the probabilistic features we optimize the strategies also provides the range of input tests values, but it’s
error rate and increase the bug detection. Using this newly not give the reliability to select or filter the input values
proposed model the failures rates of software in the near future among multiple values which placed within the range that
are strictly optimum. This model also suitable for following which input values have to be selected for tests data, because
major applications of Software Engineering Issues such as: it’s impossible to tests software systems with all of its input
values so we have not any mechanisms to select a desired
 Real time and Simulation based Software Systems input values for tests which behavior are either positive (Best)
tests. or negative (worst). But In this newly proposed model we use
 Some Automated framework for complex software the Gaussian Distribution to select or filter input values based
projects. on their probability, because distribution is the perfect
 API testing methodology. methodology for selecting and filtering input values among
 Checked Quality Assurance and Quality Control of the set of multiple values. Now a one reason behind the use of
Software Management issues. Gaussian distribution is that it’s a continuous probabilistic
distribution that works on the theorem of the Central of
There are various automated software testing software
Tendency because it’s equally distributed values along the
such as Selenium, HPE Unified Functional Testing, Test
means according to their respective probability values. So
Complete etc. which pick value from in the interval set in the
using the mechanisms of probability we easily select or filter
Randomized manner, so using or to applying Gaussian
inputs for tests data among the set of multiple input values.
Distribution gives a better results for to design tests cases and
tests suits. This model works with the modular approach of software
system architecture because this model equally applies on
To design test cases and tests suits for any software
each module of the software systems. Some previously Black
systems are really a very complex and typical tasks, because
Box testing methodologies such as Equivalent Partitioning and
tests cases or tests suits defined the overall mechanisms and
Boundary Value Analysis (BVA) provides the range of the
functionality of the software systems, and black box testing
input values so we filter or to select value from these range
play’s a major role for designing tests cases, but various
through Randomization or Randomly because it’s very typical
previously developed traditional black box testing strategies
and complex to tests all input values and impossible to tests a
such as: Equivalent Partitioning, and Boundary Value analysis
whole software systems with each nook and corner so we
divide or partitioned some set of numbers into similar classes
don’t achieve desired output or we don’t easily examine the
or to do check software behavior at extreme ends respectively,
behavior of the software systems with it’s input tests values,
but here both these model picks value in the Randomized
because we take tests data as a input through the mechanisms
manner, but in this newly proposed model we use some
of randomization, but In this newly proposed model we take
distributions mechanisms such as Gaussian or normal
tests data with the mechanisms of Gaussian Distribution.
distribution whereas each number defined based on their
probability and we select particular input based on their Gaussian Distribution is a continuous probabilistic
probability, because these probability defines the occurring approach based distribution, which equally distributes values
patterns of particular input, which provides reliability to select based on their probabilities, so for filtering or to selection of
input and tests the behavior of software systems on that input. any input values through their occurrence probability gives
So there are many improvisations have to be done in this better results as compare to Randomization mechanisms. So
model in the upcoming future and this model is very with the use of Gaussian Distribution this model becomes a
significant to release a bug free software product. fundamental Black Box testing method, which tests or to
explore each aspects of software systems in a very reliable
V. CONCLUSION
manner. This model also categorized data into different
This newly proposed model comes under the category of categories such as Numeric data, Textual data or some other
Black Box Testing methodology, because this newly proposed files data generally called the multimedia data such as image,
model examine and compile the functionalities of the software audio, video or some document files etc. so for numeric data
systems rather path coverage or code coverage or some we easily finding range using the Interval Arithmetic, but for
branching and lopping, but In the previously developed Black some file data such as image, or some textual data we assign a


unique integer value for each data item using the mechanisms [8] J.Irena. ―Software Testing Methods and Techniques‖,
of Randomization, then further we apply Interval Arithmetic 2008, pp. 30-35.
on these values. So here categorization of data also a biggest [9] Guide to the Software Engineering Body of
milestone for successfully tests software systems with whole Knowledge, Swebok, A project of the IEEE Computer
aspects. We design test cases also for invalid inputs, so at the Society Professional Practices Committee, 2004.
end we also compare our actual resultant data set with the [10] E. F. Miller, ―Introduction to Software Testing
expected data set and compare the accuracy and correctness of Technology‖, Software Testing & Validation
our tests. After done the detailed analysis of data sets at the Techniques, IEEE, 1981, pp. 4-16
end we validate the test cases. So this mechanisms applying on [11] M. Shaw, ―Prospects for an engineering discipline of
each of the module of the software systems, so it’s also work software,‖ IEEE Software, November 1990, pp.15-24
with module level. Using this model for tests any software [12] D. Nicola et al. "A grey-box approach to the functional
systems functionalities we tests software product in a very testing of complex automatic train protection systems."
reliable and accurate manner. Dependable Computing-EDCC 5. Springer Berlin
Heidelberg, 2005. 305-317. 181
ACKNOWLEDGEMENT [13] J. A. Whittaker, ―What is Software Testing? And Why
Is It So Hard?‖ IEEE Software, 2000, pp. 70-79.
We are very much grateful to all respected professors of
[14] N. Jenkins, ―A Software Testing Primer‖, 2008, pp.3-
DIET Rishikesh for their kind help, lasting encouragement,
15.
valuable suggestion throughout the entire period of our project
[15] Luo, Lu, and Carnegie, "Software Testing
work. We are highly indebted to his astute guidance, sincere
TechniquesTechnology Maturation and Research
support and boosting confidence to make this Research
Strategies’, Institute for Software Research
successful. The acknowledgment will be incomplete if we fail
International-Carnegie Mellon University, Pittsburgh,
to express our obligation and reverence to our family members
Technical Report, 2010.
and friends whose moral support is great factor in doing this
[16] Krithikadatta J, Valarmathi S. Research Methodology
research.
in dentistry: Part II — The relevance of statistics in
FUNDING research. J Conserv Dent 2012;15:206-213.
[17] M. S. Sharmila and E. Ramadevi. "Analysis of
This study was not funded by any other profitable or non- performance testing on web application." International
profitable organisations. Journal of Advanced Research in Computer and
Communication Engineering, 2014.
CONFLICT OF INTEREST [18] S.Sampath and R. Bryce, Improving the effectiveness
of Test Suite Reduction for User-Session-Based
The authors declare that they have no conflict of interest.
Testing of Web Applications, Elsevier Information and
REFERENCES Software Technology Journal, 2012.
[19] B. Pedersen and S. Manchester, Test Suite
[1] P. Ron. Software testing. Vol. 2. Indianapolis: Sam’s, Prioritization by Cost based Combinatorial Interaction
2001. Coverage International Journal of Systems Assurance
[2] S. Amland, "Risk-based testing:" Journal of Systems Engineering and Management, SPRINGER, 2011.
and Software, vol. 53, no. 3, pp. 287–295, Sep. 2000. [20] S. Sprenkle et al., "Applying Concept Analysis to User-
[3] Redmill and Felix, ―Theory and Practice of Risk-based sessionbased Testing of Web Applications", IEEE
Testing‖, Software Testing, Verification and Transactions on Software Engineering, Vol. 33, No. 10,
Reliability, Vol. 15, No. 1, March 2005. 2007, pp. 643 – 658
[4] B. Agarwal et al., ―Software engineering and testing‖. [21] C. Michael, ―Generating software test data by
Jones & Bartlett Learning, 2010. evolution, Software Engineering‖, IEEE Transaction,
[5] K. Bogdan. ―Automated software test data generation‖. Volume: 27, 2001.
Software Engineering, IEEE Transactions on 16.8 [22] A.Memon, ―A Uniform Representation of Hybrid
(1990): 870-879. Criteria for Regression Testing‖, Transactions on
[6] Jacobson et al. The unified software development Software Engineering (TSE), 2013.
process. Vol. 1. Reading: Addison-Wesley, 1999. [23] R. W. Miller, ―Acceptance testing‖, 2001, Data
[7] Everett et al., ―Software testing: testing across the retrieved from
entire software development life cycle‖. John Wiley & (http://www.dsc.ufcg.edu.br/~jacques/cursos/map/recur
Sons, 2007. sos/Testin g05.pdf)


[24] Infosys, ―Metric model‖, white paper, 2012. Data [27] M. I. Babar, ―Software Quality Enhancement for value
retrieved from (http://www.infosys.com/engineering- based systems through Stakeholders Quantification‖,
services/whitepapers/Documents/comprehensive- 2005, pp.359-360. Data retrieved from
metrics-model.pdf) (http://www.jatit.org/volumes/Vol55No3/10Vol55No3.
[25] B. Boehm, ―Some Future Trends and Implications for pdf)
Systems and Software Engineering Processes‖, 2005, [28] R. Ramler, S. Biffl, and P. Grünbacher, "Value-based
pp.1-11. management of software testing," in Value-Based
[26] R. Bryce, ―Test Suite Prioritisation and Reduction by Software Engineering. Springer Science Business
Combinational based Criteria‖, IEEE Computer Media, 2006, pp. 225– 244.
Society‖, 2014, pp.21-22. [29] D. Graham, "Requirements and testing: Seven missing-
link myths," Software, IEEE, vol. 19, 2002, pp. 15-17.
Citation of this Article:
Shivankur Thapliyal, Renu Bahuguna, ―A Newly Proposed Ambidextrous Software Testing Model Based on Conventional
Black Box Testing Strategy Using the Applications of Gaussian Distribution‖ Published in International Research Journal of
Innovations in Engineering and Technology - IRJIET, Volume 5, Issue 8, pp 94-101, August 2021. Article DOI
*******

© 2021. Notwithstanding the ProQuest Terms and Conditions, you may use
this content in accordance with the associated terms available at
https://irjiet.com/about_open_access

A Newly Proposed Ambidextrous Software Testing Model Based On Conventional Black Box Testing Strategy Using The Applications of Gaussian Distribution

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

A Newly Proposed Ambidextrous Software Testing Model Based On Conventional Black Box Testing Strategy Using The Applications of Gaussian Distribution

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Newly Proposed Ambidextrous Software Testing Model Based On Conventional Black Box Testing Strategy Using The Applications of Gaussian Distribution

Uploaded by

Copyright:

Available Formats

International Research Journal of Innovations in Engineering and Technology (IRJIET)

ISSN (online): 2581-3048