Research Trends in Software Development Effort Estimation: September 2023
Research Trends in Software Development Effort Estimation: September 2023
Research Trends in Software Development Effort Estimation: September 2023
net/publication/375135875
CITATIONS READS
0 46
3 authors:
SEE PROFILE
All content following this page was uploaded by Yulia Swandari on 27 December 2023.
Abstract— Developing a software project without the SDEE, with no confirmation to support model estimates above
appropriate amount of effort would significantly impede and expert estimations.
even fail the project, putting the software developer's quality at
risk. Therefore, software development effort estimation (SDEE) Machine learning (ML) technique in SDEE was
is the most critical activity in software engineering. SDEE has investigated by Wen et al. [1] who determined different ML
seen extensive research, resulting in a massive rise in the techniques, their prediction precision, and the comparison
literature in a relatively short period. In this regard, it is crucial between other models and estimation contexts from 84
to identify the significant study topics in software development primary studies. Based on their research, they disclosed that
effort estimation that will assist researchers in understanding eight different ML techniques had been used in SDEE and
and recognizing research trends. This research applied a concluded that ML models offer more precise predictions than
systematic literature review (SLR) to compile all journals from non-ML models.
the predefined search directory about software development
effort estimation thoroughly and unbiasedly from 2018 to 2022. Effort estimation in agile software development (ASD)
This review was a prelude to further research activities in was reviewed by Usman et al. [5] based on 20 papers selected.
software development effort estimation. Five research topics out The study showed that most methods rely on expert
of 71 papers have been revealed, including the machine learning judgement; extreme programming (XP) and scrum are the
approach, algorithmic technique, expert judgement, dataset only two agile methods identified in the primary analysis.
analysis, and evaluation metric. With 27 journals, deploying a
machine learning approach for SDEE is the most discussed Idri et al. [6] reviewed 65 papers published on analogy-
research topic. The potential research described in this study based software effort estimation (ASEE) from 1990 to 2012
can be investigated further in software development effort and revealed that the research's primary focus is feature and
estimation field. case subset selection. They identified that ASEE methods
outperform eight strategies and produce more accurate results
Keywords— software development, effort estimation, literature when combined with fuzzy logic (FL) or genetic algorithms
review, research trends (GA).
I. INTRODUCTION Research patterns in SDEE were summarized by Sehra et
al. [3] through a generative statistical method called Latent
Software development effort estimation (SDEE) predicts Dirichlet Allocation (LDA) from a large set of SDEE research
the effort needed to develop a software application [1]. SDEE articles published from 1996 to 2016. They found twelve core
can also be shortened to software effort estimation (SEE) or research areas and sixty research trends based on a library of
several synonymous words: cost estimation, cost prediction, 1178 articles.
time estimation, and effort prediction. The number of man-
hours dedicated to software development over time, from This research used a systematic literature review (SLR) to
specification up to delivery, is generally measured by SDEE update the summarization of all papers from the predefined
[2]. But not limited to that, SDEE can also define the skill search directory concerning software development effort
requirements that must be possessed and the number of costs, estimation from 2018 to 2022. An SLR is an approach to
time, and assets needed to work on the project system. discovering, analyzing, and comprehending all accessible
Overestimation wastes resources, whereas underestimation research on a specific topic area [7]. Research trends identified
results in schedule or budget overruns and quality through this review can assist early researchers in finding and
compromises [2]. Software professionals and researchers updating potential research areas in the software development
highly value accurate estimation of software effort. effort estimation.
Many effort estimation methods have been suggested in The rest of this research is structured as follows. Section 2
the SDEE field since the 1980s [1]. Both benefits and describes the research method, including the planning and
drawbacks of the models and approaches created by conducting stage. Section 3 discusses key findings, and in
researchers are well acknowledged. The vast availability of Section 4 challenges to validity are presented. Finally, Section
published literature on the domain challenged researchers to 5 provides a conclusion.
analyze and determine the optimal path for their research [3].
II. RESEARCH METHOD
The review article on expert estimation of SDEE
conducted by Jørgensen [4] reviewed 15 studies for validating We planned and conducted the SLR following
the conformance to twelve expert estimation 'best practices' Kitchenham & Charter [7] guidelines. All the steps are
and found expert estimation to be the dominant approach for explained in the following subsections.
We used the following search string to include the Intending to answer RQ1, we have identified the five
keywords and filters applied: TITLE ( software AND ( research topics presented in Table 3, and the most widely
effort* OR cost* OR siz* ) AND ( estimat* OR predict* discussed research topic is the machine learning approach,
OR forecast* ) AND NOT defect* ) AND ( LIMIT-TO ( with 27 journals.
OA , "all") ) AND ( LIMIT-TO ( PUBYEAR , 2022 ) OR
LIMIT-TO ( PUBYEAR , 2021 ) OR LIMIT-TO (
PUBYEAR , 2020 ) OR LIMIT-TO ( PUBYEAR , 2019 )
OR LIMIT-TO ( PUBYEAR , 2018 ) ) AND ( LIMIT-TO (
LANGUAGE , "English") ). This search string was required
to restrict the search results based on the following selection
parameters:
• the publications on effort estimation with the option
of using another relevant phrase and different
spellings in the titles
• not in the defect prediction domain
• only open-access documents
• published during the five years from 2018 to 2022
• written in English
The search process resulted in 136 publications, and then we
exported the results in a CSV file to facilitate further
elimination of literature. When compiling this review using Fig. 1. Paper publication year-wise.
Zotero [8], these results were exported into RIS file format
TABLE III. FIVE RESEARCH TOPICS
(.ris extension) to manage bibliographic references.
As indicated in Table 1, the search results from the search Id Research Topics
Paper
Refs
strategy were manually reselected by title and abstract using Count
the inclusion and exclusion selection criteria. Only journal T1 Machine learning approach 27 [9]–[35]
papers were selected because they had undergone peer review T2 Algorithmic technique 22 [36]–[57]
by the journal's publisher.
T3 Expert judgement 11 [58]–[68]
Table 2 displays the number of papers based on the search
results and the manual selection procedure. The final steps T4 Dataset analysis 10 [69]–[78]
are data extraction and data synthesis. The goal of data
T5 Evaluation metric 1 [79]
extraction is to gather all the information required to respond
626
Fuzzy logic is also applied in single and hybrid methods:
fuzzy regression tree [9], [32], neuro-fuzzy [25], fuzzy delphi
in test effort estimation [29], and fuzzy-neuro-genetic [35].
Furthermore, researchers also explored the techniques in the
context of open-source software [10], dynamic cross-
company mapped model learning (Dycom) [23], and
unstructured software project descriptions [21]. Aside from
the research trends stated above, some contemporary trends
have evolved from the study, including adaptive neuro-fuzzy
inference system (ANFIS) [16], extreme learning machine
(ELM) [17], feature engineering [24], deep belief network
[31], and random forest [34].
Carvalho et al. [17] applied the ELM technique and
compared it with the literature models, which resulted in the
error estimate rates tending to be reduced. For future studies,
ELM and other machine learning methods can be optimized
with metaheuristic algorithms, such as particle swarm
optimization (PSO), to enhance their accuracy. Besides, there
is the opportunity for additional study of:
• Deep learning models improve with fully automatic
element integration.
• Design the transformation of the artificial neural
network model to a web service and deploy it to a
cloud computing platform.
B. Algorithmic Technique
Various trends have emerged in the "algorithmic
technique" (T2), with the number of papers for the last five
years being 5, 3, 2, 7, and 5. One such trend is the
metaheuristic algorithm, namely particle swarm optimization
[36], [39], [48], [56], differential evolution [41], antlion
optimization [42], firefly algorithm [43], dolphin algorithm
[49], artificial bee colony algorithm [57], and hybrid bat
algorithm [55]. Other trends identified are the morphological
approach [54], local data approach [37], regression analysis
[38], use case point-based method [40], [44], global software
Fig. 2. Research trends of five topics.
development project [45], class diagram-based estimation
[46], software rework index [47], use case reuse [50], safety-
Further, Fig. 2 highlights the research trends for each critical software [51], function point analysis [52], and linear
topic. The trends in research topics and the potential for programming [53].
further research are described in the subsequent section to Four papers improved and modified PSO to enhance its
answer RQ2. accuracy for effort estimation. Ardiansyah et al. [36] proposed
A. Machine Learning Approach modified chaotic particle swarm optimization with uniform
particle initialization (MUCPSO) to improve the
Applying "machine learning approach" (T1) to estimate comprehensive performance of standard PSO by presenting
software effort has garnered numerous studies and uncovered three additional schemes. Venkataiah et al. [39] proposed
several research trends. The paper counts from 2018 to 2022 chaotic linear increasing inertia weight and diversity-
for this research topic were 2, 12, 1, 7, and 5, respectively. The improved mechanism to enhance the variety of PSO. Alanis-
most widely explored are artificial neural networks (ANN), Tamez [48] et al. presented the application of PSO for
deep learning (DL), regression models, and ensemble improving the parameters of statistical regression equations
methods. Some ANN, DL, and regression models combined (SRE). Khuat et al. [56] proposed an improved algorithm
with several techniques then become a hybrid or ensemble combining the advantages of the artificial bee colony (ABC)
method, such as genetic algorithm and neural network in test and PSO algorithms. The variant improvements of PSO can
effort estimation [11]; deep learning and metaheuristic be compared and combined with another method, such as a
algorithm [12], [19]; bayesian optimization and ensemble machine learning approach, in future work to enhance the
learning [13]; deep learning and random forest [14]; fuzzy- effort estimation accuracy. Another research can be
neural network and metaheuristic algorithm [15]; ANN and conducted:
Taguchi method [20]; support vector regression (SVR) and
feature selection [22]; metaheuristic algorithm and regression • Investigate the impact of different methodologies,
model [26]; SVR and metaheuristic algorithm [27]; SVR and languages, datasets, and environments on algorithm
neural network [28]; feature selection and multilayer performance.
perceptron [30]; bayesian and synthetic bootstrap [33]; and
deep learning and function point [18].
627
C. Expert Judgement that related papers will be missed. We rigorously examined in
"Expert judgement" (T3) reveals the research trends, two steps to ensure the relevance of the selected publications
including analogy-based estimation [58], [59], [63]–[65], in this study. Another threat to validity refers to the naming of
[67], estimation in agile projects [61], [62], [66], [68] and research topics. It is an issue because of the subjectivity and
blockchain-based models [60]. The paper counts for this topic bias involved. To get around this constraint, we read the
in 2018, 2020, 2021, and 2022 were 3, 2, 1, and 5, journal's full text to determine appropriate and significant
respectively, and none in 2019. research topics.
628
Approach,” Comput. Intell. Neurosci., vol. 2022, 2022, doi: Models,” Turk. J. Electr. Eng. Comput. Sci., vol. 27, no. 2, pp. 1308–
10.1155/2022/3145956. 1324, 2019, doi: 10.3906/elk-1809-129.
[13] R. Marco, S. S. S. Ahmad, and S. Ahmad, “Bayesian Hyperparameter [31] H. M. Premalatha and C. V. Srikrishna, “Effort Estimation in Agile
Optimization and Ensemble Learning for Machine Learning Models Software Development Using Evolutionary Cost- Sensitive Deep
on Software Effort Estimation,” Int. J. Adv. Comput. Sci. Appl., vol. Belief Network,” Int. J. Intell. Eng. Syst., vol. 12, no. 2, pp. 261–269,
13, no. 3, pp. 419–429, 2022, doi: 10.14569/IJACSA.2022.0130351. 2019, doi: 10.22266/IJIES2019.0430.25.
[14] A. G. Priya Varshini, K. Anitha Kumari, and V. Varadarajan, [32] A. B. Nassif, M. Azzeh, A. Idri, and A. Abran, “Software Development
“Estimating Software Development Efforts Using A Random Forest- Effort Estimation Using Regression Fuzzy Models,” Comput. Intell.
Based Stacked Ensemble Approach,” Electron. Switz., vol. 10, no. 10, Neurosci., vol. 2019, 2019, doi: 10.1155/2019/8367214.
2021, doi: 10.3390/electronics10101195. [33] L. Song, L. L. Minku, and Y. A. O. Xin, “Software Effort Interval
[15] A. Karimi and T. J. Gandomani, “Software Development Effort Prediction via Bayesian Inference and Synthetic Bootstrap
Estimation Modeling Using A Combination of Fuzzy-Neural Network Resampling,” ACM Trans. Softw. Eng. Methodol., vol. 28, no. 1, 2019,
and Differential Evolution Algorithm,” Int. J. Electr. Comput. Eng., doi: 10.1145/3295700.
vol. 11, no. 1, pp. 707–715, 2021, doi: 10.11591/ijece.v11i1.pp707- [34] A. Zakrani, M. Hain, and A. Namir, “Software Development Effort
715. Estimation Using Random Forests: An Empirical Study and
[16] Z. R. Mohsin, “Investigating the Use of An Adaptive Neuro-Fuzzy Evaluation,” Int. J. Intell. Eng. Syst., vol. 11, no. 6, pp. 300–311, 2018,
Inference System in Software Development Effort Estimation,” Iraqi doi: 10.22266/IJIES2018.1231.30.
J. Comput. Sci. Math., vol. 2, no. 2, pp. 18–24, 2021, doi: [35] P. Rijwani and S. Jain, “Level of Confidence in Software Effort
10.52866/ijcsm.2021.02.02.003. Estimation by An Intelligent Fuzzy - Neuro - Genetic Approach,” Int.
[17] H. D. P. De Carvalho, R. Fagundes, and W. Santos, “Extreme Learning J. Adv. Comput. Sci. Appl., vol. 9, no. 9, pp. 568–576, 2018, doi:
Machine Applied to Software Development Effort Estimation,” IEEE 10.14569/ijacsa.2018.090972.
Access, vol. 9, pp. 92676–92687, 2021, doi: [36] A. Ardiansyah, R. Ferdiana, and A. E. Permanasari, “MUCPSO: A
10.1109/ACCESS.2021.3091313. Modified Chaotic Particle Swarm Optimization With Uniform
[18] K. Zhang, X. Wang, J. Ren, and C. Liu, “Efficiency Improvement of Initialization for Optimizing Software Effort Estimation,” Appl. Sci.
Function Point-Based Software Size Estimation With Deep Learning Switz., vol. 12, no. 3, 2022, doi: 10.3390/app12031081.
Model,” IEEE Access, vol. 9, pp. 107124–107136, 2021, doi: [37] Y. Alqasrawi, M. Azzeh, and Y. Elsheikh, “Locally Weighted
10.1109/ACCESS.2020.2998581. Regression With Different Kernel Smoothers for Software Effort
[19] M. S. Khan, F. Jabeen, S. Ghouzali, Z. Rehman, S. Naz, and W. Abdul, Estimation,” Sci. Comput. Program., vol. 214, 2022, doi:
“Metaheuristic Algorithms in Optimizing Deep Neural Network 10.1016/j.scico.2021.102744.
Model for Software Effort Estimation,” IEEE Access, vol. 9, pp. [38] A. Sharma and N. Chaudhary, “The Combined Model for Software
60309–60327, 2021, doi: 10.1109/ACCESS.2021.3072380. Development Effort Estimation Using Polynomial Regression for
[20] N. Rankovic, D. Rankovic, M. Ivanovic, and L. Lazic, “A New Heterogeneous Projects,” Radioelectron. Comput. Syst., vol. 2022, no.
Approach to Software Effort Estimation Using Different Artificial 2, pp. 75–82, 2022, doi: 10.32620/reks.2022.2.06.
Neural Network Architectures and Taguchi Orthogonal Arrays,” IEEE [39] V. Venkataiah, M. Nagaratna, and R. Mohanty, “Application of
Access, vol. 9, pp. 26926–26936, 2021, doi: Chaotic Increasing Linear Inertia Weight and Diversity Improved
10.1109/ACCESS.2021.3057807. Particle Swarm Optimization to Predict Accurate Software Cost
[21] T. Kangwantrakool, K. Viriyayudhakorn, and T. Theeramunkong, Estimation,” Int. J. Electr. Electron. Res., vol. 10, no. 2, pp. 154–160,
“Software Development Effort Estimation From Unstructured 2022, doi: 10.37391/IJEER.100218.
Software Project Description by Sequence Models,” IEICE Trans. Inf. [40] H. L. T. K. Nhung, V. Van Hai, R. Silhavy, Z. Prokopova, and P.
Syst., vol. E103D, no. 4, pp. 739–747, 2020, doi: Silhavy, “Parametric Software Effort Estimation Based on Optimizing
10.1587/transinf.2019IIP0014. Correction Factors and Multiple Linear Regression,” IEEE Access,
[22] A. Zakrani, M. Hain, and A. Idri, “Improving Software Development vol. 10, pp. 2963–2986, 2022, doi: 10.1109/ACCESS.2021.3139183.
Effort Estimation Using Support Vector Regression and Feature [41] S. P. Singh, V. P. Singh, and A. K. Mehta, “Differential Evolution
Selection,” IAES Int. J. Artif. Intell., vol. 8, no. 4, pp. 399–410, 2019, Using Homeostasis Adaption Based Mutation Operator and Its
doi: 10.11591/ijai.v8.i4.pp399-410. Application for Software Cost Estimation,” J. King Saud Univ. -
[23] L. L. Minku, “A Novel Online Supervised Hyperparameter Tuning Comput. Inf. Sci., vol. 33, no. 6, pp. 740–752, 2021, doi:
Procedure Applied to Cross-Company Software Effort Estimation,” 10.1016/j.jksuci.2018.05.009.
Empir. Softw. Eng., vol. 24, no. 5, pp. 3153–3204, 2019, doi: [42] M. A.-A. Alabajee, N. A. AL-Saati, and T. R. Alreffaee, “Parameter
10.1007/s10664-019-09686-w. Tuning of Software Effort Estimation Models Using Antlion
[24] S. P. Pillai, T. Radharamanan, and S. D. Madhukumar, “Feature Optimization,” Telkomnika Telecommun. Comput. Electron. Control,
Engineering for Enhanced Model Performance in Software Effort vol. 19, no. 3, pp. 817–828, 2021, doi:
Estimation,” Int. J. Recent Technol. Eng., vol. 8, no. 3, pp. 6053–6063, 10.12928/TELKOMNIKA.v19i3.16907.
2019, doi: 10.35940/ijrte.C5602.098319. [43] A. Moradbeiky, V. K. Bardsiri, and M. Jafari, “Open Hybrid Model:
[25] S. Rama Sree, S. N. S. V. S. C. Ramesh, and C. Prasada Rao, “An A New Esemble Model for Software Development Cost Estimation,”
Effective Software Effort Estimation Based on Functional Points Comput. Inform., vol. 39, no. 6, pp. 1148–1171, 2021, doi:
Using Soft Computing Techniques,” Int. J. Innov. Technol. Explor. 10.31577/CAI_2020_6_1148.
Eng., vol. 8, no. 10, pp. 3729–3733, 2019, doi: [44] R. Silhavy, P. Silhavy, and Z. Prokopova, “Using Actors and Use
10.35940/ijitee.J9674.0881019. Cases for Software Size Estimation,” Electron. Switz., vol. 10, no. 5,
[26] S. W. Ahmad and G. R. Bamnote, “Whale–Crow Optimization pp. 1–21, 2021, doi: 10.3390/electronics10050592.
(WCO)-Based Optimal Regression Model for Software Cost [45] J. A. Khan, S. U. R. Khan, T. A. Khan, and I. U. R. Khan, “An
Estimation,” Sadhana - Acad. Proc. Eng. Sci., vol. 44, no. 4, pp. 1–15, Amplified COCOMO-II Based Cost Estimation Model in Global
2019, doi: 10.1007/s12046-019-1085-1. Software Development Context,” IEEE Access, vol. 9, pp. 88602–
[27] R. Marco, N. S. Herman, and S. S. S. Ahmad, “Optimizing Software 88620, 2021, doi: 10.1109/ACCESS.2021.3089870.
Effort Estimation Models Based on Metaheuristic Methods: A [46] M. Daud and A. A. Malik, “Improving the Accuracy of Early Software
Proposed Framework,” Int. J. Adv. Trends Comput. Sci. Eng., vol. 8, Size Estimation Using Analysis-to-Design Adjustment Factors
no. 1.5 Special Issue, pp. 294–304, 2019, doi: (ADAFs),” IEEE Access, 2021, doi: 10.1109/ACCESS.2021.3085752.
10.30534/ijatcse/2019/5181.52019. [47] T. A. Khalid and E.-T. Yeoh, “Enhancing Software Development Cost
[28] V. K. Attri and J. S. Bal, “An Advanced Mechanism for Software Size Control by Forecasting the Cost of Rework: Preliminary Study,”
Estimation Using Combinational Artificial Intelligence,” Int. J. Intell. Indones. J. Electr. Eng. Comput. Sci., vol. 21, no. 1, pp. 524–537,
Eng. Syst., vol. 12, no. 4, pp. 255–263, 2019, doi: 2021, doi: 10.11591/ijeecs.v21.i1.pp524-537.
10.22266/ijies2019.0831.24. [48] M. D. Alanis-Tamez, C. López-Martín, and Y. Villuendas-Rey,
[29] M. M. Rejab, N. F. M. Azmi, and S. Chuprat, “Fuzzy Delphi Method “Particle Swarm Optimization for Predicting the Development Effort
for Evaluating HyTEE Model (Hybrid Software Change Management of Software Projects,” Mathematics, vol. 8, no. 10, pp. 1–21, 2020, doi:
Tool With Test Effort Estimation),” Int. J. Adv. Comput. Sci. Appl., 10.3390/math8101819.
vol. 10, no. 4, pp. 529–535, 2019, doi: 10.14569/ijacsa.2019.0100465. [49] A. A. Fadhil, R. G. H. Alsarraj, and A. M. Altaie, “Software Cost
[30] E. Ertuğrul, Z. Baytar, Ç. Çatal, and C. Muratli, “Performance Tuning Estimation Based on Dolphin Algorithm,” IEEE Access, vol. 8, pp.
for Machine Learning-Based Software Development Effort Prediction 75279–75287, 2020, doi: 10.1109/ACCESS.2020.2988867.
629
[50] K. Rak, Ž. Car, and I. Lovrek, “Effort Estimation Model for Software [66] M. Usman, K. Petersen, J. Börstler, and P. Santos Neto, “Developing
Development Projects Based on Use Case Reuse,” J. Softw. Evol. and Using Checklists to Improve Software Effort Estimation: A Multi-
Process, vol. 31, no. 2, 2019, doi: 10.1002/smr.2119. Case Study,” J. Syst. Softw., vol. 146, pp. 286–309, 2018, doi:
[51] L. P. dos Santos and M. Ferreira, “Applying COCOMO II for A DO- 10.1016/j.jss.2018.09.054.
178C Safety-Critical Software Effort Estimation,” J. Aerosp. Technol. [67] Ardiansyah, M. M. Mardhia, and S. Handayaningsih, “Analogy-Based
Manag., vol. 11, 2019, doi: 10.5028/jatm.v11.1031. Model for Software Project Effort Estimation,” Int. J. Adv. Intell.
[52] P. Silhavy, R. Silhavy, and Z. Prokopova, “Categorical Variable Inform., vol. 4, no. 3, pp. 251–260, 2018, doi: 10.26555/ijain.v4i3.266.
Segmentation Model for Software Development Effort Estimation,” [68] M. Usman, R. Britto, L.-O. Damm, and J. Börstler, “Effort Estimation
IEEE Access, vol. 7, pp. 9618–9626, 2019, doi: in Large-Scale Software Development: An Industrial Case Study,” Inf.
10.1109/ACCESS.2019.2891878. Softw. Technol., vol. 99, pp. 21–40, 2018, doi:
[53] F. Sarro and A. Petrozziello, “Linear Programming as A Baseline for 10.1016/j.infsof.2018.02.009.
Software Effort Estimation,” ACM Trans. Softw. Eng. Methodol., vol. [69] S. Sharma, P. C. Pendharkar, and Karmeshu, “Learning Component
27, no. 3, 2018, doi: 10.1145/3234940. Size Distributions for Software Cost Estimation: Models Based on
[54] S. Bilgaiyan, K. Aditya, S. Mishra, and M. Das, “A Swarm Intelligence Arithmetic and Shifted Geometric Means Rules,” IEEE Trans. Softw.
Based Chaotic Morphological Approach for Software Development Eng., vol. 48, no. 12, pp. 5136–5147, 2022, doi:
Cost Estimation,” Int. J. Intell. Syst. Appl., vol. 10, no. 9, pp. 13–22, 10.1109/TSE.2021.3139216.
2018, doi: 10.5815/ijisa.2018.09.02. [70] V. V. Hai, H. L. T. K. Nhung, Z. Prokopova, R. Silhavy, and P.
[55] D. Nandal and O. P. Sangwan, “Software Cost Estimation by Silhavy, “A New Approach to Calibrating Functional Complexity
Optimizing COCOMO Model Using Hybrid BATGSA Algorithm,” Weight in Software Development Effort Estimation,” Computers, vol.
Int. J. Intell. Eng. Syst., vol. 11, no. 4, pp. 250–263, 2018, doi: 11, no. 2, 2022, doi: 10.3390/computers11020015.
10.22266/ijies2018.0831.25. [71] S. S. Gautam and V. Singh, “Adaptive Discretization Using Golden
[56] T. T. Khuat and M. H. Le, “A Novel Hybrid ABC-PSO Algorithm for Section to Aid Outlier Detection for Software Development Effort
Effort Estimation of Software Projects Using Agile Methodologies,” Estimation,” IEEE Access, vol. 10, pp. 90369–90387, 2022, doi:
J. Intell. Syst., vol. 27, no. 3, pp. 489–506, 2018, doi: 10.1515/jisys- 10.1109/ACCESS.2022.3200149.
2016-0294. [72] V. Van Hai, H. L. T. K. Nhung, Z. Prokopova, R. Silhavy, and P.
[57] S. Yigit-Sert and P. Kullu, “Software Cost Estimation Using Enhanced Silhavy, “Toward Improving the Efficiency of Software Development
Artificial Bee Colony Algorithm,” Int. J. Adv. Comput. Sci. Appl., vol. Effort Estimation via Clustering Analysis,” IEEE Access, vol. 10, pp.
9, no. 4, pp. 67–70, 2018, doi: 10.14569/IJACSA.2018.090412. 83249–83264, 2022, doi: 10.1109/ACCESS.2022.3185393.
[58] E. M. De Bortoli Fávero, D. Casanova, and A. R. Pimentel, “SE3M: A [73] P. Rai, D. K. Verma, and S. Kumar, “A Hybrid Model for Prediction
Model for Software Effort Estimation Using Pre-Trained Embedding of Software Effort Based on Team Size,” IET Softw., vol. 15, no. 6, pp.
Models,” Inf. Softw. Technol., vol. 147, 2022, doi: 365–375, 2021, doi: 10.1049/sfw2.12048.
10.1016/j.infsof.2022.106886. [74] P. Silhavy, R. Silhavy, and Z. Prokopova, “Spectral Clustering Effect
[59] A. Kaushik, P. Kaur, N. Choudhary, and Priyanka, “Stacking in Software Development Effort Estimation,” Symmetry, vol. 13, no.
Regularization in Analogy-Based Software Effort Estimation,” Soft 11, 2021, doi: 10.3390/sym13112119.
Comput., vol. 26, no. 3, pp. 1197–1216, 2022, doi: 10.1007/s00500- [75] K. Ono, M. Tsunoda, A. Monden, and K. Matsumoto, “Influence of
021-06564-w. Outliers on Estimation Accuracy of Software Development Effort,”
[60] M. Ahmed et al., “Blockchain-Based Software Effort Estimation: An IEICE Trans. Inf. Syst., vol. E104D, no. 1, pp. 91–105, 2021, doi:
Empirical Study,” IEEE Access, vol. 10, pp. 120412–120425, 2022, 10.1587/transinf.2020MPP0005.
doi: 10.1109/ACCESS.2022.3216840. [76] M. Gan, Z. Yücel, A. Monden, and K. Sasaki, “Empirical Evaluation
[61] P. Sudarmaningtyas and R. Mohamed, “Significant Factors in Agile of Mimic Software Project Data Sets for Software Effort Estimation,”
Software Development of Effort Estimation,” Pertanika J. Sci. IEICE Trans. Inf. Syst., vol. E103D, no. 10, pp. 2094–2103, 2020, doi:
Technol., vol. 30, no. 4, pp. 2851–2878, 2022, doi: 10.1587/transinf.2019EDP7150.
10.47836/pjst.30.4.30. [77] I. Hassan, A. Latif, K. Khan, F. Hassan, and M. Saeed, “Accurate Cost
[62] L. Cao, “Estimating Efforts for Various Activities in Agile Software Estimation for Software With Volatile Requirements,” Int. J. Adv.
Development: An Empirical Study,” IEEE Access, vol. 10, pp. 83311– Appl. Sci., vol. 7, no. 3, pp. 1–8, 2020, doi:
83321, 2022, doi: 10.1109/ACCESS.2022.3196923. 10.21833/ijaas.2020.03.001.
[63] M. Dashti, T. J. Gandomani, D. H. Adeh, H. Zulzalil, and A. B. M. [78] M. F. Bosu and S. G. Macdonell, “Experience: Quality Benchmarking
Sultan, “LEMABE: A Novel Framework to Improve Analogy-Based of Datasets Used in Software Effort Estimation,” J. Data Inf. Qual.,
Software Cost Estimation Using Learnable Evolution Model,” PeerJ vol. 11, no. 4, 2019, doi: 10.1145/3328746.
Comput. Sci., vol. 7, 2021, doi: 10.7717/PEERJ-CS.800. [79] H. Huynh Thai, P. Silhavy, M. Fajkus, Z. Prokopova, and R. Silhavy,
[64] M. A. Shah, D. N. A. Jawawi, M. A. Isa, M. Younas, A. Abdelmaboud, “Propose-Specific Information Related to Prediction Level at x and
and F. Sholichin, “Ensembling Artificial Bee Colony With Analogy- Mean Magnitude of Relative Error: A Case Study of Software Effort
Based Estimation to Improve Software Development Effort Estimation,” Mathematics, vol. 10, no. 24, 2022, doi:
Prediction,” IEEE Access, vol. 8, pp. 58402–58415, 2020, doi: 10.3390/math10244649.
10.1109/ACCESS.2020.2980236.
[65] V. Resmi and S. Vijayalakshmi, “Analogy-Based Approaches to
Improve Software Project Effort Estimation Accuracy,” J. Intell. Syst.,
vol. 29, no. 1, pp. 1468–1479, 2020, doi: 10.1515/jisys-2019-0023.
630
View publication stats