On Fuzzy and Case-Based Dynamic Software Development Process Modeling and Simulation Approach

Comments to the Author

Authors’ answers

Literature review is exhaustive, and the authors seem to have reviewed a lot of recently  published journal articles. HOWEVER, the authors seem to miss the research gaps. What is lacking in the works of previous researchers

The purpose of literature review is to show that similar approaches are existing but none of them are suitable for HF impact prediction.

In line 236:
Existing approaches have their advantages and disadvantages, some of which are presented above. Moreover, there are difficulties in integrating the analyzed models, because, as the aforementioned author states, they are not described clearly enough. As such, additional research is needed to investigate process simulation – or, more precisely, SDP simulation.

Some disadvantages of the analyzed approaches are also presented in the text, like:

Line 200:

COCOMO also has its disadvantages, such as its ignorance of requirements and documentation, hardware issues, personnel turnover levels, and its dependence on the amount of time spent on each phase [30].

Line 213:

The main disadvantage of the traditional Monte Carlo method is that it is suitable for static data only, i.e., it is not suitable for run-time adaptations during simulation; moreover, it is difficult to conduct large-scale maneuvers in Monte Carlo [16].

Line 225:

The main disadvantage of the suggested approach is that the HF (stress level, motivation, experience, etc.) is not considered.

why you use fuzzy in the SDP Model? what is addvantage of fuzzy in your study

We are using fuzzy logic in the SDP model: “Because calculating and predicting the influence of the HF is difficult and can differ according to subject areas, in this research we are going to use fuzzy logic to express HFs when simulating the SDP”. Please see line 68.

figure 2 and 3 are not clear.

We have enlarged Fig. 2

We have enlarged Fig. 3

Information about data and dataset must be clearly discussed in the research method

We have added sub-section 4.7 to describe the dataset (please see line 446)

in ANFIS Validation you use MSE and RMSE. in fact both of them have same logic. MSE=RMSE^2. please add the The coefficient of determination (R²) of model.

We have added (R²) in Table 6.

The conclusion section should be divided into several paragraphs. The assumptions and limitations should be discussed. Future research related to the results of your paper is suggested to be added in this paper.

We have added sub-section 6.1, that helps us to divide Discussions and Conclusions sections. In sub-section 6.1 we have discussed limitations and future works.

Comments to the Author

Authors’ answers

The title of the paper is not clear. It should be improved.

The title of the paper is improved.

The words used in the title should not be used as keywords. It will improve the searchability of articles in search engines.

We have changed our keywords that it would not double title.

In the first line itself references number is [48], which is not correct. The references should be serially arranged.

The list of references is serially arranged.

Few sentences are not clear like page 3 line 98  “The novelties of the research presented in this paper are in its domain, the use of 98 fuzziness, and the combination of CMMN and ANFIS into one simulation model.

We have changed the novelties of the research to be clearer. Please see line 98: “The novelty of the current research is the combination of fuzzy inference applying ANFIS with case-based process modelling using CMMN into one simulation model”.

The legends and text are not visible in Figure 1.

We have enlarged text in Fig. 1

The ANFIS is used in the study. This section may be improved with similar studies like - 10.3390/batteries9010013 and https://doi.org/10.3390/pr8030312. You may also use the article "Recent advances in machine learning research for nanofluid-based heat transfer in renewable energy system." 

Thank you. We reviewed these studies and added following text in line 228: “Other authors already used machine learning methodologies for optimizing and modelling processes in different domains. For example, authors in [35] use machine learning for engineering challenges in non-linearity and complexity areas. Other authors in [36] use data-driven machine learning techniques for forecasting thermophysical features and hear transfer rate. Authors reviewed literature to highlight current advances in machine learning for their domains. They had concluded that machine learning techniques can accurately predict nanofluid characteristics.”

Equatiosn used in section 4 should be cited.

We have added all equations in section 4.

The novelty of the study should be clearly shown in last section of the introduction.

The novelty is presented in line 99: “The novelty of the current research is the combination of fuzzy inference applying ANFIS with case-based process modelling using CMMN into one simulation model”

How was the model prevented from overfitting?

In the current research during experiments, we have not got the overfitting of the model. However, we are going to investigate this behavior in the future research. Moreover, we have added the following text in line 738: “Investigate the overfitting behavior of the model.”

What is the meaning of possible deviation in Table 3. This is only input data. 

The explanation is added in line 533: “These possible deviations are values that show how the initial parameter values (in columns 1-3) could vary because of HF influence. For example: if employee motivation, experience and availability are low, possible deviation value from estimated task execution time is 1.”

Figure 8 is not adding any value to the study. It may be removed.

The idea of adding Fig. 8 in the paper is presenting training and testing results with selected parameters.

Provide more context and background information to help readers better understand

We have added explanatory text to provided comments. Also, we have improved our paper text for better understanding according to another reviewer comments.

However, if there are still unclear text, we will be happy to improve it.  

The conclusion should summarize primary outcome of the study.

We have added sub-section 6.1, that helps us to divide Discussions and Conclusions sections. In sub-section 6.1 we have discussed limitations and future works. In conclusions we have left only summary of primary outcome.