A Test Management System to Support Remote Usability Assessment of Web Applications

Round 1

Reviewer 1 Report

The authors present a web platform for remote usability tools to help moderators to collect and analyze data. There is a hot topic in informatics and web systems. This topic has a relevant interest in the community to help the developers get the users' behavior. 

 

Overall the paper is interesting and well presented, but I would like to point out some issues that I will list below:

 

- The abstract needs to be revised and written in one block, not in 3. 

 

- The authors should review the template.

 

- The authors should adjust the image quality and increase their size. Maybe an architecture or a flowchart would improve the work. Figures 3, 5, 6, 10, 11, and 15 must improve the quality and translate

 

- The references do not have the correct format. They should be revised and must include de DOI. 

Author Response

We would like to express our sincere gratitude to the reviewer for the constructive comments and valuable remarks on our manuscript and for the time and effort put in it.
Below, we outline how we have handled each of the received suggestions, providing a point-by-point response.

Point 1: The abstract needs to be revised and written in one block, not in 3.

Response 1: We revised and refined the contents of the abstract, rewriting it in one block as you suggested. We also described as concisely as possible the purpose of the study, the methodology proposed, the results obtained, and the conclusions drawn.

Point 2: The authors should review the template.

Response 2: The template has been revised.

Point 3: The authors should adjust the image quality and increase their size. Maybe an architecture or a flowchart would improve the work. Figures 3, 5, 6, 10, 11, and 15 must improve the quality and translate.

Response 3: We adjusted the image quality and increased the size of the suggested images. We added a flowchart describing the frame collection and processing. We removed the image with Italian text.

Point 4: The references do not have the correct format. They should be revised and must include de DOI.

Response 4: The references have been revised and updated. They now inlcude the DOI.

Reviewer 2 Report

The authors proposed a deep learning-based system to support remote usability assessment of web applications, which is a interesting topic. However, further improvements are needed before publication.

1. The author's description of the method is not clear enough. For example, the author proposed a framework based on deep learning to support the remote availability assessment of Web applications. In any case, I have not seen the specific composition of the deep learning framework used by the author. The author should detailed the deep learning framework in the form of a flowchart and analyzes its each composition.

2. The abstracts need to be further refined. In general, the abstract needs to describe as concisely as possible the purpose of the study, the methodology proposed, the results obtained, and the conclusions drawn. 

3. The conclusion is too cumbersome, it should be more concise. The conclusion should show a summary of the method and conclude the results obtained based on the method. Finally, the problem that the current method cannot be solved should be explained.

4. References need to be further updated. Many references are relatively old and should present as much as possible in recent years.

Author Response

We would like to express our sincere gratitude to the reviewer for the constructive comments and valuable remarks on our manuscript and for the time and effort put in it.
Below, we outline how we have handled each of the received suggestions, providing a point-by-point response.

Point 1: The author's description of the method is not clear enough. For example, the author proposed a framework based on deep learning to support the remote availability assessment of Web applications. In any case, I have not seen the specific composition of the deep learning framework used by the author. The author should detailed the deep learning framework in the form of a flowchart and analyzes its each composition.

Response 1: The framework has been described through two architectural diagrams (Figures 1 and 3) and citations have been given to the articles that specifically describe the behavior of each module of the framework. To improve the understanding of the data flow for the Deep Learning component, a flowchart has been added as suggested by the reviewer (Figure 2).

Point 2: The abstracts need to be further refined. In general, the abstract needs to describe as concisely as possible the purpose of the study, the methodology proposed, the results obtained, and the conclusions drawn.

Response 2: We revised and refined the contents of the abstract, rewriting it in one block as you suggested. We also described as concisely as possible the purpose of the study, the methodology proposed, the results obtained, and the conclusions drawn.

Point 3: The conclusion is too cumbersome, it should be more concise. The conclusion should show a summary of the method and conclude the results obtained based on the method. Finally, the problem that the current method cannot be solved should be explained.

Response 3: We revised the conclusions as suggested by the reviewer, summarizing the concepts in two parts: in the first we described the tool with respect to the obtained results, and in the second we described the weaknesses and possible future developments.

Point 4: References need to be further updated. Many references are relatively old and should present as much as possible in recent years.

Response 4: Most references have been revised and updated. Moreover, they now inlcude the DOI.

Round 2

Reviewer 2 Report

I agree to accept this paper for publication in the journal, but I still suggest the author to further improve the article. The author did not respond to my concern: what is the deep learning model used? What is the operating mechanism of model? Training parameters of the model? Is the model a CNN architecture or something else? How many layers does the model have? Training loss and learning rate? Configuration of input layer, hidden layer and output layer? .....

Author Response

Point 1: The author did not respond to my concern: what is the deep learning model used? What is the operating mechanism of model? Training parameters of the model? Is the model a CNN architecture or something else? How many layers does the model have? Training loss and learning rate? Configuration of input layer, hidden layer and output layer?

Response 1: Although it is foregrounded in the title (which has now been changed by removing the word "Deep Learning"), the purpose of our work is to propose a framework based on Deep Learning technologies, but without going into the technical details behind the CNN training. This decision was made primarily to avoid burdening the article with details that have already been published and that we felt were unnecessary to report again. However, reference was made to the use of Convolutional Neural Networks. For further details, we refer the reviewer to our articles (references 51 and 53 in this paper).

• A. Generosi, S. Ceccacci, S. Faggiano, L. Giraldi and M. Mengoni, "A Toolkit for the Automatic Analysis of Human Behavior in HCI Applications in the Wild," Advances in Science, Technology and Engineering Systems Journal, vol. 5, no. 6, pp. 185--192, 2020, doi: 10.25046/aj050622.
• A. Talipu, A. Generosi, M. Mengoni and L. Giraldi, "Evaluation of Deep Convolutional Neural Network architectures for Emotion Recognition in the Wild," in 2019 IEEE 23rd International Symposium on Consumer Technologies (ISCT), 2019 , pp. 25-27. doi: 10.1109/ISCE.2019.8900994.

Please see the attachment with our already published articles.

Author Response File: Author Response.pdf