A Novel Approach for Web Service Recommendation Based on Advanced Trust Relationships
Round 1
Reviewer 1 Report
Tha research is interesting and well written. In the following line I am making some recommendation for improving the current article.
The authors should correct the following typos:
-line 29, after users the word that should be added;
-line 33, the word means should be changed to mean;
-line 41, the acronym QoS should be explained;
In introduction the authors should give more insights about the keyword web services as readers might think the article is about architectural principle of web service (and is not). As such, I recommend them to add a sentence/short paragraph where they should explain web service recommendation.
In line 293 the authors specify that 150 users are divided into two groups but they fail to explain why they distribute them unevenly across the two sets(80% for training set and 20% for test set). Please explain!
Also, they are talking about 100 web services but do not give details about it. They should.
In the results sections, the authors should present and detail the test tool/s that they have used for performing the test (included in table 1 for example). Besides the physical platform that they describe in lines 276-278 (and is not relevant) they should mainly present details about the tool (such as Apache JMeter, for e.g.) used for running the tests.
Also, what are the criterias are taken into account for evaluating/analysing the service recommendations? In the research is not specified.
The authors should clearly state the instrument used for data and charts analysis (SPSS, STATA, EXCEL, etc).
The authors should include the source for all tables/figures? If created by the authors, should state: "Source: Authors own projection".
The Conclusions are far too narrow and they should be more elaborated.
The authors should also include the limitations of their study.
Author Response
Response to Reviewer 1 Comments
Point 1: The authors should correct the following typos:
-line 29, after users the word that should be added;
-line 33, the word means should be changed to mean;
-line 41, the acronym QoS should be explained;
Response 1: We have corrected the above typos.
Point 2: In introduction the authors should give more insights about the keyword web services as readers might think the article is about architectural principle of web service (and is not). As such, I recommend them to add a sentence/short paragraph where they should explain web service recommendation.
Response 2: We have added a long sentence to explain web service recommendation in lines 33-35.
Point 3: In line 293 the authors specify that 150 users are divided into two groups but they fail to explain why they distribute them unevenly across the two sets(80% for training set and 20% for test set). Please explain!
Response 3: The algorithm proposed in this paper belongs to the category of machine learning in essence. In the algorithm test of machine learning, in order to avoid over-fitting or under-fitting, 80% of the data will be usually used for training and the remaining 20% for testing.
Point 4: Also, they are talking about 100 web services but do not give details about it. They should.
Response 4: We have added a table (Table1) to give details about the web services in lines 302-305.
Point 5: In the results sections, the authors should present and detail the test tool/s that they have used for performing the test (included in table 1 for example). Besides the physical platform that they describe in lines 276-278 (and is not relevant) they should mainly present details about the tool (such as Apache JMeter, for e.g.) used for running the tests.
Response 5: We have presented details about the tools used for running the tests in lines 291-293.
Point 6: Also, what are the criterias are taken into account for evaluating/analysing the service recommendations? In the research is not specified.
Response 6: Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) are chosen as the measurement metrics for evaluating service recommendations in this paper. We have given the reason in lines 319-325.
Point 7: The authors should clearly state the instrument used for data and charts analysis (SPSS, STATA, EXCEL, etc).
Response 7: We have stated that EXCEL is the instrument used for data and charts analysis in line 332.
Point 8: The authors should include the source for all tables/figures? If created by the authors, should state: "Source: Authors own projection".
Response 8: We have added a statement about all tables and figures in lines 427-429.
Point 9: The Conclusions are far too narrow and they should be more elaborated. The authors should also include the limitations of their study.
Response 9: We have modified the Conclusions in lines 411-414, 421-423, 425-426.
Thank you for your valuable comments！
Author Response File: 
Author Response.docx
Reviewer 2 Report
The paper is well written and it deals with an interesting topic. Although, I am not an expert for recommender systems, content is comprehensible for me. Formula and algorithm seems to be fine
Author Response
Response to Reviewer 2 Comments
Point 1: The paper is well written and it deals with an interesting topic. Although, I am not an expert for recommender systems, content is comprehensible for me. Formula and algorithm seems to be fine
Response 1: Thank you for your affirmation and wish you all the best！
Author Response File: Author Response.docx
Reviewer 3 Report
- In Related Work section, the authors should more analyst about differences between the proposed method (based on true relationship) and related papers.
- In Formalization of Trust Relationship and Recommendation Mechanism sections, the authors should present the useful and proposed reasons of the formulas and the algorithm. Besides, some examples should be present to clarify the ideas.
- In Experiment and Analysis section, the datasets need to be written clearer. Why do the authors select the datasets?
Author Response
Response to Reviewer 3 Comments
Point 1: In Related Work section, the authors should more analyst about differences between the proposed method (based on true relationship) and related papers.
Response 1: We have modified the differences between the proposed method and related papers in lines 54-59.
Point 2: In Formalization of Trust Relationship and Recommendation Mechanism sections, the authors should present the useful and proposed reasons of the formulas and the algorithm. Besides, some examples should be present to clarify the ideas.
Response 2: We have added the relevant descriptions in lines 136-137, 158-159, 176-178, 194-195, 221-223, 272-274.
Point 3: In Experiment and Analysis section, the datasets need to be written clearer. Why do the authors select the datasets?
Response 3: We have added a table (Table1) to give details about the dataset in lines 302-305 and given the reason in lines 299-300.
Thank you for your valuable comments！
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors carefully approached all my suggestions. As such, the present version of the current research paper is, in my oppinion, very much improved.
Therefore, I consider the article can be published within Information Journal.
Author Response
Point 1: The authors carefully approached all my suggestions. As such, the present version of the current research paper is, in my opinion, very much improved.
Therefore, I consider the article can be published within Information Journal.
Response 1: With the help of a colleague in English major, we have further improved the expression of the manuscript. Thank you for your affirmation and wish you all the best！
Author Response File: Author Response.docx
Reviewer 3 Report
The authors should present about the complexity of SRATR algorithm.
Author Response
Point 1: The authors should present about the complexity of SRATR algorithm.
Response 1: We have added an analysis of the complexity of SRATR algorithm in lines 278-281.
Thank you for your valuable comments and wish you all the best！
Author Response File: Author Response.docx
