Analysis of a Personalized Provision of Service Level Agreement (SLA) Algorithm

Round 1

Reviewer 1 Report

In this paper, the authors have presented an analysis of the process executed when automating the task of assigning a customer SLA. The topic is interesting. Some comments are given as follows:

1. The background of service level agreement should be enriched.

2. The advantages and disadvantages should be summarized and listed.

3. More compared works should be given.

4. The Fig. 22 is not so clear. More description should be given.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

An analysis of the process executed is presented for automating the task of assigning a customer SLA. The presentation is clear with fine simulations. I recommend the acceptance of this paper. Since the proposed work is based on process mining, two papers related to the mining technology, the authors may introduce those works so as to broaden the scope of this paper.

1. Le Zhang, Su Xu, Xiangjun Li, Xiaoliang Wu and Pei-Chann Chang, “An Improved FP-Growth Algorithm Based on Projection Database Mining in Big Data”, Journal of Information Hiding and Multimedia Signal Processing, Vol. 10, No. 1, pp. 81-90, January 2019

2. Philippe Fournier-Viger, Jerry Chun-Wei Lin, Rage Uday Kiran, Yun Sing Koh, and Rincy Thomas, “A Survey of Sequential Pattern Mining,” Data Science and Pattern Recognition, vol. 1(1), pp. 54-77, 2017

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The paper presents the use of several mechanisms for the implementation of a very specific problem, i.e. the identification of needs based on the answers to many questions presented in Figure 6. The focus was on examining the effectiveness of the customer classification process in this particular problem adopted in the work.

Comments:

1) There is no technical information about the mechanisms used. So please add information, e.g. about petri nets or how to implement transitions between nodes of the graph in Figure 6 in the form of pseudocode.

2) Designation of drawings not in sequence, e.g. after figure 2 is figure 5

3) Marking Figure 5,7,8,16,17,18,19 replace with tables

4) Figure 20 is illegible - increase the resolution

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

The goal of the paper is formulated in a vague way, so it might not be properly evaluated. The paper presents no real comparison to the relevant state-of-the-art literature.

The authors employ diverse methods, including process discovery, conformance checking, and classification, using artificially generated data. However, there is no clear goal or motivation for such research.

The process model is created in an inappropriate way which does not explain the data nor provide any insight. 

In fact, it does not have any meaning, so it is hard to infer anything from it. Measuring fitness (in fact it is replay fitness) is not informative. For process mining, other metrics should be calculated as well. It is easy to create a model with 100% fitness which has no meaning at all like here. 

Some references, e.g. [1] are not related to the paper content. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 4 Report

The authors have removed the methodologically controversial parts of the paper. I do not have further suggestions.