Strategy Mining for Inferring Business Information System User Intentions
Round 1
Reviewer 1 Report
The paper presents a good and novel idea of understanding the communication messages of employees and act one them. The authors also provided access to the python code. However, this work is very poorly presented though out the paper. The readers have to reach section 4 and figure 1 until they understand what the paper is about. The paper needs to be completely rewritten as a scientific journal before it can be considered for publication. Some of the main issues with this work and how they can be improved are summarized below.
The title is too generic. Please consider something more specific.
After reading the abstract, it is not clear from the abstract what the novelty of the paper is. The authors state “The possibility to improve business information system” This wrong in terms in English as well as not clear. I assume that they mean that their work improves business information systems using user behavior data. Again, they need to be more specific on what they improve as well as what they mean by user behavior data. In addition, the abstract does not contain any results. Please rewrite the abstract and highlight the novelties of the work as well as the results.
The introduction is written as bullet points. Please rewrite it as paragraphs with flow. Each paragraph must contain its own introduction and conclusion. Sections 1.1 to 1.4 are meanness. Please remove them. After reading the introduction, it is still not clear to me what the paper is about. The introduction needs to be completely rewritten. Please include a figure as an overview in the introduction, explain the proposed approach. Something similar to figure 1 but more high level.
In the related work section, authors are expected to provide a summary of other research works that are trying to solve the same problem. In the same section, ideally, with a table, they are expected to summarize the related work and explain the differences and novelties of their work. This will allow the reviewers to decide if the paper is worth publishing or not. Please take this one into account and rework the related work. We are a looking for small section of text explaining the similar approaches. Not bullet points.
Section 4 is one of the best sections of the paper. The approach is very clearly explained and easy to follow. The authors also provided the python code. Please create a new reference for the github link in line 221, add it as a new reference in the references section and cite it in the text as a normal reference.
In addition, I can still see the same problem of writing in bullet points. It looks like the work was copy pasted from a power point presentation. Some material can be moved to the appendix section at the end of the paper and be replaced with explanations in English.
The conclusion section needs to include a summary of the work in some detail, the results, the limitations and the future work.
Author Response
Please see the attachment
Author Response File: 
Author Response.docx
Reviewer 2 Report
The article presents a proposal for improving business information systems with an approach based on user behavior (from user strategies).
The research addresses a current and relevant topic.
Some improvement comments:
- Many topics in the text (including the Introduction) are presented in bullets; I suggest turning them into plain text to make it easier for the reader to understand.
I congratulate the authors on their research.
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 3 Report
This paper describer an approach to improve business information systems based the user strategies and behavior.
The Abstract should also detail the lines of the approach, how it was evaluated, and what were the conclusions of the experiment.
“ create a KB for any business” seems to us to be an overly broad statement.
Does the proposed database allow for an acceptable time to work, when the volume of documents runs into the millions?
The writing quality is not uniform... it varies across sections.
Sentences such as the one bellow (Abstract) show that the text needs a very deep revision in writing.
“... this method is developed from the activities (strategies) contains in a knowledge
base all businesses in general, and the strategies contained in an event log a specific business”
The main references mentioned in section 2 are from 2013 and 2014... No more recent work?
“From differences between the current process and real process of the business the user's strategies are identified”
What about if a user follows all the current process steps? Will this approach detect the (non-original) strategy and user intention?
Section 2.1 has too much text taken from a previous post...
Section 5 doesn't exactly present results, but rather a recap of resources and strategy…
At the end of this section, the reader does not know whether there is an objective assessment of the proposed approach.
Section 6.1 seems redundant. A critical analysis of the results is expected... not a recap of resources and where to get them.
In Discussion, it would be important to mention the correctness that the NLP tasks have for the texts/phrases used in this work.
Section 7 has sentences without punctuation marks.
How many sentences about the concrete domain of sales did this work use?
How many records does the system database have? Will it scale to a real system?
We believe there is little detail about the analysis of logs and the NLP and mining process.
* Other minor Comments/Remarks on writing:
We've consider
→ We have considered
29
the daily activities are develop
the daily activities are developed
36
reach the aims business
reach the business aims
65 and 77
In an item enumeration, all but the last one end with ".", but the last one ends with the text without "."
118 – there is an unexpected font size/type change
140 “ISE” was already defined?
283
The text is divide into
The text is divided into
341
An unconfirmed order or a quote compromises the stock for a certain time
An unconfirmed order or quote compromises the stock for a certain time
# but authors should avoid the “order or quote” repetition
351, 464, 528, and 543:
items: maybe you could end the last with “.”, and the previous items with “;”
421
From differences between the current process and real process of the business (Figure
2 and Figure 3); the user's strategies are identified;
From differences between the current process and real process of the business (Figure
2 and Figure 3) the user's strategies are identified;
# here, we suggest not to split that sentence
423
Based on the user's strategies, we infer their intentions in the development of busi-
ness activities, the experimental results are presented below.
# split sentences:
Based on the user's strategies, we infer their intentions in the development of busi-
ness activities. The experimental results are presented below.
422
answer the re-search question
answer the research question
429
From news (one million articles), randomly form it is takes the 200 articles
please rewrite
514
the user intentions are infer based
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 4 Report
Dear Authors,
The manuscript presents an interesting topic of Intention Mining of Business Information System Users. the work can be useful for those interested in Intention Mining approaches. I acknowledge the effort made by the authors to prepare the manuscript. However, in my opinion, major revisions are needed, before its consideration for possible publication in Applied Sciences journal.
Recommendations:
- Abstract: the concept and the goal of your work is not addressed correctly, try to enhance this section.
- What is the novelty of this work (i.e. scientific contribution in this field)?
- Before the authors explain their concepts in the introduction section, a clear context about their work should be stated.
- There are some problems with English language, I attach a file with some improving suggestions
Good luck.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors addressed all the comments. The paper can be published.
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 3 Report
In section 5, Results, there is still content on materials and methods, which normally comes first.
Other details:
"7. Conclusions."
or
"7. Conclusions" # headings don't usually end with punctuation
"In the absence of data on a specific business. The data that were used..."
# It's one of those examples where we thought it would be better to tweak the text:
"In the absence of data on a specific business, the data used..."
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 4 Report
I thank the authors for their efforts to improve the manuscript. There is a noticeable improvement .However , the grammatical quality of the manuscript must be greatly enhanced . English proofreading is mandatory.
Good luck.
Author Response
Please see the attachment
Author Response File: Author Response.docx
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
Dear Authors,
thank you for the article. The authors performed an analysis of the problem area and summarized the obtained results. Providing more details about the used machine learning algorithm its settings, learning process, evaluating and obtained results would be beneficial for the readers.
The manuscript contains few typos but these do not affect the overall quality of the article.
There are some comments that I hope, can be beneficial to the article:
- 1/45-50 (page 1/line 45 to 50) – it would be appropriate to present in more detail the conditions under which the used articles were selected
- 9/270 – it would be appropriate to specify in more detail the training process in more detail, e.g. number of used records, dividing of the training set into training, testing and validation subsets, achieved results, accuracy, algorithm used ...
- 11/357 – it would be appropriate to specify in more detail used machine learning algorithm, its setting, learning process, obtained results...
- 2/55 – “The chosen operating sys-tem 55 was Ubuntu…” – dash in the word system
- 2/59 – “…we chose Python programming language for coding and.”– unfinished sentence
- 4/129 – “The cod of this process is shown in the Appendix B.” – the missing letter “e” in the word code
- 4/139 – “The cod of this process is shown in the Appendix C.” – the missing letter “e” in the word code
- 4/148 – “The cod of process POS tagging is shown in the Appendix D.” – the missing letter “e” in the word code
- 5/164 – “The cod of this process is shown in the Appendix F.” – the missing letter “e” in the word code
- 5/173 – “The cod of this process is shown in the Appendix G.” – the missing letter “e” in the word code
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 2 Report
In this paper the authors have proposed a mining method to mine the intentions of the users of the business information system; this method is developed from a supervised machine learning algorithm that is trained with the knowledgebase for a business in general. The algorithm is tested with the strategies and activities contained in an event log of a particular sales business;
The paper is easy to read but it has to undergo a major revision before it is in a publishable form
- The figure 2 can be redrawn as a business process model using the BPMN notation. In the current form the figure is hard to understand about the control flow between the activities.
- Which machine learning (supervised) algorithm is used here. There is no discussion about it.
- The authors have not discussed about what are the internal and external threats to validity of this study.
- The entire paper just uses the sales business process as an example. Can this work (or the results obtained here) be generalized to the business process models from other domain
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 3 Report
The authors should add a section with the description of similar works.
The authors should compare their methodology with similar works.
The authors should explain why the proposed methodology seems to work well and present some information about the time efficiency of their method.
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 4 Report
The research topic is relevant and has a high practical application. However, the presented material does not meet the requirements for scientific articles. Scientific Soundness is extremely low.
1. There are no relevant references reviews.
2. There are no goals and scientific hypothesis.
3. The methods are not described, only the algorithm and instruments for its realization were introduced.
4. The results should be restructured as a scientific paper, but now it is a programming instruments report. The results must be formalized, for example, using the mathematical apparatus
5. It is necessary to clarify what the authors mean by the categories "Activity", "Strategy", "Strategy name", "rules of the business", "Business information system".
6. There are no conclusions. It is not shown what results are obtained, what is developed by the author and what effect the business system can get from the implementation of the proposals.
The paper may be reconsider only after major revision or it should be fully rewrites
Author Response
Please see the attachment
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
My concerns have been addressed and the paper can be accepted
Reviewer 3 Report
Some kind of cross validation should be used for the experiments.
There are not results (evaluation metrics) about the application of supervised machine learning in the manuscript.
Reviewer 4 Report
There are no major revision. The article has serious flaws. It has no scientific results. The results of a single experiment are poorly presented. An additional experiments are needed for other objects or its scientific generalization should be done.
