Using Machine Learning to Compare the Information Needs and Interactions of Facebook: Taking Six Retail Brands as an Example
Round 1
Reviewer 1 Report
The manuscript Using machine learning to compare the information needs 2 and interactions of social media: Take retail brands as an example is valuable research paper that I recommend for publication in this esteemed journal. The introduction is written concisely and includes the comprehensive literature review of all key concepts such facebook data mining, brand image and information cues.
First sentence of abstract is too long.
Since the study is focused on Facebook data mining I would suggest title change since other social media are not covered by the study. Also since only 6 brands are included in the study that should be clear from the title as well.
The hypotheses and research on which the hypotheses are based as well as research model are clearly and comprehensively presented.
It would be easier to read if the chapters were divided into results, hypothesis verification and discussion.
Fig. 3.1. - 3.18 are visually informative but it is harder to track the text and it is complicated to read, so if possible I would recommend to transfer it to the paper supplement
It is excellent that clear theoretical and practical implications have been included which will be useful in future research especially to the researches that will test the model presented.
Research recommendations and limitations are included.
The article can be a useful example and model for researching consumer behaviour on social networks through available published data that can be systematized and researched with specific goals. Data minig of published content on social media is certainly a research-valuable source of data that can show the trends of various phenomena and such research needs to be conducted in the future
Author Response
Thanks for the detailed suggestions provided by the reviewer committee; I have made the following revisions according to the suggestions.
The title and abstract of the article have been revised.
The structure of the article has been renewed with paragraphs such as results, hypothesis verification, and discussion.
The author has put Figure 3.1- 3.18 in the attachment.
Reviewer 2 Report
The authors will present a very fashionable topic related to social networks and machine learning. The authors decided to use machine learning to analyze the information needs of users on the example of retail brands.
The authors' research is part of the broad field of research related to social networks and the application of this research to create an appropriate response in emarketing. The conducted research could also be used in methods of combating the impact of fake news and creating false trends.
Although the theses put forward by the authors are obvious to anyone who deals professionally with social networks in emarketing, the use of machine learning can significantly facilitate inference and confirm the theses.
Although the authors narrowed their research to only a certain fragment of the topic, the method and research itself can be successfully used for a wider spectrum of cases.
In my opinion, the content of the article is worth publishing.
I have a major caveat to the format of the article. The authors do not use the IMRAD structure. There is no discussion and conluclusions. Formatting also makes the article read unfriendly to the reader.
In tables, authors should use '0.000' instead of '.000'.
Although the list of references is extensive, it has not been numbered and there are no relevant references in the text.
It definitely needs to be improved.
Author Response
Thanks for the detailed suggestions provided by the reviewer committee; I have made the following revisions according to the suggestions.
The structure of the article has been renewed with paragraphs such as results, hypothesis verification, and discussion.
The author revised the correct presentation of the data in the tables and figures.
The author has revised the reference format.
