A Social Media Mining and Ensemble Learning Model: Application to Luxury and Fast Fashion Brands

Round 1

Reviewer 1 Report

I am pleased to have the opportunity to review this research paper. This study attempted to explore a Social Media Mining and Ensemble Learning Model. Although the topic of this research study is interesting and fits within the journal scope, I think authors should apply the comments indicated below to increase the quality of research justification, contributions and findings.

 

First of all, paper research gap. Please improve this part in introduction section: This part is very general and lacked alignment to the research findings, no discussion was provided to derive the implication from. Theoretical and pragmatics implication are vague and need to be better aligned with this paper theoretical underpinnings and proposed process. Furthermore, there is insufficient support and weak arguments in support of the objective that is proposed as well as the model developed. In the final part of introduction, the manuscript structure should be summarised as well as the objetives proposed, originality and gap that would be covered. Also how the author will perform the methodology.

 

What is the originality of this research?  Paper research gap and originality should be better presented at the end of introduction section.

 

Please consider this structure for manuscript final part.

Conclusion

Managerial Implication

Practical/Social Implications

Limitations and future research

 

There is no Discussion section. This section needs to be a coherent and cohesive set of arguments that take us beyond this study in particular, and help us see the relevance of what authors have proposed.  Author need to contextualise the findings in the literature, and need to be explicit about the added value of your study towards that literature. Also other studies should be cited to increase the theoretical background of each of the method used. Findings should be contextualised in the literature and should be explicit about the added value of the study towards the literature. Please use this citation to copy the style and make a citation:  https://doi.org/10.1016/j.ijinfomgt.2021.102331

 

Questions to be answered:

What practical/professional and academic consequences will this study have for the future of scientific literature (theoretical contributions)?

Why is this study necessary? Again, the authors should make clear arguments to explain what is the originality and value of the proposed model. This should be stated in the final paragraphs of introduction and conclusion sections.

Author Response

Thanks for the detailed suggestions provided by the reviewer committee; I have made the following revisions according to the suggestions.

• The introduction and abstract have been modified.

Abstract: This research proposes a framework for the fashion brand community to explore public participation behaviors triggered by brand information and to understand the importance of key image cues and brand positioning. In addition, it reviews different participation responses (likes, comments, and shares) to build systematic image and theme modules that detail planning requirements for community information. The sample includes luxury fashion brands (Chanel, Hermès, and Louis Vuitton) and fast fashion brands (Adidas, Nike, and Zara). Using a web crawler, a total of 21,670 posts made from 2011 to 2019 are obtained. A fashion brand image model is constructed to determine key image cues in posts by each brand. Drawing on the findings of the ensemble analysis, this research divides cues used by the six major fashion brands into two modules, image cue module and image and theme cue module, to understand participation responses in the form of likes, comments, and shares. The results of the systematic image and theme module serve as a critical reference for admins exploring the characteristics of public participation for each brand and the main factors motivating public participation.

Introduction

……While studies have proposed various theories related to information analysis, such as relations marketing, service introduction, and community analysis, this research examines public participation from a different theoretical perspective. It focuses on information image and public participation to examine the information effects of media, public perception, and the main factors motivating participation. The results of the integrated approach, including artificial intelligence data analysis and community content exploration, verify the effectiveness of the proposed framework. This research combines social media exploration and artificial intelligence data analysis to verify cues in brand information and the relationship between images and themes. It also examines the interactive characteristics of various luxury and fast fashion brands to discuss the different types of key image cues.

• The author has readjusted the structure of the discussion in Chapter 6, and described it in the order of Conclusion, Finding and discussion, Theoretical implications, Practical implications, Research limitations and suggestions for future research.

• The author also revised the format of the reference paper.

• The author has strengthened the description of the academic and practical contributions of this paper.

……This research performs a social media exploration and artificial intelligence data analysis to verify information cues in brand pages and the relationship between images and themes. Social media provides marketers with the opportunity to increase brand exposure and promotes continuous interaction between brands and consumers irrespective of time, location, and medium . Information interaction enhances the relationship between brands and the public , and active public participation affects the perceived value of information . Information communication that is positive and frequent tends to have a stronger influence on consumers’ brand associations and attitudes . Therefore, brands should conduct accurate evaluations of information benefits to mitigate any misunderstanding or prejudice and enhance brand value through key information positioning . This research combines artificial intelligence data analysis and community content exploration, and the results successfully verify the effectiveness of the model.

This study examines the interactive characteristics of various luxury and fast fashion brands as well as the different types of key image cues to theoretically verify cognitive, emotional, and behavioral stimulations. Importantly, this research explores public participation from a different theoretical perspective . While the information analysis literature addresses relationship marketing , service introduction , and social analysis , this study focuses on public cognition, emotions, and behaviors  to examine the information effects of media on public perception and the motivation to participate.

Reviewer 2 Report

 

line 6-7 in the abstract " few  offer  effective  suggestions 6on how brands can use information to promote community interactions. " This is not true. There are numerous papers that discuss this.

line 8 Your models would not classify as 'theoretical' as they do not propose how the interactions arise. They fit the 'extant' data gathered. This is applied and not theoretical since you do not model their synthesis.

For line 59, "Can artificial intelligence data analysis and machine learning techniques", you are not applying or exploring 'artificial intelligence'. You work with machine learning methods to explore data. Re-phrase.

line62, again, you are not actually proposing a theoretical model.

line 87, can you please also discuss emoticons and memes? How do those relate?

lines 142-149, discuss the identity and content relationship for the viewer please.

lines 150-153, what is the mechanism for trust establishment or growth for a user?

line170, discuss how those cognitive effects are produced and stimulated.

line203,  "artificial intelligence data analysis " change that, as you are using machine learning methods. It is not as if you explore to any significant extent the descriptive statistics for data analysis

Please put lines226 to 247 in a list

lines 274-276, " Finally, this study integrates  computer  science,  text  mining,  big  data  analysis,  and  machine  learning  aproaches to conduct multiple analyses on the results of the social media exploration." This does not need to be said

5.1, you apply factor analysis but do not describe your data structures? How was the data organized? factor analysis was applied to which variables?

What is the verdict column for table 1, it makes no sense

The Figure2, that block diagram is wrong. Those methods are not a subcomponent of 'Brands'. What are you trying to predict? The caption for Fig2 needs to be expanded.

Did you segment the data in a preprocessing step to produce the digrams? If so, state it.

In figure 3, in the main caption describe what the blue lines and the orange lines represent, as it will make it easier.

Should section 6 not be 'Discussion' or 'Conclusions'?

line 384 what is “here, basketball, now,” ? Is it a feature extracted from the data? Is it a predictor?

 

you can also cite these papers to support your approach to monitoring aspects of social media content that amplifies the influence

[Laflin, Peter, et al. "Discovering and validating influence in a dynamic online social network." Social Network Analysis and Mining 3.4 (2013): 1311-1323.] [Mantzaris, Alexander V. "Uncovering nodes that spread information between communities in social networks." EPJ Data Science 3 (2014): 1-17.] [Laflin, Peter, et al. "Dynamic targeting in an online social medium." International Conference on Social Informatics. Springer, Berlin, Heidelberg, 2012.]  

Author Response

Thanks for the detailed suggestions provided by the reviewer committee; I have made the following revisions according to the suggestions.

• The introduction and abstract have been modified.

Abstract: This research proposes a framework for the fashion brand community to explore public participation behaviors triggered by brand information and to understand the importance of key image cues and brand positioning. In addition, it reviews different participation responses (likes, comments, and shares) to build systematic image and theme modules that detail planning requirements for community information. The sample includes luxury fashion brands (Chanel, Hermès, and Louis Vuitton) and fast fashion brands (Adidas, Nike, and Zara). Using a web crawler, a total of 21,670 posts made from 2011 to 2019 are obtained. A fashion brand image model is constructed to determine key image cues in posts by each brand. Drawing on the findings of the ensemble analysis, this research divides cues used by the six major fashion brands into two modules, image cue module and image and theme cue module, to understand participation responses in the form of likes, comments, and shares. The results of the systematic image and theme module serve as a critical reference for admins exploring the characteristics of public participation for each brand and the main factors motivating public participation.

Introduction

……While studies have proposed various theories related to information analysis, such as relations marketing, service introduction, and community analysis, this research examines public participation from a different theoretical perspective. It focuses on information image and public participation to examine the information effects of media, public perception, and the main factors motivating participation. The results of the integrated approach, including artificial intelligence data analysis and community content exploration, verify the effectiveness of the proposed framework. This research combines social media exploration and artificial intelligence data analysis to verify cues in brand information and the relationship between images and themes. It also examines the interactive characteristics of various luxury and fast fashion brands to discuss the different types of key image cues.

• The author has made more detailed explanations and modifications for content such as "content relationship for the viewer", "trust establishment or growth for a user", "cognitive effects are produced and stimulated".

2.1 Brand community participation

……Moreover, when there are higher levels of trust, members are more likely to view other members as reliable and active sharing and dialogue becomes easier. Such word-of-mouth promotion enhances brand image and loyalty. The information cueing effect is based on possible crowd behaviors and interactions in response to text information shared in the community . To conceptualize information cues , the research transforms vague information cues into definite text concepts , a problem emphasized in numerous image studies evaluating the individual attributes of information to obtain specific factors composing an image . Information cues are commonly defined as potential ideal information in the minds of the public that may be transformed into a specific image or concept .

2.2 Information image and cues

……Studies suggest that the motivations underpinning information searches include satisfaction , participation , and the gaining of trust . Consumers read information to understand a brand , analyze product characteristics , and make purchase decisions. The value of brand fan pages depends on whether the information drives fans toward active participation . Fan pages are considered a reliable source of brand information and can be used to gain consumer trust, making it easier to encourage participation and purchases. Trust is a fundamental factor motivating a community to share and exchange opinions.

• The author has redefined and modified figures and tables.

• The author has readjusted the structure of the discussion in Chapter 6, and described it in the order of Conclusion, Finding and discussion, Theoretical implications, Practical implications, Research limitations and suggestions for future research.

• The author has added relevant papers suggested by the review committee to the articles and references.

Round 2

Reviewer 1 Report

The authors have made the indicated changes. This reviewer has no additional comments. 

Author Response

Thanks for the detailed suggestions provided by the reviewer committee.

Reviewer 2 Report

This study runs a series of natural language processing (NLP) programs to examine the unstructured data . The unstructured data are transformed, and separable sentences are tokenized and converted into words and punctuations .

-> you must expand upon what NLP methods are use and how they are used. Is tokenization the only transformation? If so state that.

 

While the information analysis literature addresses relationship marketing , service introduction , and social analysis , this study focuses on public cognition, emotions, and behaviors to examine the information effects of media on public perception and the motivation to participate.

->List the literature on what is addressed, which publications say what exactly and how yours differs. It must be organized so that your work in comparison to the work of others can be compared for key differences.

Author Response

Thanks for the detailed suggestions provided by the reviewer committee; I have made the following revisions according to the suggestions.

• The model implementation focuses on content posted by each brand on their Facebook fan pages. First, a data mining program is used to gather post data. This study runs a series of natural language processing (NLP) programs to examine the unstructured data ^{94, 95}. The unstructured data are transformed, and separable sentences are tokenized and converted into words and punctuations ⁹⁶. It entails editing, organizing, and analyzing expansive data and offers in-depth information on, for example, representative indicators. Thus, numerous companies have employed community mining to define various services, interact with the public, analyze competition, and transform data into references for decision-making processes. Moreover, studies have found that data mining simplifies procedures involving large-scale data. The distributed vertical frequent mode, in particular, applies an array method to process large amounts of data and target variables. This mode can be used to optimize problems in the original groups of a data warehouse, and thus, is widely applied in social network analyses to mine consistent characteristics from social interaction content.
  
  Verifying data from actual chat records can help create a framework for a community interaction model to collect data from a software and calculate the relationship and minimum distance between each node. The three most commonly used exploration techniques are mass data, clickstream, and classification analyses. Exploration processes also include data cleaning and pre-processing. An overthrow feature can be used when datasets are balanced and weighting does not produce noise after data mining; however, this feature does not apply until the dataset is balanced, which requires repeated weighting to eliminate noise. Community enhancement services are another approach to understand the benefits of such services, and thus, improve users’ cloud experience. Therefore, this study adjusts the content to a community service enhancement model and references information enhancement models available on various community websites to determine judgment strength. If the process reveals that the target variable (variable to be predicted) is a discrete value, a classification algorithm can be used to redefine the information (e.g., new vs. old or strong vs. weak).

• This study examines the interactive characteristics of various luxury and fast fashion brands as well as the different types of key image cues to theoretically verify cognitive, emotional, and behavioral stimulations. Importantly, this research explores public participation from a different theoretical perspective ^{10, 107}. Information organization theory and information processing research have repeatedly demonstrated the relationship between consistent information and imagery and higher memory recall. In addition, clearer and more explicit content contributes to long-term brand memory and achieving a successful brand link. Emotional identification with brands can be used to determine if a crowd positively or negatively perceives a brand and critically influences subjective impression. While the information analysis literature addresses relationship marketing ¹⁰⁸, service introduction ⁷⁹, and social analysis ¹⁰⁹, this study focuses on public cognition, emotions, and behaviors ^{23, 110}to examine the information effects of media on public perception and the motivation to participate.