Quantifying Opinion Strength: A Neutrosophic Inference System for Smart Sentiment Analysis of Social Media Network

Round 1

Reviewer 1 Report

The manuscript is presenting the Quantifying Opinion Strength: A Neutrosophic Inference System for Smart Sentiment Analysis of Social Media Network. 

The authors claim that:

The main objectives of this study are to reduce social bias through algorithmically certain and more accurate OM classification results befitting of trust. The proposed model can improve the process of OM in OSNs by highlighting the most irritating technical problems that are caused by the social tendency of the application domain.

 

1. Equations require further explanation. 

2.  Add a clear proposed approach in the abstract section. 

3. Re-organize the related work section. It's so confusing the what are the major topics authors focused on, in the literature review. 

 

 

 

 

Author Response

To The Editor-in-chief and the Associate Editor,

Applied Science

 

Manuscript ID:  applsci-1771385

 

Title:  Quantifying Opinion Strength: A Neutrosophic Inference System for Smart Sentiment Analysis of Social Media Network

 

Dear Sir,

 

 

Thanks for giving us the opportunity to revise the paper for the possible publication in your interesting journal. We also thank the reviewers for their constructive comments. We have significantly revised the manuscripts based the comments received and the details are summarized below.

 

Reply to the comments:

In the revised version, what was modified and added was written in yellow

 

(a) Equations require further explanation.

• Reply: thank you so much for your review. We updated the manuscript by providing the required information as follows:

1. Adding new sentences and/or figures that can help in increasing the understanding of the equations.
2. Select only one representation of users in the equations. In an in-depth clarification of the applied actions, the before-and-after findings are as follows:

In the proposed model section:

In subsection B. Users Weighting Process: Action (a):

Before:

In such a graph, users are represented as nodes, and the relations between them are edges with directions, so-called directed graph .

After:

In such a graph, a user is represented as node, symbolized as, and his relations between other users are edges, symbolized as  per each, with directions, so-called directed graph.

Action (b) for the first equation:

Before:

where,  is the social network graphically represented,   is the is the set of nodes representing users in the graph network, and  is the set of edges representing directed relations between the graphically represented nodes.

After:

where,  is the social network graphically represented, as shown in Fig. 3.   is the is the set of nodes representing users in the OSN: V = {u₁, u₂, u₃, …, u_n}, and  is the set of edges representing directed relations between the graphically represented nodes in the OSN: E= {e₁, e₂, e₃, …, e_m}. If two users (e.g. ₁ and ₂) follow each other, then a directed edge between the two users is constructed:  ∈ .

Action (c) for the 2^nd equation:

Before:

 a set of neighbors of node  that  is connected to

After:

 is a set of nodes that node  is connected to.

 

Action (d) for the 3^rd equation:

Before:

                               (3)

where,  is the betweenness centrality of node , denotes the number of shortest paths from node  to node and  denotes the number of the shortest paths from node  to node  through node.

After:

                               (3)

where,  is the betweenness centrality of node , denotes the number of shortest paths from node  to node and  denotes the number of the shortest paths from node  to node  through node.

Action (e) for the 4^th equation:

Before:

                               (4)

where is the closeness centrality of node ,  is the number of nodes in , and   is the length of the shortest paths between the ^th node and the rest of it in a network.

After:

                               (4)

where is the closeness centrality of node ,  is the number of nodes in , and   is the length of the shortest paths between the ^th node and the rest of the network.

 

In subsection D. Neutrosophic-based OM Classification: Action (a) for equations 5-7

Before:

where,  , , and  are the overall true, indeterminate, and falsity scores of the opinion text , respectively.  represents the polarity score assigned to the opinion sentence  by the observer of the highest influence level.

After:

where,  , , and  are the overall true, indeterminate, and falsity scores of the opinion text , respectively. andrepresent the polarity scores assigned to the same opinion sentence  by the observers of both the highest and the lowest influence level, respectively. For example is the true component of the opinion sentence  assigned by the first highest influence observer.

Action (b) for the 8^th equation:

Before:

NL can effectively deal with this case by setting a confident value for the truth component (). Using this value, one can determine the significance of  components for a given opinion’s score. If  is greater than the confidence value (i.e., 0.5 based on [9]), then the corresponding  components can be considered insignificant [9]. All the above mentioned purposes can be achieved using the following steps:

 

                      (8)

 

After:

NL can effectively deal with this case by setting a confident value for the truth component (). Fig. 8 shows an example of a graphically represented neutrosophic set and how a confident value could be set on it. Using this value, one can determine the significance of  components for a given opinion’s score. If  is greater than the confidence value (i.e., 0.5 based on [9]), then the corresponding  components can be considered insignificant [9]. All the above mentioned purposes can be achieved using the following steps:

 

                      (8)

FIGRE 8: A confident value setting on a neutrosophic set.

 

Action (c) for the 9^th equation:

Before:

            (10)

 

After:

 

            (10)

 

 

1. b) Add a clear proposed approach in the abstract section.

• Reply: we did clearly explain our proposed approach, in the abstract section.

Action:

Before:

The contemporary speed at which opinions move on social media makes it an undeniable force in the field of opinion mining (OM). This may cause the OM challenge to become more social than technical. This is when the process is able to represent everyone to the degree they are worth. Moreover, pondering the existence of opinion indeterminacy and dynamicity can effectively improve OM on social media. This study proposes a neutrosophic–based OM approach for Twitter that handles perspectivism and uncertainty. For perspectivism, influence weighting of users is performed using an artificial neural network (ANN) and social network analysis (SNA). SNA was conducted using popular SNA tools (e.g., Graphistry). The initiative adoption of neutrosophic logic (NL) to integrate users’ influence with their OM scores is to deal with both the indeterminacy of opinion and dynamicity. Thus, it provides new uncertainty–less OM scores that can reflect everyone. The OM scores needed for integration are done with the aid of TextBlob. The results show the ability of NL to improve the OM process and consider the innumerable degrees in an accurate way. This will eventually aid in a better understanding of people’s opinions, helping OM in social media to become a real pillar of many applications.

After:

The contemporary speed at which opinions move on social media makes them an undeniable force in the field of opinion mining (OM). This may cause the OM challenge to become more social than technical. This is when the process can determinately represent everyone to the degree they are worth. Nevertheless, considering perspectivism can result in opinion dynamicity. Pondering the existence of opinion dynamicity and uncertainty can provide smart OM on social media. This study proposes a neutrosophic–based OM approach for Twitter that handles perspectivism, its consequences, and indeterminacy. For perspectivism, social network analysis (SNA) was conducted using popular SNA tools (e.g., Graphistry). An influence weighting of users is performed using an artificial neural network (ANN) based on the SNA provided output and people's reactions to the OM analyzed texts. The initiative adoption of neutrosophic logic (NL) to integrate users' influence with their OM scores is to deal with both the opinion dynamicity and indeterminacy. Thus, it provides new uncertainty–less OM scores that can reflect everyone. The OM scores needed for integration are generated using TextBlob. The results show the ability of NL to improve the OM process and accurately consider the innumerable degrees. This will eventually aid in a better understanding of people's opinions, helping OM in social media to become a real pillar of many applications, especially business marketing.

 

 

1. c) Re-organize the related work section. It's so confusing what are the major topics authors focused on, in the literature review

• Reply: We updated the manuscript by doing as follows:

1. The literature works are re-arranged in time series approach (i.e. from old to recent)
2. highlight the importance of the provided works in a story telling approach in a more clear way

Action (a):

Before:

In marketing applications, business owners and researchers are concerned about SNA, particularly when attempting to find influencers [10]. In line with that, Oueslati et al. [7] highlighted the ….

After:

In marketing applications, business owners and researchers are concerned about SNA, to find influencers [10]. In line with that, in 2017, at the time that most research used to find OSNs’ influencers based on graph theory by measuring network topology (e.g., degree of centrality, betweenness, etc.) [4] [11]. Jianqiang et al. [12] proposed measuring …

Action (b):

Before:

Jin et al. [13] designed a new sentiment link analysis using a graph network. They …

After:

It was found that the isolated application of one analysis whether it was structural or content based could cause information incompleteness then essential patterns and knowledge loss [4]. This caused a few researchers to combine both analyses for different application purposes. In 2021, Jin et al. [13] attempted

Action (c):

Before:

Chauhan et al. [14] highlighted the importance of predicting election result….

After:

In the same year, Chauhan et al. [14] highlighted the importance of predicting election result

Action (d):

Before:

Consequently, a few researchers have started to suggest an implementation method for applying NL in practice [9]. Bhutani et al. [19] highlighted the ….

After:

Consequently, a few researchers have started to suggest an implementation method for applying NL in practice [9]. In 2012, Ansaria et al. [9] mentioned …….. In the same direction, in 2016, Basha et al. [20] suggested the utilization  …….

Action (e):

Before:

The authors in [19] followed the same concept and steps as in [20]. Basha et al. [20] suggested the utilization of a knowledge-based rule–based

After:

In 2018, Bhutani et al. [19] highlighted the importance of rule-base classification using fuzzy logic for its ability to handle interpretation and its weakness in handling uncertainty. The authors in [19] followed the same concept and steps as in [20]. They constructed …

Action (f):

Before:

In dealing with OM, Kandasamya et al. [8] suggested a refined process for SA polarity results from TextBlob to handle the problem of indeterminacy in polarity. They ….

After:

In dealing with the uncertainty and OM, in 2017, Smarandache et al. [18] attempted to express the importance of the NL in dealing with uncertainties and how it can handle ideas’ dynamicity. They applied NL in an election ….. In 2019, Smarandache et al. [21] addressed the problem of word similarity….. In 2020, Kandasamya et al. [8] suggested

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript entitled "Quantifying Opinion Strength: A Neutrosophic Inference System for Smart Sentiment Analysis of Social Media Network" proposes a neutrosophic–based opinion mining (OM) approach for Twitter that handles perspectivism and uncertainty. Using an artificial neural network (ANN) and social network analysis (SNA) to manage the active users on different online social networks (OSNs) is very interesting. According to my observation the reported materials are proper for publication in e-prime.  However, some points should be corrected before publication. I would recommend that this manuscript needs to be revised (minor) before considering its publication. My concerns are:

  1. Include limitations and future scope of this work

2. Abbreviations have to be provided in full name the first time they are used with the short text in brackets. Please check and apply across the document

3. Check the Grammatical mistakes carefully

5. Check the all cited papers should be in proper order

6  Figure 5 needs to be explained properly

Author Response

To The Editor-in-chief and the Associate Editor,

Applied Science

 

Manuscript ID:  applsci-1771385

 

Title:  Quantifying Opinion Strength: A Neutrosophic Inference System for Smart Sentiment Analysis of Social Media Network

 

Dear Sir,

 

 

Thanks for giving us the opportunity to revise the paper for the possible publication in your interesting journal. We also thank the reviewers for their constructive comments. We have significantly revised the manuscripts based the comments received and the details are summarized below.

 

Reply to the comments:

In the revised version, what was modified and added was written in yellow

 

1. a) Include limitations and future scope of this work

• Reply: done by including both the limitations and the future work in the section of conclusion

1. We first renamed this section to be conclusions, limitations, and future work.
2. We highlighted the limitations and future work at the end of the section.

Action:

Before:

…… In its first of its type, encouraging application, NL effectively achieved the goal of solving the problem of opinion dynamicity and identifying undecided polarities that in turn reflected on the process of OM and its dependable results. Optimizing error performance of influencers’ classification and including non-textual posts are our future work aims.

After:

1. CONCLUSIONS, LIMITATIONS, AND FUTURE WORK

….. In its first of its type, encouraging application, NL effectively achieved the goal of solving the problem of opinion dynamicity and identifying undecided polarities that in turn reflected on the process of OM and its dependable results. The limited training sets for the ANN and the adoption of more state-of-art- approaches for ranking users (e.g. Deep Learning) are the limitation of this study. Optimizing error performance of influencers’ classification and including non-textual posts are our future work aims.

 

 

 

1. b) Abbreviations have to be provided in full name the first time they are used with the short text in brackets. Please check and apply across the document

• Reply: done check and we updated the manuscript according to this.

In the related work

Action:

Before:

… their texts using ANN.

After:

… their texts using an artificial neural network (ANN).

In the proposed model

Action (a):

Before:

.. and the UCINET was conducted …

After:

 

.. and the university of california at Irvine network (UCINET)) was conducted …

Action (b):

Before:

TextBlob is a NLP Python …

After:

TextBlob is a natural language processing (NLP) Python ..

Action (c):

Before:

- The removal of URLS, …

After:

- The removal of the uniform resource locators (URLS), ….

In the experimental results

Action (a):

Before:

….. – and (MATLAB) libraries ..

After:

….. – and the matrix laboratory (MATLAB) libraries ..

 

 

 

 

 

1. c) Check the Grammatical mistakes carefully

• Reply: we updated the manuscript by correcting the grammatical mistakes using grammarly.

1. d) Check the all cited papers should be in proper order

• Reply: done check

1. e) Figure 5 needs to be explained properly

• Reply: we updated the manuscript according to your comment.

 

Action:

Micro Influencers: … Thus, Micro influencers could possess low reactions on their writings and low centrality measures as shown in Fig. 7.

Macro Influencers: … Thus they can possess high reactions on their writings as shown in Fig.7. They can become mega influencers with time based on their ability to gain audience trust.

Mega Influencers: … Thus, they possess high centrality measures as shown in Fig. 7. They record an active presence on social media, providing service, doing marketing, advertising, etc.

A-Listers … Thus, they possess both high reactions and centrality measures as shown in Fig.7.

Author Response File: Author Response.pdf

Reviewer 3 Report

 

The paper presents an opinion mining approach for Twitter that handles perspectivism and uncertainty using tools such as an artificial neural network, social network analysis, and neutrosophic logic. The authors indicate that the main objectives of the paper are related to reduce social bias through algorithmically certain and more accurate opinion mining classification results. In the 3^rd Section, the authors present “a summary of the problems to be solved in this study”. I suggest that authors describe in conclusion whether the highlighted problems were solved or not. Moreover, the figures and tables should be placed as close to the reference of figures / tables as possible (for example, the reference related to Figure 1 is on page 2 (line 87), whereas Figure 1 is placed on page 4). In the abstract, the authors write “This will eventually aid in a better understanding of people’s opinions, helping OM in social media to become a real pillar of many applications.“; please, illustrate some examples of mentioned applications. Moreover, the title of paper includes the phrase “Smart Sentiment Analysis”; please, explain in what context is this analysis “smart”.

Author Response

To The Editor-in-chief and the Associate Editor,

Applied Science

 

Manuscript ID:  applsci-1771385

 

Title:  Quantifying Opinion Strength: A Neutrosophic Inference System for Smart Sentiment Analysis of Social Media Network

 

Dear Sir,

 

 

Thanks for giving us the opportunity to revise the paper for the possible publication in your interesting journal. We also thank the reviewers for their constructive comments. We have significantly revised the manuscripts based the comments received and the details are summarized below.

 

Reply to the comments:

In the revised version, what was modified and added was written in yellow

 

1. a) I suggest that authors describe in conclusion whether the highlighted problems were solved or not.

• Reply: we updated the manuscript by rephrasing the conclusions section according to your comment.

Action:

Before:

In this study, we proposed a neutrosophic-based OM process suitable for the uncertainty nature of OSNs. In the proposed model, a new element was integrated with the classical opinion elements using NL. The new element is the user’s weight, which indicates the user’s importance or influence level in the social network. The concept was inspired by the primary use of SNA on OSNs to identify influencers and use them to motivate people to buy specific products by increasing positive feedback. It depends on measuring the influence of users—using SNA centrality measure and recording reactions to their expressed opinions—in the data collection phase, then classifying them into four main influencer types—using ANN. A comparative analysis of different SNA tools was conducted and ended in choosing Graphistry for the mission. Different polarity scores than the regularly applied techniques—TextBlob in our case—are expected due to the perspectivism factor of users. This could result in different scores for the same text when shared by users of different influence levels. NL was implemented for the integration purpose of all opinion elements and for its ability to handle the resulted opinion dynamicity from considering perspectivism, and the classification indeterminacy.

The proposed model succeeded in classifying users of the dataset into their influence types using ANN with promising error performance. On top of that, it emphasized the fact of readapting OM while being applied to OSNs texts for high quality-based information suitable for accurate decision making. It dealt with opinion dynamics and indeterminacy in an emulating way of humans using NL. In its first of its type, encouraging application, NL effectively achieved the goal of solving the problem of opinion dynamicity and identifying undecided polarities that in turn reflected on the process of OM and its dependable results. Optimizing error performance of influencers’ classification and including non-textual posts are our future work aims.

 

After:

In this study, we proposed a neutrosophic-based OM process suitable for the uncertain nature of OSNs. In the proposed model, a new element was integrated with the classical opinion elements using NL. The new element is the user's weight, which depends on measuring the influence of users—using SNA centrality measure and recorded reactions to their expressed opinions, then classifying them into four main influencer types—using ANN. A comparative analysis of different SNA tools was conducted and ended in choosing Graphistry for the mission. Different polarity scores than the regularly applied techniques—TextBlob in our case—were generated due to the perspectivism factor of users. This resulted in different scores for the same text when shared by users of different influence levels. NL was implemented for the integration purpose of all opinion elements, to provide hybrid classification, and for its ability to handle the resulted opinion dynamicity by considering perspectivism, and the classification indeterminacy. The proposed model succeeded in including a smart property for the SA process in social media by classifying users of the dataset into their influence types using ANN with promising error performance. On top of that, it emphasized the fact of readapting OM while being applied to OSNs texts for high quality-based information suitable for accurate decision making. It dealt with opinion dynamics and indeterminacy in an emulating way of humans using NL. In its first of its type, encouraging application, NL effectively achieved the goal of solving the problem of opinion dynamicity and identifying undecided polarities that in turn reflected on the process of OM and its dependable results. The limited training sets for the ANN and the adoption of more state-of-art- approaches for ranking users (e.g. Deep Learning) are the limitation of this study. Optimizing error performance of influencers’ classification and including non-textual posts are our future work aims.

 

1. b) The figures and tables should be placed as close to the reference of figures / tables as possible (for example, the reference related to Figure 1 is on page 2 (line 87), whereas Figure 1 is placed on page 4).

• Reply: we updated the manuscript according to your comment.

1. c) In the abstract, the authors write “This will eventually aid in a better understanding of people’s opinions, helping OM in social media to become a real pillar of many applications.“; please, illustrate some examples of mentioned applications.

• Reply: thank you for your valuable comment. Although examples of the applications are already mentioned in the first paragraph in the related work section, we found it useful to mention the application this study is mostly interceded in. thus, we updated the manuscript according to your comment.

In the abstract

Action:

Before:

This will eventually aid in a better understanding of people's opinions, helping OM in social media to become a real pillar of many applications.

After:

This will eventually aid in a better understanding of people's opinions, helping OM in social media to become a real pillar of many applications, especially business marketing.

 

1. d) The title of paper includes the phrase “Smart Sentiment Analysis”; please, explain in what context is this analysis “smart”.

• Reply: it is smart as it can provide some metal abilities like considering perspectivism and indeterminacy. Thus, we updated the manuscript according to your comment.

In the abstract

Action:

Before:

Moreover, pondering the existence of opinion indeterminacy and dynamicity can effectively improve OM on social media.

After:

Pondering the existence of opinion dynamicity and uncertainty can provide smart OM on social media.

In the introduction

Action:

Before:

Furthermore, NL is malleable in terms of accepting different values from observers. In NL, each object …

After:

Furthermore, NL is malleable in terms of accepting different values from observers. Thus, NL can acquire some mental ability for the OM process to become smart. In NL, each object

 

Author Response File: Author Response.pdf