Sentiment Analysis in Twitter Based on Knowledge Graph and Deep Learning Classification
Round 1
Reviewer 1 Report
Authors conducted this research in the title of "Sentiment Analysis in Twitter based on Knowledge Graph and Deep Learning Classification".
The paper’s subject could be interesting for readers of journal. Therefore, I recommend this paper for publication in this journal but before that, I have a few comments on the text that should be addressed before publication:
Comments:
1) Abstract: In the line 5 of abstract authors used the word "We". It is not common to use subjects or pronouns as we or us . It should be corrected.
2)Again in the line 13 of Abstract authors used the word "Us". Authors should fix this problem.
3)L22: Authors used this word "Twitter" without any definition. Authors should emend that.
4)The space of between bullets and lines are too much and it does not look good.
5)In line of 102 authors used this "Section ??". What is it? What are these quation marks for?
6)Title of section 2 is not right. This title should be corrected as Related Words. Related work is not correct.
7)Table 2: This table should be moved to the right direction.
8)Line 400: This equation lacks of a number. Every equation has to have a number.
9)Figure 9: The title of this Figure is too long. It should be shorted by authors.
10)Conclusion section lacks of conflict of intrests part.
11)In the Conclusion section, there is nothing about research fund.
12) Since recently it has been proved that artificial intelligence (AI) and machine learning has a numerous applications in all of engineering fields, I highly recommend the authors to add some references in this manuscript in this regard. It would be useful for the readers of journal to get familiar with the application of AI in other engineering fields. I recommend the others to add all the following references, which are the newest references in this field of electrical engineering [1], civil engineering [2], petroleum engineering [3]
[1] Design and Modeling of a Compact Power Divider with Squared Resonators Using Artificial Intelligence. Wireless Personal Communications. 2021 Apr;117(3):2085-2096, doi:10.1007/s11277-020-07960-5.
[2] Nazemi, B.; Rafiean, M. Forecasting house prices in Iran using GMDH. Int. J. Hous. Mark. Anal. 2021, 14, 555–568.
[3] Roshani, M.; Sattari, M.A.; Ali, P.J.M.; Roshani, G.H.; Nazemi, B.; Corniani, E.; Nazemi, E. Application of GMDH neural network technique to improve measuring precision of a simplified photon attenuation based two-phase flowmeter. Flow Meas. Instrum. 2020, 75, 101804
Author Response
Following we list the reviewer's comments and its respective answer
1) Comment: Abstract: In the line 5 of abstract authors used the word "We". It is not common to use subjects or pronouns as we or us . It should be corrected.
Answer:
The phrase was changed to:
In this work, the proposed model is a new hybrid approach ...
Also line 8 had a "we" so it was changed from: "We represent the tweets using graphs" to: "Graphs are data structures that represents tweets"
Also line 10 had a "we" so it was changed from: "as we integrate" to: "thanks to the integration of"
The Abstract and the rest of the article does not use personal pronouns.
2) Comment: Again in the line 13 of Abstract authors used the word "Us". Authors should fix this problem.
Answer:
This was changed from "allow us" to "allows".
Also line 14 had a "we" so it was changed from: "We describe each phase of our proposed..." to: "The phases of the proposed approach are..."
Also line 16 had a "we" and "our" so it was changed from: " We compare our proposal with character n-gram embeddings based Deep Learning models" to "There is a comparison between the proposed model and a character..."
Again, the focus was to change the whole article to attend this problem.
3) Comment: L22: Authors used this word "Twitter" without any definition. Authors should emend that.
Answer:
A definition of what Twitter is given. At the moment the Introduction section includes a definition of what Twitter is.
4) Comment: The space of between bullets and lines are too much and it does not look good.
Answer:
The format of the paper is changed so now there's no space between bullet points and lines. As a result, the space between the lines and the bullets is small.
5) Comment: In line of 102 authors used this "Section ??". What is it? What are these quation marks for?
Answer:
This is a lost reference for one section of the paper (The Related Work section). It has been fixed now, since the label was allocated and therefore there are no more "??".
6) Comment: Title of section 2 is not right. This title should be corrected as Related Words. Related work is not correct.
Answer:
I'm afraid I don't agree with this, perhaps the reviewer is confused. The section is the work of other people. This is the related to the work we developed, and I think it should be called related work.
7) Comment: Table 2: This table should be moved to the right direction.
Answer:
The format of the paper is changed so now all the directions and metainformation of the paper has been adjusted.
8) Comment: Line 400: This equation lacks of a number. Every equation has to have a number.
Answer:
It was changed to be an equation and now it's number 10.
9) Comment: Figure 9: The title of this Figure is too long. It should be shorted by authors.
Answer:
The information was reduced to only include what the figure is about. The rest of the information that was in the caption was already in the explanaition/text.
10) Comment: Conclusion section lacks of conflict of intrests part.
Answer:
The authors show no conflict. This is now specified at the end of the Conclusion.
11) Comment: In the Conclusion section, there is nothing about research fund.
Answer:
The project was funded, this is specified at the end of the conclusion.
12) Comment: Since recently it has been proved that artificial intelligence (AI) and machine learning has a numerous applications in all of engineering fields, I highly recommend the authors to add some references in this manuscript in this regard. It would be useful for the readers of journal to get familiar with the application of AI in other engineering fields. I recommend the others to add all the following references, which are the newest references in this field of electrical engineering [1], civil engineering [2], petroleum engineering [3]
[1] Design and Modeling of a Compact Power Divider with Squared Resonators Using Artificial Intelligence. Wireless Personal Communications. 2021 Apr;117(3):2085-2096, doi:10.1007/s11277-020-07960-5.
[2] Nazemi, B.; Rafiean, M. Forecasting house prices in Iran using GMDH. Int. J. Hous. Mark. Anal. 2021, 14, 555–568.
[3] Roshani, M.; Sattari, M.A.; Ali, P.J.M.; Roshani, G.H.; Nazemi, B.; Corniani, E.; Nazemi, E. Application of GMDH neural network technique to improve measuring precision of a simplified photon attenuation based two-phase flowmeter. Flow Meas. Instrum. 2020, 75, 101804
Answer:
There is a paragraph at the beginning of the Introduction section that explains the importance of AI.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In this paper, authors propose a new hybrid approach for Sentiment Analysis based on Knowledge Graphs and Deep Learning techniques, to identify the sentiment polarity (positive or negative) in short documents, such as the posts in Twitter. However, there are some issues that should be addressed before this work can be accepted. The detailed comments are given as follows:
- It needs careful proofreading in terms to English grammar, spelling, and sentence structure for the clarity to the reader. For example, “these interpretability model are classified in:” -> “these interpretability models are classified in:”(on line 60), “in certain way” -> “in a certain way”(on line 84), “this understanding of results provide insights into the model” -> “this understanding of results provides insights into the model”(on line 87), what is “??” on line 102, and “4rd step” -> “4th step”(on line 417).
- In the introduction, the significance and motivation of the design are not well explained.
- The reference list can be enhanced. Especially, many recent strategies have potential to deal with the problems mentioned in the paper, such as
[1]. Co-LSTM: Convolutional LSTM model for sentiment analysis in social big data[J]. Information Processing & Management, 2021, 58(1): 102435.
[2]. ABCDM: An attention-based bidirectional CNN-RNN deep model for sentiment analysis[J]. Future Generation Computer Systems, 2021, 115: 279-294.
[3]. Attention-emotion-enhanced convolutional LSTM for sentiment analysis[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021.
[4] Enhancing Learning Efficiency of Brain Storm Optimization via Orthogonal Learning Design, IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2020, http://dx.doi.org/10.1109/TSMC.2020.2963943.
[5] An Adaptive Localized Decision Variable Analysis Approach to Large-Scale Multi-objective and Many-objective Optimization, IEEE Transactions on Cybernetics, 2021,http://dx.doi.org/10.1109/TCYB.2020.3041212.
These references should be included in the paper.
- The format of Table 2 is incorrect.
- on line 199, it is not right-aligned, and on line 233, the end of the line should be a “.” instead of a "/".
- Why do you choose containment similarity measurement, maximum common sub-graph similarity measurement and maximum common sub-graph number of edges as the measures of graph similarity? Please give explanation.
- In the experiment, the state-of-the-art emotion analysis methods can be added for comparison.
- The main contributions of the research work need to elaborated in detail.
- Please unify the format of formulas in this paper.
Author Response
Following we would like to list the reviewer's comment and then our answer:
1) Comment: It needs careful proofreading in terms to English grammar, spelling, and sentence structure for the clarity to the reader. For example, “these interpretability model are classified in:” -> “these interpretability models are classified in:”(on line 60), “in certain way” -> “in a certain way”(on line 84), “this understanding of results provide insights into the model” -> “this understanding of results provides insights into the model”(on line 87), what is “??” on line 102, and “4rd step” -> “4th step”(on line 417).
Answer:
All these were corrected, furthermore, all sections of the paper were revised.
2) Comment: In the introduction, the significance and motivation of the design are not well explained.
Answer:
The Introduction was changed to include motivations on knowledge graphs (lines 49-60), and also there was extra information added. Knowledge Graphs are important in this research and we express this very well in the corresponding sections.
3) Comment: The reference list can be enhanced. Especially, many recent strategies have potential to deal with the problems mentioned in the paper, such as
[1]. Co-LSTM: Convolutional LSTM model for sentiment analysis in social big data[J]. Information Processing & Management, 2021, 58(1): 102435.
[2]. ABCDM: An attention-based bidirectional CNN-RNN deep model for sentiment analysis[J]. Future Generation Computer Systems, 2021, 115: 279-294.
[3]. Attention-emotion-enhanced convolutional LSTM for sentiment analysis[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021.
[4] Enhancing Learning Efficiency of Brain Storm Optimization via Orthogonal Learning Design, IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2020, http://dx.doi.org/10.1109/TSMC.2020.2963943.
[5] An Adaptive Localized Decision Variable Analysis Approach to Large-Scale Multi-objective and Many-objective Optimization, IEEE Transactions on Cybernetics, 2021,http://dx.doi.org/10.1109/TCYB.2020.3041212.
These references should be included in the paper.
Answer:
The first reference was added at the level of explaining the LSTM, these other paper contains explanaitions of the activation function sigmoid, that are also relevant for the article. This is a doi reference for the journal: Information Processing Management 2021, and it's bibliography 60 in the article.
The second reference was added as a related work.
The thrid reference was added in the definition of memory mechanism of LSTM.
The fourth and fifth article was added to the introduction
4) Comment: The format of Table 2 is incorrect.
on line 199, it is not right-aligned, and on line 233, the end of the line should be a “.” instead of a "/".
Answer:
The format of the table was changed by the editor, the "/" was replaced by "."
6) Comment: Why do you choose containment similarity measurement, maximum common sub-graph similarity measurement and maximum common sub-graph number of edges as the measures of graph similarity? Please give explanation.
Answer:
The motivation is that they are equivalent to the graph comparison (they allow to compare graphs) and they reduce the data structure repetition. This is mentioned at the end of the explainaition of the metrics section 4.5
7) Comment: In the experiment, the state-of-the-art emotion analysis methods can be added for comparison.
Answer:
This is already done in the article, in the result section we compare the results with state-of-the-art techniques under the same dataset.
8) Comment: The main contributions of the research work need to elaborated in detail.
Answer:
This is specified in the related work section but also is added in the results section.
9) Comment: Please unify the format of formulas in this paper.
Answer:
Done
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
All the comments have been addressed correctly.
Reviewer 2 Report
My main conerns have been addressed.
