All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
The author(s) have addressed the critical comments raised by the reviewer(s). The paper can be accepted for publication in PEERJ Computer Science.
[# PeerJ Staff Note - this decision was reviewed and approved by Arkaitz Zubiaga , a 'PeerJ Computer Science' Section Editor covering this Section #]
The authors of the paper addressed all of my remarks in the first review. Together with the modifications, the paper improved a lot. The literature review is now comprehensive.
The paper fits into the scope of the journal. The methods are described in sufficient detail.
The procedure for summarizing the original literature is comprehensible and plausible.
All previous queries are addressed well
All previous queries are addressed well
All previous queries are addressed well
All previous queries are addressed well
In revized version of the manuscript, deep learning methods with CNN, LSTM, and Transformers are sufficiently written and discussed. Roadmaps and related sections are now suitable.
subject (in) dependent,and related ML and DL approaches are well- organized.
The approaches on EEG emotion recognition are sequentially included, and discussed with their advantageous and disadvantageous.
Maybe, the year 2024 in fig1 can be removed.
Most of the reviewers have given critical comments on your paper. Their main concern was that the paper did not follow the required standard to meet the expectations of the review paper. So, I suggest the authors consider all the reviewer's comments carefully and substantially improve the quality of the manuscript to improve its standard.
**PeerJ Staff Note:** Please ensure that all review and editorial comments are addressed in a response letter and that any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
**Language Note:** The review process has identified that the English language must be improved. PeerJ can provide language editing services - please contact us at [email protected] for pricing (be sure to provide your manuscript number and title). Alternatively, you should make your own arrangements to improve the language quality and provide details in your response letter. – PeerJ Staff
This article is well written and easy to understand. Its structure is appropriate and it contains the sources expected for the topic. As a literature review, it fits in with the scope of the journal and covers an existing gap: The background of the topic as well as the existing literature (including reviews) is named and it is pointed out what the current gap is.
The Survey Methology chapter describes the procedure for selecting the relevant sources. However, the exact meaning of the criteria "demonstrating high research value and innovation" and how they were operationalized are not discussed here. It is suggested that this point be explained in more detail before publication.
The procedure for summarizing the original literature is comprehensible and plausible.
The authors provided a comprehensive overview of recent advances in EEG-based emotion classification techniques and proposed a systematic classification framework for subject-dependent and subject-independent approaches. The authors also provided a detailed analysis of datasets in the EEG-based emotion recognition area. The paper is well-written and readable. The quality of the article may further improve if the following queries are addressed.
1. The authors presented the literature review well, but it would be more beneficial for the readers if a summary table were given for such review papers.
2. Are there any specific reasons the authors did not consider papers from PubMed, and Web of Science platforms, as they also cover EEG-based emotion recognition papers?
3. The authors say that the sourced papers were screened initially. However, it is not clear what the metrics used for screening are.
4. The papers were sourced using a set of keywords and then classified. What are the specific keywords used (In Figure 1, it is said …, so it means that many more keywords than what is given in the paper were used.
5. Some of the details are not clear. For example, how many papers have been sourced using the keyword search for the last five years? After the screening, how many were selected? And under each classification category, how many papers were listed? This information helps the readers to know the recent research direction. Also, the discussion section may be strengthened using such details.
6. The discussion section is missing, and the conclusion section is lengthy.
EEG based emotion recognition is an extensively studied topic. First, it has started with basic ML and signal processing methods, and then followed by advanced signal processing methods as feature extraction. After DL, Convolutional, time-series, sequence modellings are studied using basic or processed inputs.
From this point of view, the road maps in this paper are not sufficient to be review paper. The recent and the most comprehensive apporaches in the field of DL are not included. Thus, it can not motivate the readers aboout the recent developments.
Finally, not suitalbe to be review paper.
Only, subject (in) dependent and cross-subject were written, but the combination of the past and the recent studies are not written. Thus, tecnically insufficient, especially classic ML and new aspects of the DL methods in this area.
in fact, I could not find valid things.
1. In the author's paper, mention of "physiological signals such as EEG, facial expressions, and speech signals" appears in line 39-42. However, it should be noted that facial expressions and voice signals are considered non-biological signals.
2. Lines 48-58 of the paper discuss the ability of convolutional neural networks and recurrent neural networks to enhance EEG features, and the potential impact on future research directions. While these networks are still commonly used, it is important to acknowledge the relevance of transformer networks. Recent literature should be consulted to demonstrate the importance of incorporating this method into future research endeavors.
3. The author introduces the content of a reference in lines 65-71, but fails to specify the prospective direction and primary findings of the cited material. Additionally, the paper overlooks the potential challenges and future research directions highlighted in lines 72-81.
4. The description of the DEAP dataset in lines 191-196 is inaccurate, as the baseline of the DEAP dataset is actually 3 seconds, not 5 seconds. Furthermore, in line 210, the author incorrectly states the position of Table 1.
5. Lastly, the manuscript contains numerous grammatical errors which could be rectified through a thorough proofreading by a fluent English language speaker.
no comment
no comment
no comment
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.