Research on the Knowledge Structure and Sustainable Development Pathways of Artificial Intelligence from the Perspective of Technological Science
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
By extending Qian’s theory of technological science, the authors have categorised Artificial Intelligence into 10 domains and three knowledge structures. These are; machine learning and deep learning (basic theoretical knowledge); fuzzy theory, intelligent systems, data intelligence, computer vision, swarm intelligence (applied basic knowledge); control systems, image processing and human-computer intelligence (applied knowledge). Two main sustainable development paths for VR/AR technology have also been given. These directions are expected to provide value for promoting sustainable technological development.
Some comments on the paper are as follows.
1. Though it is understandable that identifying/ categorizing various AI domains into 3 categories will assist the sustainable development of knowledge growth, the benefit of the proposed structured approach in the field of AI needs to be discussed in the paper, especially in the discussions/conclusion. For example, researchers in the field of AI know these domains – how would it assist researchers, engineers, stakeholders, and policymakers in using the newly proposed structure in this paper?
2. State clearly what is “State Council” in line 27.
3. The “terms” of formula (2) are missing in the explanations – line 191.
4. Figure 10 axes are not readable– and there are Chinese Characters - suggest to only to use English letters.
5. Texts in Figures 4, 6 and 15 are not sharp; please correct these.
6. In Figure 15, there are repetitive terms, e.g., computer vision and machine vision, repeated multiple times under the bigger tab of “Computer Vision”.
7. Neuroscience and Behaviour are not represented in Figure 12 (under Basic Science) as mentioned in the text.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors
This study creates a system for the maintainable advancement of information
in specialized sciences, centering on fake insights. It categorizes information
spaces utilizing assorted strategies like information mapping and quotation
investigation, and recognizes feasible improvement ways, giving profitable
experiences for progressing AI advances and cultivating intrigue and collaboration. So consider the following comments to enhance your article.
Fig 2 Not clear
It would be useful to include a visual representation of this system to enhance understanding.
The case study on VR/AR technology offers viable experiences. It would be useful to expand on the suggestions of these discoveries for future inquiries about virtual reality and clever modeling.
The discourse of feasible improvement ways for AI advances is especially pertinent. More detail on how these ways can be actualized in hone would upgrade the viable suggestions of your findings.
The utilization of various strategies, such as information mapping and characteristic dialect handling, is an excellent approach. A detailed explanation of how these strategies were implemented would enhance clarity for readers, making them feel more informed and knowledgeable.
The distinguishing proof of the two economic improvement methods is a solid viewpoint. Extending how these methods can be operationalized in hone would give specialists profitable experiences.
Comments on the Quality of English LanguageMinor editing of English language required.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThank you for your submission. I thoroughly enjoyed reading the manuscript and believe it offers important and insightful guidance on sustainable development pathways for AI researchers.
I have a few comments that I think will help improve the quality of the manuscript:
- All figures should be replaced with higher-resolution images for better clarity.
- The authors used the K-means++ algorithm for clustering to divide the knowledge domains in the AI field. I wonder if K-means++ is indeed the optimal method for this task. Could you provide more evidence or justification for its use?
- The authors applied the Transformer model (NLP) to encode paper titles into sentence vectors for knowledge representation. I believe more details on the Transformer model and how it was effectively used for keyword clustering would strengthen the manuscript.
- Is there any open data or open-source code available? I think the methods proposed could potentially be applied to fields beyond AR/VR, which would benefit from such resources.
- Please check for typos throughout the manuscript.
- The conclusion feels weak. What is the main contribution of this paper? Is it presenting a new methodology, or is it more focused on highlighting insights gained from the methodology? If it's the latter, I find the conclusion lacking, as it only provides a single example, such as in the AR/VR field. Expanding on the broader implications or findings would help strengthen the conclusion.
1. Figures should be changed. The pictures look blurry.
2. A few typos need to be fixed.
Author Response
Please see the attachment.
Author Response File: Author Response.PDF
