Embedding Uncertain Temporal Knowledge Graphs
Round 1
Reviewer 1 Report
see attached review and markup
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
The work is interesting and meaningful, there are only a couple of issues to solve:
1. Section 2 is too short, please merge it with Section 3
2. The sub-sections, in Section 3, are hard to follow, consider adding a table at the end of each section with a summary of the methods described above
3. Did you only use a dataset from Wikidata from an arxiv work? Given that arxiv is not peer-reviewed, it should be avoided whenever possible. Also, the evaluation benefits from a public available dataset.
4. The evaluation used a non-traditional data split (.85, .7, .8) any reason for this? Please consider the usage of temporal cross validation
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
In this manuscript, authors proposed an embedding model for uncertain temporal KGs called the confidence score, time, and ranking information embedding jointly model (CTRIEJ), which aims to preserve uncertainty, temporal and structural information of relation facts in the embedding space. The model is novel and creative, and the experiment design is rational. The authors need to improve their writing for a broad audience.
1. Line 16, CTRIEJ shows effectiveness in capturing uncertain and 15 temporal knowledge by achieving promising results, and it consistently outperforms baselines on 16 these tasks. The author needs to give out scientific measurement results in the abstract.
2. Line 113, "For a temporal KG G". Is "KG G" a typo?
3. Line 343, We can further explore higher performance 343 score functions based on semantic matching and translation distance for integration into 344 our framework in the future. The future work should be discussed in the discussion section.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
This is a review of the revisions (mathematics-2164163-peer-review-v2). It appears that the more straightforward suggestions have been incorporated. The discussion of related work was modified by removing unnecessary discussion of neural networks. This improves the presentation since the authors do not go into detail on how NNs are incorporated in their methods. The request to explain notations a little more at the beginning of Sect. 4.3 (now Sect 3.3) was addressed and also improves the presentation.
On the negative side, my question about computational complexity-which is relevant to accuracy in computational results insofar as there might be a tradeoff-was not addressed. Some other minor points were addressed while others were not. This is not a major concern.
One new typo was introduced with the change in Sect. 2 organization: there is no longer a Section 6 so th reference to it on line 105 should be changed to "Section 5."
Overall the manuscript is moderately improved. I do not see a need for a further review and it can be accepted if the editors find that it fits in the scope of the Special Issue on Graph Theory and Applications.
Author Response
Thank you very much for the reviewer for the comments and suggestions. We have revised the chapter introduction section in the Introduction; see lines 103 and 105. Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
My main concerns were addressed
Author Response
Thank you very much for the reviewer for the comments and suggestions. We have revised the chapter introduction section in the Introduction; see lines 103 and 105. Please see the attachment.
Author Response File: Author Response.pdf
