Multi-Feature Fusion Event Argument Entity Recognition Method for Industrial Robot Fault Diagnosis
Round 1
Reviewer 1 Report
The formulas need to be corrected and aligned as it was not correct in the text as well as the sequence R in line 244.
The similarity report revealed a 14%.
Fig 2 a and b aspect ration could be adjusted to read better.
I am questioning if the word "charged" in Fig 2b is the correct term in relation to the objective.
Overall a good paper representing this study.
Author Response
The formula has been aligned and corrected.
The similarity has decreased.
The "charge" in Figure 2 (b) is not a problem.
Reviewer 2 Report
The submitted manuscript develops a multi-feature fusion event argument entity recognition method for industrial robot fault diagnosis. This topic is interesting and well matches the scope of the journal. However, this manuscript still needs to be improved before considering for publication. My suggestions on the manuscript are listed as follows:
1. An abstract generally consists of background/motivation, method, results/findings and conclusions. Please strengthen the background/motivation, results/findings and conclusions of the current abstract.
2. The overview of the related work is not comprehensive enough. It neglects some important work. Please strengthen this section with more literature. In addition, please identify the research gaps and clarify the novelties/contributions of the work over the existing work in the related work section.
3. The performance of the developed method does not have direct relationships with ontology. Why do you use it? What are the benefits of using it in the work? These need to be justified in the manuscript.
4. From table 1 and table 3, there is no obvious difference among the performance of the comparison methods. Based on this, are there any other advantages of the presented method over the existing ones?
5. The conclusion section needs to be improved with a description of the findings/results of the work and a discussion of the limitations of the work.
6. The list of references is weak. More recent related papers need to be added.
Author Response
The summary has been modified.
Relevant work has been supplemented.
Building the knowledge atlas requires more fine-grained knowledge, so it is necessary to build a more fine-grained ontology and extract fine-grained knowledge through event argument entity recognition methods.
The performance of the fault diagnosis data set has been improved, and it also performs well for the Weibo public data set.
Other changes have also been made.
Round 2
Reviewer 2 Report
No further comments.
