Next Article in Journal
The Effect of Load Carrying on Gait Kinetic and Kinematic Variables in Soldiers with Patellofemoral Pain Syndrome
Next Article in Special Issue
An Investigation into Audio–Visual Speech Recognition under a Realistic Home–TV Scenario
Previous Article in Journal
Research on the Applicability of Transformer Model in Remote-Sensing Image Segmentation
Previous Article in Special Issue
Multi-Hypergraph Neural Networks for Emotion Recognition in Multi-Party Conversations
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Applied Sciences
Volume 13
Issue 4
10.3390/app13042262
Review Report
applsci-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
Article
Peer-Review Record

DIR: A Large-Scale Dialogue Rewrite Dataset for Cross-Domain Conversational Text-to-SQL

Appl. Sci. 2023, 13(4), 2262; https://doi.org/10.3390/app13042262
by Jieyu Li, Zhi Chen, Lu Chen *, Zichen Zhu, Hanqi Li, Ruisheng Cao and Kai Yu *
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Appl. Sci. 2023, 13(4), 2262; https://doi.org/10.3390/app13042262
Submission received: 5 January 2023 / Revised: 30 January 2023 / Accepted: 7 February 2023 / Published: 9 February 2023
(This article belongs to the Special Issue Audio, Speech and Language Processing)

Round 1

Reviewer 1 Report

(1) Relational database requires normalizations, which means data is stored across several tables. Information retrieval from relational database involves joining tables that need additional concerns. What is the accuracy (standard) of the query results? Similar SQL could lead to different query results.

(2) In section 6.1.1, Setup RAT-SQL, Question Exact Match (QEM), Concat RAT-SQL and Oracle RAT-SQL are separated terms. What's the relationship among those terms? They need to be reorganized. 

(3) In Tables 4 and 5, the performance of oracle RAT-SQL implied the contribution in this research. Because in this work the pretrained model is considered as the state-of-the-art, the result of BERT with concatenated data might be better than the method (dataset) proposed in this work. The authors need to verify this. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Authors create their own large-scale multi-domain dialogue rewrite dataset - DIR, and investigate the efficiency of the two-stage framework on conversational text-to-SQL. The paper is nicely written, and I found it interesting. However, the author needs to address the following concern:

1. Specify the abstract. Give the results obtained in numerical terms and add information about the methods by which they were obtained. So the reader will be able to evaluate the contribution of your research to the subject area.

2. The Related Work section does not reflect the current state of the subject area at the moment; add more modern research. In addition, a full review should include results, methods, and a dataset for each work.

3. Discuss any publicly available datasets, if any are available, and their limitations.

4. Provide justification for using Concat RAT-SQL, Seq2Seq RAT-SQL, RUN RAT-SQL, and Oracle RAT-SQL models.

 

5. The conclusion is very brief and does not support all of the achieved results and findings. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors have indicated the reason for not using accuracy as the performance metric and have revised the article for other comments as suggested.

Back to TopTop