search

THU-BPM / ISESL-SQL

The source code of paper "Semantic Enhanced Text-to-SQL Parsing via Iteratively Learning Schema Linking Graph" in KDD2022.
14 stars 1 forks source link

Question about oracle schema linking #2

Closed jc-ryan closed 2 years ago

jc-ryan commented 2 years ago

Nice work！ I have noticed that the oracle schema linking information could bring a huge improvement for Spider and Spider-SYN in section3.6, I wonder that where does the oracle schema information come from for both of two datasets. Could you offer a link please? (I guess that the oracle schema information of Spider come from SLSQL(EMNLP2020), and how about Spider-SYN? )

Besides, could you please offer more details about how you directly change the implicit graph matrix A(𝑡) in Eq. 7 based on the oracle information, becaucse I want to try this by myself. Thanks a lot.

Best wishes!

exlaw commented 2 years ago

Thanks for your interest in our work. We get the oracle schema information from SLSQL(EMNLP2020). The spider-syn is generated by synonym substitution for the spider development set, so we use the same schema linking information as the spider development set.

Since the implicit graph matrix A(𝑡) in Eq. 7 is a matrix of weights between text and schema, so you only need to set the weight to 1 if there are oracle schema linking occurrences and 0 otherwise.

jc-ryan commented 2 years ago

Thanks for your interest in our work. We get the oracle schema information from SLSQL(EMNLP2020). The spider-syn is generated by synonym substitution for the spider development set, so we use the same schema linking information as the spider development set.

Since the implicit graph matrix A(𝑡) in Eq. 7 is a matrix of weights between text and schema, so you only need to set the weight to 1 if there are oracle schema linking occurrences and 0 otherwise.

Okay, I got it, thanks!