Hi, I’m experiencing a couple of issues with your project, and I would appreciate your guidance to help clarify these points.
-
Ground Truth Tables Missing:
We have observed that the four operations—opendata union, opendata join, webtable union, and webtable join—are showing a significant number of ground truth tables that are not found in either the candidate tables or the query tables. The proportion of missing ground truth tables ranges from approximately 1% to 7%.
-
Reproduction of Starmie Results:
We are experiencing considerable differences between our results and those presented in the Starmie paper. To better understand these discrepancies, could you please provide the specific parameters used when running query.py? Additionally, we would like to confirm whether all query tables were included in the data lake during your Starmie runs, as many ground truth tables appear to be located in the query tables rather than in the data lake.
Thank you for your assistance in addressing these issues.
Hi, I’m experiencing a couple of issues with your project, and I would appreciate your guidance to help clarify these points.
Ground Truth Tables Missing:
We have observed that the four operations—opendata union, opendata join, webtable union, and webtable join—are showing a significant number of ground truth tables that are not found in either the candidate tables or the query tables. The proportion of missing ground truth tables ranges from approximately 1% to 7%.
Reproduction of Starmie Results:
We are experiencing considerable differences between our results and those presented in the Starmie paper. To better understand these discrepancies, could you please provide the specific parameters used when running query.py? Additionally, we would like to confirm whether all query tables were included in the data lake during your Starmie runs, as many ground truth tables appear to be located in the query tables rather than in the data lake.
Thank you for your assistance in addressing these issues.