Clarification on PaperAI scores

[LeoBrait]

Greetings!

In my recent attempt to utilize the tool for aquifer-related investigations, I executed a query on "groundwater samples" within my database, specifying a score parameter of 0.7. While the tool successfully retrieved relevant sentences from certain articles, I noticed that some conspicuously pertinent matches were not captured.

I would appreciate it if you could shed light on the reasons behind this occurrence. I initially assumed that the paper in question would yield a score closer to 1, given its relevance to the query.

For your convenience, I have attached the PDF and XML files used in the SQL database, adhering to the recommended protocols from GROBID and Paper.ETL.

5.pdf
5.txt

Thank you for your time and assistance. I look forward to your insights on this matter.

[davidmezzetti]

I know this has been a long time. But there isn't a real way to answer this without knowing more, such as the model used.