applicaai / kleister-nda Goto Github PK

Hi, thanks a lot for putting together this dataset and sharing with the community, great work!

However, I couldn't find expected.tsv file in this repo to evaluate my solutions, would really appreciate if you could add it to your repo, thanks in advance.

Thanks for providing this dataset. In the kleister dataset paper(https://arxiv.org/pdf/2003.02356v2.pdf), it proposes a pipeline technique to solve the extraction problem. However, it is a bit hard to locate the answer spans of the answers, because the answers may not exists in the pdf parser results. Can the implementation of the baseline pipeline be open-sourced? Or could you please share more details about the implementation of the auto-tagging procedure?

Hello,

Would you please let me know (or add to the README) about the annotation tool that was used and if it is available publicly?

Thank you

Hi, firstly thanks for putting together this interesting dataset.

I believe I have figured out how to run the metrics for the test set by taking a fork of the repo - pushing up results and then running
./geval -S -t test-A/
However I am prompted for a token which I cannot find in the repo, if I pass no value for the token I get timeouts.

Which license if any is the data released under? I've had a look at the Edgar website but can't find anything specific. Thanks