Confusion about evidence concatenating

[Haelles]

Dear author,     Thanks for your great work. However, I'm confused about the concatenating code(https://github.com/Raldir/FEVEROUS/blob/059cb63d483abe1d25e07bca4d4e0501022777dd/src/feverous/utils/prepare_model_input.py#L100). It seems that the wiki title is not concatenated in the evidence since the silce always begins with idx 1?

    And the format of evid is [PAGE ID]_[EVIDENCE TYPE]_[NUMBER ID], when I run the code. wiki_page.get_context(evid) doesn't return the context. Should it be wiki_page.get_context('_'.join(evid.split('_')[1:]))?

    Looking forward to your reply!

[Raldir]

Hi Haelles,

Thanks for the message and the find! You are right regarding the title concatenation. I will check again if this affects performance negatively and will get back to you on that.

Regarding the context: Since the extracted context is dependent on the type of evidence, the id field needs to be [EVIDENCE TYPE]_[NUMBER ID]. See https://github.com/Raldir/FEVEROUS/blob/059cb63d483abe1d25e07bca4d4e0501022777dd/src/feverous/utils/wiki_page.py#L181