Research paper about entity linking feature

[Joozty]

Hi is there any research paper/in-depth explanation of entity linking functionality. It's a really nice feature and I would really like to understand how it works under the hood.

I found a video of Sofie Van Landeghem giving a presentation but I would like to know even more about it.

Thanks for help :)

[svlandeg]

Hi, happy to hear the Entity Linking work is useful to you!

There's no paper, and I think the underlying methodology is actually quite straightforward right now.

I'd be happy to give you some additional pointers that may help you with practical applications though.

• We created a video on how to use the EL algorithm in spaCy v2 (on a toy dataset): https://www.youtube.com/watch?v=8u57WSXVpmw. The corresponding code can be found here: https://github.com/explosion/projects/tree/master/nel-emerson/
• If you're interested in working with (the release candidate of) spaCy v3, an updated version of that code can be found here: https://github.com/explosion/projects/tree/v3/tutorials/nel_emerson
• With spaCy v3, it becomes much easier to swap out the default ML model for a custom one, and even to define a custom candidate generation algorithm get_candidates, see also here.

[Joozty]

Thanks, Sofie, your presentation was really helpful and awesome at the same time! I will study more the links you provided. It's definitely more than I found through Google 😄