7 comments

[ 2.8 ms ] story [ 26.7 ms ] thread
(comment deleted)
The comment is now deleted, but someone asked for some examples on nanopublications. I just found out about nanopublications myself, but here is what I found:

- The website itself [1].

- This paper from 2018 has an example (and it shows some broader scientific context about this idea) [2].

[1] http://nanopub.org/wordpress/?page_id=8

[2] https://arxiv.org/pdf/1809.06532.pdf

I asked for examples in context. E.g., how might one use nanopublications to document an assertion on a webpage. The examples page on the website seems to be for someone already very familiar with Linked Data/semantic web/etc.

I deleted the comment because I thought I could figure it out on my own, but someone here could probably save me the time.

This comment is on the same point: https://news.ycombinator.com/item?id=23899048

I tried to search through and figure it out before I posted the comment, figuring if I saw a few examples I'd get it, but I just got more confused. This sounds like a really interesting concept for building a essentially a human knowledge base with individual assertions built on eachother, maybe linked and networked with other assertions via the publications bit so you can start tracing where the knowledge came from/the background etc. But I couldn't figure out how it's supposed to do anything or how scientists are supposed to interact with the standard, and if I was even understanding the capabilities of this correctly. Is there a central repository for these, or a user interface to create/peer review/publish/read nanopublications? Who is supposed to implement the standard, who is supposed to use it, and how is it supposed to be used? I was hoping looking through examples would clear that up but I just ended up more confused and am hoping someone more knowledgeable can bail me out on that.
I got an answer from an expert. I asked:

1. How might one use nanopublications to document an assertion on a webpage?

2. Are these meant to be parsed by human readers? Who are these for and what are people supposed to do with them?

The answer:

(1) there's no real difference in a nanopub that originates in a webpage or elsewhere. I think the best to do here is to document the origin of assertions in the provenance graph (through e.g. prov:wasDerivedFrom or prov:used predicates that have URL/webpage entities has objects).

(2) That's a very open question, but I'd say they're mostly for machine use; you can get a nice overview at https://arxiv.org/pdf/1809.06532.pdf

What if all scientific publications were composed by linking together a cluster of nanopublications?