What is the best thing to do when you find a stash of 200,000 recent papers with complete meta data like author name, tags, date of publication, author affiliation, conference or journal, name where published, abstract and every other detail.
I have several ideas.
1) Train an ANN to write research papers 2) Analyze the distribution of research in the countries around the world. Which countries publish more in conference and which in journals, the impact factor of each university and country. And publish these findings in a paper. 3) See what domains are the hot topic and which domains get ow much attention. Publish this research.
Please discuss the feasibility of these ideas and give more ideas.
Put all the research papers into a self-learning neural network
Jordan Hughes
bump
Grayson Cook
It would be nice if you just extracted some simple metadata and put it into a CSV.
> List of titles of publications grouped by university > Impact by university > Extract influential papers > etc
You can't trust university rankings because they are based off of bullshit criteria such as diversity. This would be very useful data.
> 1) Train an ANN to write research papers
That would never work.
I doubt you could get one to write a convincing children's book.
Logan Mitchell
two, and three, and this
Austin Parker
You give them back, Jamal.
Grayson Phillips
# why not. let's get a fine idea of what to do with them, then I would be happy to do it together.
Zachary Allen
#
> That would never work. If you train the AI character by character, then I am sure it would be next to impossible. But if we scan the papers, we'll have some sort of dictionary. Then if we train the AI to arrange those words, it may be somewhat doable.
What do you think?
Ryder Richardson
Are you the world's leading expert in artificial intelligence? Maybe in five years with millions in funding plus a team of researchers working under you.
No? You're fucked.
Asher Diaz
If you are going to do this, say where you will posts them
Noah Rodriguez
Here's what you will most likely get from your AI : isotropic.org/papers/chicken.pdf It's fine if you don't want it to make much sense, but you have to remember that it will just spout gibberish.
Zachary Martin
First one won't happen, last two are unpublishable because people already published them years ago
T. Almetric and Bibliometric researcher
Parker Nguyen
Give us the metadata. Though running your own University Ranking site like is a good idea.
Jose Kelly
Post some of the more interesting ones
Zachary Gutierrez
>you find a stash.. You are not the only one. Does your "stash" include retractions and yet-to-be-published entries?