Etymo Workshop

You can download 243 papers from ArXiv converted using pdfminer from our S3 bucket (around 4MB)

If you want, you can also download the original PDFs too from our S3 bucket (around 400MB).

You can download the metadata and keywords of 10000 processed papers at this github repo or by typing:

git clone git@github.com:EtymoIO/OpenData.git

Etymo background

Description of who we are and what we do.

Text extraction

Main methods of extraction, the main problems and which one looks the best.

Main problems:

• Whitespace
• Equation conversions
• Figures
• References

pdf2txt

pdfminer

tesseract

Pypdf2

How to download a bundle of text conversions.

Keyword extraction

Overview of main methods

RAKE

TF-IDF

Graph Theory approach

Institution extraction

University search (Heuristic 1 and 2)

TODO: make heuristic better

Briefly explain how it works in general terms, show examples

Build up method from very simple to more complicated.

Can we use the author's name?

Deep learning

Deep learning? Object recognition

The data

How to download a bundle of PDFs. Brief look at some examples.