Hackathon, extracting meaningful, machine-interpretable data from scholarly publications

Printer-friendly version

Hi all, we would like to organize a Hackathon while at the ESWC.

We want to do it on Monday, May 27 in Montpillier, France.

We are probably going to use https://www.hackerleague.org/ to manage signups


The ability to extract meaningful, machine-interpretable data from

scholarly publications in PDF form is a big challenge.  Several open

source libraries exist that attempt to automate this process, but work

needs to be done on them to improve accuracy and reliability.  Some

specific and relevant  challenges include:

Ability to automatically identify and tokenize citations from the PDF

(or more accurately, from a string of text)

Ability to automatically identify those blocks of text that represent

the narrative in a PDF.

Ability to identify references within the narrative, extract their

scope, and associate them with citation information in the PDF.

Anybody interested is welcome to join us, we will announce more

presice details later on this week. Also, we are looking for someone who co-organizes the meeting, ideally

someone who is local to Montpellier or to France.