We propose a method for creating co-word networks starting from keyphrases automatically extracted from the full text of scientific papers.
We release the following resources, available at this link https://github.com/dhfbk/Keyphrase_Co-Occurrence_Networks :

  • an R script to convert the output of KD in a list of undirected edges that can be imported in Gephi;

  • two folders containing the newtorks in Sigma.js generated from the abstracts of AIUCD 2015 and AIUCD 2018: to visualize and search in the networks, open the index.html files with FIREFOX.

Please cite: Sprugnoli, R., Moretti, G.. 2019. Discovering Research Themes in Scientific Research: from Keyphrase Extraction to Co-occurrence Networks. In the book of abstracts of AIUCD 2019.