slowwavesleep / BabelNetExtractor

A scala tool to extract data from local BabelNet indices

BabelNetExtractor

Scala tool to extract data from local BabelNet indices

Currently, the tool does the following:

• reads nouns from nouns_no_header.csv
• queries local BabelNet with these nouns (the expected results are also nouns in English)
• stores BabelNet ID, candidate FrameNet ID, BabelNet definition, candidate FrameNet definition, and all the edges for each synset in bn_entries.csv and bn_edges.csv

And that's basically it so far.

Steps to reproduce

• Clone the repository

git clone https://github.com/slowwavesleep/BabelNetExtractor.git

• Install IntelliJ IDEA (Community version will suffice) if you don't have it

• Install Scala plugin

• Open the cloned repository with IDEA (File -> Open)

• Add BabelNet API to the libraries:

  • File -> Project Structure -> + -> Java

  

  • Navigate to BabelNet API and select the corresponding .jar file (such as /BabelNet-API-3.6.1/babelnet-api-3.6.1.jar)
  • Add lib directory the same way (such as /BabelNet-API-3.6.1/lib)
  • The list of libraries should look like this:

  

• Copy config directory from BabelNet API to the project

• Point to local indices in config/babelnet.var.properties (for example, babelnet.dir=/home/user/BabelNet/BabelNet-3.6) and comment out the rest of parameters in this file (unless you know what you're doing)

• Run src/main/scala/Extract.scala using IDEA (right click on it, then Run 'Extract')

• After a few minutes bn_entries.csv and bn_edges.csv should appear in src/main/resources