Skip to content

Latest commit

 

History

History

benchmark_system


SemEval18 Irony Shared Task Example Scripts

This repository contains a benchmark system for the SemEval-2018 Task 3 Irony detection in English tweets. The script example.py contains Python example code for simple featurisation and classification using the Scikit-learn library.

The script is written in Python 3 but is Python 2 compatible.

Dependencies

  • Python 3.6.x
  • NLTK 3.2.4
  • numpy 1.13.1
  • scikit-learn 0.19.0
  • scipy 0.19.1
  • six 1.10.0

Python module dependencies can be found in requirements.txt

Install with pip: pip install -r requirements.txt

Usage

Run python example.py from terminal.

NOTE: the path to the input file and the task (A or B) has to be hard-coded in the script.