https://www.aclweb.org/anthology/S17-2089/
Author - Oleksii Sliusarenko. The solution is implemented in python notebook - solution.ipynb. You can see explanations, comments and results there. Note that some files used here are available for research purposes only, for example MaxDiff Twitter Sentiment Lexicon from this resource: https://www.svkir.com/resources.html#manual_lexicons_BWS.
Execute these steps to download and install anaconda python:
wget https://repo.anaconda.com/archive/Anaconda3-5.3.1-Linux-x86_64.sh
bash Anaconda3-5.3.1-Linux-x86_64.sh -b -p ~/anaconda
Execute these steps to create and activate a separate working environment:
. ~/anaconda/bin/activate
conda create -n sentiment_analysis python=3.6
. activate sentiment_analysis
Install python libraries or requirements:
# for macos only:
export CFLAGS=-stdlib=libc++
pip install -r requirements.txt
Launch python notebook and open solution.ipynb in your browser:
jupyter notebook