b4msa 0.1.0

Baselines for Multilingual Sentiment Analysis

A Baseline for Multilingual Sentiment Analysis (B4MSA)

Predict a training set using B4MSA

Let us assume that B4MSA is installed and on the PATH, then the stratisfied k-fold can be computed as follows:

b4msa -k 2 text.json

Note: it is recommended to install the following package

pip install tqdm

Install B4MSA

Let us download python (from conda distribution), install it, and include python in the PATH.

wget http://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh
chmod 755 miniconda.sh
./miniconda.sh -b
export PATH=/home/$USER/miniconda3/bin:$PATH

B4MSA needs the following dependencies.

pip install coverage
pip install numpy
pip install scipy
pip install scikit-learn
pip install gensim
pip install nose

In order to install B4MSA use the following:

* Clone the repository

git clone  https://github.com/INGEOTEC/b4msa.git
cd b4msa

• Install the package as usual

  python setup.py install