Sentiment Analysis
The current implementation uses the rudamentary functionality
available in the TextBlob
python package. The embed
function
uses polarity to assign scores to strings which then can be used in
conjunction with clustering or other techniques.
Sentiment module in woffle
The sentiment part of woffle currently contains deterministic
clustering, sentiment embed based on TextBlob
and deterministic
parsing.
All of the following can be imported in python as:
from woffle.sentiment import cluster, embed, parse
For an example pipeline please refer to sentiment.py
in
the examples
directory.
Structure of the example file:
Firstly it imports the relevant parts of the pipeline as described in the previous section.
It then loads in the example_comments.txt
from the data
folder
then parses
them and embeds
them.
These scores are then clustered and a rudamentary plot of the outcome of the clustering is presented.