jaeyeun97 / sentiment

Sentiment classification system with mutliple classifiers and stemmers; Pang et al. + alpha

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Intro

This repository contains an implementation of Naive Bayes classifier, utilization of scikit-learn's SVM implementation and multiple bag-of-words based feature extractions for replicating the results of Pang et al. 2002, for the NLP course at Cambridge.

I do realise this code is highly inefficient and has fairly straightforward changes that may improve the speed of the classifications significantly, but I had no problem just running it overnight while I sleep (and the purpose of this code is to get the results so whatever).

If you want to tweak the number of classifications to run, just change the __main__.py. It will iterate through everything by default.

Run

At the root directory of the repository:

$ pip install -r requirements.txt
$ python -m sentiment

About

Sentiment classification system with mutliple classifiers and stemmers; Pang et al. + alpha


Languages

Language:Python 100.0%