Sentiment Analysis

This project aims to study the sentiment of Indian languages:

Example: haan bhai I will be there don't worry

As we can see the sentence is written in english but is bilingual in nature of speech. This project aims to classify the sentiments of the social media text into +ve, -ve and neutral using ML techniques.

Models

one feature model
three feature model
four feature model
six featur model

Algorithms

Count Model
K-Nearest Neighbours
SVM Models: Linear, kernel
Tree Models: DT - Decision Trees, RF - Random Forest, GBDT - Gradient Boosted Dcision trees
Naive Bayes model

Improved Model

Modularized and improved model here: https://github.com/gitvivekgupta/Sentiment-Improved-II