README.md

Sentiment Analysis of Reviews

This is my first experiment with Machine Learning, in which the main objective is to train the machine using Naive Bayes algorithm to classify a review, informing whether it's a positive or negative review. Thus, the problem statement is: "A review received is positive or negative?"

The code and results can be found in this workbook: sentiment-analysis.ipynb

Machine Learning Workflow

The following steps were followed to achieve the goal of this project:

Problem statement

Defined the problem statement: "A review received is positive or negative?"

Feature

Selected the data set collected on UCI archive to be used and represented it using numerical attributes

Model training

Trained the model with the training data and choose the traning algorithm Naive Bayes.

Test

Tested the model with test data

How to run locally

1. Download and install Anaconda
2. Clone this repo
3. Open Anaconda Navigator and launch jupyterlab
4. Open the workbook sentiment-analysis.ipynb on jupyterlab
5. Have fun! :)

References

Sentiment Labelled Sentences Data Set