Twitter Sentiment Analysis Project

Overview

This project focuses on sentiment analysis of tweets from the Twitter US Airline Sentiment Dataset, which includes a training dataset with 10,980 entries and a testing dataset with 3,660 entries. The goal is to predict the sentiment of each tweet as either positive, negative, or neutral.

File: train.csv

Contains X train and Y train data for model training (10,980 entries).

File: test.csv

Predictions are required for this dataset (3,660 entries).

Submission Guidelines

Submission Format:

Have submited a CSV file with predictions for the test data. File have one column: predictions, without any headers.

Note:

The submission is in CSV format and follows the specified structure.

Evaluation:

The evaluation was conducted by Coding Ninjas. The accuracy was determined based on the number of accurate predictions. This model achieved an accuracy of 80% in sentiment analysis.

Acknowledgements

This project utilizes the Twitter US Airline Sentiment Dataset provided by Coding Ninjas.