preloader

Twitter- US Airlines Sentiment Analysis

Analysing sentiments of passengers regarding US airlines.

Project Details

  • Kaggle : Kaggle Challenge
  • Tools Used : fast.ai(ULMFiT), NLTK
  • Skills : NLP / Analysis

Twitter US Airlines Sentiment Analysis

Analyze how travelers in February 2015 expressed their feelings on Twitter

This repository contains the analysis and classfication model for the dataset taken from Kaggle’s Twitter US Airline Sentiment.

The dataset contains following columns:

  • tweet_id: Tweet ID
  • airline_sentiment: Sentiment of each Tweet (Positive, Neutral, Negative)
  • airline_sentiment_confidence: Information not given
  • negativereason: The reason of each negative comment
  • negativereason_confidence: Information not given
  • airline: The name of the airline company
  • airline_sentiment_gold: Information not given
  • name: The username of each Twitter account
  • negativereason_gold: Information not given
  • retweet_count: The number of re-posting of each Tweet
  • text: The content of each Tweet
  • tweet_coord: Information not given
  • tweet_created: The exact time each tweet was posted
  • tweet_location: Information not given
  • user_timezone: The time zone that each user was in

The goal is to classify whether the sentiment of the text(present in the form of a tweet by someone) is negative, neutral or positive.

Respository