dixboi / bajs

A collection of files from the British Airways Job Simulation

DATA SCIENCE JOB SIMULATION by BRITISH AIRWAYS from THE FORAGE

A collection of files from the British Airways Job Simulation

Scenario

Bristish Airlines is interested about their customers' reviews and also about their booking purchases.

Tasks

As the data scientist of the team, they wanted me to analyze their data with regards to reviews and booking purchases.

Task 1

Analyze customer reviews based from the https://www.airlinequality.com website.

Task 2

Predict from the customer's data whether they will complete a booking purchase or not.

Action

By using webscraping tools, I got the review data from the website and analyze them using NLP techniques. As for the other task, the data was already prepared so I just needed to apply machine learning models to it.

Results

These are my findings and what happened

Task 1

I found out that 51% of reviews says something positive about the airline while 49% says something negative. The most common words are good, best, worse, and worst.

Task 2

I compared three machine learning models for classification and the one with the best performance is Random Forest Classifier.