Flight price prediction website : development of a flight prediction website using web scraping and neural networks

Abstract

This thesis presents a flight price prediction system designed to assist travelers in making decisions about cost-effective flight options and planning their trips efficiently. The primary objective of this work was to develop a system that integrated web scraping techniques, machine learning algorithms, and web application development to predict flight prices accurately.

The methodology employed involved three main stages: web scraping, data preparation and modeling, and web application development. Flight information was extracted from Google Flights using Python libraries. The scraped data was then pre-processed and utilized to train a neural network model for predicting flight prices, given some parameters. Afterwards, a web application prototype was developed using Flask, providing users with an intuitive interface to input their flight preferences and receive predicted prices.

The main results of this work included the definition of requirements for the system development, and a prototype of it in all its stages. Overall, this research contributed to the advancement of predictive systems in the domain of travel and tourism, offering practical solutions for optimizing travel expenditures and enhancing user satisfaction.