=================
Business Context
=================

:Authors:
    Cao Tri DO <cao-tri.do@keyrus.com>
:Version: 2025-04

.. admonition:: Objectives
    :class: important

    This article is intended to provide a comprehensive overview of the business context in which the project is being developed.

source: https://www.kaggle.com/datasets/ahsan81/hotel-reservations-classification-dataset/data

With the rise of **online hotel booking platforms**, the hospitality industry has undergone a profound transformation in **customer booking behavior**. Digital channels have made it easier than ever for guests to compare offers, modify their plans at the last minute, or cancel reservations effortlessly.  

While this flexibility benefits customers, it poses a significant **economic and operational challenge** for hotels: a large proportion of bookings are **canceled before arrival** or result in **no-shows** (guests who do not show up without prior notice).  

The main reasons behind cancellations include:

- changes in personal or professional plans;
- scheduling conflicts;
- misjudgment of dates or availability;
- attraction to competing offers or better deals;
- flexible or low-cost cancellation policies that encourage risk-free booking behavior.  

These patterns introduce **operational uncertainty** for hotels, leading to:

- difficulty predicting the **actual occupancy rate**;
- potential **revenue loss** due to late cancellations;
- complex **optimization of staffing, inventory, and dynamic pricing**.  

In this context, the central question of the use case is: **Can we predict whether a customer will honor their reservation or cancel it?**

.. image:: context/hotel.png
    :width: 300 px
    :align: center