drukeroni/airline-satisfaction-analysis

---
title: "Airline Passenger Satisfaction Analysis"
license: "mit"
tags:
- dataset
- airlines
- satisfaction
- analysis
- kaggle
---

# Airline Passenger Satisfaction – EDA Report

This project analyzes the **Airline Passenger Satisfaction Dataset**, containing 103,904 rows and 25 columns describing passenger demographics, flight information, and service ratings.
The goal is to understand **which factors influence satisfaction**, identify important service features,
and compare satisfaction between different traveler types and flight classes.

---

## Dataset Overview

The dataset includes:

- Passenger demographics (age, gender, type of travel)
- Flight information (distance, delays, class)
- Service ratings (seat comfort, entertainment, food & drink, online boarding, etc.)
- Satisfaction label (Satisfied / Neutral or Dissatisfied)

---

## Key Questions

1. What is the overall satisfaction level?
2. Does type of travel (Personal vs Business) affect satisfaction?
3. Does flight class matter?
4. Which service factors most strongly predict satisfaction?
5. Do delays influence satisfaction?

---

# Key Visualizations

---

## **1. Passenger Satisfaction Distribution**

**Insight:**
Most passengers report *neutral or dissatisfied*.
This highlights the importance of understanding what drives satisfaction.

---

## **2. Satisfaction by Type of Travel**

**Insight:**
Business travelers have **much higher satisfaction**, while personal travelers are mostly dissatisfied.
This shows that purpose of travel strongly affects experience.

---

## **3. Satisfaction by Flight Class**

**Insight:**
- **Business Class → highest satisfaction**
- **Eco Class → lowest satisfaction**

This reinforces the strong connection between service level and satisfaction.

---

## **4. Correlation of Numeric Features with Satisfaction**

**Insight:**
Top predictors of satisfaction:

1. **Online boarding**
2. **Inflight entertainment**
3. **Seat comfort**
4. **On-board service**
5. **Leg room service**

Weak predictors:

- Flight distance
- Delays (almost no correlation)

This indicates satisfaction is driven by *service quality*, not flight length or delay.

---

## Summary of Key Findings

- Overall satisfaction levels are low.
- Travel purpose plays a MAJOR role in satisfaction.
- Business class delivers significantly better customer experience.
- Best predictors are **online boarding**, **seat comfort**, and **entertainment**.
- Delays surprisingly do NOT strongly affect satisfaction.
- Therefore, improving service experience has the largest impact.

---
## Video Presentation
You can watch my short presentation here:
https://youtu.be/2sE4KeQiqIg

## Technologies Used

- Python
- Pandas
- NumPy
- Matplotlib
- Seaborn

## Author

Roni Druker
Reichman University – Introduction to Data Science