Video delivery dataset analysis in Python using Jupyter notebook

Dataset Information

This dataset data.csv is measuring the data downloaded through P2P (peer to peer) and through the CDN (Content Delivery Network) by the viewers.

Content

Each data point has the following dimension, represented with 6 variables:

Name	Description
#stream	ID of each stream
isp	Name of the Internet Service Provider
browser	Browser name
connected	Boolean value, true if the user is connected to the backend during his session
p2p	Data downloaded through P2P (peer to peer)
cdn	Data downloaded through CDN (Content Delivery Network)

Goal

The goal of this notebook is to explore the dataset and give recommendations as to where the service should be improved.

Technologies used

• Python (version 3.7.0) - because the dataset is not too large to require some Big Data framework;
• Jupyter notebook (version 5.7.2) - because of the readability and explanation of the steps;
• Pandas library (version 0.23.4) - for data manipulation, analysis and plotting;
• Seaborn library (version 0.9.0) - data visualization library based on matplotlib.

Approach

1. Understanding the dataset in terms of columns and their values;
2. Visualizing which values affect connectivity to the backend;
3. How much data is sent in which stream and over which ISP/browser;
4. Visualizing correlation between ISP, browser, connectivity and bandwidth;
5. Trying to see if there are points where the service could be improved.