This project is about sparkify music streaming service which is a hypothetical concept of music streaming platform, user's can listen music on their devices, it is similar kind of service like spotify, pandora etc. Our main goal of the project is to find the user's who left the service or the user who is going to churn.
Read my blog on this project as well here
- Project Motivation
- Project Summary
- Project Components
- File Descriptions
- Installation & Run
- Results
I recently gained some knowladge about Spark and it's libraries, I found out this project can help me build an overall intution about spark & it's environment. I applied these skills to analyze the data which proveded by the udacity for this project.
This project is about sparkify music streaming service which is a hypothetical concept of music streaming platform, user's can listen music on their devices, it is similar kind of service like spotify, pandora etc. we will be analyzing it's data which is pretty interesting. The data is about user's interaction with their service, like user's data contains their gender, location, which songs they listen to, which are the pages they visit, did they upgrade or downgrade & so on, again music data like songs we don't have much data related to songs, again we have log informatin of the user like, when did a user registered with the serive, sessions, is the errors they encounter then what are those errors, timestamp of each action, when did a perticular song was played and all, like this we have a lot explore from this data in this project.
Our main goal of the project is to find the user's who left the service or the user who is going to churn. For that we have all the information almost or we don't, we have to find that in this project pretty much exciting han? So, without the further go let's dive into it & start working on the data. To make our path easy we have set small milestones as follows which helps us understand the high level of the process.
This project basically divided into sections as follows:
- Business Understanding
- Data Understanding
- Data Preprocessing
- Feature Engineering
- Data preparation for Modeling
- Modeling
- Evaluating Model
- Displaying Result
- Justification & Conclusion
-
Data
| - data/data_link.txt : link of the data is provided
| - Project_Sparkify.ipynb : main notebook file.
*README.md : this files has all description of the project.
This project needs the spark installed on your machine if you are planning to run on your local machine along with that anaconda distribution for notebook functionality & that's all for the run pyspark & you are good to go on local env. I have also included google colab code if you are trying to run it on cloud, you will only need the data file.