/chat-explore

Primary LanguageJupyter NotebookMIT LicenseMIT

Chat Explore

This is a script that generates an infographic that has statistics and plots between any two users whatsapp chat export.

  • Limitation 1: Currently only works with two users chat, Group chats doesn't work.
  • Limitation 2: Currently only works for Android exports. iOS not supported.

Getting Started

1. Install Python 3+

If you don't already have Python 3+ installed, grab it from https://www.python.org/downloads/. You'll want to download install the latest version of Python 3.x. As of 2019-11-22, that is Version 3.8.

2. Get Your Whatsapp Chat Export Data

Here you can find out how to export your whatsapp chat between two users: https://faq.whatsapp.com/en/android/23756533/

To use this script, you need to export the chat history between yourself and any other user. That exported chat (.txt) needs to be present in the same system which runs the code mentioned below.

3. Clone This Repository

On https://github.com/kartheekpnsn/chat-explore, click the green "Clone or Download" button at the top right of the page. If you want to get started with this script more quickly, click the "Download ZIP" button, and extract the ZIP somewhere on your computer.

4. Install Dependencies

In a command prompt or Terminal window, navigate to the directory containing this repository's files. Then, type the following, and press enter:

pip install -r requirements.txt

wordcloud installation problem

Note: If you are facing issues installing wordcloud on Windows10. The follow the below procedure.

pip install <location_of_wordcloud_whl_file>

Example:

pip install "C:\Users\Testuser\Desktop\wordcloud-1.6.0-cp38-cp38-win32.whl"

5. Run the Script

In the same command prompt or Terminal window, type the following, and press enter:

python run.py -f <file_path>

Example:

python run.py -f "C:\Users\Testuser\Desktop\WhatsApp Chat with Kartheek.txt"

6. Sample Output

Sample 1:

Sample 1 Text

Sample 2:

Sample 2 Text

7. Update Patch

  • 16Jan2020 Added First to text feature in the plots
  • 17Jan2020: Sentiment over period of time, Usual Sentiment through out the 24 hours
    • Can help in understanding if people are showing positive emotions in the morning and negative in the evenings.
  • TODO: Parser for iOS exports.

8. Contributors

9. Idea Credits

Acknowledgement:

  • This entire idea is inspired from a reddit post (links posted below):