Social media analytics

Social media corporations are collecting and utilizing user data to enhance various kinds of algorithms. Collected user data need to be available to the user and can be retrieved via a manual request from their website.
This data can be further analysed with the help of social media analytics tool, which outputs a PDF with a brief insight into the texting behavior of the participants.

Description

Currently supported sources:

• instagram,
• facebook,
• whatsapp.

Only peer to peer conversation is supported (you vs friend).
Data export from a specific source usually consist of multiple folder named as namesurname_RandomID. If this is not the case, you need to rename conversation folder manually.

Supported use cases

Message throughout the history.shows overall texting trend.

Favorite emojis.shows favorite use of emojis (reactions are omitted).

Response time.shows response time per source. Response time is recorded only during daytime (07:00 - 23:00), where response time can not be longer than 24h (people often forget to replay).

Message throughout the day.shows person texting habit during the day.

Getting Started

Dependencies

• Python 3.9
• packages: os, csv, glob, json, re, pandas, collection, emoji, plotly fpdf

Installing

• Collected data should be put into data/raw/SOCIAL_MEDIA_SOURCE

Structure sample:

├── data/
    ├── raw/
        ├── facebook/
        │   │
        │   ├── name1surname1_randomid/
        │           ├── gifs/
        │           ├── photos/
        │           └── message_1.json
        │   └── name2surname2_randomid/
        │
        │
        ├── instagram
        └── whatsapp

Config sample:

{
  "my_username": "mynamemysurname",
  "friend_username": "name1surname1",
  "my_custom_name": "Me",
  "friend_custom_name": "Friend",
  "sources": [
    "whatsapp",
    "instagram",
    "facebook"
  ]
}

Executing program

• How to run the program

cd socialmedia_analytics
python src/main.py

• Results

Generated PDF can be found in pdf/results folder (see example).

Authors

Blaž Košenina

Version History

• 1.0 Version
  • Initial Release

License

This project is licensed under the MIT License