emoji_analysis

Topic : sentiment analysis using WhatsApp emojis.

Platform : Python3.6

To run:

• Clone the reposiory and cd into code

Syntax:

python3 get_score.py 'screenshot_name' 
python3 get_graph.py 'screenshot_name'
python3 get_diag.py 'screenshot_name' 

Example:

python3 get_score.py 'Subhasish'

To collaborate

• To contribute in this project, here is a small guide : grab a copy

Directory layout

• code : core , utils, notebooks containing .py files and .ipynb notebooks
• lib : pre-constructed dictionary
• resource : data, screenshots and other resources

The workflow :

TODO : this is dirty, make it organized

• Data collection & pre-processing
  • Collection of whatsapp recent used emojis (observed samples) : user screenshots
  • Collection of labelled whatsapp emojis : emoji_name & unicode representation
  • Collection of sentiment labelled emojis : see reference
• Image Processing (template matching)
  • implement template matching using CV2
  • obtain emojis for a perticular user
• Sentiment analysis(easy)
  • compute consolidated sentiment score
• Sentiment analysis(advanced)
  • classify different sentiment classes
  • propose a bijective mapping of sentiment score* and sentiment class

Data source :

• Click to view data source

Sentiment Classification :

• Click to attempt the Questionnaire

References:

• Sentiment of Emojis, click here

Contact:

• Contact me