suspicious-tweets-detection

Machine learning models for suspicious tweets detection

Dataset

{
  "dataset_name": "tweets_suspect.csv",
  "dataset_url": "https://drive.google.com/file/d/1US0luOWPOeVPpUQnpyxr41zrBmeg4Gjk/view?usp=share_link",
  "dataset_language": "english",
  "dataset_size": "60_000 rows X 2 Columns",
  "dataset_columns_description":
    {
      "message": "the tweet message",
      "label": "message label, 1 is suspicious, 0 non suspicious" 
    }
}

Project Workflow

Setup

1. Clone the project

  git clone git@github.com:abdoulfataoh/suspicious-tweets-detection.git
  cd suspicious-tweets-detection

2. Install Dependancies with poetry

  poetry install

3. Enable virtual env

  poetry shell

4. Check everything is OK

  make test

5. Run train and test

  make train_test

Output

see trained models at assets/models