Skip to content

Latest commit

 

History

History
35 lines (20 loc) · 1.56 KB

README.md

File metadata and controls

35 lines (20 loc) · 1.56 KB

ReverseRecruiter API

An API that can be used to detect emails from recruiters.

Motivation

Software recruiters use their own set of automated tools, which enables them to send a lot of email to potential candidates (which can get spammy). This API wraps around a custom trained machine learning model that detects messages that are likely generated by recruiters.

Resources

This project has the following key dependencies:

Dependency Name Documentation Description
FastAPI https://fastapi.tiangolo.com high performance, easy to learn, fast to code, ready for production
scikit-learn https://scikit-learn.org/stable/index.html Machine Learning in Python
NLTK https://www.nltk.org/ A leading platform for building Python programs to work with human language data
Docker https://www.docker.com/ We help developers and development teams build and ship apps

Machine Learning Model

  • The private dataset used for training and testing is a set of 600 real emails sent to me and frequent collaborator, Andres Tavio.
  • Each email was manually labeled as either a recruiter or not.
  • RigdeClassifier-TfidfVectorizer was chosen as it was one of the highest performing models that was tested.

Quickstart

  • Clone the repo locally
  • $ docker-compose build
  • $ docker-compose up