Project for the GE GhostRed Hackathon February 2017


List of libraries used: BeautifulSoup (bs4) NLTK NLTK dataset movie_reviews Word_Cloud tldextract textract Django numby scipy sklearn

Features: Django Framework tf/idf analysis on job descriptions and resumes Resumes and job positions taken from Dice using API Github data pulled using Github API Sentiment analysis performed on personal web pages

Functionality: Our app functions as a webapp operating on a django framework over an apache server. To use, you visit the url of the web server (which on our computer is localhost:8080). Then, select either seeking a job or hiring, and you will be redirected to a webpage where you will either input your name, github username, and resume if seeking a job, or email, company culture, and job description if a company looking to hire. Once submitted, the system will then map you to the appropriate page. If you're a job seeker, you'll be matched with a list of companies and positions considered most appropriate for you. If you're a company, then you'll ideally be matched with employees most suited towards both your culture and your job description. This is done through text analysis of the resume, job description, github profile commits and contributions, and personal webpage.

Potential Impact: This app will enable GE and many other companies to more quickly locate and sort applicants based on how well they fit with the company according to our metrics. These metrics can be improved and refined over time, as well as being scalable due to the selection of algorithms and the would-be improved organization of the data. By quickly locating and sorting applicants, company can spend less money and fewer resources towards these endeavors, saving money in the long run. In addition, employees which are a better fit with the company will be more likely to be noticed, thus improving the workforce of the company and the general enjoyment of the employees within and soon-to-be hired.

Built With

Share this project: