A content based recommendation system which recommends job applications based on professional experience, location and category.
This software contains two major components.
- Offline Preprocessor
- Online Processor
All preprocessing and data gathering for content based filtering are done offline and stored in DB for a quicker response while recommending. This software is powered by GLASSDOOR and all contents are obtained from www.glassdoor.com Preprocessor consists of following components
- Job Listing Crawler
- Given a seed of URLS per location and sector, it crawls to actual Job Listings and gathers more detailed information about the advertisement.
- Employer Sentiment Analyzer
- Given location and sector, crawls to gather anonymous user reviews and performs sentiment analysis on them and drops employers below a threshold.
- Employer portfolio Crawler
- Given location and sector, crawls to gather information about various employers and industry information
Software presents user with a web interface for entering location, category and professional experience summary and responses with top 10 job listing that matches or best fits user.
- create a python virtual environment [ preferred ]
- clone this git repository and navigate to top folder.
- pip install -r requirements.txt
- python main.py , would launch the backend webserver
- open http://localhost:5000/index using any web browser