Findit: Job recommendation system
The aim of the project is to explore the use of recommender systems. The application is built as a recommendation system for people in search of jobs. Users can register a profile, login and view all the current jobs being advertised. Users can also add other users as friends, view each other's profiles, make reviews of jobs and post articles to the home news feed. Once the user has reviewed a number of jobs, they can have job recommendations selected for them - specifically the top five picks along with a predicted rating. The technologies used include: Python, Django and MySQL.
A recommender system is a technology that is deployed in an environment and attempts to predict the preference or rating that a user would give an item. Recommendation systems are integrated into sites like Amazon and Spotify where items or songs and artists are suggested to the user. The aim of this project is to create a recommendation system for people in search of jobs. Users can create profiles listing details like previous experience, what they are looking for, and particular skills they have obtained. Once a profile has been created, users can begin reviewing jobs that have either caught their interest, or completely missed what they are looking for. The recommender system aims to use the information obtained from the reviews to learn about the preferences of the user and to recommend a list of jobs that might interest them. The application was coded using Python inside the Django web framework that connects to SQL database.
‘Findit’ aims to create a social and personal platform for job hunting. Users have the chance to review jobs while also getting a sense of community by interacting with other users. The objective of the project was to provide users with job recommendations based on previous job reviews. In order to investigate the most successful way of developing this system, a range of different machine learning algorithms were tested, each of which used a different approach. It was discovered that Pearson's correlation was best suited for this type of project, providing the most efficient and accurate results. Users were able to view job recommendations along with a predicted rating that they might give.