Abstract
The digital world has evolved from finding job positions in everyday newspapers to online platforms for turning in job applications. With more than thousands of job opportunities getting created every day and more than hundreds of applications received for a single position, recruiting the right candidates and landing on the right job for the skillset has become a tedious and stressful process. One way to ease this strenuous process is to filter out and recommend the most suitable positions to candidates and most suitable candidates to recruiters. While traditional recommendation systems aim to resolve these issues, they run into conditions like cold start and rating matrix problems when they have little or no information about user preferences and hence affect the accuracy of the recommender systems. Hence, in this project we propose an efficient job recommendation system using Knowledge Graphs (KG) that could recommend and explain the reason behind recommending the most suitable choices. This job recommendation system will consist of three phases namely the data extraction phase, recommendation phase and visualization phase. During the data extraction phase candidate resumes are parsed using deep learning techniques to obtain the keywords and skills acquired. A dataset containing job positions and related information are also extracted. The data extracted is built into a knowledge graph to perform prediction, classification, and recommendation tasks. Finally, different visualizations are provided to different user types to help them plan during this tedious process of job/candidate search.