Job Vacancy Recommendation System using JACCARD Method On Graph Database

In the rapidly evolving digital era, recommendation systems play a crucial role in helping users discover relevant information aligned with their preferences. PT Nirmala Satya Development, a company engaged in psychology and human resource development, faces challenges in utilizing big data consisting of 500 applicants, 500 job postings, and 500 job applications to generate accurate and relevant job recommendations. This study develops a job recommendation system using the Jaccard Coefficient method to measure similarity between users based on their job application history, implemented within a Neo4j graph database. The system models the relationships between entities through nodes and edges, allowing dynamic analysis using the Cypher Query Language. Testing on 237 users demonstrated that the majority received at least one relevant recommendation, with recall values often reaching 1.0, especially among users who had a single job target. The system achieved precision values ranging from 10% to 20%, which is considered acceptable given that ten recommendations are generated per user. The highest F1-score reached 0.33, although some users received F1 = 0 due to limited application history or unique preferences. Overall, the system effectively delivers personalized and efficient job recommendations, particularly for active users. This research also proves that combining the Jaccard Coefficient with a graph database structure is a powerful approach to representing and analyzing complex relationships between users and job postings in a modern recruitment platform.

Abstract licence: CC BY-SA

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.