원문정보
초록
영어
Often, a worker’s occupational domain and educational qualifications mismatch, leading to a discrepancy with their actual skill set. This can lead to either underperformance or overperformance, as the competencies required by the company may differ from those possessed by potential employees. This research aims to develop a system for matching user profiles and job vacancies called the Job Matching System. The system can help individuals find jobs that suit their educational background and skills. By collecting large amounts of data from the Jobstreet.co.id website, the system automatically identifies job opportunities in four job categories, which are data analyst, digital marketing, developer, and machine learning. We used the Latent Dirichlet Allocation approach to analyze the gathered data and identify potential topics within large data sets. Based on this information, the database correlates educational background with appropriate job classifications. The "Job Matching" website has a built-in decision-making tool. Prior knowledge of the worker’s background is required to aid in the matching procedure and facilitate the alignment of competencies with the worker’s preferred industry throughout the matching process.
목차
Ⅰ. Introduction
Ⅱ. Literature Review
2.1. Big Data in Human Resource Management
2.2. Text Mining (TM) and Latent Dirichlet Allocation (LDA)
Ⅲ. Methodology
3.1. Data Collection: Web Scrapping
3.2. Data Preprocessing
3.3. Data Processing
3.4. Decision System Design
Ⅳ. Findings and Disc
4.1. Data Collection
4.2. Data Preprocessing and Processing
4.3. Decision System Design
4.4. Discussion
Ⅴ. Conclusion and Future Research Directions
Acknowledgement