APPLICATION OF ENSEMBLE LEARNING TO ONLINE RECRUITMENT FRAUD DETECTION

• Main Author: Zaky, Muhammad
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/83146
• Institution: Institut Teknologi Bandung

The development of internet usage has influenced many recruitment processes, but it has also brought negative impacts in the form of an increase in online recruitment fraud cases. To address this challenge, this research aims to develop a machine learning-based Online Recruitment Fraud Detection System. The approach taken involves the use of an ensemble learning model that combines the CatBoost Classifier algorithm and the BERT Natural Language Processing (NLP) model. The two algorithms will be trained separately, then the probability results will be combined using the weighted voting method to obtain the final probability. The development of this system follows the CRISP-DM (Cross-Industry Standard Process for Data Mining) methodology, which ensures systematic steps in model development. The main objective of this research is to compare the performance of each algorithm with the performance of the ensemble model, and determine the best approach in building an Online Recruitment Fraud Detection System. The expected result is the creation of a model with a high level of precision and a tolerable recall value in identifying fake job vacancies online. This research is expected to contribute significantly to the development of effective solutions to the challenges of recruitment fraud in the online context.