DETEKSI LOWONGAN KERJA FIKTIF MENGGUNAKAN ALGORITMA NAÏVE BAYES DAN TF-IDF PADA PLATFORM REKRUTMEN KERJA ONLINE DI INDONESIA

Abstract

The development of information technology has made the job recruitment process easier to carry out online. However, this convenience also brings new problems, namely the increasing number of fake job vacancies circulating on various recruitment platforms in Indonesia. These kinds of vacancies are often used for fraud, theft of personal data, or asking applicants for money. Many job seekers become victims because it is difficult for them to distinguish which vacancies are genuine and which are fake. Based on this problem, this research was conducted to build a fake job vacancy detection model using the Naïve Bayes algorithm and TF-IDF text weighting method.

This study uses text data from job vacancy examples collected from several online platforms. Each data entry is then labeled as “genuine” or “fake” based on common characteristics that usually appear in fraud cases. After the data cleaning process and feature extraction using TF-IDF, the data is trained using the Naïve Bayes algorithm to see how well the model can recognize patterns in fake job vacancies. The evaluation is carried out using accuracy, precision, recall, and F1-score metrics.

The results show that the combination of Naïve Bayes and TF-IDF is able to provide fairly good performance in detecting fake job vacancies. The model can identify certain words or patterns that often appear in fake vacancies, such as unreasonable salary offers, instant selection processes, or the presence of required fees. These findings are expected to help job seekers be more careful when applying for jobs online. In addition, this research can be used as a basis for developing an automatic detection system that can be implemented by job recruitment platforms in Indonesia. This study still has limitations in terms of the amount of data, but it can be further developed using a larger and more diverse dataset.