SIGNATURE WORK
CONFERENCE & EXHIBITION 2023

The COVID-19 pandemic has resulted in a surge of unemployment, making job seekers vulnerable to fake job postings in social media. The proliferation of fake job postings in social media has become a significant problem, leading job seekers to be misled with false information and reducing the credibility of companies. To address this issue, this study proposes two methods for predicting fake job postings. One approach involves utilizing topic modeling and clustering techniques, while the other applies the bag of words model and classification model. While the topic modeling methods produced some accuracy, additional numerical data and classification models were required to improve the prediction accuracy. Various classification models were employed to further predict the fakeness of job postings, with KNN model achieving a high prediction rate of 98 percent. Overall, this research demonstrates the significance of employing multiple models to accurately predict fake job postings and combat their proliferation in social media.