Abstract:

With the advancements of modern technology and social communication, the topic of advertising new job offers has been very common in today's world. Thus, predicting fake job offers will be a huge problem for anyone. Similar to many other classification tasks, there are numerous challenges in predicting fake job offers. In this paper, we intend to predict whether a job offer is genuine or not by using various data mining techniques and classification algorithms such as ANN, decision tree, support vector machine, naive Bayes classifier, random forest classifier, multi-layer perceptron, and deep neural network. We experimented with the Employment Scam Aegean Dataset (EMSCAD), which contains 18,000 examples. A deep neural network is the best classifier for this classification task. We used three dense layers for this deep neural network classifier. The trained classifier gives a classification accuracy of around 98% (DNN) in predicting fraudulent job advertisements.

Description:

.