Comparison Study of Different Neural Network Models for Assessing Employability Skills of IT Graduates

Determining suitable jobs for IT graduates based on their skills and interests is challenging. This paper explores neural networks to predict job placement for IT graduates based on their qualifications. Various neural network architectures like Text-CNN, Bi-GRU-LSTM-CNN, and Bi-GRU-CNN with pre-trained word embedding are compared on an IT job dataset. An ensemble model combining these networks is also proposed. The ensemble model achieved the best F1-score of 72.71%, indicating neural networks show promise for job placement prediction. The paper provides insights on applying deep learning to this task and suggestions for future work