An improved deep neural network model for job matching
Yu Fu Deng, Hang Lei, Xiaoyu Li, Yiou Lin · 2018
Job matching which benefit job seekers, employees and employers is very important today. In this work, a deep neural network model is proposed to predict an employee's future career details, which includes position name, salary and company scale based on the online resume data. Like most NLP tasks, the input features are multi-field, non-sparse, discrete and categorical, while their dependencies are mostly unknown. Previous works were mostly focused on engineering, which resulted in a large feature space and heavy computation. To solve this task, we use embedding layers to explore feature interactions and merge two automatically learned features extracted from the resumes. Experimental results on over 70,000 real-word online resumes show that our model outperforms shallow models, like SVM and Random Forests, in effectiveness and accuracy.