A network embedding approach to a fair, efficient and fulfilling job market.

The development of a model, based on Conditional Network Embedding, that represents job seekers, vacancies and metadata without sensitive information, like gender, in low-dimensional embeddings. Downstream tasks that aim for more job market efficiency, like interest prediction between job seekers and vacancies, are consequently fairer. This model can also provide suggestions to job seekers and employers to widen their reach.