JobBERT: Understanding Job Titles through Skills
This work addresses the need for better job title understanding in HR processes, such as online recruitment and internal organization, but it is incremental as it builds on existing language models with domain-specific enhancements.
The paper tackled the problem of accurately modeling job titles for HR tech applications by proposing JobBERT, a neural representation model that augments a pre-trained language model with skill label co-occurrence information, resulting in considerable improvements in job title normalization compared to generic sentence encoders.
Job titles form a cornerstone of today's human resources (HR) processes. Within online recruitment, they allow candidates to understand the contents of a vacancy at a glance, while internal HR departments use them to organize and structure many of their processes. As job titles are a compact, convenient, and readily available data source, modeling them with high accuracy can greatly benefit many HR tech applications. In this paper, we propose a neural representation model for job titles, by augmenting a pre-trained language model with co-occurrence information from skill labels extracted from vacancies. Our JobBERT method leads to considerable improvements compared to using generic sentence encoders, for the task of job title normalization, for which we release a new evaluation benchmark.