Blog Aus- und Weiterbildung | 7th OJA Forum (Part 5) – A Hybrid Methodology for Job Ad Title Normalization

7th OJA Forum (Part 5) – A Hybrid Methodology for Job Ad Title Normalization

18. September 2024 / Jana Fingerhut, Ibrahim Rahhal

In this blog series, international experts on the topic of analyzing online job advertisements present their work. The presentations were given at the 7th OJA Forum, which is jointly organized twice a year by the Federal Institute for Vocational Education and Training and the Bertelsmann Stiftung. Topics at the 7th OJA Forum included the visualization and use of data using dashboards and the assignment of job titles to online job advertisements using various methodological approaches. This article is about the contribution of Ibrahim Rahhal from the International University of Rabat.

A Hybrid Methodology for Job Ad Title Normalization

Occupation identification is an important step in structuring job ads and gaining insights from them. Using traditional approaches requires extensive labeled datasets, limiting their applicability. Ibrahim Rahhal proposes a hybrid methodology using BERT for sector classification and similarity measures for job title identification. He introduces a novel document embedding strategy to enhance accuracy. Results show that incorporating document embedding-based methods, such as weighting and noise removal, boosts accuracy by 23.5% compared to Bag of Words models surpassing 85% accuracy in some sectors. Evaluation demonstrates superiority or parity with state-of-the-art methods, highlighting efficacy in identifying emerging occupations in the Moroccan job market.