Mitigating Language Bias in Cross-Lingual Job Retrieval: A Recruitment Platform Perspective
This work addresses language bias in job-matching systems for recruitment platforms, offering an incremental improvement through a unified approach.
The paper tackled the problem of disjointed text analysis in cross-lingual job retrieval by proposing a unified sentence encoder with a multi-task dual-encoder framework, which outperformed state-of-the-art models despite a smaller size and introduced a novel metric (LBKL) that demonstrated significant bias reduction and superior cross-lingual performance.
Understanding the textual components of resumes and job postings is critical for improving job-matching accuracy and optimizing job search systems in online recruitment platforms. However, existing works primarily focus on analyzing individual components within this information, requiring multiple specialized tools to analyze each aspect. Such disjointed methods could potentially hinder overall generalizability in recruitment-related text processing. Therefore, we propose a unified sentence encoder that utilized multi-task dual-encoder framework for jointly learning multiple component into the unified sentence encoder. The results show that our method outperforms other state-of-the-art models, despite its smaller model size. Moreover, we propose a novel metric, Language Bias Kullback-Leibler Divergence (LBKL), to evaluate language bias in the encoder, demonstrating significant bias reduction and superior cross-lingual performance.