Author

Malek Essayeh

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Taxonomy-Aware Hybrid Retrieval with Bounded LLM Reranking for TalentCLEF 2026 Task A

The results suggest that a strong retrieval backbone makes bounded LLM reranking safer, with the largest benefit in the cross-lingual setting, and that ESCO graph evidence works best in the submitted system as a calibration signal within that retrieval-first design.

Arij Riabi, Malek Essayeh · 1 citation