Wolfgang Nejdl
Wolfgang Nejdl wrote 2 of the papers Mova reads, in 2025–2026, most of them on Named entity recognition; often with Hailay Kidu Teklehaymanot, Debela Desalegn Yadeta, Dren Fazlija.
OpenAlex · ORCID · h-index 63 · 18275 citations · 627 works — openalex-snapshot, 2026-10-14
Topics: Named entity recognition 1
Papers
- Expanding the Lexicon of Ge'ez Based African Languages: A Comparative Study of Amharic and Tigrinya (2026)
XLM-R's Latin-centred subword vocabulary splits Amharic and Tigrinya words, written in the Ge'ez script, into many pieces: on the authors' held-out text, 2.07 tokens per word for Amharic and 3.13 for Tigrinya. - MoVoC: Morphology-Aware Subword Construction for Ge'ez Script Languages (2025)
The paper argues that BPE subwords cut Ge'ez-script languages (Amharic, Tigrinya, Ge'ez, Tigre) at the wrong places, splitting stems and affixes.
Co-authors (3)
- Hailay Kidu Teklehaymanot ×2
- Debela Desalegn Yadeta
- Dren Fazlija