Bashar Alhafni

4 paper cards in Mova's library · 2026

Bashar Alhafni wrote 4 of the papers Mova reads, in 2026, most of them on Dependency parsing; often with Nizar Habash, Go Inoue, Junhong Liang.

The profile is the library's match of the name; two people with one name are not always told apart.

Papers

Arabic Morphosyntactic Tagging and Dependency Parsing with Large Language Models (2026)
The authors test whether instruction-tuned large language models can produce explicit Arabic grammatical analysis: 14 morphological features per token (part of speech, aspect, mood, voice, person, gender, number, case, state, four proclitics and one enclitic), and labelled CATiB dependency trees.
Arabic Sentence Segmentation Across Genres and Punctuation Conditions (2026)
Arabic sentence boundaries are hard to find because punctuation is sparse, inconsistent or absent in many texts, and clauses are often joined by conjunctions such as wa- and fa-.
Do Diacritics Matter? Evaluating the Impact of Arabic Diacritics on Tokenization and LLM Benchmarks (2026)
The paper asks how Arabic diacritics (vowel marks and related signs) change tokenization and the scores of large language models.
When Similar Means Different: Evaluating LLMs on Arabic–Hebrew Cognates (2026)
Arabic and Hebrew share many words with similar spellings. Some are true cognates with the same meaning, some are false friends with different meanings, and some are loanwords from a third language. The authors build SemCog Bench: 1,858 validated word pairs (858 true cognates, 364 false friends, 636 loanwords)…