--- title: "Hamdy Mubarak — papers, page 3 — Elamur Research" description: "Hamdy Mubarak, Hamad bin Khalifa University: 67 papers on Elamur Research (2014–2026), h-index 30 — affiliations, topics, co-authors and…" canonical: "https://research.elamur.ai/free/en/authors/hamdy-mubarak/page/3/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/hamdy-mubarak/page/3/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / [Hamdy Mubarak](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/) / page 3 # Hamdy Mubarak · page 3 67 papers in our library · 2014–2026 · page 3 of 3 [The profile, affiliations and co-authors →](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/) ## Papers The most cited first (OpenAlex and our library). [← prev](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/page/2/) [1](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/) [2](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/page/2/) **3** 51. **[From RAG to Agentic RAG for Faithful Islamic Question Answering](https://md.research.elamur.ai/free/en/papers/bhatia-2026-rag-agentic-rag-faithful-islamic/)** *(2026 · 8 citations)* LLMs are increasingly used to answer Islamic questions, where an unsupported answer can have real religious consequences. 52. **[ArabGend: Gender Analysis and Inference on Arabic Twitter](https://md.research.elamur.ai/free/en/papers/mubarak-2022-arabgend-gender-analysis-inference-arabic/)** *(2022 · 8 citations)* The paper tackles the shortage of resources for analysing and predicting user gender on Arabic Twitter. The authors collected Arabic tweets in two separate phases. An earlier batch of about 4.35M tweets was used only to build a lexicon of gendered self-description words. A later batch of about 100M tweets was then… 53. **[AraSafe: Benchmarking Safety in Arabic LLMs](https://md.research.elamur.ai/free/en/papers/mubarak-2025-arasafe-benchmarking-safety-arabic-llms/)** *(2025 · 6 citations)* Arabic has lacked a large safety benchmark built from genuine user input. Existing resources are mostly English or Chinese, and the Arabic ones are small, translated, or limited to a few labels such as hate or abuse. AraSafe addresses this gap. Recruited speakers from several Arab countries interacted with Arabic… 54. **[Diacritization of Maghrebi Arabic Sub-Dialects](https://md.research.elamur.ai/free/en/papers/abdelali-2018-diacritization-maghrebi-arabic-sub-dialects/)** *(2018 · 6 citations)* Arabic is usually written without short vowels, and systems such as text-to-speech must restore them. Most diacritization research has targeted Modern Standard Arabic (MSA), so the authors address two Maghrebi sub-dialects, Moroccan (MOR) and Tunisian (TUN). Their data are two dialectal New Testament translations of… 55. **[Advancing Arabic Diacritization: Improved Datasets, Benchmarking, and State-of-the-Art Models](https://md.research.elamur.ai/free/en/papers/mohamed-2025-advancing-arabic-diacritization-improved-datasets/)** *(2025 · 5 citations)* Written Modern Standard Arabic usually omits diacritics, the marks for short vowels and grammatical endings. As a result, one consonant skeleton can stand for several different words. The authors improve automatic restoration of these marks in three ways: better training data, better evaluation, and better models. 56. **[Beyond Orthography: Automatic Recovery of Short Vowels and Dialectal Sounds in Arabic](https://md.research.elamur.ai/free/en/papers/kheir-2024-beyond-orthography-automatic-recovery-short/)** *(2024 · 5 citations)* Arabic dialects contain sounds that Modern Standard Arabic spelling does not show (such as the g of Gulf and Egyptian loanwords, or p and v), and the written text often lacks short vowels. 57. **[Arabic Curriculum Analysis](https://md.research.elamur.ai/free/en/papers/mubarak-2020-arabic-curriculum-analysis/)** *(2020 · 5 citations)* Arabic has few NLP tools for checking whether school textbooks meet their learning goals, for example in the vocabulary they teach. 58. **[Answer Selection in Arabic Community Question Answering: A Feature-Rich Approach](https://md.research.elamur.ai/free/en/papers/belinkov-2015-answer-selection-arabic-community-question/)** *(2015 · 5 citations)* The paper tackles answer selection in Arabic community question answering. Given a forum question and its attached comments, each comment must be labelled DIRECT, RELATED or IRRELEVANT. The authors use the SemEval-2015 Task 3 Arabic setup, built on the Fatwa corpus of religious questions answered by scholars. Every… 59. **[UL2C: Mapping User Locations to Countries on Arabic Twitter](https://md.research.elamur.ai/free/en/papers/mubarak-2021-ul2c-mapping-user-locations-countries/)** *(2021 · 4 citations)* Twitter lets people type any free-text location in their profile. These strings are messy: they include nicknames, informal spellings, mixed scripts, emojis and made-up places. That makes it hard to tell which country an Arabic-speaking user is from, and very few tweets carry geotags. Mubarak and Hassan built UL2C, a… 60. **[Nahw: A Comprehensive Benchmark of Arabic Grammar Understanding, Error Detection, Correction, and Explanation](https://md.research.elamur.ai/free/en/papers/mubarak-2026-nahw-comprehensive-benchmark-arabic-grammar/)** *(2026 · 3 citations)* The paper builds Nahw, a benchmark for Modern Standard Arabic grammar taken from school grammar material. It has four tasks in a ladder: grammar understanding (5K multiple-choice questions, Nahw-MCQ, grades 6 to 12), error detection and correction in 100 short passages with about five errors each (Nahw-Passage, 511… 61. **[WASIL: In-the-Wild Arabic Spoken Interactions with LLMs](https://md.research.elamur.ai/free/en/papers/ali-2026-wasil-wild-arabic-spoken-interactions/)** *(2026 · 2 citations)* Voice assistants built as a speech-recognition stage followed by an LLM are hard to diagnose. When a user dislikes an answer, the cause could be a misrecognized utterance, a request that was unclear or impossible from the start, or a weak LLM. WASIL is an Arabic resource designed to pull these causes apart. 62. **[QVoice: Arabic Speech Pronunciation Learning Application](https://md.research.elamur.ai/free/en/papers/kheir-2023-qvoice-arabic-speech-pronunciation-learning/)** *(2023 · 2 citations)* QVoice is a mobile app for practising Arabic pronunciation. It targets two groups: non-native learners, and native speakers whose regional dialect affects their Modern Standard Arabic (MSA). The authors present it as a system demo. 63. **[ArabicWeb-Edu: Educational Quality Data for Arabic LLM Training](https://md.research.elamur.ai/free/en/papers/hawasly-2025-arabicweb-edu-educational-quality-data/)** *(2025 · 1 citation)* Arabic is heavily underrepresented in LLM training data, and the Arabic text that exists on the web is often noisy or repetitive. 64. **[DialG2P: Dialectal Grapheme-to-Phoneme. Arabic as a Case Study](https://md.research.elamur.ai/free/en/papers/hawasly-2025-dialg2p-dialectal-grapheme-phoneme-arabic/)** *(2025 · 1 citation)* The paper addresses grapheme-to-phoneme (G2P) conversion for dialectal Arabic, using Egyptian (EGY) as the test case. 65. **[ArGuard Shared Task: Harmful Content Detection in Arabic Memes and LLM Prompts](https://md.research.elamur.ai/free/en/papers/alam-2026-arguard-shared-task-harmful-content/)** *(2026)* ArGuard is a 2026 shared task on detecting harmful Arabic content in two settings that earlier work treated separately: multimodal memes and prompts sent to LLMs. 66. **[Beyond Cultural Knowledge: Evaluating Arabic Cultural Appropriateness of Large Language Models](https://md.research.elamur.ai/free/en/papers/altinisik-2026-cultural-knowledge-evaluating-arabic-cultural/)** *(2026)* Most cultural benchmarks for LLMs check what a model knows. This paper instead measures how a model behaves when it gives advice, opinions or guidance in Arabic. The authors built AraBehave, a benchmark of open-ended prompts. The prompts come mainly from user-flagged chat logs, plus tester-written seeds and expert… 67. **[Automatic Expansion and Retargeting of Arabic Offensive Language Training](https://md.research.elamur.ai/free/en/papers/mubarak-2021-automatic-expansion-retargeting-arabic-offensive/)** *(2021)* The paper tackles a gap in Arabic offensive-language detection. Generic classifiers often miss abuse aimed at a particular entity, because such abuse can rely on deliberately twisted spellings of the target's name or on ordinary words used with hostile intent. The authors propose growing the training data… [← prev](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/page/2/) [1](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/) [2](https://md.research.elamur.ai/free/en/authors/hamdy-mubarak/page/2/) **3**