Papers by Younes Samih
Multilingual Nonce Dependency Treebanks: Understanding how Language Models Represent and Process Syntactic Structure (2024.naacl-long)
Copied to clipboard
| Challenge: | a number of studies have focused on making explicit the linguistic information encoded in language models (LMs) however, this method has been criticized for various reasons. |
| Approach: | They introduce a framework for creating nonce treebanks for multilingual UD corpora . they investigate word co-occurrence statistics and show how nonce data affects the performance of syntactic dependency probes. |
| Outcome: | The proposed framework satisfies syntactic argument structure and ensures grammaticality via language-specific rules. |
German and French Neural Supertagging Experiments for LTAG Parsing (P18-3)
Copied to clipboard
| Challenge: | Lexicalized Tree Adjoining Grammars are a linguistically motivated grammar formalism that allows parsers to express linguistic generalizations that are not captured by statistical parsing. |
| Approach: | They propose a supertagging approach combined with deep learning to extract LTAG supertags from the French Treebank and propose n-best supertailing for German and French. |
| Outcome: | The proposed supertagging approach is able to extract LTAG supertags from the French Treebank and n-best supertracking for German and German. |
Multi-Dialect Arabic POS Tagging: A CRF Approach (L18-1)
Copied to clipboard
Kareem Darwish, Hamdy Mubarak, Ahmed Abdelali, Mohamed Eldesouki, Younes Samih, Randah Alharbi, Mohammed Attia, Walid Magdy, Laura Kallmeyer
| Challenge: | Existing work on dialectal POS tagging is rather scant with POS tags for most dialects being nonexistent or of limited availability. |
| Approach: | They propose a dataset of POS-tagged Arabic tweets in four major dialects and a tagging guideline for each dialect. |
| Outcome: | The proposed model can tag four different dialects with an average accuracy of 89.3%. |
A System for Diacritizing Four Varieties of Arabic (D19-3)
Copied to clipboard
| Challenge: | Short vowels, aka diacritics, are omitted when writing different varieties of Arabic . diacritization is essential for language learning and text-to-speech applications . |
| Approach: | They propose a system for recovering diacritics in Arabic without short vowels . they use a character-based sequence-to-sequence deep learning model . |
| Outcome: | The proposed system beats all previous SOTA systems for Arabic varieties . it uses a character-based sequence-to-sequence deep learning model . |
Probing for Constituency Structure in Neural Language Models (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Using standard probing techniques, we examine whether contextual neural language models implicitly learn syntactic structure. |
| Approach: | They investigate to which extent contextual neural language models implicitly learn syntactic structure. |
| Outcome: | The proposed model is able to represent constituents of different categories within the neuron activations of a LM such as RoBERTa with high performance even on manipulated data. |
AraVQA: Building a New Arabic Factoid Visual Question Answering Dataset from Wikipedia (2026.acl-long)
Copied to clipboard
| Challenge: | Existing Arabic VQA datasets focus on culturally-specific and dialect-aware domains. |
| Approach: | They propose a pipeline that leverages Wikipedia template tags to extract relevant information for each image and utilize it to generate a new visual question answering dataset. |
| Outcome: | The proposed pipeline can enhance existing VLMs on Arabic VQA tasks by leveraging Wikipedia template tags. |
A Few Topical Tweets are Enough for Effective User Stance Detection (2021.eacl-main)
Copied to clipboard
| Challenge: | Recent work that employs unsupervised classification has shown that user stance detection is highly accurate on vocal Twitter users, but fails for less vocal users, who may have only authored a few tweets about a target. |
| Approach: | They propose to perform user stance detection on vocal Twitter users using contextualized embeddings to represent tweets and then perform unsupervised classification of the user. |
| Outcome: | The proposed method outperforms two baseline methods and achieves 89.6% accuracy and 91.3% macro F-measure on eight controversial topics. |
Multilingual Multi-class Sentiment Classification Using Convolutional Neural Networks (L18-1)
Copied to clipboard
| Challenge: | a new language-independent model for sentiment analysis is proposed for social media . a sentiment dictionary cannot list all the possible ways people can express their opinions . |
| Approach: | They propose a language-independent model for multi-class sentiment analysis using a neural network architecture. |
| Outcome: | The proposed model does not rely on language-specific features such as ontologies, dictionaries, or morphological or syntactic pre-processing. |
From Multiple-Choice to Extractive QA: A Case Study for English and Arabic (2025.coling-main)
Copied to clipboard
Teresa Lynn, Malik H. Altakrori, Samar M. Magdy, Rocktim Jyoti Das, Chenyang Lyu, Mohamed Nasr, Younes Samih, Kirill Chirkunov, Alham Fikri Aji, Preslav Nakov, Shantanu Godbole, Salim Roukos, Radu Florian, Nizar Habash
| Challenge: | Recent years have brought about very fast developments in Natural Language Processing (NLP), but many other languages are overlooked due to limited resources. |
| Approach: | They propose to repurpose a multilingual BELEBELE dataset for a task of extractive QA in the style of machine reading comprehension. |
| Outcome: | The proposed approach could be used to extract QA in the style of machine reading comprehension. |
Linear Semantic Segmentation for Low-Resource Spoken Dialects (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing models for semantic segmentation are primarily developed and evaluated on high-resource written text, limiting their effectiveness on low-resourced conversational varieties. |
| Approach: | They propose a multi-genre benchmark for semantic segmentation in Arabic, focusing on dialectal discourse. |
| Outcome: | The proposed model outperforms baselines on dialectal non-news genres while performing well on high-resource written text. |
Highly Effective Arabic Diacritization using Sequence to Sequence Modeling (N19-1)
Copied to clipboard
| Challenge: | Arabic text is written without short vowels (or diacritics) their presence is essential for properly verbalizing Arabic . |
| Approach: | They propose a character-level sequence-to-sequence deep learning model that recovers both types of diacritics without the use of explicit feature engineering. |
| Outcome: | The proposed model outperforms all previous state-of-the-art models on overlapping windows of words . it achieves a word error rate (WER) of 4.49% compared to the state- of-the art systems . |
Comprehensive Layer-wise Analysis of SSL Models for Audio Deepfake Detection (2025.findings-naacl)
Copied to clipboard
| Challenge: | Existing algorithms for audio deepfake detection are based on layer-wise analysis of self-supervised learning (SSL) models. |
| Approach: | They conduct a layer-wise analysis of self-supervised learning (SSL) models for audio deepfake detection across diverse contexts. |
| Outcome: | The proposed models achieve competitive equal error rate (EER) scores even when employing a reduced number of layers. |
Cultural Benchmarking of LLMs in Standard and Dialectal Arabic Dialogues (2026.acl-long)
Copied to clipboard
Muhammad Dehan Al Kautsar, Saeed Almheiri, Momina Ahsan, Bilal Elbouardi, Younes Samih, Sarfraz Ahmad, Amr Keleg, Omar El Herraoui, Kareem Elzeky, Abed Alhakim Freihat, Mohamed Anwar, Zhuohan Xie, Junhong Liang, Mohammad Rustom Al Nasar, Preslav Nakov, Fajri Koto
| Challenge: | Most benchmarks focus on short text snippets in Modern Standard Arabic (MSA), overlooking cultural nuances that naturally arise in dialogues. |
| Approach: | They propose a culturally grounded conversational dataset covering 13 Arabic-speaking countries, in both Modern Standard Arabic (MSA) and each country’s respective dialect, spanning 12 daily-life topics and 54 fine-grained subtopics. |
| Outcome: | The proposed model performs worse on all three tasks than the MSA benchmark. |