Papers by Walid Magdy

18 papers
Pearl: A Multimodal Culturally-Aware Arabic Instruction Dataset (2025.findings-emnlp)

Copied to clipboard

Challenge: Mainstream large vision-language models (LVLMs) inherently encode cultural biases, highlighting the need for diverse multimodal datasets.
Approach: They propose to construct a large-scale Arabic multimodal dataset and benchmark explicitly designed for cultural understanding.
Outcome: The proposed dataset covers ten culturally significant domains covering all Arab countries and includes two evaluation benchmarks (PEARL and PEARL-LITE) and a specialized subset (PearL-X).
DLAMA: A Framework for Curating Culturally Diverse Facts for Probing the Knowledge of Pretrained Language Models (2023.findings-acl)

Copied to clipboard

Challenge: a few benchmarking datasets have been released to evaluate the factual knowledge of pretrained language models.
Approach: They propose a framework for curating factual triples from Wikidata that are culturally diverse.
Outcome: The proposed framework is built of factual triples from three pairs of contrasting cultures with 78,259 triples.
Culture Matters in Toxic Language Detection in Persian (2025.acl-long)

Copied to clipboard

Challenge: Toxic language detection is crucial for creating safer online environments and limiting the spread of harmful content.
Approach: They compare different methods for toxic language detection in Persian to fine-tune, enrich data, and cross-lingual transfer learning.
Outcome: The language of a country with cultural similarities to Persian yields better results in transfer learning.
Urban Dictionary Embeddings for Slang NLP Applications (2020.lrec-1)

Copied to clipboard

Challenge: a new set of word embeddings is released to improve word embedment performance . word embeds provide useful representations of meanings of words in vectors .
Approach: They present a set of word embeddings trained on Urban Dictionary . they show they have high performance across a range of common word embeding evaluations .
Outcome: The first set of word embeddings trained on Urban Dictionary has high performance . the embeddables perform better on a range of common word evaluation tasks .
Part-of-Speech Tagging for Arabic Gulf Dialect Using Bi-LSTM (L18-1)

Copied to clipboard

Challenge: Part-of-speech (POS) tagging is one of the most important building blocks in many natural language processing (NLP) applications.
Approach: They propose to use a POS tagger for Arabic Gulf dialect to improve POS tagging accuracy.
Outcome: The proposed POS tagger improves POS tagging accuracy for the Arabic Gulf dialect from 75% accuracy to 91% accuracy using a bi-LSTM labeler.
Exploring Author Context for Detecting Intended vs Perceived Sarcasm (P19-1)

Copied to clipboard

Challenge: Existing studies on textual sarcasm detection use manual labelling and tag-based distant supervision to detect sarcasm.
Approach: They define author context as the embedded representation of their historical tweets and suggest neural models that extract these representations.
Outcome: The proposed models achieve state-of-the-art on two datasets labelled manually and via tag-based distant supervision indicating a difference between intended and perceived sarcasm .
AX-MABSA: A Framework for Extremely Weakly Supervised Multi-label Aspect Based Sentiment Analysis (2022.emnlp-main)

Copied to clipboard

Challenge: Aspect Based Sentiment Analysis is a dominant research area with potential applications in social media analytics, business, finance, and health.
Approach: They propose a weakly supervised multi-label Aspect Category Sentiment Analysis framework which does not use any labelled data.
Outcome: The proposed framework outperforms weakly supervised baselines on four benchmark datasets and is able to generate multiple aspect category-sentiment pairs per review sentence.
Alexandria: A Multi-Domain Dialectal Arabic Machine Translation Dataset for Culturally Inclusive and Linguistically Diverse LLMs (2026.acl-long)

Copied to clipboard

Challenge: Arabic is a highly diglossic language where most daily communication occurs in regional dialects rather than modern standard Arabic (MSA).
Approach: They propose a large-scale, community-driven, human-translated dataset to bridge this gap . Alexandria covers 13 Arab countries and 11 high-impact domains . it provides unprecedented granularity by associating contributions with city-of-origin metadata .
Outcome: The Alexandria dataset covers 13 Arab countries and 11 high-impact domains . it provides unprecedented granularity by associating contributions with city-of-origin metadata . Alexandria is a training resource and a rigorous benchmark for evaluating MT and LLMs based on the Alexandria dataset .
iSarcasm: A Dataset of Intended Sarcasm (2020.acl-main)

Copied to clipboard

Challenge: Existing methods for detecting intended sarcasm have shown low performance compared to previous studies.
Approach: They propose a dataset of tweets labeled for intended sarcasm by their authors . they aim to encourage future NLP research to develop methods for detecting sarkasmus in text as intended by the authors of the text .
Outcome: The proposed model shows that existing methods are biased or obvious and sarcasm could be understudied.
Validating Automatic Evaluation of Controllable Counterspeech Generation: Rankings Matter More Than Scores (2026.eacl-long)

Copied to clipboard

Challenge: Existing methods for evaluating attributes of counterspeech are limited and the validity of such evaluations is questionable when the classifiers themselves have only modest performance.
Approach: They examine the automatic evaluation of counterspeech attributes using a multi-attribute counterseech dataset containing 2,728 samples.
Outcome: The proposed model can be trusted by classifier validation, and it can rank models with confidence.
Revisiting Common Assumptions about Arabic Dialects in NLP (2025.acl-long)

Copied to clipboard

Challenge: Existing assumptions about Arabic dialect variation are not quantitatively verified.
Approach: They extend and analyze Arabic dialects to assess their validity using a multi-label dataset . they find that the assumptions oversimplify reality and are not always accurate .
Outcome: The proposed methods oversimplify reality and are not always accurate, the authors argue . they show that the proposed assumptions oversimply represent reality and may hinder future work .
Multi-Dialect Arabic POS Tagging: A CRF Approach (L18-1)

Copied to clipboard

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%.
Sarcasm Detection is Way Too Easy! An Empirical Comparison of Human and Machine Sarcasm Detection (2022.findings-emnlp)

Copied to clipboard

Challenge: sarcasm detection datasets focus on intended, rather than perceived sarcasm, but there is no comparison between human and machine performance.
Approach: They collect author-annotated sarcasm datasets that focus on intended, rather than perceived sarcasticism . they compare human-level benchmarks to that of state-of-the-art sarkasmatic detection systems .
Outcome: The proposed datasets compare human and machine performance on sarcastic tasks in English and Arabic.
ALDi: Quantifying the Arabic Level of Dialectness of Text (2023.emnlp-main)

Copied to clipboard

Challenge: Existing work on Dialect Identification (DI) on the sentence level has focused on binary tasks, whereas ALDi treats the task as binary.
Approach: They propose a dataset which contains 127,835 sentences manually labeled with their level of dialectness.
Outcome: The proposed model can identify dialectness on a range of other corpora, providing a more nuanced picture than traditional DI systems.
Casablanca: Data and Models for Multidialectal Arabic Speech Recognition (2024.emnlp-main)

Copied to clipboard

Challenge: despite recent advances in speech processing, the majority of world languages and dialects remain uncovered.
Approach: They propose to collect and transcribe a new Arabic dataset for eight dialects . they also develop strong baselines exploiting the new dataset .
Outcome: The proposed dataset covers eight Arabic dialects, including Algerian, Egyptian, Emirati, Jordanian, Mauritanian, Moroccan, Palestinian, and Yemeni.
Chandler: An Explainable Sarcastic Response Generator (2021.emnlp-demo)

Copied to clipboard

Challenge: sarcasm generators assume intended meaning is opposite of literal meaning . sarcastically generated responses are more specific and coherent to input .
Approach: They propose a system that generates sarcastic responses to a given utterance . they ground their generation process on a formal theory that unambiguously differentiates .
Outcome: The proposed system generates sarcastic responses to a given utterance.
Should a Chatbot be Sarcastic? Understanding User Preferences Towards Sarcasm Generation (2022.acl-long)

Copied to clipboard

Challenge: sarcasm generation research focused on creating more human-like interactions . previous research focused only on how to generate text that people perceive as sarkastic .
Approach: They propose a theory-driven framework for generating sarcastic responses that allows us to control linguistic devices included during generation.
Outcome: The proposed framework allows us to control the linguistic devices included during generation.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations