Saad Ezzini

King Fahd University of Petroleum and Minerals · Saudi Arabia

Unclaimed · AI-structured profileORCID OpenAlex

Arabic NLP researcher at KFUPM building shared-task benchmarks for genre classification and dialectal Arabic MT

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-13) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.

Research profile

Output—

No output score is available yet. This axis would measure total publication volume and its trajectory over time; OpenAlex logs 60 works in the trailing three years, but a validated output assessment has not been computed.

Impact—

No impact score is available yet. This axis would measure citation influence such as field-normalised citation rates; only h-index 9 and i10-index 8 are on record, with no percentile or venue-level citation data to support a score.

Focus0/100

The focus score of 0 reflects wide dispersion rather than a single track: the field split is an even 50/50 between Arabic NLP and machine translation, and OpenAlex topic counts span topic modeling (47), NLP techniques (38), software engineering research (27), software engineering practices (17) and sentiment analysis (16). The two reviewed 2026 shared tasks sit squarely in the Arabic-language core of this spread.

Collaboration54/100

The collaboration score of 54 rests on 14 distinct coauthors. Both reviewed 2026 papers follow the shared-task consortium pattern — long author lists with Ezzini in a middle, non-corresponding position. Recurring partners include Mo El-Haj, Mustafa Jarrar, Zaid Alyafeai, Abdellah El Mekki, AbdelRahim A. Elmadany and Samar M. Magdy, plus Shadi Abudalfa, indicating a network centred on Arabic-language benchmark building.

Activity100/100

The activity score of 100 is the profile's strongest signal. OpenAlex per-year counts climb from 1 work in 2018 to 3 in 2021, 5 in 2022, 10 in 2023, 12 in 2024, 19 in 2025 and 29 in 2026, with 60 works recorded across the trailing three years — a steep, uninterrupted acceleration.

Reproducibility55/100

The reproducibility score of 55 is a mean readiness rating across 2 reviewed papers, both rated medium. AraGenre 2026 pairs a sound hierarchical, definition-guided design with honest results but thin gold-label validation; AlexandriaX 2026 delivers a broad three-subtask benchmark whose findings are credible, though a transductive S2 winner and single-reference spBLEU ranking weaken headline claims.

Research summary

Saad Ezzini is an Arabic NLP researcher at King Fahd University of Petroleum and Minerals (KFUPM), Saudi Arabia, whose OpenAlex-linked profile (ORCID 0000-0001-7657-4738) splits evenly between computational linguistics / Arabic NLP and natural language processing / machine translation (50/50 field distribution). The Evolve catalog reviews two of his 2026 co-authored papers, both shared-task overviews: AraGenre 2026, a hierarchical definition-guided Arabic genre classification task built around zero-shot label generalisation, and AlexandriaX 2026, the first shared task on dialectal Arabic machine translation, with subtasks covering cross-dialect translation, MT error detection and quality estimation. In both he appears as a middle author rather than corresponding author, consistent with the large-consortium benchmark style of the Arabic NLP community. The bibliometric record shows a steep activity ramp: 1 work in 2018 rising to 29 in 2026, with 60 works in the trailing three years and an activity score of 100. Reported citation figures are h-index 9 and i10-index 8, while output and impact dimensions remain unscored for lack of validated data. Topic counts (topic modeling 47, NLP techniques 38, software engineering research 27) indicate breadth beyond Arabic MT. Evolve's reproducibility rating averages 55 across two papers rated medium. The profile is name-resolved from public bibliographic data.

Collaboration pattern

Ezzini's collaboration score is 54 with 14 distinct coauthors recorded. Both reviewed 2026 papers follow the shared-task consortium pattern: long author lists in which he holds a middle, non-corresponding position. Recurring partners include Mo El-Haj, Mustafa Jarrar, Zaid Alyafeai, Abdellah El Mekki, AbdelRahim A. Elmadany and Samar M. Magdy, alongside Shadi Abudalfa — a network centred on Arabic-language benchmark building.

Research keywords

Arabic NLPArabic genre classificationDialectal Arabic machine translationShared tasks and benchmarksZero-shot label generalisationDefinition-guided learningHierarchical classificationTopic modelingSentiment analysis and opinion miningCross-dialect translationMachine translation quality estimationSoftware engineering research

Highlights

AraGenre 2026: definition-guided Arabic genre classification
A hierarchical shared task built on definition-guided learning and zero-shot label generalisation across Arabic genres. Evolve's verdict: sound design and honest results, but gold-label validation stays thin; reproducibility rated medium.
AlexandriaX 2026: the first dialectal Arabic MT shared task
A three-subtask benchmark covering cross-dialect translation, MT error detection and quality estimation. Evolve finds it broad and useful, while noting a transductive S2 winner and single-reference spBLEU ranking weaken headline claims; reproducibility rated medium.
Twin 2026 shared tasks anchor Arabic evaluation infrastructure
Within a single year Ezzini co-authored overviews for two community benchmarks — Arabic genre classification and dialectal Arabic MT — matching the 50/50 field split in his OpenAlex record and the 2026 peak of 29 recorded works.

Timeline

2018First OpenAlex-recorded work appears: 1 publication, the starting point of a profile that later accelerates sharply.
2022Output reaches 5 recorded works, up from 3 in 2021 — the ramp begins.
202412 works recorded, following 10 in 2023 — growth is now sustained year over year.
202519 works recorded, a new annual high at the time.
2026Peak year: 29 works recorded; Evolve reviews two co-authored shared-task overviews, AraGenre 2026 and AlexandriaX 2026.

Papers in the Evolve catalog

Frequent co-authors

Mo El-Haj×2 Mustafa Jarrar×2 Shadi Abudalfa×1 Nguyen Minh Chi×1 Nguyen Minh Quan×1 Abdellah El Mekki×1 AbdelRahim A. Elmadany×1 Samar M. Magdy×1 Saad Ezzini×1 Mo El-Haj×1 Mustafa Jarrar×1 Zaid Alyafeai×1

About this profile

This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.

Bibliometric figures come from an OpenAlex author record (identity: openalex) matched via ORCID 0000-0001-7657-4738; the profile is name-resolved from public bibliographic data.
Output and impact dimensions are unscored; only h-index 9 and i10-index 8 are reported, with no citation percentiles or validated output counts available.
The reproducibility mean of 55 covers just 2 Evolve-reviewed papers; reproducibility beyond these is not assessed.
Timeline entries before 2026 are derived from OpenAlex per-year work counts, not from papers individually reviewed by Evolve.