Walid Abdelhalim

Walid Abdelhalim

orcid

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 February 2026 in Asian-Pacific Journal of Second and Foreign Language Education
Verified
Journal Article
Open access
PDF file
Mohamed Amin MekheimerMohamed Amin Mekheimer
Mohamed Amin Mekheimer
WAWalid Abdelhalim
Walid Abdelhalim
For non-Anglophone scholars, high-level foreign language proficiency (FLP) may serve as a form of agency that alleviates the identity conflict present in global academia. This research quantitatively analyzes the relationship among foreign language proficiency (FLP), global academic integration, and identity constructs through a survey of 496 Egyptian educational researchers. Through correlation and multiple regression analysis, we found that elevated FLP significantly predicts a more robust sense of global academic integration and, most crucially, lower identity dissonance. A strong local identity was the main factor linked to higher identity dissonance. However, a higher FLP was the only factor that was linked to lower dissonance. Our results suggest that learning the academic lingua franca is not just a matter of fitting in. Instead, it can be a strategic tool that gives scholars the rhetorical confidence to cross academic borders and take part in global scholarly discourse in a more authentic way. It therefore reframes proficiency as an important part of scholarly agency. As these conclusions are based on self-reported, cross-sectional data, they should be interpreted as highlighting associations rather than causal links.

20 April 2026 in British Educational Research Journal
Verified
Journal Article
Closed access
WAWalid Abdelhalim
Walid Abdelhalim
Kayaty AshourKayaty Ashour
Kayaty Ashour
Mohamed Amin MekheimerMohamed Amin Mekheimer
Mohamed Amin Mekheimer

11 September 2026 in Scientific Reports
Verified
Journal Article
Open access
PDF file
BABunder Sebail Alshammari
Bunder Sebail Alshammari
KAKhaled Nasser Alfraidi
Khaled Nasser Alfraidi
MAMohammad Omar Aljudaiey
Mohammad Omar Aljudaiey
WAWalid Abdelhalim
Walid Abdelhalim
The rapid development of large language models has intensified the need for reliable methods that distinguish human-authored Arabic text from AI-generated text while remaining interpretable across orthographic, dialectal, and register variation. This study evaluates baseline, transformer, ensemble, and hybrid architectures for Arabic authorship attribution and AI-generated text detection using the AraGenEval corpus and a 10,000-text dialectal Arabic supplement. It combines in-domain evaluation, robustness testing, bootstrap confidence intervals, paired statistical comparisons, feature-based explainability, prediction-level error analysis, and probability-calibration diagnostics. The prediction-level matrix was used as the controlling source for exact confusion counts, class-level metrics, Brier scores, expected calibration error, and error-code distributions. The ensemble of independently fine-tuned AraBERT v2 and XLM-RoBERTa Large produced the strongest in-domain result among the tested systems (accuracy = 94.7%, weighted F1 = .947, macro F1 = .947, MCC = .894, 95% bootstrap CI [.932, .960]), but performance declined on diacritic-stripped text and sharply under dialectal/social-media shift. Exact calibration analysis showed that the Brier score worsened from .046 in-domain to .167 in the dialectal condition, while 10-bin expected calibration error increased to .163 for dialectal data. The hybrid AraBERT-stylometric model numerically exceeded AraBERT in the primary in-domain matrix (weighted F1 = .936 vs. .921), but the paired comparison was not statistically significant (p = .199); consequently, handcrafted stylometric features are interpreted as explanatory aids rather than evidence of reliable predictive superiority. The findings support calibrated, dialect-aware, and uncertainty-reported Arabic AI-text detection.

7 August 2026 in Discover Sustainability
Verified
Journal Article
Open access
PDF file
Mohamed Amin MekheimerMohamed Amin Mekheimer
Mohamed Amin Mekheimer
WAWalid Abdelhalim
Walid Abdelhalim
This study examines how university faculty perceive and navigate the tradeoff between AI-enabled research productivity and the maintenance of academic integrity, a central challenge for sustainable scholarly ecosystems. Using an explanatory sequential mixed-methods design, the study surveyed 478 faculty members across six universities in Upper Egypt and then used 118 individual follow-up interviews to contextualize the quantitative patterns. Descriptively, respondents reported moderate integrity concerns (M = 3.81 on a 1–5 scale) alongside neutral-to-slight endorsement of AI productivity benefits (M = 3.06), indicating that integrity concerns remain salient amid emerging instrumental use. MANOVA revealed small but statistically significant group differences by discipline (eta_p squared = 0.023) and academic rank (eta_p squared = 0.031): early-career and STEM faculty reported comparatively higher AI adoption and stronger productivity orientations, whereas senior and humanities/social-science faculty reported comparatively higher integrity concerns. Theory-guided hierarchical regression showed that institutional context was positively associated with perceived productivity benefits after accounting for discipline, career stage, and AI-use frequency, while AI-use frequency was negatively associated with integrity-risk perceptions in the corresponding associational model. Because AI Adoption & Usage and Productivity Perception were very highly correlated ( r = .914), the results are interpreted as construct-proximate co-variation rather than causal evidence. Harman’s single-factor check did not indicate a single dominant method factor at item level, but common-method inflation remains a limitation. Qualitative findings show that faculty managed this tradeoff through principled resistance, pressure-driven adoption, pragmatic task segmentation, innovation-oriented experimentation, active risk management, and discipline-specific boundary setting. The study supports explicit AI-literacy training, disclosure guidance, and verification frameworks for responsible AI integration in academic publishing.