Rida Khatoun

—

Unclaimed · AI-structured profilename-matched identity

Applied ML researcher spanning AI-driven cybersecurity (malware, email threats) and LLM-based social simulation.

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

—

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-14) 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 publication-volume metric is available in the current snapshot — only the three reviewed 2026 preprints are on record. This axis would measure total paper counts and output trends across venues and years, and is left unscored here.

Impact—

Citation counts, h-index, and downstream uptake are absent from the input data, so research impact cannot be scored. This axis would measure citations, citing works, and scholarly influence of the three reviewed papers.

Focus0/100

Focus scores 0, reflecting maximal dispersion: the field distribution splits exactly into thirds — ML malware detection, ML email threat detection, and LLM-based social simulation each take 0.333 of the reviewed output. With only three 2026 preprints, no single theme dominates yet; the breadth spans security applications and computational social science, suggesting an exploratory phase rather than a settled specialty.

Collaboration48/100

Collaboration scores 48, a mid-range value based on 10 distinct coauthors across the three reviewed papers. The recurring-partner list is led by Omran Berjawi (listed twice), alongside Bijied Brahimi, Vincent Cohadon, Gabriel Glazman, Rayan Al Mohaize, Walid Fahs, Giuseppe Fenza, and Sherali Zeadally — a mix spanning both security and social-simulation collaborators.

Activity—

No temporal activity metrics — publication cadence, submission frequency, or year-over-year trends — are available beyond the three 2026 preprints. This axis would measure research output over time and remains unscored until a longer publication window exists.

Reproducibility55/100

Reproducibility scores 55, the mean over three reviewed papers (reproPapers: 3), each rated medium readiness. Evolve's reviews note concrete gaps: AURA lacks a no-routing baseline, Delphi Scanner shows dataset accounting inconsistencies and misstated FP/FN rates, and the digital-twins paper released no code. Efficiency claims are concrete, but artifact availability limits replication.

Research summary

Rida Khatoun is an applied machine-learning researcher whose reviewed work concentrates on two fronts: deep-learning cybersecurity and LLM-driven computational social science. Evolve's catalog holds three 2026 arXiv preprints: AURA (adaptive uncertainty-routed email threat detection), Delphi Scanner (efficient static malware detection via API sequence modeling), and Digital Twins for Opinion Dynamics (a generative LLM framework for social networks). The field distribution splits exactly into thirds (0.333 each), and Khatoun occupies the last (senior) author position on AURA and Delphi Scanner, with a middle position on the digital-twins paper. Reported results are concrete — out-of-distribution F1 of 0.944–0.950 for AURA, 95.35% accuracy at 1.53MB with sub-millisecond inference for Delphi Scanner, and a claimed >50% MAE improvement over classical opinion models — though Evolve's reviews flag missing baselines, dataset-accounting inconsistencies, and absent code. Radar gives collaboration 48 (10 distinct coauthors) and reproducibility 55 (mean across 3 papers, all medium readiness); output, impact, and activity are unscored pending more data.

Collaboration pattern

Collaboration scores a mid-range 48, drawn from 10 distinct coauthors on three 2026 preprints. The provided frequent-coauthor list is led by Omran Berjawi (appearing twice) and includes Bijied Brahimi, Vincent Cohadon, Gabriel Glazman, Rayan Al Mohaize, Walid Fahs, Giuseppe Fenza, and Sherali Zeadally. Khatoun takes the last (senior) author slot on two of three papers and a middle slot on the third, consistent with small-team, hands-on collaboration; the network spans both security and social-simulation work.

Research keywords

phishing detectionspam filteringuncertainty routingentropy-based early exitmultimodal fusionLoRA fine-tuningmalware detectionstatic analysisAPI call sequencesopinion dynamicsgenerative agentsdigital twin

Highlights

AURA: entropy-routed multimodal email threat detection
Last-author 2026 preprint proposing adaptive uncertainty routing with entropy-based early exit, multimodal fusion, and LoRA fine-tuning for phishing and spam detection; Evolve credits strong out-of-distribution F1 of 0.944–0.950 but notes the routing gain lacks a no-routing baseline and the Nazario OOD set's dating is dubious.
Delphi Scanner: lightweight CNN malware detection on API sequences
Senior-author static-analysis work modeling Windows PE API call sequences with a CNN, reporting 95.35% accuracy, a 1.53MB model, and sub-millisecond inference with MITRE ATT&CK mapping; Evolve flags dataset accounting inconsistencies and misstated false-positive/false-negative rates.
LLM digital twins for opinion dynamics on social networks
Middle-author 2026 framework using generative agents (Mistral-7B) as digital twins of social-network users; Evolve notes the claimed >50% MAE improvement over classical opinion models on two Twitter datasets is promising, but rests on small networks, sentiment-proxy ground truth, and no released code.

Timeline

2026Published (arXiv) AURA, an adaptive uncertainty-routed framework for email threat detection, as last author.
2026Co-authored Digital Twins for Opinion Dynamics, a generative LLM framework simulating social networks with Mistral-7B agents, in a middle author position.
2026Published (arXiv) Delphi Scanner, efficient interpretable static malware detection via API sequence modeling, as last author.
2026All three reviewed works appear as arXiv preprints within a single year, split evenly across three research fields.

Papers in the Evolve catalog

Frequent co-authors

Omran Berjawi×3 Bijied Brahimi×1 Vincent Cohadon×1 Gabriel Glazman×1 Rayan Al Mohaize×1 Omran Berjawi×1 Rida Khatoun×1 Walid fahs×1 Giuseppe Fenza×1 Sherali Zeadally×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.

Radar scores are computed from a small bibliometric snapshot covering only the three Evolve-reviewed 2026 arXiv preprints; output, impact, and activity remain unscored (dimMissing).
No citation counts, h-index, venue history, or publication-volume statistics were available; none are inferred.
Institution, country, and ORCID are absent; the profile is name-resolved from public bibliographic data, and disambiguation from other researchers named Rida Khatoun is not guaranteed.
All quoted results (F1, accuracy, model size, MAE gains) come from Evolve's review notes, not independently verified source papers.