ML researcher spanning LLM multi-agent opinion dynamics, GAN-hardened DDoS detection, and web browsing behavior mining
7
papers here
7
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
19
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-15) 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—
Insufficient data: this axis measures publication volume and consistency over time (papers per year, venue spread). The catalog currently holds only three 2026 arXiv entries, so no output trend can be scored yet.
Impact—
Insufficient data: this axis measures citation-based influence (citation counts, h-index, field-normalized attention). No citation or h-index figures are available for this profile, so impact is left unscored rather than estimated.
Focus0/100
Focus scores 0 because the three reviewed papers split evenly — one-third each across LLM-based multi-agent simulation, adversarial machine learning for network security, and web mining and user modeling. This is genuine breadth with no dominant specialization yet; keyword clusters (opinion dynamics, WGAN-GP, GraphSAGE) show each paper in a different methodological home, so concentration should emerge only as more work accumulates.
Collaboration56/100
Collaboration scores 41 from a footprint of 7 distinct coauthors across the reviewed papers. Recurring partners include Omran Berjawi, Walid Fahs, Giuseppe Fenza, Makram Chehayeb, Amina Rizk, and Ralph Elsaghbini. Authorship positions vary — last author twice, middle author once, no corresponding role — suggesting a compact, possibly multi-group network rather than one large lab.
Activity100/100
Insufficient data: this axis measures ongoing publishing cadence and momentum. All three reviewed papers carry 2026 arXiv dates, too short a window to establish whether activity is rising, holding, or slowing over time.
Reproducibility50/100
Reproducibility averages 43 across the 3 papers with recorded assessments. The behavioral fingerprinting and multi-agent bias studies are rated medium — transparent setups but weak baselines or single-model designs — while the GAN-based DDoS framework is rated low, with Evolve citing ambiguous evaluation details and naive threat modeling. The mid-range mean reflects consistently documented methods undermined by thin validation rigor.
Research summary
Rida Khatoun is an early-stage, name-resolved profile built from Evolve's reviewed catalog: three arXiv papers, all dated 2026, spread evenly across three distinct areas — LLM-based multi-agent simulation in computational social science, adversarial machine learning for network security, and machine learning applied to web mining and user modeling. The field distribution is an exact one-third split across each, which drives a focus score of 0: the reviewed work shows breadth rather than a consolidated specialty. Within the catalog Khatoun appears as last author on the multi-agent bias paper and the behavioral fingerprinting paper, and a middle author on the GAN-based DDoS detection work, with no corresponding-author role recorded. The reviewed output sits in a small collaboration footprint — 7 distinct coauthors, collaboration score 41 — with Omran Berjawi, Walid Fahs, Giuseppe Fenza, Makram Chehayeb, Amina Rizk, and Ralph Elsaghbini recurring. Evolve's reproducibility assessment averages 43 across the 3 papers: the DDoS framework is rated low and the other two medium. Verdicts are constructively critical throughout — a biased LLM minority (≤20%) shifting opinions beyond a Friedkin-Johnsen baseline, ~90% user identification and ~80% next-domain accuracy, and WGAN-GP augmentation improving detector robustness are all judged plausible, but single-model designs, naive threat modeling, and missing statistical tests are flagged.
Collaboration pattern
The reviewed work shows a small-team pattern: 7 distinct coauthors and a collaboration score of 41 indicate limited but real collaborative reach. Omran Berjawi, Walid Fahs, Giuseppe Fenza, Makram Chehayeb, Amina Rizk, and Ralph Elsaghbini recur across the catalog. Khatoun takes last-author positions on the multi-agent bias and web-fingerprinting papers and a middle position on the DDoS work, with no corresponding-author credit recorded — consistent with senior-but-not-corresponding roles on compact, arXiv-first teams.
Research keywords
Highlights
A small biased LLM minority reshapes opinion dynamics
WGAN-GP augmentation hardens DDoS detectors in synthetic tests
Short web sessions identify ~90% of users, predict next domain at ~80%
Timeline
- 2026Last-author arXiv paper on bias amplification in LLM multi-agent opinion dynamics
- 2026Middle-author arXiv paper proposing a GAN-based robust DDoS detection framework
- 2026Last-author arXiv paper on behavioral fingerprinting and next-domain prediction in web browsing
- 2026All three Evolve-catalogued reviews land in a single year, spanning three distinct fields
Papers in the Evolve catalog
Digital Twins for Opinion Dynamics: A Generative LLM Framework for Social Networks
AURA: Adaptive Uncertainty-Routed Analysis for Email Threat Detection
Delphi Scanner: efficient and interpretable static malware detection via API sequence modeling
Behavioral Fingerprinting and Navigation Prediction in Web Browsing
A GAN-Based Framework for Robust DDoS Attack Detection
Bias Amplification in Multi-Agent Network: How Biased Agents Shape Opinions and Rhetoric
Digital Persuasion: Understanding the Impact of Online Influencers on Public Opinion
Frequent co-authors
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.
- Profile is name-resolved from public bibliographic data; no institution, country, or ORCID is provided, so identity cannot be fully confirmed.
- Bibliometrics are OpenAlex-backed; citation counts, h-index, and venue-level impact data are absent, leaving output and impact unscored.
- The catalog covers only three arXiv preprints dated 2026; peer-reviewed publications and earlier career history are not represented.
- All quantitative claims in this profile (7 coauthors, 43 reproducibility mean, one-third field split) come solely from Evolve's reviewed snapshot.