Khuong Nguyen-An

Vietnam National University Ho Chi Minh City

Unclaimed · AI-structured profileORCID OpenAlex

VNU-HCMC researcher on honest, reproducible evaluation in legal-compliance NLP and encrypted-traffic security

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

—

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-02) 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—

Output data is missing: the snapshot lists no publication counts or venue history beyond the two reviewed papers. This axis would measure cumulative publication volume and its trajectory over time; it cannot be scored from available data.

Impact—

Impact data is missing: no citation counts, h-index, or uptake metrics are provided. This axis would measure how widely the reviewed work is cited and adopted; no score is estimated here.

Focus0/100

Focus scores 0 because the field distribution is an even 50/50 split between two distinct areas: natural language processing for legal-compliance information extraction, and computer security spanning encrypted traffic analysis and network intrusion detection. The zero reflects breadth across two parallel research lines rather than weak output — each line produced one Evolve-reviewed, high-reproducibility paper in 2026.

Collaboration36/100

Collaboration scores 36 from a raw count of 5 distinct co-authors in the snapshot. Both reviewed papers place Khuong in the last author position (non-corresponding), a slot conventionally associated with senior or supervisory contribution. The frequent-coauthor list — To Duy Hinh, Nguyen Le Quoc Anh, Phan Van Tri, Hoang-Huy Nguyen-Huu, Van-Tri Phan — suggests a compact, stable local team rather than a wide international network.

Activity—

Activity data is missing: the snapshot contains only 2026 papers, with no multi-year cadence or recency trend. This axis would measure ongoing research pace and continuity; a single-year sample cannot support a score.

Reproducibility100/100

Reproducibility scores 100, the radar maximum: reproMean is 100 across 2 papers (reproPapers: 2), each rated 'high' for reproducibility readiness by Evolve. The RoPA work pairs its benchmark with an honest three-level evaluation, and the C2 study is explicitly leakage-controlled. This is the profile's strongest axis — evaluation rigor and openly specified methodology are central to both research lines.

Research summary

Khuong Nguyen-An is a researcher at Vietnam National University Ho Chi Minh City whose profile, resolved from public bibliographic data (ORCID 0000-0002-9910-6387), currently rests on two 2026 arXiv papers in the Evolve catalog, both with last-author placement. The work spans two equally weighted fields — natural language processing for legal-compliance information extraction and computer security in encrypted traffic analysis — a 50/50 split that yields a focus score of 0 on the radar. The first paper builds the first Vietnamese benchmark for Records of Processing Activities (RoPA), combining hybrid RAG with locally deployed large language models; Evolve's verdict credits an honest three-level evaluation in which local models match cloud ones on roughly 55% coverage, while precision remains unmeasured and the data are synthetic. The second is a leakage-controlled measurement study of multi-view fusion for encrypted C2 detection, arguing that fusion's clean-condition gain is only 0.022 F1 while evaluation artifacts dominate headline results. Methodological honesty is the connecting thread: both papers carry 'high' reproducibility-readiness ratings, giving a reproducibility score of 100 across 2 papers — the profile's strongest signal. Collaboration is compact, with 5 distinct co-authors (score 36). Output, impact, and activity axes lack underlying data and are reported as missing rather than estimated.

Collaboration pattern

Collaboration centers on a compact Vietnam-based team: 5 distinct co-authors appear in the snapshot (radar collaboration score 36). Frequent co-authors include To Duy Hinh, Nguyen Le Quoc Anh, Phan Van Tri, Hoang-Huy Nguyen-Huu, and Van-Tri Phan — the last two likely name-order variants of one collaborator, which the raw count may not deduplicate. On both reviewed papers Khuong is the last-listed, non-corresponding author, consistent with a supervisory role within the same group.

Research keywords

RoPApersonal data protectionRAGhybrid retrievallocally deployed LLMsVietnamese benchmarkencrypted traffic analysiscommand and control (C2)Cobalt Strikedata leakageTLS fingerprinting (JA3/JA3S)evaluation methodology

Highlights

First Vietnamese RoPA benchmark with a three-level evaluation
Introduces the first Vietnamese Records of Processing Activities (RoPA) benchmark, evaluated under an honest three-level protocol using hybrid RAG and locally deployed large language models; Evolve notes local models match cloud ones on roughly 55% coverage, while precision remains unmeasured and the data are synthetic.
Leakage-controlled measurement study of encrypted C2 detection
Argues multi-view fusion's clean-condition gain is 0.022 F1 while evaluation artifacts dominate performance claims; the paper documents evaluation pitfalls in encrypted traffic analysis (Cobalt Strike C2, JA3/JA3S TLS fingerprinting) with controlled data-leakage treatment.
Both 2026 releases rated high reproducibility
Evolve rates both 2026 arXiv papers 'high' on reproducibility readiness, producing the profile's top score: reproducibility 100 (reproMean over 2 papers). The RoPA benchmark ships with a three-level evaluation protocol, and the C2 study is explicitly leakage-controlled — evaluation rigor is the connective thread across both fields.

Timeline

2026arXiv release: 'Automated Extraction of Records of Processing Activities (RoPA) Using Hybrid RAG and Locally Deployed Large Language Models' — first Vietnamese RoPA benchmark; last-author placement.
2026arXiv release: 'Multi-View Fusion for Encrypted C2 Detection: A Leakage-Controlled Measurement Study of Evaluation Pitfalls' — reports fusion's clean-condition gain of 0.022 F1.
2026Both 2026 papers enter the Evolve catalog with 'high' reproducibility-readiness ratings, yielding a reproducibility score of 100 across 2 papers.

Papers in the Evolve catalog

Frequent co-authors

To Duy Hinh×1 Nguyen Le Quoc Anh×1 Phan Van Tri×1 Hoang-Huy Nguyen-Huu×1 Van-Tri Phan×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 derive from a public bibliometric snapshot; only two Evolve-reviewed papers (both 2026 arXiv preprints) are available, and no peer-reviewed venue status is provided.
output, impact, and activity dimensions have no underlying data in this snapshot — no publication counts, citation metrics, or multi-year cadence — so they are reported as missing, not scored.
Profile is name-resolved from public bibliographic data anchored to ORCID 0000-0002-9910-6387; independent identity confirmation is advised before citing.
The frequent-coauthor list contains apparent name-order variants (Phan Van Tri / Van-Tri Phan), so the 5-distinct-coauthor count may overcount actual individuals.