A*STAR NLP researcher working across LLM code intelligence, financial report generation, and human-AI collaboration
3
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
55
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-17) 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 score yet: this axis measures per-work publication volume and venue-type mix, but the per-work output data needed to compute it are absent from the current snapshot; only aggregate yearly counts are available.
Impact—
Not yet scored. This axis would measure per-paper citation impact, field-normalized. Aggregate indicators exist (h-index 55, i10-index 165), but the per-work citation distribution needed for scoring is missing from the snapshot.
Focus0/100
Focus scores 0/100, reflecting exceptional breadth rather than narrow specialization. The recent field distribution splits evenly (0.5 each) between human-AI collaboration and the future of work, and software engineering (LLM code intelligence), and the three 2026 catalog papers each sit in a different field. Historical topic mass is likewise spread across topic modeling (171), NLP techniques (121), information retrieval (86), text analysis (62), and multimodal ML (41).
Collaboration66/100
Collaboration scores 66/100. Over the last three years the profile records 68 works with 26 distinct coauthors — roughly 2.6 works per partner — indicating a broad but partially repeated collaborator pool. Recurring names include Yajing Yang, Yunshan Ma, Kelvin J. L. Koa, Jiaying Wu, Caleb Ziems, Raymond Chan, and Nancy F. Chen, alongside the group entity CIVIC-AI Collaboration. All three catalog papers are middle- or last-authored, none corresponding.
Activity100/100
Activity scores 100/100, the axis's maximum. The snapshot logs 68 works in the last three years, and yearly output reached its recorded peaks late in the series: 29 works in 2023, 26 in 2024, 19 in 2025, and 23 already in 2026 — well above the roughly 5–19 works per year logged from 1998 through 2022.
Reproducibility60/100
Reproducibility scores 60/100, computed from two assessed papers (reproMean 60). The over-editing benchmark study rates high readiness (controlled benchmark, validated metrics), AnalysisBank rates medium (LLM-judged metrics, possible benchmark overlap), and the workflow-framework whitepaper rates low as an empirically untested conceptual framework. One of the three catalog papers is not yet assessed.
Research summary
Min-Yen Kan (profile resolved from public OpenAlex bibliographic data; ORCID 0000-0001-8507-3716) is a researcher at the Agency for Science, Technology and Research (A*STAR), Singapore, with a long record in natural language processing. The snapshot attributes an h-index of 55 and an i10-index of 165, with topic mass concentrated in topic modeling (171 works), NLP techniques (121), information retrieval and search behavior (86), advanced text analysis (62), and multimodal machine learning (41). Activity is high and rising: 68 works in the last three years, with yearly counts peaking at 29 (2023), 26 (2024), and 23 (2026) — the activity score is 100/100. The Evolve catalog holds three 2026 arXiv papers spanning LLM code intelligence and human-AI collaboration (field distribution 0.5/0.5), where Kan appears as middle or last author; the reproducibility read averages 60/100 across two assessed papers. None of the catalog papers lists Kan as corresponding author, consistent with a senior collaborative role. With 26 distinct recent coauthors (collaboration 66/100) and wide topical spread (focus 0/100), this is a broad, senior, heavily collaborative profile. Per-work output and impact axes remain unscored pending citation-level data.
Collaboration pattern
Recent collaboration is broad but anchored: 68 works over three years with 26 distinct coauthors, roughly 2.6 works per partner. Recurring partners include Yajing Yang, Yunshan Ma, Kelvin J. L. Koa, Jiaying Wu, Caleb Ziems, Raymond Chan, and Nancy F. Chen; the CIVIC-AI Collaboration entry suggests at least one group-authored line of work. In the 2026 catalog papers Kan takes middle or last authorship and is never the corresponding author, consistent with a senior collaborative role.
Research keywords
Highlights
A controlled benchmark exposes LLM over-editing in code repair
AnalysisBank lifts LLM financial-report insight rates 1.7–3.7×
Six workflow-level conditions for when AI genuinely augments work
Timeline
- 2026Co-authored (middle position) the workshop whitepaper 'When Does AI Augment Work?', a workflow-level framework with six conditions for genuine AI augmentation.
- 2026Senior-authored (last position) 'When Models Edit Too Much', a controlled over-editing benchmark for LLM code repair rated high for reproducibility readiness.
- 2026Senior-authored (last position) 'AnalysisBank', whose expert analysis pattern library lifts LLM insight rates 1.7–3.7× in financial report generation.
Papers in the Evolve catalog
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.
- Bibliometric indicators (h-index, i10-index, topic counts, yearly counts, coauthor counts) come from an OpenAlex snapshot only; no other database was consulted.
- The output and impact radar axes are unscored because per-work publication and citation data are absent from the snapshot.
- The reproducibility read averages 60/100 over just two assessed papers; the third catalog paper is unassessed.
- All three catalog papers are 2026 arXiv preprints rather than peer-reviewed versions, and the yearly counts include a stray 1963 record that likely reflects a metadata artifact.