Hua Lu

Aalborg University

Unclaimed · AI-structured profileOpenAlex

Representation learning for urban and trajectory data: Hua Lu, spatio-temporal ML researcher at Aalborg University

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

—

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-16) 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—

Not yet scored: this snapshot lacks a verified publication count across indexed venues, so total output, venue mix, and annual volume cannot be assessed. This axis would measure aggregate publications and their distribution over venues and years.

Impact—

Insufficient data: citation counts and downstream uptake for this researcher's papers are not yet available, so scholarly influence cannot be scored. This axis would measure citations received, referencing works, and the reach of the reviewed contributions.

Focus0/100

Scores 0: the observed record divides evenly across three field labels — machine learning for intelligent transportation systems, machine learning / urban computing, and spatio-temporal data mining — each holding exactly one-third of the three catalogued papers. The labels describe adjacent spatio-temporal ML problems, so part of the spread likely reflects labeling granularity rather than scattered interests, but this small sample yields no dominant field.

Collaboration50/100

Scores 50, with 11 distinct co-authors in the observed window. The frequent-coauthor list is dense with Aalborg University names — Sean Bin Yang, Zongyi Xu, Jilin Hu, Bin Yang, Christian S. Jensen, Ying Sun, Kristian Torp, Hao Miao, Xiangmeng Wang, Tung Kieu, and Torben Bach Pedersen — indicating team-embedded work in a spatio-temporal data-management environment. All three catalogued papers place this researcher in a middle author position rather than leading it.

Activity—

Not yet scored: no complete, date-stamped publication history is available, so publishing cadence and continuity in recent years cannot be evaluated. This axis would measure output frequency and engagement over time.

Reproducibility70/100

Scores 70, the mean readiness rating across all 3 papers Evolve has reviewed. DGCPath rates high, while CURE and REFINE rate medium. The verdicts document unusually broad experiments — 10 baselines on 3 trajectory datasets for DGCPath, 4 datasets for REFINE, 3 cities for CURE — yet repeatedly flag absent significance testing and thin statistics, which keeps the score below the top band.

Research summary

Hua Lu is a researcher at Aalborg University working on machine learning for spatio-temporal and urban data. The Evolve catalog holds three 2026 arXiv papers, each with Hua Lu in a middle author position: CURE on confounder-aware multi-view urban region representation learning, DGCPath on distribution-aware generative contrastive path representation, and REFINE, the extended version of a KDD'26 paper on trajectory representation learning via closed-loop transcription. The methodological signature is consistent across all three: self-supervised objectives — contrastive learning, masked reconstruction, diffusion-generated views — fused over graphs, road networks, and trajectories. Evolve's verdicts credit unusually broad empirical coverage (strong gains on 3 cities for CURE, wins over 10 baselines on 3 trajectory datasets for DGCPath, wins on 4 datasets for REFINE) while flagging thin statistical testing throughout and, for CURE, deconfounding that lacks causal rigor. Reproducibility scores 70, the mean across all 3 reviewed papers (DGCPath high; CURE and REFINE medium). Collaboration scores 50 with 11 distinct co-authors, including Sean Bin Yang, Zongyi Xu, Bin Yang, Christian S. Jensen, Ying Sun, Kristian Torp, and Torben Bach Pedersen — a cluster consistent with Aalborg's spatio-temporal data-science environment. Output, impact, and activity are unscored for insufficient data.

Collaboration pattern

Collaboration scores 50 with 11 distinct co-authors. The recurring list — Sean Bin Yang, Zongyi Xu, Jilin Hu, Bin Yang, Christian S. Jensen, Ying Sun, Kristian Torp, Hao Miao, Xiangmeng Wang, Tung Kieu, and Torben Bach Pedersen — is anchored at Aalborg University's spatio-temporal data group, and the same cluster recurs across all three 2026 papers. Hua Lu appears in a middle author position on each catalogued paper, consistent with a contributing role inside a senior-led team.

Research keywords

urban region representation learningtrajectory representation learningpath representation learningself-supervised learningcontrastive learningmulti-view learningmasked reconstructiondiffusion modelsgraph neural networksspatio-temporal data miningconfounder-aware learningintelligent transportation systems

Highlights

CURE: confounder-aware multi-view urban region embeddings
The paper adds heuristic shared-component residualization to multi-view urban embeddings and reports strong gains on 3 cities; Evolve's verdict notes the deconfounding lacks causal rigor and statistical testing.
DGCPath: diffusion-based distributional contrastive path representation
Diffusion-generated views combined with Jensen-Shannon distributional contrast outperform 10 baselines on 3 trajectory datasets; it is the only catalogued paper rated high for reproducibility readiness, though Evolve flags thin novelty and no significance tests.
REFINE: closed-loop transcription unifies trajectory representation learning
This KDD'26 extended paper unifies contrastive and generative trajectory representation learning through closed-loop transcription, with broad wins on 4 datasets; Evolve finds the theory conditional and the statistical support thin.

Timeline

2026CURE, a confounder-aware multi-view urban region representation learning framework, appears on arXiv.
2026DGCPath, a distribution-aware generative contrastive framework for self-supervised path representation learning, is released on arXiv in an extended version.
2026REFINE, the extended version of a KDD'26 paper, unifies contrastive and generative trajectory representation learning via closed-loop transcription on arXiv.

Papers in the Evolve catalog

Frequent co-authors

Sean Bin Yang×3 Zongyi Xu×3 Jilin Hu×3 Bin Yang×3 Christian S. Jensen×2 Ying Sun×2 Kristian Torp×2 Hao Miao×1 Xiangmeng Wang×1 Tung Kieu×1 Torben Bach Pedersen×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.

Bibliometrics are derived from OpenAlex-based data only; output, impact, and activity lack sufficient underlying data and are left unscored.
'Hua Lu' is a common name; this profile is name-resolved from public bibliographic data and reflects only the three 2026 arXiv papers reviewed by Evolve.
All three catalogued papers list this researcher in a middle author position, so first- or corresponding-author leadership is not observable in this sample.
The three field labels overlap heavily (all spatio-temporal machine-learning variants), so the focus score of 0 may partly reflect labeling granularity rather than genuinely dispersed research.