Bin Yang

East China Normal University

Unclaimed · AI-structured profileOpenAlex

East China Normal University researcher in spatio-temporal ML — trajectory, path, and urban region representation.

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—

Output is unscored in this snapshot: no verified publication count or venue list is available. The axis would measure total scholarly production — number of papers, venues, and authorship roles — once a name-resolved publication record exists.

Impact—

Impact is unscored: citation counts, h-index, and downstream usage are absent from this snapshot. The axis would measure how far this work propagates — citations, adoption by later methods, and venue visibility — once citation data is linked.

Focus0/100

The focus score is 0 because the reviewed output splits evenly — 0.33 each — across three labeled fields: machine learning for intelligent transportation systems, machine learning / urban computing, and machine learning for spatio-temporal data mining. In practice these labels are thematically adjacent: all three catalog papers apply self-supervised representation learning to trajectory, path, or urban-region data, so the low score reflects taxonomy splitting more than scattered research interests.

Collaboration50/100

The collaboration score is 50, computed from 11 distinct coauthors. All three cataloged papers place this researcher as a middle author without corresponding role, consistent with a specialist contributor inside larger group-led projects. The frequent-coauthor list spans 11 names, including Sean Bin Yang, Zongyi Xu, Jilin Hu, Hua Lu, Christian S. Jensen, Ying Sun, Kristian Torp, Hao Miao, Xiangmeng Wang, Tung Kieu, and Torben Bach Pedersen.

Activity—

Activity is unscored: the snapshot lacks year-by-year publication timing, so recent productivity cannot be judged. The axis would measure publishing cadence and sustained engagement across recent years once a complete, dated publication record is resolved.

Reproducibility70/100

Reproducibility scores 70 as the mean readiness across 3 reviewed papers. DGCPath rates high, while the CURE and REFINE extended versions rate medium. Recurring strengths are broad benchmark coverage — 3 cities, 3 trajectory datasets, and 4 datasets respectively — and self-supervised setups that ease replication; recurring gaps are missing statistical significance tests and, for CURE, a deconfounding step that lacks causal rigor.

Research summary

Bin Yang (East China Normal University) works on machine learning for spatio-temporal and urban data, spanning three closely related lines captured in this profile: trajectory representation learning, path representation learning, and urban region representation learning. All three cataloged papers are 2026 arXiv entries in which Bin Yang appears as a middle author without corresponding role, indicating a contributor position within larger research teams. REFINE (a KDD'26 extended version) unifies contrastive and generative trajectory representation learning via closed-loop transcription with wins across 4 datasets; DGCPath combines diffusion-generated views with Jensen–Shannon distributional contrastive objectives, beating 10 baselines on 3 trajectory datasets; CURE adds heuristic shared-component residualization to multi-view urban embeddings, reporting strong gains on 3 cities. Evolve's review scores reproducibility at a mean of 70 across the 3 papers, with DGCPath rated high and the other two medium; common weaknesses are thin statistical significance testing and, in CURE's case, deconfounding without causal rigor. The collaboration score is 50 with 11 distinct coauthors; the frequent-coauthor list includes Sean Bin Yang, Zongyi Xu, Jilin Hu, Hua Lu, Christian S. Jensen, Ying Sun, Kristian Torp, Hao Miao, Xiangmeng Wang, Tung Kieu, and Torben Bach Pedersen. Output, impact, and activity are unscored because no citation or publication-count data is available in this snapshot.

Collaboration pattern

With 11 distinct coauthors, Bin Yang works inside a dense, stable team: the frequent-coauthor list includes Sean Bin Yang, Zongyi Xu, Jilin Hu, Hua Lu, Christian S. Jensen, Ying Sun, Kristian Torp, Hao Miao, Xiangmeng Wang, Tung Kieu, and Torben Bach Pedersen. All three cataloged papers list Bin Yang as a middle author without corresponding role, suggesting a specialist contributor within group-led projects rather than a solo line of work.

Research keywords

spatio-temporal data miningtrajectory representation learningpath representation learningurban region representation learningself-supervised learningcontrastive learningdiffusion modelsmulti-view learningconfounder-aware learninggraph neural networksvariational inferenceintelligent transportation systems

Highlights

REFINE: closed-loop transcription unifies contrastive and generative trajectory learning
The KDD'26 extended version integrates contrastive and masked-reconstruction signals in one closed-loop framework, with broad wins across 4 datasets; Evolve notes the theory is conditional and significance testing is thin.
DGCPath: diffusion-generated views with distributional contrast
Generates views with a diffusion model and applies Jensen–Shannon distributional contrastive learning, beating 10 baselines across 3 trajectory datasets; reproducibility is rated high, though novelty is judged thin.
CURE: confounder-aware multi-view urban region embeddings
Adds heuristic shared-component residualization to multi-view urban region representations and reports strong gains on 3 cities; Evolve flags that the deconfounding step lacks causal rigor and statistical support.

Timeline

2026REFINE extended version (KDD'26): unifies contrastive and generative trajectory representation learning via closed-loop transcription, with broad wins across 4 datasets.
2026DGCPath extended version: diffusion-generated views with Jensen–Shannon distributional contrastive learning, outperforming 10 baselines on 3 trajectory datasets.
2026CURE: confounder-aware multi-view urban region representation learning, reporting strong gains across 3 cities.

Papers in the Evolve catalog

Frequent co-authors

Sean Bin Yang×3 Zongyi Xu×3 Jilin Hu×3 Hua Lu×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.

All bibliometrics and the collaboration network are derived from public bibliographic data only; no ORCID is provided, and 'Bin Yang' is a common name, so identity resolution carries some uncertainty.
Output, impact, and activity axes are unscored because this snapshot contains no citation counts, publication counts, or year-by-year publication record.
The Evolve catalog covers only 3 reviewed papers (all 2026 arXiv entries); the overall publication record and peer-reviewed venue status are outside this snapshot.
The standard Chinese characters of the researcher's name are unverified; Chinese text keeps the romanized name.