World Bank researcher pairing Earth-observation transfer learning and NLP to measure welfare in displacement settings
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
18
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—
The output axis is unscored: recent publication volume is too sparse to support a reliable measure. This axis would track annual papers, preprints, and venue types across the record; only 2 recent three-year works are visible in the current data.
Impact—
Impact is unscored: field-normalized citation data is not yet available for this record. The axis would measure raw and field-normalized citation uptake; h-index (18) and i10-index (23) exist in the raw data but carry no score.
Focus0/100
Focus scores 0, the minimum, reflecting dispersion rather than concentration. The reviewed field distribution splits exactly 50/50 between Earth-observation socioeconomic estimation and NLP information extraction, while OpenAlex topics spread across malaria research (14 works), zoonotic disease and public health (8), mosquito-borne control (7), animal behavior and reproduction (7), and vector-borne infectious disease (7) — five clusters without a dominant thread.
Collaboration41/100
Collaboration scores 41, grounded in a compact network of 7 distinct coauthors in the raw record. The frequent-coauthor list comprises Steven Ndung'u, Adel Daoud, Ismael Yacoubou Djima, Hai-Anh H. Dang, Rafael Macalaba, Aivin V. Solatorio, and Olivier Dupriez. On both reviewed papers Brock is a non-corresponding author — last author on the transfer-learning paper and a middle author on the extraction paper.
Activity32/100
Activity scores 32, consistent with the yearly counts: output peaked at 8 works in 2017 and 7 in 2015, then fell to 1 work each in 2022 and 2023 and 2 in 2026, with recent3yWorks at just 2. The current cadence sits well below the mid-2010s peak, indicating a low, episodic recent publishing rate.
Reproducibility55/100
Reproducibility scores 55, the mean readiness across the 2 reviewed papers (reproMean 55, reproPapers 2), each rated medium. Evolve's verdicts flag the transfer-learning proof of concept for tiny labels, inflated grids, and an unvalidated 27% decline projection, and the dataset-mention framework for missing baselines that leave its gains unproven.
Research summary
Patrick Michael Brock is a researcher at the World Bank whose recent work applies machine learning to development measurement. This profile is name-resolved from public bibliographic data (OpenAlex). His reviewed record splits evenly (0.5/0.5) between Earth-observation methods for socioeconomic estimation and natural language processing for information extraction, both targeting forced-displacement and fragility settings. Two 2026 arXiv papers reviewed by Evolve illustrate the agenda: a transfer-learning study adapting a DHS-pretrained spatiotemporal vision transformer to camp-grid welfare estimation (R2 0.66, with flagged caveats of tiny labels, inflated grids, and an unvalidated 27% decline projection), and a weakly supervised, LLM-refined framework for extracting dataset mentions from forced-displacement and FCV documents. Both carry medium reproducibility ratings (mean 55 across 2 papers), matching his radar score of 55. His longer OpenAlex record is anchored in malaria and vector-borne disease research — the leading topic is 'Malaria Research and Control' (14 works) — with an h-index of 18 and i10-index of 23. Annual output peaked at 7 works (2015) and 8 (2017), thinning to 1 in 2022, 1 in 2023, and 2 in 2026 (recent3yWorks: 2), yielding a low activity score of 32. Collaboration (41) rests on a compact network of 7 distinct coauthors; both catalog papers position him as a non-corresponding, middle-to-last author.
Collaboration pattern
Collaboration is compact: 7 distinct coauthors in the raw record, matching the middling collaboration score of 41. The frequent-coauthor list comprises Steven Ndung'u, Adel Daoud, Ismael Yacoubou Djima, Hai-Anh H. Dang, Rafael Macalaba, Aivin V. Solatorio, and Olivier Dupriez. On both reviewed 2026 papers Brock is a non-corresponding author — last author on the transfer-learning paper, middle author on the extraction paper — fitting a senior-contributor rather than lead-PI pattern.
Research keywords
Highlights
Transfer learning brings DHS-pretrained Earth-observation models to displacement camps
Weakly supervised extraction of dataset mentions from FCV documents
A dual-track methods agenda in one compact team
The reviewed record splits 50/50 between Earth-observation socioeconomic estimation and NLP information extraction, both built with the same seven frequent coauthors and aimed at measuring forced-displacement and fragility settings.
Timeline
- 2008First recorded year in the OpenAlex annual counts, with 1 work.
- 2015Output rises to 7 works, the first of two mid-decade peaks.
- 2017Highest recorded output in the counts: 8 works in a single year.
- 20195 works recorded, after which the record shows a marked slowdown.
- 2023Sparse recent output continues — 1 work in 2022 and 1 in 2023.
- 2026Two arXiv preprints on machine learning for forced-displacement measurement enter the Evolve catalog, both with medium reproducibility readiness.
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 figures come from OpenAlex only (identity: openalex); the h-index (18) and i10-index (23) are quoted as provided, without independent verification.
- The output and impact radar axes are unscored due to insufficient data; no citation normalization or recent-output score is available.
- Evolve has reviewed only 2 papers, both 2026 arXiv preprints; the medium reproducibility ratings cover this small sample.
- The OpenAlex topical history (malaria, vector-borne disease) differs from the reviewed machine-learning work; the profile is name-resolved from public bibliographic data, so homonym risk cannot be fully excluded.