Advancing Large Language Models for Spatiotemporal and Semantic Association Mining of Similar Environmental Events

Anna K. Liljedahl; Fan Zhang; Lei Hu; Michael Brook; Michael Brubaker; Wenwen Li; Xiao Chen; Yuanyuan Tian

REVIEW 4 major objections 7 minor 7 references

Advancing Large Language Models for Spatiotemporal and Semantic Association Mining of Similar Environmental Events

T0 review · 4 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that adding spatial proximity, latitude, and seasonal timing to large-language-model search produces the best tested zero-shot recommendations of similar environmental events.

desk verdict The GT-R pipeline is a sensible integration with a real use case, but the headline gains are not established because the evaluation tunes on the test set and the algorithm text doesn't match the tuned weights. read the letter →

arxiv 2411.12880 v1 pith:BK6AC5UN submitted 2024-11-19 cs.IR cs.AI

Yuanyuan Tian , Wenwen Li , Lei Hu , Xiao Chen , Michael Brook , Michael Brubaker , Fan Zhang , Anna K. Liljedahl This is my paper · ORCID

classification cs.IRcs.AI

keywords Largelanguagemodelsspatiotemporalretrievaleventrecommendationgeo-timere-rankingsemanticsimilarityclimatechangeenvironmentaleventsdense

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper aims to show that large language models can automate the discovery of similar climate and environmental events from text, and that adding geography and time to semantic search does better than semantic search alone. The proposed two-stage framework first retrieves candidate events with an embedding model, then re-ranks them with a Geo-Time Re-ranking (GT-R) model that fuses semantic similarity, LLM-extracted category similarity, distance, latitude, and seasonal timing. On a corpus of about 4,000 Local Environmental Observer (LEO) Network events, the authors report GT-R's nDCG@10 of 47.4 against 41.4 for RankGPT and 44.7 for Ada-002, positioning the method as a scalable alternative to manual curation of related-event links. If the result holds up, event recommendation for climate observation networks can move from expert hand curation to automated, spatiotemporally aware matching.

What carries the argument

The central object is the GT-R (Geo-Time Re-ranking) algorithm, a rank-fusion model that re-scores the top 100 retrieved candidates by combining five ranked lists: semantic similarity from bi-encoded embeddings, category-instructed similarity from zero-shot LLM entity extraction followed by cross-encoding, Haversine great-circle distance, absolute latitude difference, and day-of-year temporal proximity. Each list is converted to ranks, boosted within thresholds of 500 km and 5 degrees of latitude, and merged with reciprocal rank fusion; after grid search, the fusion uses a semantic weight of 0.1 and a category weight of 0.9. The design lets spatial and temporal context survive the transition from retrieval to final ranking.

What would settle it

A held-out evaluation using LEO events whose human-curated related-event links were created after the model's thresholds were fixed would settle the claim: if GT-R's nDCG@10 no longer exceeds RankGPT's 41.4 and Ada-002's 44.7, the reported advantage would be tuning to the test set rather than a general property.

Watch

Extended reading notes

Core claim

The paper's central claim is that its Geo-Time Re-ranking (GT-R) framework is the best tested zero-shot method for recommending similar environmental events from the LEO Network corpus. In the reported experiments GT-R reaches nDCG@10 of 47.4, beating the RankGPT listwise re-ranker's 41.4, the pure dense retrieval of Ada-002's 44.7, and the LEO heuristic's 36.4. The authors attribute the gain to treating event similarity as a conjunction of semantic content, category-level meaning obtained by zero-shot LLM entity extraction, geographic distance, latitudinal climate similarity, and seasonal timing, so that re-ranking does not discard the spatiotemporal signal after retrieval.

Load-bearing premise

The load-bearing premise is that the LEO Network's human-curated "see also" links are a complete, unbiased ground truth for event similarity, so hyperparameters tuned on 1,000 queries drawn from that set will keep working for events that have not been curated yet.

Editorial extensions

If this is right

Structured spatiotemporal input improves retrieval: adding location and date with prefixes lifts Ada-002's Recall@100 from 84.9 to 86.8, so event retrieval systems should feed location and date explicitly rather than relying on text alone.
LLM-based category-aware entity extraction is the heaviest re-ranking input after raw semantics, with an optimal fusion weight of 0.9, so expert category schemas can be injected into reasoning without retraining the LLM.
The ablation study shows every component helps; removing semantic similarity hurts most, dropping nDCG@10 from 47.4 to 39.4, while distance, latitude, and time each contribute smaller but consistent gains.
Because the pipeline is zero-shot and uses only prompting plus frozen models, it can replace manual curation of related-event links in community observation networks without task-specific training data.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural extension is to treat the paper's 500 km and 5-degree thresholds as domain priors rather than universal constants; other event types such as floods, disease outbreaks, or air-quality episodes likely need their own calibration, and the ranking-fusion form makes such recalibration straightforward.
The paper's own example of the 2023 Canadian wildfire smoke query, where retrieval recall was zero, points to a testable improvement: detect ongoing multi-event phenomena before applying day-of-year matching, since cascading events break the seasonal-similarity assumption.
The human-in-the-loop verification of LLM entity extraction is the main remaining automation bottleneck; a fully automatic variant could be evaluated by measuring how often manual correction changes the extracted entities and how much that changes nDCG.
Because category similarity dominates the fusion weights, the method's portability to corpora without a domain category schema is uncertain; testing a semantic-only fallback or a learned weight would clarify how much of the gain depends on expert tags.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

I'll give you the short version. The paper does something real: it takes an LLM dense retriever, a GPT-4 zero-shot NER step that enriches category tags, a cross-encoder for category similarity, and then adds Haversine distance, latitude, and day-of-year features with an RRF fusion. The category-instructed NER is the most interesting piece; it gives the model a way to compare events at species/topic level rather than just matching tags, and it's cheap since it's zero-shot. The LEO corpus of community-reported environmental events is a good testbed, and the authors are honest about failure cases (the wildfire smoke example with recall 0) and limitations.

What worries me is the evaluation. Section 4.3 says the semantic/category fusion weights were found by grid search, the latitude threshold by 'empirical testing' over 3/5/10 degrees, and the distance threshold/booster follow the LEO heuristic. All of this is done on the same 1,000 queries used to compute the Table 3 scores. There is no validation split, no cross-validation, and no uncertainty reported. With at least six tunable decisions, the 14% nDCG gain over RankGPT, 6% over Ada, and 30% over the LEO heuristic could be substantially inflated by test-set fitting. That's load-bearing because the paper's only evidence of superiority is this single table. No code or data are released, so the magnitude of the bias can't be checked.

There is also an internal inconsistency: Algorithm 1 defines tau_s and tau_c, but the fusion formula (line 22) sums 1/(k+r) over the ranking lists and never uses those weights. So the method as described doesn't match the grid search that supposedly found the optimal 0.1/0.9 split. Minor but telling: MRR@k in Equation 6 is written as 1/k sum 1/rank_i over k positions, which is not the standard query-averaged MRR.

None of these are fatal to the idea. The pipeline is plausible, the category enrichment is a genuine contribution, and the application is useful. But the central empirical claim — that GT-R surpasses all baselines and SOTA zero-shot methods — is not supported by the current evidence. A held-out split or nested cross-validation, plus released artifacts, would likely make this a solid paper. The ablation is also affected by the same tuning problem, so it needs to be redone under a clean protocol.

Who should read this: GeoAI and climate informatics researchers who want a practical blueprint for building a similar-event recommender. It's not ready as a validated SOTA claim. That said, I would send it to peer review — a serious referee can push for the missing validation and the paper would benefit. I just wouldn't cite the numbers as they stand.

Referee Report

4 major / 7 minor

Summary. The paper proposes a two-stage retrieval-and-reranking framework for recommending similar environmental events from the LEO Network. In the first stage, an LLM embedding model (Ada-002) retrieves the top-100 candidates per query from a corpus of about 4,000 events, using an input format that prefixes title, summary, location, and date. In the second stage, the proposed Geo-Time Re-ranking (GT-R) model reranks these candidates by fusing five features: semantic similarity, category similarity (obtained via GPT-4 Turbo zero-shot NER followed by cross-encoding), Haversine distance, latitude difference, and day-of-year temporal proximity. The paper reports retrieval results (Recall@100, Hit Rate@100) and reranking results (Hit Rate@1/3, nDCG@10, MRR@10) on 1,000 query events, and claims that GT-R outperforms several baselines, including RankGPT, Ada-002, and the LEO heuristic.

Significance. If the reported gains were reliable, the paper would make a useful contribution to spatiotemporal event recommendation: the task is real, the combination of semantic, categorical, and geotemporal signals is sensible, and the ablation study in Table 4 gives some insight into feature contributions. The use of LLM-based NER to enrich category features is an interesting idea. However, the central empirical claim is currently not supported by the evaluation protocol: hyperparameters are tuned on the same corpus used for the reported metrics, no held-out validation or cross-validation is described, no variance or confidence intervals are reported, and the algorithm as written does not implement the tuned fusion weights. The paper also does not release data or code, making independent verification impossible. The strengths of the paper are conceptual; the empirical evidence needs substantial rework.

major comments (4)

[Section 4.3 and Tables 3-4] The free parameters (semantic weight 0.1, category weight 0.9, distance threshold 500 km, distance booster 2, latitude threshold 5 degrees, latitude booster 2) are selected by grid search and empirical testing on the same 1,000-query corpus that is then used to produce Tables 3 and 4. No held-out split, cross-validation, or repeated trials are described, and no variance estimates are reported. Since the headline claim in Section 5.1.3 rests on these single-run numbers, the reported 14% nDCG gain over RankGPT, 6% over Ada-002, and 30% over the LEO heuristic are in-sample fit outcomes whose out-of-sample magnitude is unknown.
[Algorithm 1, lines 8, 11, 16, 22] Algorithm 1 is internally inconsistent with the method described in Section 3.3 and with the tuning performed in Section 4.3: lines 8 and 11 use beta_s and beta_c, which are not defined in the algorithm's input list; line 22 fuses the five rank lists by unweighted reciprocal rank fusion, so the grid-searched weights tau_s and tau_c never enter the score; and line 16 initializes the latitude ranking from the semantic ranking rather than from latitude differences. As written, the algorithm does not implement the weighted semantic/category fusion that Section 4.3 is supposed to optimize, and it is unclear which scoring function actually produced Tables 3-5.
[Section 3.2.2 and Figure 2] The category-similarity feature includes a 'Human in the Loop' step in which manual oversight refines and verifies the LLM's NER outputs. The paper does not report how much human effort was involved, whether all 1,000 query-candidate pairs were manually checked, or whether human corrections influenced the final reranking. If human intervention affected the results, the comparison with fully automated zero-shot baselines (RankGPT, cross-encoders, Ada) is not apples-to-apples, and the scalability claim in the Abstract is not supported.
[Section 4.1] The evaluation uses the LEO Network's human-curated 'see also' links as binary relevance labels, but the paper gives no statistics on the number of relevant events per query, the distribution of labels across event categories, or any assessment of label completeness. If these links are sparse or curator-dependent, the Recall@100, nDCG, and MRR numbers are systematically biased. The authors should report label statistics, consider excluding queries with very few relevant links, and, where possible, provide a manual relevance judgment set as a robustness check.

minor comments (7)

[Section 5.1.3 vs Section 7] The reported gain over the LEO heuristic is 30% in Section 5.1.3 (47.4 vs. 36.4), but the Conclusion says 'up to 40% gains over the heuristic solution'; these numbers should be reconciled.
[Tables 1 and 2] Table 1 reports BM25 Recall@100 of 76.2 for input 'Title, Summary, Location, Date', while Table 2 reports 74.7 for the same input setting (with prefix); the discrepancy should be explained.
[Tables 3 and 4] Table 4 reports Hit Rate@1 of 42.2 for the full GT-R model, whereas Table 3 reports 42.4 for the same configuration; please clarify why the numbers differ.
[Equation (6)] Equation (6) defines MRR@k as the mean over k queries, but the paper later reports MRR@10; if k is the truncation rank rather than the number of queries, the formula should be MRR@10 = (1/Q) * sum(1/rank_i) with reciprocal ranks capped at 10.
[Figure 4(b) and Section 5.2] The Figure 4(b) caption describes the query as a landslide event in Finland, while the text of Section 5.2 describes it as an event in Iceland; fix the mismatch.
[Section 3.2.3(2)] The initialization of the latitude ranking from the semantic ranking in Algorithm 1 is not explained in the prose; please clarify what the intended initialization is.
[Section 4.3] The paper does not include the exact prompt template or JSON schema used for the GPT-4 Turbo NER step; including these would improve reproducibility.

Circularity Check

1 steps flagged · score 6.0 of 10

GT-R's reported nDCG gains are partly fit outcomes: §4.3 grid-searches fusion weights and thresholds on the same 1,000-query evaluation set, so Table 3 is a selected maximum rather than an out-of-sample prediction.

fitted input called prediction [Section 4.3 (Implementation) and Section 5.1.3 / Table 3]
"Through empirical testing with thresholds of 3, 5, and 10 degrees, an absolute latitudinal threshold of 5 degrees was set, with a corresponding latitude booster factor of 2. The determination of fusion weights for semantic similarity and category similarity was conducted through grid search. The search was configured to maintain the sum of weights at unity, with increments of 0.1 within the range from 0 to 1. The optimal weight distribution was identified with a semantic similarity weight of 0.1 and a category similarity weight of 0.9, thereby achieving the best balance for the fusion."

Section 4.1 defines the evaluation corpus as '1,000 query events, each linked to related events, which were human annotated,' and no validation or test split is described anywhere in the paper. Section 4.3 then selects the latitude threshold 'through empirical testing' and the semantic/category fusion weights by 'grid search' with 'increments of 0.1,' identifying the 'optimal weight distribution.' Table 3 subsequently reports GT-R's nDCG@10 of 47.4 and the claims of 14% over RankGPT, 6% over Ada-002, and 30% over the LEO heuristic.

full rationale

The framework itself is not definitionally circular: the semantic embeddings, cross-encoder category similarity, Haversine distance, latitude and temporal features, and RRF fusion are standard components, and the comparison baselines (BM25, Ada-002, RankGPT, Cohere rerank, BGE reranker, etc.) are external. No uniqueness theorem or load-bearing self-citation forces the result; the W. Li et al. (2024) citation is motivational for latitude relevance, not an argument that the method is the only possible approach. The circular component is confined to the empirical claim: Section 4.3 tunes the fusion weights and latitude threshold by grid search and empirical testing, while Section 4.1 describes the same 1,000-query dataset as the evaluation corpus. With no validation split described, the nDCG=47.4 and the reported margins over RankGPT and Ada-002 in Table 3 are partly selected maxima on the evaluation set. That makes the headline superiority claim partially a fit outcome rather than an independent prediction. Algorithm 1's rank-fusion formula also fails to use the tuned τ_s/τ_c weights, which further weakens the link between the stated tuning procedure and the reported result.

Assumptions & free parameters 6 free parameters · 3 assumptions · 0 invented entities

No new physical entities are postulated. The framework's extra degrees of freedom are its reranking hyperparameters, all tuned on the same dataset used for evaluation, which weakens the empirical claims.

free parameters (6)

semantic similarity weight tau_s = 0.1
Set by grid search in Section 4.3 on the same LEO evaluation corpus; it controls the contribution of raw semantic rank in the RRF fusion.
category similarity weight tau_c = 0.9
Set by the same grid search; the optimal combination tau_s=0.1 and tau_c=0.9 was chosen after evaluating on the test queries, so reported numbers partly reflect this fit.
distance threshold tau_d = 500 km
Chosen following LEO Network's heuristic and empirical testing, not from a held-out validation.
distance boosting factor beta_d = 2
Adjusts distance ranking for nearby events; set with the threshold, not cross-validated.
latitude threshold tau_phi = 5 degrees
Selected by testing 3, 5, and 10 degrees on the evaluation corpus, as described in Section 4.3.
latitude boosting factor beta_phi = 2
Set alongside the threshold, based on trial on the same data.

assumptions (3)

domain assumption Tobler's first law: near things are more related than distant things.
Motivates distance relevance in Section 3.2.3(1); used to justify distance-based rank promotion.
domain assumption Latitudinal bands correspond to ecologically and climatically similar conditions.
Used to justify the latitude booster in Section 3.2.3(2).
domain assumption Day-of-year is a meaningful temporal metric for environmental events.
Used for temporal relevance in Section 3.2.3(3); assumes seasonal, not chronological, similarity is what matters.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Advancing Large Language Models for Spatiotemporal and Semantic Association Mining of Similar Environmental Events." pith.science (2026). https://pith.science/paper/BK6AC5UN

@misc{pith2026241112880,
  author       = {Pith},
  title        = {Pith review of: Advancing Large Language Models for Spatiotemporal and Semantic Association Mining of Similar Environmental Events},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BK6AC5UN}},
  note         = {Machine review of arXiv:2411.12880}
}

read the original abstract

Retrieval and recommendation are two essential tasks in modern search tools. This paper introduces a novel retrieval-reranking framework leveraging Large Language Models (LLMs) to enhance the spatiotemporal and semantic associated mining and recommendation of relevant unusual climate and environmental events described in news articles and web posts. This framework uses advanced natural language processing techniques to address the limitations of traditional manual curation methods in terms of high labor cost and lack of scalability. Specifically, we explore an optimized solution to employ cutting-edge embedding models for semantically analyzing spatiotemporal events (news) and propose a Geo-Time Re-ranking (GT-R) strategy that integrates multi-faceted criteria including spatial proximity, temporal association, semantic similarity, and category-instructed similarity to rank and identify similar spatiotemporal events. We apply the proposed framework to a dataset of four thousand Local Environmental Observer (LEO) Network events, achieving top performance in recommending similar events among multiple cutting-edge dense retrieval models. The search and recommendation pipeline can be applied to a wide range of similar data search tasks dealing with geospatial and temporal data. We hope that by linking relevant events, we can better aid the general public to gain an enhanced understanding of climate change and its impact on different communities.

Figures

Figures reproduced from arXiv: 2411.12880 by the authors.

**Figure 1.** The proposed framework for mining relevant spatiotemporal news events. 3.2 Feature construction We detail the methodologies for calculating semantic similarity, category-instructed similarity, spatiotemporal relevance, and the subsequent fusion of these features to construct the final ranking of related events. Algorithm 1 describes the formulation of the GT-R model and its annotations. Algorithm 1. GT-R: Geo-Time R… view at source ↗

**Figure 2.** An illustration of calculating category similarity in the GT-R model. The example conducts category-instructed entity extraction of a query event and a candidate event respectively using an LLM with human-in-the-loop, followed by cross-encoding for similarity calculation. To build the category similarity feature, for each pair of a query event 𝑞 and a candidate event 𝑧 from 𝑍retrieve, we initiate the process with en… view at source ↗

**Figure 3.** LEO environmental event dataset: (a) an event example with content structure explained; (b) visualization of high-frequency words in event content; (c) event category. (a) and (c) are screenshots of LEO web-based portal, (b) is generated using event title [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Visual representation of spatiotemporal event recommendations by GT-R Model. (a) a query event of spotting a dead whale in Alaska (red star) with its top ten similar events (brown crosses); (b) a query event regarding landslides in Finland (red start) with its top ten …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

7 extracted references · 2 canonical work pages

[1]

A., & Crow, D

Albright, E. A., & Crow, D. (2019). Beliefs about climate change in the aftermath of extreme flooding. Climatic Change, 155(1), 1–17. Arora, D., Kini, A., Chowdhury, S. R., Natarajan, N., Sinha, G., & Sharma, A. (2023). GAR- meets-RAG Paradigm for Zero-Shot Information Retrieval. 1–18. http://arxiv.org/abs/2310.20158 Askari, A., Aliannejadi, M., Kanoulas,...

arXiv 2019
[13]

Qin, Z., Jagerman, R., Hui, K., Zhuang, H., Wu, J., Shen, J., Liu, T., Liu, J., Metzler, D., Wang, X., & Bendersky, M. (2023). Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting. 1–12. http://arxiv.org/abs/2306.17563 Ramachandran, R., Bugbee, K., Tilmes, C., & Privette, A. P. (2016). Climate data initiative: A geocuration eff...

arXiv 2023
[37]

https://doi.org/10.1080/17538947.2012.674561 Li, W., & Hsu, C.-Y. (2022). GeoAI for large-scale image analysis and machine vision: Recent progress of artificial intelligence in geography. ISPRS International Journal of Geo- Information, 11(7),

work page arXiv 2022
[128]

A., & Vavrus, S

Francis, J. A., & Vavrus, S. J. (2012). Evidence linking Arctic amplification to extreme weather in mid-latitudes. Geophysical Research Letters, 39(6). Gao, L., Ma, X., Lin, J., & Callan, J. (2023). Precise Zero-Shot Dense Retrieval without Relevance Labels. Proceedings of the Annual Meeting of the Association for Computational Linguistics, 1, 1762–1777. ...

arXiv 2012
[385]

Li, W., Hsu, C.-Y., Wang, S., & Kedron, P. (2024). GeoAI Reproducibility and Replicability: a computational and spatial perspective. ArXiv Preprint ArXiv:2404.10108. Li, W., Hsu, C. Y., & Hu, M. (2021). Tobler’s First Law in GeoAI: A Spatially Explicit Deep Learning Model for Terrain Feature Detection under Weak Supervision. Annals of the American Associa...

work page Pith review arXiv 2024
[480]

post-truth

VanDerWal, J., Murphy, H. T., Kutt, A. S., Perkins, G. C., Bateman, B. L., Perry, J. J., & Reside, A. E. (2013). Focus on poleward shifts in species’ distribution underestimates the fingerprint of climate change. Nature Climate Change, 3(3), 239–243. Voyage AI. (2024). voyage-2. https://docs.voyageai.com/docs/embeddings Wang, L., Yang, N., Huang, X., Yang...

arXiv 2013
[496]

Li, W., Yang, C., Nebert, D., Raskin, R., Houser, P., Wu, H., & Li, Z. (2011). Semantic-based web service discovery and chaining for building an Arctic spatial data infrastructure. Computers & Geosciences, 37(11), 1752–1762. Li, X. L., & Liang, P. (2021). Prefix-tuning: Optimizing continuous prompts for generation. ArXiv Preprint ArXiv:2101.00190. Liang, ...

arXiv 2011

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

A., & Crow, D

Albright, E. A., & Crow, D. (2019). Beliefs about climate change in the aftermath of extreme flooding. Climatic Change, 155(1), 1–17. Arora, D., Kini, A., Chowdhury, S. R., Natarajan, N., Sinha, G., & Sharma, A. (2023). GAR- meets-RAG Paradigm for Zero-Shot Information Retrieval. 1–18. http://arxiv.org/abs/2310.20158 Askari, A., Aliannejadi, M., Kanoulas,...

arXiv 2019

[2] [13]

Qin, Z., Jagerman, R., Hui, K., Zhuang, H., Wu, J., Shen, J., Liu, T., Liu, J., Metzler, D., Wang, X., & Bendersky, M. (2023). Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting. 1–12. http://arxiv.org/abs/2306.17563 Ramachandran, R., Bugbee, K., Tilmes, C., & Privette, A. P. (2016). Climate data initiative: A geocuration eff...

arXiv 2023

[3] [37]

https://doi.org/10.1080/17538947.2012.674561 Li, W., & Hsu, C.-Y. (2022). GeoAI for large-scale image analysis and machine vision: Recent progress of artificial intelligence in geography. ISPRS International Journal of Geo- Information, 11(7),

work page arXiv 2022

[4] [128]

A., & Vavrus, S

Francis, J. A., & Vavrus, S. J. (2012). Evidence linking Arctic amplification to extreme weather in mid-latitudes. Geophysical Research Letters, 39(6). Gao, L., Ma, X., Lin, J., & Callan, J. (2023). Precise Zero-Shot Dense Retrieval without Relevance Labels. Proceedings of the Annual Meeting of the Association for Computational Linguistics, 1, 1762–1777. ...

arXiv 2012

[5] [385]

Li, W., Hsu, C.-Y., Wang, S., & Kedron, P. (2024). GeoAI Reproducibility and Replicability: a computational and spatial perspective. ArXiv Preprint ArXiv:2404.10108. Li, W., Hsu, C. Y., & Hu, M. (2021). Tobler’s First Law in GeoAI: A Spatially Explicit Deep Learning Model for Terrain Feature Detection under Weak Supervision. Annals of the American Associa...

work page Pith review arXiv 2024

[6] [480]

post-truth

VanDerWal, J., Murphy, H. T., Kutt, A. S., Perkins, G. C., Bateman, B. L., Perry, J. J., & Reside, A. E. (2013). Focus on poleward shifts in species’ distribution underestimates the fingerprint of climate change. Nature Climate Change, 3(3), 239–243. Voyage AI. (2024). voyage-2. https://docs.voyageai.com/docs/embeddings Wang, L., Yang, N., Huang, X., Yang...

arXiv 2013

[7] [496]

Li, W., Yang, C., Nebert, D., Raskin, R., Houser, P., Wu, H., & Li, Z. (2011). Semantic-based web service discovery and chaining for building an Arctic spatial data infrastructure. Computers & Geosciences, 37(11), 1752–1762. Li, X. L., & Liang, P. (2021). Prefix-tuning: Optimizing continuous prompts for generation. ArXiv Preprint ArXiv:2101.00190. Liang, ...

arXiv 2011