Towards a foundation model for geospatial artificial intelligence (vision paper)

Mai, Gengchen; Cundy, Chris; Choi, Kristy; Hu, Yingjie; Lao, Ni; Ermon, Stefano

doi:10.1145/3557915.3561043

Cited by 28 publications

(9 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(2) To improve model generalizability across space and time, instead of using the FSTL method as Li et al (2022) did, can we directly learn a hypernetwork to simulate how the model's parameters change based on the location and time with meta‐learning method (Tenzer et al, 2022)? (3) Given the increasing popularity of foundation models (FMs) in the natural language and vision communities, such as GPT‐3 (Brown et al, 2020), CLIP (Radford et al, 2021), PaLM (Wei et al, 2022), and DALL∙E2 (Ramesh et al, 2022), could we build a FM for GeoAI which, after pretraining, can be easily adapted to multiple symbolic GeoAI and subsymbolic GeoAI tasks, involving the use of different data modalities (Mai et al, 2022a)? (4) How can we address important ethical aspects, such as better accounting for and mitigating issues of bias, fairness, and transparency (Shin & Basiri, 2022; Zheng & Sieber, 2022), how to reduce the environmental footprint of model training, and how to better connect to communities studying ethics of technology (Goodchild et al, 2022).…”

Section: Conclusion and Next Stepsmentioning

confidence: 99%

Symbolic and subsymbolic GeoAI: Geospatial knowledge graphs and spatially explicit machine learning

Mai

Gao

et al. 2022

Transactions in GIS

View full text Add to dashboard Cite

The field of Artificial Intelligence (AI) can be roughly divided into two branches: Symbolic AI and Connectionist AI (or the so-called Subsymbolic AI). Symbolic AI focuses on research based on classical logic and higher-level symbolic (human-readable) knowledge representations. It posits the use of declarative knowledge in reasoning and learning as critical to producing intelligent behavior (Goel, 2022). Examples are logical inference, symbolic reasoning, ontol-

show abstract

Section: Conclusion and Next Stepsmentioning

confidence: 99%

Symbolic and subsymbolic GeoAI: Geospatial knowledge graphs and spatially explicit machine learning

Mai

Gao

et al. 2022

Transactions in GIS

View full text Add to dashboard Cite

show abstract

“…This success in multimodality is significant for the next generation of GeoAI models that could also be pre-trained with geo-data ranging from location descriptions to remote sensing and street-level images, and from vector data to cartographic maps. However, such models would still suffer from a lack of geographic diversity when learning latent spatial representations in a task-agnostic manner (Mai et al, 2022). Additionally, more and more geo-data could become generated by machines at scale.…”

Section: Introductionmentioning

confidence: 99%

Measuring Geographic Diversity of Foundation Models with a Natural Language–based Geo-guessing Experiment on GPT-4

Liu,

Janowicz,

Currier

et al. 2024

AGILE GIScience Ser.

View full text Add to dashboard Cite

Abstract. Generative AI based on foundation models provides a first glimpse into the world represented by machines trained on vast amounts of multimodal data ingested by these models during training. If we consider the resulting models as knowledge bases in their own right, this may open up new avenues for understanding places through the lens of machines. In this work, we adopt this thinking and select GPT-4, a state-of-the-art representative in the family of multimodal large language models, to study its geographic diversity regarding how well geographic features are represented. Using DBpedia abstracts as a ground-truth corpus for probing, our natural language–based geo-guessing experiment shows that GPT-4 may currently encode insufficient knowledge about several geographic feature types on a global level. On a local level, we observe not only this insufficiency but also inter-regional disparities in GPT-4’s geo-guessing performance on UNESCO World Heritage Sites that carry significance to both local and global populations, and the inter-regional disparities may become smaller as the geographic scale increases. Morever, whether assessing the geo-guessing performance on a global or local level, we find inter-model disparities in GPT-4’s geo-guessing performance when comparing its unimodal and multimodal variants. We hope this work can initiate a discussion on geographic diversity as an ethical principle within the GIScience community in the face of global socio-technical challenges.

show abstract

“…LLMs, such as Bidirectional Encoder Representations from Transformers (BERT) (Devlin et al 2019) and Generative Pre-trained Transformer (GPT) models (Brown et al 2020;Ouyang et al 2022), are pre-trained on large-scale textual data (e.g., all text on the Internet) in a task-agnostic manner, and can be adapted to domainspecific tasks via fine tuning, few-shot learning, or sometimes even zero-shot learning. Given their foundational roles in completing various domain-specific tasks, LLMs and other large-scale pretrained models are also called foundation models (Bommasani et al 2021;Mai et al 2022).…”

Section: Introductionmentioning

confidence: 99%

Geo-knowledge-guided GPT models improve the extraction of location descriptions from disaster-related social media messages

Hu,

Mai,

Cundy

et al. 2023

International Journal of Geographical Information Science

Self Cite

View full text Add to dashboard Cite

Social media messages posted by people during natural disasters often contain important location descriptions, such as the locations of victims. Recent research has shown that many of these location descriptions go beyond simple place names, such as city names and street names, and are difficult to extract using typical named entity recognition (NER) tools. While advanced machine learning models could be trained, they require large labeled training datasets that can be time-consuming and labor-intensive to create. In this work, we propose a method that fuses geoknowledge of location descriptions and a Generative Pre-trained Transformer (GPT) model, such as ChatGPT and GPT-4. The result is a geo-knowledge-guided GPT model that can accurately extract location descriptions from disaster-related social media messages. Also, only 22 training examples encoding geo-knowledge are used in our method. We conduct experiments to compare this method with nine alternative approaches on a dataset of tweets from Hurricane Harvey. Our method demonstrates an over 40% improvement over typically used NER approaches. The experiment results also show that geo-knowledge is indispensable for guiding the behavior of GPT models. The extracted location descriptions can help disaster responders reach victims more quickly and may even save lives.

show abstract

Towards a foundation model for geospatial artificial intelligence (vision paper)

Cited by 28 publications

References 5 publications

Symbolic and subsymbolic GeoAI: Geospatial knowledge graphs and spatially explicit machine learning

Symbolic and subsymbolic GeoAI: Geospatial knowledge graphs and spatially explicit machine learning

Measuring Geographic Diversity of Foundation Models with a Natural Language–based Geo-guessing Experiment on GPT-4

Geo-knowledge-guided GPT models improve the extraction of location descriptions from disaster-related social media messages

Contact Info

Product

Resources

About