GeoAI and Deep Learning Symposium: Generative AI - Opportunities and Challenges for GIScience Research
Type: Paper
Recording Plan:
Theme: Making Spaces of Possibility
Curated Track:
Sponsor Group(s):
Geographic Information Science and Systems Specialty Group, Spatial Analysis and Modeling Specialty Group
Organizer(s):
Junghwan Kim Virginia Polytechnic Institute & State University
Kee Moon Jang Massachusetts Institute of Technology
Jinhyung Lee Western University
Chair(s):
Junghwan Kim, Virginia Polytechnic Institute & State University
,
Call For Participation
Since ChatGPT’s public release in November 2022, the rapid and unprecedented advance in generative AI has sparked discussions among geographers about opportunities and challenges of Generative AI for GIScience and GeoAI research. Potential opportunities include (but are not limited to) synthetic data generation (e.g., simulating human daily activity-travel patterns), information retrieval and extraction, urban visual analytics (e.g., estimating how humans perceive the built environment, such as walkability, perceived safety, and bike-ability), geotagged social media analysis (e.g., conducting thematic analysis or sentiment analysis), urban planning and design solutions suggestion, automatic map creation and assistance, and AI-based learning assistants. Despite these opportunities, critical limitations and ethical issues should be considered when integrating generative AI with GeoAI and GIScience research. These challenges include (but are not limited to) inaccuracies, hallucination effects, biases - political, geographic, and sociodemographic biases, ethical and responsible uses, and challenges in traditional GIScience education.
This in-person paper session explores the potential strengths and weaknesses of integrating generative AI into GeoAI and GIScience research and education. If you are interested in presenting your paper to the session, please email your abstract information to Junghwan Kim at junghwankim@vt.edu by November 30.
Description:
Since ChatGPT’s public release in November 2022, the rapid and unprecedented advance in generative AI has sparked discussions among geographers about opportunities and challenges of Generative AI for GIScience and GeoAI research. Potential opportunities include (but are not limited to) synthetic data generation (e.g., simulating human daily activity-travel patterns), information retrieval and extraction, urban visual analytics (e.g., estimating how humans perceive the built environment, such as walkability, perceived safety, and bike-ability), geotagged social media analysis (e.g., conducting thematic analysis or sentiment analysis), urban planning and design solutions suggestion, automatic map creation and assistance, and AI-based learning assistants. Despite these opportunities, critical limitations and ethical issues should be considered when integrating generative AI with GeoAI and GIScience research. These challenges include (but are not limited to) inaccuracies, hallucination effects, biases - political, geographic, and sociodemographic biases, ethical and responsible uses, and challenges in traditional GIScience education.
This in-person paper session explores the potential strengths and weaknesses of integrating generative AI into GeoAI and GIScience research and education. If you are interested in presenting your paper to the session, please email your abstract information to Junghwan Kim at junghwankim@vt.edu by November 30.
Presentations (if applicable) and Session Agenda:
Song GUO, Tsinghua University |
Decoding the Visual Pattern Language of Urban Scenes: A Cross-Modal Approach Using Deep Learning and CLIP |
Junghwan Kim, Virginia Polytechnic Institute & State University |
Multimodal Large Language Models (LLMs) as Built Environment Auditing Tools |
Kee Moon Jang, Massachusetts Institute of Technology |
Place identity: a generative AI’s perspective |
Zachary Sherman, Virginia Polytechnic Institute & State University |
Fine-Tuning Generative AI for Geospatial Healthcare Solutions through Python Query Execution |
Shashank Karki, Virginia Polytechnic Institute & State University |
An AI-Driven Framework for Enhancing Urban Walkability using Image Analysis, Augmentation, and Community-Centric Improvements |
Non-Presenting Participants
| Role | Participant |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
GeoAI and Deep Learning Symposium: Generative AI - Opportunities and Challenges for GIScience Research
Description
Type: Paper
Contact the Primary Organizer
Junghwan Kim Virginia Polytechnic Institute & State University
junghwankim@vt.edu
Session sponsored by: