Recording Plan: 

Curated Track: 

, 

Call For Participation

Description:

Artificial intelligence has unlocked new possibilities for understanding complex geographic and socioeconomic issues, but many AI methods lack transparency, interpretability, and ethical grounding. Explainable and Ethical AI (XAI) in spatial analysis addresses these challenges, creating AI models that are both understandable and responsible. This session will explore how XAI can enhance transparency, fairness, and accountability in spatial models, with a focus on applications such as crime prediction, socioeconomic disparity analysis, and urban dynamics.

We will discuss both theoretical and practical advancements in XAI, particularly in addressing essential questions: How can we align AI models with core principles of spatial analysis, like spatial dependence and scale? In high-stakes contexts, how can XAI make predictions clearer and more ethically sound? And how can XAI reduce bias and support fairness, particularly in data-scarce regions and vulnerable communities?

We’re seeking presentations that showcase innovative approaches to integrating explainability and ethics in spatial analysis, whether through novel methodologies or real-world case studies. Topics could include, but aren’t limited to:

- Transparent modeling of spatial and temporal patterns
- Ensuring fairness and reducing bias in spatial predictions
- Examining socioeconomic and crime patterns through ethical lenses
- Interpretable models for urban mobility and community dynamics
- Explainable spatial networks and neighborhood effects
- Enhancing model reliability and fairness in low-data regions
- Ethical AI practices and privacy considerations in spatial applications
- Integrating XAI with traditional spatial analysis for greater transparency

Our aim is to foster a collaborative discussion on making AI in geospatial research both effective and ethically accountable. We welcome research from GIScience, spatial statistics, and AI, particularly those focused on explainable and responsible approaches to complex spatial challenges.

Submission Instructions:
To participate, submit your abstract through the AAG annual meeting website, and email your presenter identification number (PIN) along with your paper abstract to Daniel Kiv (dkiv2@illinois.edu) by Dec 5, 2024. Feel free to reach out with any questions—we look forward to your contributions!

Non-Presenting Participants