Fragile Earth:

Data Science for a Sustainable Planet

August 24th, 2020

Fragile Earth at KDD 2020

The Fragile Earth workshop brought together research, industry, and policy communities at KDD Virtual Conference 2020 to discuss, address and help develop the technological foundations for advancing and meeting the SDGs.

The applications and agenda of interest include food security, sustainable agricultural practices, crop yield forecasting and improvement, restoring degraded landscapes to productive landscapes, clean water management, sustainable and clean energy production, energy efficient and low waste food supply chain, and the future of intelligent technologies in tackling these topics in an ever urbanizing world.

The mission of KDD is to promote the rapid maturation of the field of knowledge discovery in data and data-mining.

Research Papers from KDD 2020

Play Video

Machine Learning for Robust Identification of Complex Nonlinear Dynamical Systems: Applications to Earth Systems Modeling

By Nishant Yadav, Sai Ravela, and Auroop R. Ganguly

Play Video

High-Resolution Air Quality Prediction Using Low-Cost Sensors

By Thibaut Cassard, Grégoire Jauvion and David Lissmyr

Play Video

Online Learning Algorithm for Hurricane Intensity Prediction

By Ding Wang, Boyang Liu and Pang-Ning Tan

Play Video

People-Centered Climate Hazard Impact Assessment using Machine Learning: A Drought Risk Perspective

By Markus Enenkel and Molly Brown

Play Video

The Promise of Causal Reasoning in Reliable Decision Support for Wind Turbines

By Joyjit Chatterjee and Nina Dethlefs

Play Video

Mapping New Informal Settlements using Machine Learning and Time Series Satellite Images: An Application in the Venezuelan Migration Crisis

By Isabelle Tingzon, Niccolo Dejito, Ren Avell Flores, Rodolfo De Guzman, Liliana Carvajal, Katerine Zapata Erazo, Ivan Enrique Contreras Cala, Jeffrey Villaveces, Daniela Rubio and Rayid Ghani

Play Video

Trust and Transparency in Contact Tracing Applications

By Stacy Hobson, Michael Hind, Aleksandra Mojsilovic and Kush Varshney

Play Video

Optimizing crop cut collection for determining field-scale yields in an insurance context

By Ritvik Sahajpal, Inbal Becker-Reshef and Sylvain Coutu

Play Video

Resilient In-Season Crop Type Classification in Multispectral Satellite Observations using Growth Stage Normalization

By Hannah Kerner, Ritvik Sahajpal, Sergii Skakun, Inbal Becker-Reshef, Brian Barker, Mehdi Hosseini, Estefania Puricelli and Patrick Gray

Play Video

Towards a Global Species Dataset by Fusing Remote Sensing and Citizen Science Data with Graph Neural Networks

By Kenza Amara, David Dao and Bjoern Luetjens

Play Video

Leveraging traditional ecological knowledge in ecosystem restoration projects utilizing machine learning

By Bogdana Rakova and Alexander Winter

Areas

Two key technological challenges posed by the Sustainable Development Goals are (a) how to achieve accurate, robust and scalable modeling on physical, environmental, system and societal data, and (b) how to ensure that the obtained models are socially acceptable for use in the associated policy and decision making support.

A key technological enabler for the former is theory-guided data science and scientific discovery, which by augmenting data driven modeling with domain physics and constraints, realizes both accuracy and flexibility in modeling. For the latter, leveraging the emerging techniques of trustworthy machine learning and artificial intelligence to attain the interpretability, accountability, fairness and privacy required for social adoption would be key, along with explicit consideration and inclusion of the viewpoints of policy makers.

Applications of interest include but are not limited to:

  1. Food security, sustainable agricultural practices, crop yield forecasting and improvement, energy efficient and low waste food supply chain.
  2. Fostering degraded landscapes to productive landscapes, clean water management, sustainable and clean energy production.
  3. The future of intelligent technologies in tackling these topics in an ever urbanizing world.

Methodological contributions of interest include but are not limited to:

  1. The integration of physics into data driven environmental modeling and use of advanced machine learning techniques to enhance or speed up physical models.
  2. Addressing interpretability of theory guided and data driven models of environment, e.g. by incorporation of physics into causal explanation of models.
  3. Privacy aware schemes for data sharing in agriculture and food systems and addressing fairness of benefit and credit assignment in data sharing for sustainability.

Additionally, we welcome cross domain and policy and paradigmatic topics:

  1. Paradigms for enhancing scientific discovery through theory guided data science.
  2. Data-informed Food/Energy/Water/Earth Sciences policy discussions.
  3. Frameworks for helping the scientific and KDD communities to work together.

Accepted papers will be allocated to three themed sessions and a poster session. We expect presentations to last 15-20 minutes (including questions), and will prioritize submissions based on relevance, scientific rigour, and potential for societal impact.

Workshop papers will not be published as a part of the SIGKDD Conference proceedings.

The EXTENDED submission deadline is June 20th, 2020!

News & Updates

  • Best paper awards and registration fee waivers for early career authors will be provided. This is made possible through generous funding from Cargill, Inc.
  • Please check back here regularly for updates on travel fund availability/applications for PhD students or those with special circumstances. Please reach out to with any questions.

Organizers

KDD 2020

Naoki Abe

Distinguished Research Staff Member, IBM Research AI

KDD 2020

Kathleen Buckingham

Research Manager, World Resources Institute

KDD 2020

Bistra Dilkina

Associate Professor, CS, USC; Associate Director, USC Center for AI in Society (CAIS)

KDD 2020

Emre Eftelioglu

Applied Scientist at Amazon US

Auroop Ganguly

Auroop R. Ganguly

Professor at Northeastern University in Boston, Director of the Sustainability and Data Sciences Laboratory (SDS Lab). A co-founder and the chief scientific adviser of risQ Inc.

KDD 2020

James Hodson

CEO, AI for Good Foundation, Chief Science Officer, Cognism, Inc. Researcher, Jozef Stefan Institute, Artificial Intelligence Lab

KDD 2020

Ramakrishnan Kannan

CEO, AI for Good Foundation, Chief Science Officer, Cognism, Inc. Researcher, Jozef Stefan Institute, Artificial Intelligence Lab