Toward Responsible Data Stewardship

Erik Johnston

Erik Johnston

Abstract: This talks looks at the use of data in modeling and is grounded in the lessons learned from a series of major, collaboratively developed research projects at the Decision Theater. An emphasis is placed on the value of collaborative modeling, translating evidence to practice, and the ethical obligations of data stewards and researchers.

Biography: Associate Professor Erik Johnston, PhD. Dr. Johnston is an Associate Professor in the School for the Future of Innovation in Society and the Director of Policy Informatics at the Decision Theater at Arizona State University. Johnston’s research focuses on open governance and policy informatics, the study of how computational and communication technology is leveraged to specifically understand and and realize innovations in communities, governance processes, and information interventions. At its simplest, his work tries to reduce the gaps between knowledge creation and use. Johnston earned a Ph.D. in Information and Complex System Certificate from the University of Michigan where he was a two-time NSF IGERT fellow. He holds an M.B.A. and an M.S. in Information Technology as well as a B.S. in Psychology and Computer Science from the University of Denver. His work has been supported by the National Science Foundation, MacArthur Foundation, Robert Wood Johnson Foundation, Helios Foundation, Arizona Board of Regents, American Academy of Diplomacy, US Army, and Virginia G. Piper Trust.

AI for biodiversity data collation and reuse

Samantha Cheng

Samantha Cheng

Title: Finding the needle in the evidence haystack – AI for biodiversity data collation and reuse

Abstract: Scientific research is growing at exponential rates, generating potentially useful datasets at a faster pace than humans can practically find, understand, and use it to make decisions about natural ecosystems. While there has been significant progress in systematically and transparent harnessing this data and synthesizing key insights around biodiversity patterns, threats, and mechanisms driving changes in diversity, this is still a particularly time-consuming task. Artificial intelligence approaches such as machine learning and natural language processing present exciting opportunities to automate, or semi-automate the processing of this information to maximize the full potential of the entire universe of research data in near real-time. In order for these approaches to perform, we require greater clarity around common data structures for biodiversity information (incl. taxonomies, geographies, biomes, and conservation approaches). This presentation will discuss new applications of AI approaches to “smart-sort” relevant information from the “evidence universe” and highlight areas where improvement in needed.

Biography: By training, Samantha Cheng is a population geneticist and conservation scientist, with experience both in the field working in tropical coral reefs, cephalopod fisheries, seafood sustainability - and in the policy sphere, engaging with diverse stakeholders in organizations, governments, and academic institutions to develop evidence-based solutions for conservation and human well-being outcomes. Cheng's research aims to improve understanding of the process of using scientific evidence in conservation planning and decision-making using a multi-disciplinary approach by systematically examining the role of evidence, methodology, interactions, and stakeholders to determine effective pathways from information to action to outcome. A project Cheng is pursuing at the Center for Biodiversity Outcomes along with Drs. Gerber and Anderson, is examining the public value of conservation, specifically from products of specific mechanism to move from evidence to outcomes, knowledge partnerships. Additionally, Cheng is also exploring the role that data science and technology can play in improving the use of evidence in conservation, leading to developing partnerships with data scientists and developers to design and deploy two apps that facilitate and democratize evidence uptake and use machine learning to help find evidence that matters, respectively. Cheng is a former Fulbright Fellow to Indonesia. At the Center for Biodiversity Outcomes, Cheng aims to contribute insight on best practices for pursuing evidence-informed policy and generating tenable, applicable, and sound science for fisheries management.