Cindy Kaiying Lin

Cindy Kaiying Lin is a Postdoctoral Fellow at the Cornell Atkinson Center for Sustainability, affiliated with the Department of Information Science. She is also a Digital Life Initiative Visiting Fellow at Cornell Tech. In Fall 2022, she will be an assistant professor at Pennsylvania State University’s College of Information Sciences and Technology. At present, Cindy is working on a book project that examines ground truth within the history of machine learning as a shifting political and scientific category. This book asks how ground truth and its claim to accuracy and evidentiality have shaped and been shaped by transnational exchanges of mapping, surveying, and computing expertise between Southeast Asia and the United States.

Title and Abstract:

Clustering Sameness: Data Science and the Politics of Peatland Fires in Rural Indonesia

This talk centers on the use of data science techniques to predict subterranean peatland fires in rural Indonesia. It traces a public-private partnership between industry data scientists and government peatland experts and computer engineers aimed at developing a nationwide fire prediction system. This system is designed to help firefighters and military personnel fight fires and patrol human actors responsible for fires in rural South Sumatra. Since one of Indonesia’s largest peatland fires in South Sumatra in 2015, the central government has tasked government researchers to investigate the causes of peatland fires, with Indonesia's minister officials and international development agencies putting blame on the swidden cultivation practices of “asli” (Indonesian language) or native smallholder farmers. This blame however fails to capture the legacies of race, land ownership, and settlement in rural Indonesia. It also falls short of considering how peatland itself has transformed over the years with industrial palm oil plantation expansion. Sometimes glossed as a static landscape in which humans act upon, peatlands have become dried out and fire-prone, producing erratic and unpredictable fires since the construction of massive water canals for palm oil cultivation in the 1990s. But peatlands are also agents that shape data science activities in diverse ways. Industry data scientists and government researchers have begun experimenting with data mining techniques to locate potential sources of peatland fire. As they prototype with these techniques, they learn to use data science to make ethical claims on who should be (and can be) held responsible for peatland fires. To this end, data science in rural Indonesia does not simply abstract and quantify social relations held between population and rural settlement, but also puts industry and government data scientists and engineers in a position to challenge the legacies of race and land ownership in rural peatland settlement.