The one day agenda will be packed with insightful talks from experts and practitioners of spatial data science from academia and industry. Lightning talks will keep the day moving at a fast-pace, and interspersed breaks and happy hour afterwards will provide plenty of opportunity to network and learn more from peers.
Be inspired by keynote speakers and their vision for the future of geospatial data science.
Learn first-hand about pioneering projects in the fields of geospatial analysis, data visualization, analytics, and data science.
Exchange ideas & valuable insights with other professionals during coffee breaks and unconference sessions.
Marc has deep expertise in both the technical and cultural aspects of how data can be harnessed to transform the way that organizations operate. In addition, Marc is a fellow at the Columbia School of Journalism’s Brown Institute for Media Innovation, collaborating with investigative journalists to develop new approaches to data-driven storytelling. Marc holds a doctorate in cultural anthropology from the University of California, where his research examined the social consequences of new forms of data analytics. Marc has received numerous awards for his work, including fellowships from the National Science Foundation and Intel Labs.
Kevin directs SharedStreets, a non-profit organization that builds tools for public-private collaboration around transport data. Blending technology and policy, we're building standards, digital infrastructure, and governance models to support new ways of managing and sharing data about our transport systems. Prior to launching SharedStreets, Kevin served in leadership roles at the open government/open data non-profits OpenPlans and Sunlight Foundation, and Alphabet's Sidewalk Labs.
Kevin Stofan is a Customer Facing Data Scientist at DataRobot and an Adjunct Instructor at Pennsylvania State University where he teaches a graduate level GIS course. In his role as a Data Scientist with DataRobot he works with commercial and public sector clients throughout the full customer lifecycle to ensure success with implementing machine learning using DataRobot. DataRobot’s customers include organizations in a wide variety of private and public market sectors including healthcare, banking, and insurance. He has over 19 years of experience managing data science teams and projects.
Shan is a senior data visualization engineer at Uber. She is a coder, a designer, and a data artist. Shan is the founding member of Uber's data visualization team and creator of kepler.gl. At Uber, she builds data tools and platforms to make business intelligence easy to access, creates exploratory data visualizations to facilitate data analysis and modeling, and publishes public facing graphics and animations to tell data stories about Uber. Prior to joining Uber, Shan studied at MIT and received a master in design computation while conducting research at the MIT Senseable City Lab as data visualization specialist.
At the intersection of technology, government, data, and civic engagement, Andrew has been called an “innovative thinker and skilled technologist”. He currently leads JHU GovEx's data practices efforts, working with governments across the globe to treat their data as a strategic asset. He led former Mayor Michael R. Bloomberg’s NYC OpenData and Governor Andrew M. Cuomo’s Open NY programs – both widely recognized for their ground-breaking policy frameworks, excellent data management practices, and powerful public engagement strategies. Andrew has worked in government for 20 years, leading small teams, medium-sized departments, and executive governance groups. His broad technology experience covers areas such as data management, standards collaboration, information security, infrastructure operations, application lifecycle management, project management, resource planning, procurement, and more. Andrew, an avid motorcyclist, rode back and forth across North America in 2009. He and Cindy love games, gadgets and geekery, and proved it by getting married on Ultimate Pi(e) Day (3/14/15).
Title: Chief Data Officer Affiliation: Louisville Metro Government Bio: In the Office of Civic Innovation and Technology under mayor Greg Fischer, Louisville’s CDO helps employees, residents, and partners get the data they need to to their jobs an innovate. This includes our internal cross-departmental Data Governance team focusing on data culture, training, standards, and inventories, our external open data platform, and data sharing agreements with private entities (Waze, Bird, D4D, CfA), universities (UPenn, UofL), and other cities (Harvard CAN). His passion is collaboration with governments, public private partnerships, and resource pooling to create products that benefit multiple entities at once through the Open Government Coalition at www.GovInTheOpen.com.
Jeff Ferzoco is a designer and mapper who works closely with CARTO clients on the Customer Success team to help them create and manage beautiful projects. His background is a blend of information design, technology, urban planning and information journalism. Jeff teaches urban data and mapping at Pratt’s SAVI program. He loves coding, cartography and city simulation games. He went to California College of the Arts and is a lead organizer for GeoNYC, the NYC meetup for all things geo.
Jamie Saxon is a postdoctoral fellow with the economics group of the Harris School of Public Policy and the Center for Spatial Data Science, at the University of Chicago. He has studied compactness as a tool for districting reform (gerrymandering), the local graph structure of American neighborhoods, and the accessibility of public resources -- mainly health and parks. In practice, this means clustering algorithms, distributed computing, and some graph theory.
Dan Snow is a graduate student, research assistant, and teaching assistant at the University of Chicago Harris School of Public Policy. His research focuses on healthcare and spatial accessibility. He also works as a data analyst and project manager at a nonprofit organization on Chicago’s north side.
Alex Singleton is a Professor of Geographic Information Science at the University of Liverpool, where he was appointed as a Lecturer in 2010. Previously he held research positions at University College London, where he was also awarded a PhD in 2007. He completed a BSc in Geography at the University of Manchester, graduating with a First-class honours degree in 2003.
Xiaojiang is a Postdoctoral Associate at MIT Senseable City Lab. His current research focuses on developing and applying geospatial analyses and data-driven approaches in the domain of urban studies. He has proposed to use Google Street View for urban environmental studies and developed the Treepedia project, which aims to map street greenery for cities around the world. He is also working on using human trace data to study human activities and investigate the connection between urban environments and human activities. He received his Ph.D. in the Department of Geography, University of Connecticut.
Lindsay Poirier, PhD, is a cultural anthropologist trained in the field of Science and Technology Studies and Information Technology and Web Science. Her research focuses on digital expertise and data cultures. At BetaNYC, she is researching data practices and needs in New York City’s community boards, designing tools and curriculum to promote open data accessibility, and advising city officials on how city data resources should be structured and published to best support civic engagement.
Vipassana Vijayarangan is the Transport Data Scientist for the Urban Mobility team under the WRI Ross Center for Sustainable Cities. As a part of the Open Transport Partnership, she uses data to empower transport agencies to develop evidence-based solutions to mobility challenges. Prior to joining WRI, Vipassana worked as a consultant with the NYU Marron Institute of Urban Management mapping and studying the informal transit network in Bogota, Colombia. Vipassana started her professional career as the first employee of a ride-sharing startup in Bangalore, India. Vipassana holds a bachelor’s degree in Computer Science and Engineering and a master’s degree in Applied Urban Informatics from the NYU Center for Urban Science and Progress.
Dongjie is a Data Scientist at CARTO. He has a background in mathematics & statistics and received his master’s degree in urban informatics from New York University. Dongjie previously worked as a research assistant at NYU, where he researched on applying statistical modeling and data science technology to solve urban problems like transportation and public health issues. In his spare time, Dongjie enjoys hiking and visiting museums. Also, you may find him at a sports bar celebrating the victory of his favorite soccer team.
Laurel Donaldson is a Senior Urban Planner at WXY architecture + urban design, where she specializes in generating innovative urban policy and development strategies to support industrial job growth, clean technology deployment, and creative place-making initiatives for public and private sector clients. Prior to WXY, Laurel served as a Policy Innovation Fellow for the Boston Mayor's Office of New Urban Mechanics and led research management for an international political risk consulting firm, the Eurasia Group. Laurel holds an M.A. in City Planning from the Massachusetts Institute of Technology and a B.A. in Political Science from Duke University.
Alyssa Bianco is a Data Analyst at Dstillery, where she uses behavioral and location data to help brands understand and engage with their current and prospective customers. Her work includes developing behavioral audiences for audience insights and activation in advertising campaigns. In her recent work, she is involved in researching methods to use observed consumer behaviors to report on real-time economic trends. Prior to joining Dstillery’s Data Science team, Alyssa held positions in advertising operations and content management. She holds a bachelor’s degree in writing from Loyola University Maryland.
Lori’s research interests focus on chronic disease prevention and maintenance and healthcare delivery system transformation. In light of the US’ increasingly challenging chronic disease burden and her own community activities, Lori’s approach to health encompasses our built and social environment and the systematic and generational inequalities that shape risk factors and opportunities for health. Lori holds degrees from NYU Wagner School of Public Service and Carnegie Mellon University. She bicycle commutes whenever she can.
Xi is a fourth-year PhD Candidate in Geography (GIScience focus) and Social Data Analytics at Penn State. He is a Graduate Associate at the Friendly Cities Lab and the GeoVISTA Center. His research focuses on developing methods to model human activities in the urban space and conducting empirical studies to reveal the latent patterns in cities. His goal is to make cities friendlier and better connected through geographic theories and data science. He also worked on topics related to urban life as a Data Science Research Intern at Bell Labs and a Software Engineering Intern at Google.
Stephanie Lackner is a postdoctoral researcher at the Program in Science, Technology, and Environmental Policy (STEP) in the Woodrow Wilson School of Public and International Affairs at Princeton University. Her research interests revolve around natural hazards, the socioeconomic impacts of disasters, and geospatial data analysis. Before joining STEP, she completed a PhD in Sustainable Development at Columbia University's School of International and Public Affairs and an MSc in Mathematics at the Vienna University of Technology.
Emily Goldman works at BetaNYC and co-directs the Civic Innovation Lab in the Office of the Manhattan Borough President. She has a M.A. in Historic Preservation Planning and Ph.D. in City & Regional Planning from Cornell University and has served on BetaNYC's leadership committee four four years, since October 2014. She believes in the power of newly-available information to help address intractable societal problems.
Mary John is a project manager and business consultant at Arcadis. Mary works with public agencies to derive more value from their data, striving to make every strategy and policy decision data driven. She is endlessly curious about how cities function and always looking for ways to gain insights from the vast amount of data cities produce. Mary holds a Bachelor’s degree in Biological and Environmental Engineering from Cornell University.
Yuan Shi is an Urban Data Scientist in Arcadis, specifically focused in leveraging data for urban solutions and mobility analytics. She is passionate about utilizing data for well-being of citizens and improve the quality of life. Yuan holds a Master’s degree of Urban Informatics from New York University.
Check this section regularly, as new sessions and topics will be added on a frequent basis.
We begin with a review of traditional methods for territory management (or measuring scarcity). We then motivate a "Rational Agent Access Model" (RAAM) whereby individuals balance costs of transit time and congestion at the resource. It is implemented efficiently in c++ and is free to use. We evaluate our model on a nation-wide, tract-level, multimodal OD matrix, itself built using open source tools distributed on AWS. We compare our results to the outcomes of existing methods and present practical conclusions. We conclude by looking forward to measurements of realized access with cell phone traces.
The street is the basic unit of the city and a focal point of human activity, acting as the foundation for transportation and information exchange. A thorough quantification and understanding of the physical streetscape (i.e., features and dynamics) would offer great utility to those investigating the urban environment, its physical social interactions, and implications on human well-being. The combination of globally available geo- tagged street-level images and the recent progress in deep learning provides a powerful tool for us to understand cities and solve urban issues. This presentation will talk about using Google Street View to map and quantify street greenery, shade provision, and sun glare occurrence.
Bicycle commuting has been associated with increases in physical activity and reduced risk for chronic diseases and all-cause mortality. While the evidence for health effects as a result of bicycle commuting is strong, the evidence for health effects from bike shares is more equivocal. Because data on socioeconomic status and physical activity for bike share users were not available, our study utilized spatial data to explore bike share access and participation for neighborhoods. In this talk, we’ll explore how spatial data and mapping empowered us to explore new public health questions and opportunities to reduce health inequities.
yt (yt-project.org) is a python toolkit for analyzing and visualizing volumetric data. While yt was traditionally used only for analyzing computational astrophysics simulations we have begun work on adding support in yt for georeferenced data. This includes modifications both on the visualization side as well as in the yt core to support new data formats and meshes that are commonly used in the hydrology, oceanography, geophysics, and meteorology communities.
This project investigates the spatial and temporal relationship between remotely sensed nightlight data and socioeconomic variables on population, employment, and infrastructure at a high spatial resolution. Utilizing a unique data set for Austria, we analyze what forms of social use of space are emitting the most light and evaluate the relationship between temporal changes in the variables. The results suggest that a general concept of “social use of space” can explain most of the observed correlations. Furthermore, we test whether established relationships between the variables are a function of spatial resolution.
Cities are the habitat of people's daily lives. Besides the roads, buildings, and residents in a city, there are also many intangible properties, such as social connections, neighborhood culture, and residential interior spaces, which were hard to quantify. However, those properties are directly related to people's wellbeing and quality of life. In this talk, we will introduce three case studies that measure the intangible in cities with user-generated content using spatial analysis, crowdsourcing, and deep learning techniques. The case studies provide new perspectives on both the spatial data science theory and how we can improve the urban life of people.
The movement of commercial waste in and out of New York City is an incredibly complex system, involving hundreds of companies and hundreds of thousands of businesses. Combining city-specific best practices with advanced analytics capabilities, Arcadis is helping the city optimize waste removal for a safe and efficient collection system that advances the City’s zero waste goals.
An extended coffee break and unconference session
At Uber, location data is our biggest asset. The data visualization team recently open sourced its own geospatial application - kepler.gl for visual exploration of large-scale geolocation datasets in the browser. Powered by deck.gl, kepler.gl can render millions of points representing thousands of trips and perform spatial aggregations on the fly. In this talk, I will talk about how did we build kepler.gland how do we smoothly rendering millions of points in the browser leveraging GPU calculation and deck.gl's instance rendering mechanism.
Automated machine learning (AML) and feature engineering are emerging topics in predictive analytics. Many of the existing techniques in AML address universal feature engineering tasks such as missing value imputation, one-hot encoding, or binning. However, automated techniques specific to feature engineering of spatial data is lacking. Spatial Data Scientists are uniquely qualified to address the peculiarities of working with spatial data in machine learning workflows. This talk will discuss techniques for automating spatial feature engineering and selection in zonal data using open source Python libraries such as PySAL, GeoPandas, and scikit-learn. We will focus on a technique to create a large set of spatially lagged features from an input feature set and conduct feature selection on the derived features for input into machine learning models. The technique will be demonstrated on a classic real estate prediction problem using residential home sales data in St. Petersburg, Florida.
This talk presents SLA Mapper (SLAM) - a tool BetaNYC designed in collaboration with NYC community boards to support their processing of liquor license applications and renewals. When reviewing licenses, NYC community boards gather a great deal of information - such as the establishment’s previous licenses, its certificate of occupancy, its restaurant health grades, and noise complaints made to 311 about the establishment. I will discuss the reasons why it is so difficult to aggregate and query this information from a single information system, and I will describe how leveraging geolocation has improved the capacity to pull information into a single view.
This talk will review the approach and outcomes of two data-driven pilot programs that WXY architecture + urban design developed as a consultant for the New York City Department of Education and the City of New Rochelle. The first project leverages a unique, proprietary NYCDOE dataset to interrogate the factors that drive racial and socio-economic segregation in NYC public schools as students matriculate from the elementary to middle school grade levels. The second project proposes a novel method for quantifying the impacts of new residential development on public school infrastructure in New Rochelle. In both cases, our team analyzed and visualized geospatial data that highlight the dynamics of school feeding patterns and school placement trends/bias, which can inform new policies to promote school diversity and manage development-induced enrollment growth.
This talk presents Tenants Map - a tool to help track buildings where rent-stabilized tenants are experiencing unsafe conditions. By joining open and non-open datasets, this tool shows us when rent-stabilized tenants are using 311 to log their concerns about housing safety, dilapidated conditions, remiss landlords, and/or tenant harassment in real time. It is meant to help community boards and elected officials hold landlords accountable to their tenants and provide safe housing. More broadly, it also conveys the use of 311 as a tool that New Yorkers are using as a call for help. Next steps may include creating alerts to activate the tool and require less exploration on the user-side.
With data harvested by Dstillery, we can complement data collected by government agencies and organizations advocating for entrepreneurship and economic growth. We find that the concentration of a region’s visits to website resources for entrepreneurship and business development are statistically related to business start-up and, particularly, growth activity. This analysis points to the potential of this data source to nowcast business formation and growth at a regional level.
The inaugural Spatial Data Science Conference took place in December 2017 at the CARTO offices in the Bushwick neighborhood of Brooklyn. Check out the agenda and videos here.
Thank you for your interest in speaking at the Spatial Data Science Conference. Our talk submissions are now closed. We look forward to seeing you at the conference!