The First International Geocoding Conference took place in December at the Esri campus in Redlands, CA. Participants Daniel Goldberg from the University of Southern California and Geoffrey Jacquez of BioMedware, Inc. offer this report.
The world of geocoding descended on the Esri Campus in Redlands, CA for the First International Geospatial Geocoding Conference (IGGC) on the 6th and 7th of December, 2011. This first-of-its-kind event was sponsored by Esri, NAVTEQ, the University of Southern California Spatial Sciences Institute, and the Centers for Disease Control and Prevention’s (CDC) National Center for Environmental Health (NCEH) and The Agency for Toxic Substances and Disease Registry (ATSDR). The conference sought to bring together the world’s leading minds on geocoding for a two-day meeting to discuss the past, present and future of geocoding technology and its use in scientific applications, global industry and development, and policy making. This conference presented an opportunity for geocoding users to meet and mingle with geocoding developers to discuss how and where challenges exist; for scientists to present studies of how geocoding accuracy impacts research results and conclusions to policy makers; and for geocoding researchers and developers to hear where efforts are most pressingly needed to build geocoding tools and data structures from the diverse geocoding user community. By all accounts, the event delivered in all respects.
The event drew 200+ attendees from around the globe including such diverse geographic regions as the United Kingdom, Brazil, Australia, Canada and the United States. In attendance were geocoding software developers from industry, academia and non-profits; geocoding users from a variety of domains including health, insurance and logistics; and researchers and policy makers from local, state and federal governments and agencies including the U.S. Census Bureau, HUD and the CDC. Although widely varying in terms of level of interaction with geocoding tools and data – ranging from daily use of geocoding tools to strategic planning and policy making – this diverse group of participants focused their attention on the art and science of geocoding for the entirety of the two-day event. During this time, attendees shared experiences, identified consistent challenges and emerging opportunities, and charted a roadmap for where geocoding technology needs to go and how the community of geocoding researchers, developers and practitioners can make strides down this path.
The deliberations were as far-ranging as the attendees’ countries of origin, spanning the full gamut from the reference data layers and algorithms used to produce geocoded data, to their application in scientific studies for investigating such topics as disease propagation and health service delivery, to the implications geocoding accuracy may have on policy making at all levels of government. The conference opened with a keynote session by Donald Cooke followed by two tracks of presentations. In his keynote, Cooke – by many accounts one of the inventors of what we know today as geocoding – recounted his personal history and involvement with geocoding, starting from the creation of the original Census DIME files (the predecessor of today’s TIGER files), to the development of the first deterministic matching algorithms, to the creation of modern probabilistic matching and scoring techniques, to the founding of GDT. This historical view of the long road that geocoding development has taken through the years provided the perfect context and motivation for the conference.
This was followed by presentations roughly structured to fit within special issues of two academic journals: Transactions in GIS, edited by Dr. John P. Wilson of the University of Southern California, and Spatial and Spatio-temporal Epidemiology, edited by Dr. Andrew Larson of the Medical University of South Carolina. The former is comprised of articles that present advances in the science, development and engineering of geocoding approaches, which are technical in nature, while the latter contains articles that describe advances in the use of geocoded data in the health sciences, which are applied in nature. These academic papers were intermingled with a series of industry and government presentations across six sessions with themes covering novel techniques for geocoding non-standard data such as Twitter messages, methods and techniques for quantifying and propagating geocoding error, temporal geocoding, impacts of geocoding error on health investigations, methods to perform analysis under uncertain geocoding conditions, and perspectives from large industry, government and international geocoding attempts, infrastructures and applications. The day closed with an energetic reception hosted by Esri and NAVTEQ, featuring live music, superb appetizers and a healthy dose of conversation.
The second day of the conference opened with a keynote by Mark Greninger, geographic information officer for Los Angeles County. Opportunities for an increased level of participant interaction through focused breakout sessions, interactive lightning talks and a panel discussion with leaders in the field of geocoding followed. In his keynote, Greninger impressed the audience with the scope and magnitude of the geocoding and address management systems in place at Los Angeles County, an administrative unit often cited as the “eighth largest state in the country” when measured across a number of metrics including population, land area and GDP, among others. Through a series of illustrative case studies, Greninger demonstrated how address management and geocoding go hand-in-hand to serve civil society across the diverse population of LA County, and how when either fails to produce accurate results, dramatic consequences may occur including the loss of life and property.
Next, a series of informative and exciting lightning talks by a diverse group of academics, government officials and industry speakers led the way into two parallel tracks of breakout sessions focused on topics including address standardization, confidentiality and ethics, volunteered geographic information, and regional differences in geocoding. The motivation for the breakout sessions was to enable face-to-face discussion among all levels of the geocoding user pipeline – developers, users, researchers and analysts, policy makers and data providers – so that each group could identify and prioritize the most pressing challenges and best opportunities to address them. Each session produced and reported on a set of action items for each topic that will be used to drive research and development priorities across industry, academic and government agendas.
These breakout session reports by the session chairs were used to spark a discussion among a panel of geocoding experts from a broad range of disciplines and roles in the geocoding community. Led by Christophe Charpentier of Esri, this panel represented the interests of data producers (Dan Gibbons – NAVTEQ), industry groups (Don Cooke – Esri), the health community (Dave Stinchcomb – Westat), federal agencies (Ama Danso – Census; Jon Sperling – HUD) and local governments (Mark Greninger – LA County). Discussions covered topics including the views of each participant on the most pressing geocoding challenge as well as what each considered to be the most opportune area for future research and development. Responses to these questions drew from the diverse backgrounds and expertise of the group and shed interesting light on the priorities of each respective group. However, despite these differences, the methods, techniques and data sources for using geocoding to address the needs of the developing world was a consistent theme, as was the need to develop means to better quantify and represent the accuracy of geocoded data.
The organizing committee for this conference consisted of representatives from industry, academia and government, all of whom wanted to see this conference take place for a variety of necessarily selfish reasons including but not limited to: gathering a captive group for discussing the latest and greatest geocoding tools available for purchase; making the case for the need for continued research funding; and providing a rationale for the existence of labs, programs, hefty geocoding budgets and research agendas. However, underlying all of these perspectives was the overarching and unifying goal of achieving the next great advances in geocoding that will serve to help tackle society’s most pressing challenges. This conference made substantial steps forward by facilitating an environment where geocoding users, developers, analysts and policy makers could gather to chart the next generation of geocoding tools and techniques; providing opportunities to discuss methods to utilize these data to their fullest potential in scientific research and policy making; and organizing a diverse community of practice to follow the path pioneered by luminaries such as Cooke and his colleagues at the U.S. Census Bureau in the 1970s.
With these goals in mind, the First International Geospatial Geocoding Conference could not have been more successful. The support of the CDC, Esri, NAVTEQ and USC provided the opportunity for more than 200 participants – including 12 travel scholarship winners consisting of students and researchers from underrepresented groups in science – to gather for a two-day intensive meeting where ideas were exchanged, concerns voiced and addressed (or at least acknowledged) and research collaborations formed. These interactions may not bear immediate fruit today, but will certainly do so tomorrow and down the road as participants take what they learned into practice within their careers; as funding agencies look to peer-reviewed publications as a measure of the importance of and need to fund a field of study; as commercial organizations prioritize where development efforts are placed; and as reviewers of scientific findings scrutinize results and conclusions that have the potential to affect policy decisions.
However, as noted in the papers and presentations of the Best Paper award winner Geoffrey M. Jacquez of BioMedware, Inc. and reiterated by the Best Paper runner-up Paul Zandbergen of the University of New Mexico, bringing these long-term benefits into reality requires continued effort on the part of the geocoding community. It is up to you, the reader, to take the outputs of this conference – special issues of two academic journals, online versions of presentations, webinars of the sessions – and run with them. If you are a developer, build better tools; if you are a consumer, ask tougher questions; if you are an analyst, scrutinize your geocoded data as if your entire analysis depends on it – because it does. Without a user base continually demanding improvements, we run the risk of stagnating at our current level of geocode quality – which although light-years beyond what was initially envisioned, can by all accounts still be improved. And last but not least, make plans to attend the Second International Geospatial Geocoding Conference (date TBD)!
The IGGC and this publication were supported in part by Cooperative Agreement Number 1H13EH000793-01 from the Centers for Disease Control and Prevention. The contents are solely the responsibility of the authors and do not necessarily reflect the official views of the Centers for Disease Control and Prevention.