GeoAI 2019 Workshop ReportThe 3nd ACM SIGSPATIAL International Workshop on

GeoAI: AI for Geographic Knowledge DiscoverySeattle, WA, USA - November 5, 2019

Song Gao1, Shawn Newsam2, Liang Zhao3, Dalton Lunga4, Yingjie Hu5, Bruno Martins6,Xun Zhou7, Feng Chen8

1University of Wisconsin, Madison; 2University of California, Merced;3George Mason University; 4Oak Ridge National Lab;

5University at Buffalo; 6University of Lisbon;7University of Iowa; 8The University of Texas at Dallas.

Nowadays artificial intelligence (AI) is bringing tremendous opportunities and challenges to geospatial re-search. Big data enable computers to observe and learn the world from many different perspectives, whilehigh performance machines support the development, training, and deployment of AI models within reasonableamount of time. Recent years have witnessed significant advances in the integration of geospatial study and AI inboth academia and industry. There have already been many successful studies for both physical environment andhuman society. Focusing on modeling the physical nature, research has shown that deep learning can improvethe representation of clouds that are smaller than the grid resolutions of climate models. Examining the humansociety, AI and natural language processing methods, such as word embeddings, help quantify changes in stereo-types and attitudes toward women and ethnic minorities over 100 years in the United States. There are also manyother applications that effectively integrate AI with problems in geospatial studies, such as vehicle trajectoryprediction, high-definition mapping and navigation, historical map digitizing, gazetteer conflation, geographicfeature extraction, and place understanding. The 3nd International Workshop on AI for Geographic KnowledgeDiscovery (GeoAI 2019) builds on the success of the previous workshops in 2017 and 2018. GeoAI is bring-ing together geoscientists, computer scientists, engineers, entrepreneurs, and decision makers from academia,industry, and government to discuss the latest trends, successes, challenges, and opportunities in the field ofartificial intelligence for data mining and geographic knowledge discovery.

GeoAI 2019 received 25 paper submissions in total. After a rigorous peer-review process by the programcommittee, in total 17 papers (5 full papers, 10 short papers, and 2 vision papers) were accepted by the workshopand selected for presentations. Dr. Xin Chen, the Director of Engineering at HERE Technologies, gave akeynote on “HD Live Map for Automated Driving: An AI Approach” (industry keynote), and Dr. Raju Vatsavai,Professor at North Carolina State University, gave a keynote on “Geospatial AI for Monitoring Crops to NuclearProliferation Using Global Earth Observations” (academic keynote). This year’s workshop also featured onewrap-up discussion on “How GeoAI will progress?” Many participants contributed constructive ideas, such asencouraging future submissions with share of datasets for common benchmark testing, engagement betweendifferent domains of people, enhance data privacy and data security issues, transparency and reproducibility,and explicit spatial constraints on AI models. Participants also discussed the design of core skills for studentswho want to build expertise in GeoAI. There are 43 participants who officially registered to GeoAI’19. Theworkshop attracted in average 50 participants, with a maximum of about 80 participants in the room.

We sincerely thank our program committee for their time and efforts in reviewing the submitted papers. We

hope that the proceedings of GeoAI’19 can stimulate new ideas and make a contribution to this growing field.

Figure 1: GeoAI’19 Workshop in Chicago

3rd International Workshop on GeoAI 8:00 am 9:00 am Breakfast 9:00 am 9:05 am Welcome and announcements 9:10 am 10:00 am Keynote: HD Live Map for Automated Driving: An AI Approach

Dr. Xin Chen, Director of Engineering, HERE Technologies. 10:00 am 10:20 am An Unsupervised Approach to Geographical Knowledge

Discovery Using Street Level and Street Network Images Stephen Law Universitys College London London, UK Daniel Mateo Neira Alvarez University College London London, UK

10:20 am 10:30 am A Deep Learning Based Illegal Parking Detection Platform Zhengcong Yin Texas A&M University; Haoyi Xiong University of Iowa; Xun Zhou University of Iowa; Daniel Goldberg Texas A&M University; Dave Bennett University of Iowa; Chong Zhang Esri Inc.

10:30 am 11:00 am Coffee break 11:00 am 11:20 am ChangeNet: Learning to Detect Changes in Satellite Images

Ying Chen, Xu Ouyang, Gady Agam Illinois Institute of Technology

11:20 am 11:40 am Keras Spatial: Extending Deep Learning Frameworks for Preprocessing and On-The-Fly Augmentation of Geospatial data Aiman Soliman, Jeffrey Terstriep University of Illinois at Urbana-Champaign

11:40 am 12:00 pm Assessing the Placeness of Locations through User-contributed Content Xiaoyi Yuan and Andrew Crooks George Mason University

12:00 pm 12:20 pm Urban Flood Mapping with Residual Patch Similarity Learning Bo Peng, Xinyi Liu, Zonglin Meng, Qunying Huang University of Wisconsin - Madison

12:20 pm 2:00 pm Lunch (not provided) 2:00 pm 2:50 pm Keynote: Geospatial AI for Monitoring Crops to Nuclear

Proliferation Using Global Earth Observations Professor Raju Vatsavai, North Carolina State University.

2:50 pm 3:00 pm A Machine Learning Approach to Estimate Median Income Levels of Sub-Districts in Thailand Using Satellite and Geospatial Data Ugyen Jigten Dorji, Anon Plangprasopchok, Navaporn Surasvadi, Chaiyaphum Siripanpornchana; NECTEC, Thailand

3:00 pm 3:10 pm A Traffic Sign Discovery Driven System for Traffic Rule Updating Tengfei Xing, Yang Gu, Zhichao Song, Zhihui Wang, Pengfei Xu, Runbo Hu, Hua Chai Didi Chuxing

3:10 pm 3:20 pm City-level Geolocation of Tweets for Real-time Visual Analytics Luke Snyder Purdue University; Morteza Karimzadeh University of Colorado Boulder; Ray Chen Purdue University; David Ebert Purdue University

3:20 pm 3:30 pm Estimating the Spatial Resolution of Very High-Resolution Overhead Imagery Haolin Liang and Shawn Newsam University of California, Merced

3:30 pm 4:00 pm Coffee break 4:00 pm 4:10 pm Kartta Labs: Unrendering Historical Maps

Sasan Tavakkol and Yao-Yi Chiang Google Research, New York; Tim Waters Waters Geospatial, Leeds, UK Feng Han Google Research, New York, NY Kisalaya Prasad Google, New York, NY Raimondas Kiveris Google Research, New York, NY

4:10 pm 4:20 pm Land Use Inference from Mobility Traces John Krumm Microsoft Research, Redmond, WA Kora Krumm Seattle Pacific University, Seattle, WA

4:20 pm 4:30 pm Mapping Miscanthus Using Multi-Temporal Convolutional Neural Network and Google Earth Engine Yanan Xin The Pennsylvania State University, University Park, PA Paul R. Adler United States Department of Agriculture, Agricultural Research Service (USDA-ARS) University Park, PA

4:30 pm 4:40 pm Multi-scale Graph Convolutional Network for Intersection Detection from GPS Trajectories Yifang Yin National University of Singapore, Singapore Abhinav Sunderrajan, Xiaocheng Huang, Jagannadan Varadarajan, Guanfeng Wang GrabTaxi Holdings, Singapore Dhruva Sahrawat, Ying Zhang, Roger Zimmermann, See-Kiong Ng National University of Singapore, Singapore

4:40 pm 4:50 pm STIETR: Spatial-temporal Intelligent E-Taxi Recommendation System Using GPS Trajectories Ke Mei, Wei Tu, Qingquan Li, Tianhong Zhao Shenzhen University Haoyu Ye, Yatao Zhang Wuhan University

4:50 pm 5:00 pm Contextualizing Space and Time for GeoAI JITAIs (Just-in-Time Adaptive Interventions) Jiue-An Yang and Marta Jankowska UC San Diego, San Diego, CA

5:00 pm 5:10 pm Imitation Learning from Human-Generated Spatial-Temporal Data (Vision Paper) Yanhua Li and Weixiao Huang Worcester Polytechnic Institute, USA

5:10 pm 5:30 pm Wrap up discussion – Future Direction and Collaboration Opportunities

i

GeoAI 2019 Proceedings of the 3nd ACM SIGSPATIAL International Workshop on

AI for Geographic Knowledge Discovery

(GeoAI 2019)

Nov. 5th, 2019, Chicago, IL, USA

Editors:

Song Gao, University of Wisconsin, Madison, WI, USA

Shawn Newsam, University of California, Merced, CA, USA

Liang Zhao, George Mason University, VA, USA

Dalton Lunga, Oak Ridge National Lab, TN, USA

Yingjie Hu, University at Buffalo, NY, USA

Bruno Martins, University of Lisbon, Portugal

Xun Zhou, University of Iowa, IA, USA

Feng Chen, The University of Texas at Dallas, TX, USA

ii

The Association for Computing Machinery, Inc. 1601 Broadway, 10th Floor New York, NY 10019-7434

ACM COPYRIGHT NOTICE.

Copyright © 2019 by the Association for Computing Machinery, Inc (ACM). Permission to make digital or hard copies of portions of this work for personal or classroom use is granted without fee provided that the copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page in print or the first screen in digital media. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or a fee. Request permission to republish from: Publications Dept. ACM, Inc. Fax +1-212-869-0481 or E-mail <permissions@acm.org>. For other copying of articles that carry a code at the bottom of the first or last page, copying is permitted provided that the per-copy fee indicated in the code is paid through the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. Notice to Past Authors of ACM-Published Articles ACM intends to create a complete electronic archive of all articles and/or other material previously published by ACM. If you have written a work that was previously published by ACM in any journal or conference proceedings prior to 1978, or any SIG Newsletter at any time, and you do NOT want this work to appear in the ACM Digital Library, please inform permissions@acm.org, stating the title of the work, the author(s), and where and when published.

ACM ISBN: 978-1-4503-6957-2 Additional copies may be ordered prepaid from: ACM Order Department P.O. BOX 11405 New York, NY 10286-1405 Phone: 1-800-342-6626 (USA and Canada) +1-212-626-0500 (Global) Fax: +1-212-944-1318 E-mail: acmhelp@acm.org Printed in the USA

iii

FOREWORD In today’s era of big data, advanced algorithms, and immense computational power, artificial intelligence

(AI) is bringing tremendous opportunities and challenges to geospatial research. Big data enable computers

to observe and learn the world from many different perspectives, while high performance machines support

the developing, training, and applying of AI models within reasonable amount of time. Recent years have

witnessed significant advances in the integration of geography and AI in both academia and industry. There

have already been many successful studies. Focusing on modeling the physical nature, research has shown

that deep learning can improve the representation of clouds that are smaller than the grid resolutions of

climate models. Examining the human society, AI and natural language processing methods, such as word

embeddings, are helping quantify changes in stereotypes and attitudes toward women and ethnic minorities

over 100 years in the United States. There are also many other applications that effectively integrate AI

with problems in geospatial studies, traffic rule detection, navigation, historical map digitizing, gazetteer

conflation, geographic feature extraction, land-use change detection, and place understanding.

The 3nd International Workshop on AI for Geographic Knowledge Discovery (GeoAI 2019) builds on

the success of the previous workshop. GeoAI is bringing together geoscientists, computer scientists,

engineers, entrepreneurs, and decision makers from academia, industry, and government to discuss the

latest trends, successes, challenges, and opportunities in the field of artificial intelligence for data mining

and geographic knowledge discovery.

This year we received 25 paper submissions in total. After a rigorous peer-review process by the program

committee, in total 17 papers (including 5 full papers, 10 short papers, and 2 vision papers) were accepted

by the workshop and selected for presentations. The workshop program also consists of two keynote talks

from well-known experts from academia and industry.

Song Gao, Shawn Newsam, Liang Zhao, Dalton Lunga, Yingjie Hu, Bruno Martins, Xun Zhou, Feng Chen

ACKNOWLEDGEMENTS

We would like to thank the program committee whose reviewing efforts are important for ensuring the

quality of the accepted papers. In addition, many thanks to our keynote speakers Dr. Raju Vatsavai and Dr.

Xin Chen who kindly accepted our invitations and will be sharing their research in GeoAI 2019.

iv

ORGANIZERS GENERAL CHAIRS: Song Gao, University of Wisconsin, Madison Shawn Newsam, University of California, Merced Liang Zhao, George Mason University COMMUNICATION CHAIRS: Dalton Lunga, Oak Ridge National Lab PROGRAM COMMITTEE CHAIRS: Yingjie Hu, University at Buffalo Bruno Martins, University of Lisbon Xun Zhou, University of Iowa Feng Chen, The University of Texas at Dallas PROGRAM COMMITTEE: Benjamin Adams, University of Canterbury Orhun Aydin, Esri Benjamin Bischke, German Research Center for AI Arnold Boedihardjo, Radiant Solutions Damian Borth, University of St. Gallen Dengfeng Chai, Zhejiang University Yao-Yi Chiang, University of Southern California Yanjie Fu, Missouri University of Science and Technology Jing Gao, University at Buffalo Jonathan Gerrand, Witwatersrand University Shahriar Hossain, University of Texas, El Paso Zhe Jiang, University of Alabama Asanobu Kitamoto, Japan National Institute of Informatics Kuldeep Kurte, Oak Ridge National Laboratory Xiaojiang Li, Temple University Yanhua Li, Worcester Polytechnic Institute Wenwen Li, Arizona State University Tao Liu, Oak Ridge National Laboratory Grant McKenzie, McGill University Volpi Michele, Swiss Data Science Center, ETH Zurich/EPFL Ray Dos Santos, Army Research Engineering Corp. Steven Schockaert, Cardiff University Gautum Thakur, Oak Ridge National Laboratory Devis Tuia, Wageningen University and Research Hong Wei, University of Maryland, College Park Lingfei Wu, IBM Watson Hsiuhan(Lexie) Yang, Oak Ridge National Laboratory Junsong Yuan, University at Buffalo Fan Zhang, MIT Senseable City Lab

v

Xuchao Zhang, NEC Lab Di Zhu, Peking University SPONSORS CORPORATE SPONSORS

vi

PUBLISHER SPONSOR

INSTITUTIONAL FUNDING

Table of Content Land Use Inference from Mobility Traces

John Krumm and Kora Krumm ………………………………………………..…....…......... 1-4

STIETR: Spatial-temporal Intelligent E-Taxi Recommendation System Using GPS

Trajectories

Ke Mai, Wei Tu, Qingquan Li, Haoyu Ye, Tianhong Zhao and Yatao Zhang .…..……......... 5-8

Imitation Learning from Human-Generated Spatial-Temporal Data

Yanhua Li and Weixiao Huang .…..……… .……… .……… .…..............……................... 9-10

A Machine Learning Approach to Estimate Median Income Levels of Sub-Districts in

Thailand Using Satellite and Geospatial Data

Ugyen Jigten Dorji, Anon Plangprasopchok, Navaporn Surasvadi and Chaiyaphum

Siripanpornchana .……… .…….……… .……… .……….……....................................… 11-14

Assessing the Placeness of Locations Through User-Contributed Content

Xiaoyi Yuan and Andrew Crooks .…..……… .……… .……… ...........................................15-23

ChangeNet: Learning to Detect Changes in Satellite Images

Ying Chen, Xu Ouyang and Gady Agam .…..……… .……….………................................. 24-31

A Deep Learning based Illegal Parking Detection Platform

Zhengcong Yin, Haoyi Xiong, Xun Zhou, Daniel Goldberg, Dave Bennett and Chong

Zhang….................................................................................................................................. 32-35

Multi-scale Graph Convolutional Network for Intersection Detection from GPS

Trajectories

Yifang Yin, Abhinav Sunderrajan, Xiaocheng Huang, Jagannadan Varadarajan, Guanfeng Wang,

Dhruva Sahrawat, Ying Zhang, Roger Zimmermann and See-Kiong Ng ..........................… 36-39

Urban Flood Mapping with Residual Patch Similarity Learning

Bo Peng, Xinyi Liu, Zonglin Meng and Qunying Huang …..……… .……….….……...........40-47

Kartta Labs: Unrendering Historical Maps

Sasan Tavakkol, Yao-Yi Chiang, Tim Waters, Feng Han, Kisalaya Prasad and Raimondas

Kiveris …..……… .……….….………… …..……… .……….….…… ………………........48-51

A Traffic Sign Discovery Driven System for Traffic Rule Updating

Tengfei Xing, Yang Gu, Zhichao Song, Zhihui Wang, Yiping Meng, Nan Ma, Pengfei Xu, Runbo

Hu and Hua Chai …..……… ...............…..……… .……….….………................................52-55

An Unsupervised Approach to Geographical Knowledge Discovery Using Street Level And

Street Network Images

Stephen Law and Daniel Mateo Neira Alvarez …………..……… .……….….………........ 56-65

Contextualizing Space and Time for GeoAI JITAIs (Just-in-Time Adaptive Interventions)

Jiue-An Yang and Marta Jankowska …..……… .……… .….…… .….…… .….……….....66-68

Keras Spatial: Extending Deep Learning Frameworks For Preprocessing And On-The-Fly

Augmentation of Geospatial Data

Aiman Soliman and Jeffrey Terstriep …..……… .……… .….…… .….…… .….….............69-76

Estimating the Spatial Resolution of Very High-Resolution Overhead Imagery

Haolin Liang and Shawn Newsam …..……… .……… .….…… .….…… .….…….............77-80

Mapping Miscanthus Using Multi-Temporal Convolutional Neural Network and Google

Earth Engine

Yanan Xin and Paul R. Adler …..……… .……… .….…… .….…… .….….......………......81-84

City-level Geolocation of Tweets for Real-time Visual Analytics

Luke Snyder, Morteza Karimzadeh, Ray Chen and David Ebert…..……… .……… ......… 85-88