Member Spotlight: Interview with Dr. Han Qin
Dr. Han Qin is Managing Consultant at Navigant in Washington, D.C., where he provides geospatial computing, machine learning, and custom development services. He earned his Ph.D. in 2017 from George Mason University, where he worked on quality assessment strategies and protocols in geocrowdsourcing applications. Prior to his PhD, Han earned an MS degree from Eastern Michigan University focused on Health GIS and web mapping, and an undergraduate computer software engineering from Wuhan University (Hubei, China).
Han is interviewed by Dr. Matt Rice, Past-President of CaGIS in June 2018.
Matt Rice: Tell us a little bit about yourself, and how you ended up as a Managing Consultant in Washington D.C.
Han Qin: Dr. Rice, thank you very much for this interview. As you know, I have bachelors degree in Software Engineering, master of science degree in Geographical Information System, and Ph.D. in Earth Systems and GeoInformation Science. My research focuses mainly on data quality of volunteered geographic information (VGI), and I have authored and co-authored 16 publications on related subjects. After graduation from George Mason University in Spring 2017, I joined a listed consulting company as Managing Consultant that focuses on using machine learning and advanced data analytic methods to develop frameworks and solve problems in different industries. I see my current position as a direct extension of my background in computer science, GIS, and spatial analysis.
Matt Rice: Your undergraduate degree from Wuhan is in computer software engineering. How did you end up in the GIS field?
Han Qin: I had some fundamental courses on GIScience when I was studying in Wuhan University. There was a joint program in GIS between Wuhan University and Eastern Michigan University, so I had the opportunity to join EMU with several other students. During my studies at EMU, I developed an interest in the applications of GIS in urban planning and health. I published a paper about out of hospital cardiac arrest patterns with my thesis advisor, and built a platform to visualize the analysis results and patterns. In addition, I realized public participant information could be a significant source of GIS data. I had the opportunity to join the Ph.D. program in GMU to focus on volunteered geographic information (VGI) and social media data, where I pursued this topic as a research extension of GIS and geocomputation. My technical skills and general interests in information science prompted my studies and research in the GIS field.
Matt Rice: Health applications in GIS are a big deal right now, and a clear growth area. From my perspective, your work with health applications and web mapping at Eastern Michigan is related to your work at GMU with accessibility and web-based geocrowdsourcing. Do you see these topics as a common thread? Do you work on health-related topics in your employment as a consultant?
Han Qin: Yes, health GIS is a big component in the GIS field. My studies at EMU and GMU are both about assisting public health officials and residents quickly locate and visualize risks for local communities. For example, to improve the survival rate of patients with out-of-hospital cardiac arrest (OHCA), I developed a web based application to visualize identified communities that have high risk for OHCA. These communities were targeted for CPR education outreach, priority placement of automated external defibrillators (AEDs), and other intervention activities. To reduce the risks of disabled individuals facing obstacles on their daily navigation routes, I developed a web-based testbed to collect and visualize transitory obstacles on sidewalks. The research and development combination approach is quite common in GIS, and a well-designed system framework can help people to better understand analysis results and can also improve information quality with iterative processing.
As a managing consultant, my role is a combination of data scientist and software engineer. I have joined some projects that focus on the financial side of healthcare and have developed machine learning models to help other consultants to detect bad debt and financial irregularities. Many of the data sources used in these projects contain location information, but the location information hasn’t been fully utilized to extend our analyses. I will continue to add elements of spatial analysis and GIS when our team is ready for this work.
Matt Rice: Based on your publications and Linkedin profile, your research contributions include quality assessment for crowdsourced data. While some authors suggest that geocrowdsourcing is a profound transformation in the discipline, others consider crowdsourced data to be unreliable. Can you briefly tell us about your approaches for quality assessment?
Han Qin: This is a good question. Researchers recognize that geocrowdsourced data has significantly changed the geospatial landscape, with popular web applications like OpenStreetMap, Wikimapia, and Flickr, providing sources of geographic information. Early publications on volunteered geographic information (VGI) by researchers such as Michael Goodchild, Sarah Elwood, and Mordechai Haklay, suggested that VGI will become a significant, low-cost source of data for GIS, if viable ways to assess quality could be developed to overcome this weakness. Haklay’s paper on quality assessment of OpenStreetMap data in the UK has been a persistent exemplar. One of the most comprehensive papers to come from the OSM quality assessment work is Girres and Touya’s “Quality Assessment of the French OpenStreetMap Dataset”, which looks not just at validation of position, but also at the many other aspects of data quality.
Researchers such as Matthew Zook and David Coleman look at the dynamics of organizations and people who contribute and use VGI. Scott Freundschuh, Michael Camponovo, and Giles Foody have done nice work on accuracy issues in VGI, particularly for attribute data and classification. In 2018, I think we have a much greater understanding of the “exaflood” of crowdsourced geographic information discussed in 2013 by Dan Sui, and we have ways of assessing quality using traditional GIS quality assessment methods published in the 1990s by researchers such as Howard Veregin, Gary Hunter, Steven Guptill, and Joel Morrison. There is a wealth of academic research being done on VGI and related topics. The researchers I name above have been influential in my own work, but there are hundreds of excellent theses, dissertations, and research articles being published every year.
In my own doctoral work at GMU, the limited study area and small contributor pool allowed for a moderated social quality assessment strategy, involving project moderators, who field-checked clusters of reports using web-based VGI moderation tools. Our research involved a great deal of public input and feedback, which led to iterative improvements in our quality assessment strategies. We developed and refined tools for measuring and characterizing nearly all of the facets of spatial data quality commonly used in GIS quality assessment, and actively worked with moderators to create weighting systems and criteria for evaluating quality. Work on these systems continues to the present day.
Matt Rice: George Mason University attracts some of the very brightest students from Chinese geospatial academic programs, who come and study in the US, and then continue, working in academia and industry. What general opportunities can a student gain by studying GIS in the United States?
Han Qin: The students who come over to the U.S. are typically at the very top of their academic programs in China, and are looking for opportunities to continue their education and then begin a career. In my case, the Department of Geography and Geoinformation Science at George Mason University had a strong reputation, so I applied to continue my education. While the professors and research funding at GMU attracts both domestic and international students, the location near Washington D.C. and the central role that the region plays in the geospatial world is an attractive factor. There are many well-known programs in GIS and cartography both inside the US and abroad. My advice is for students to find a program with a good fit to their interests and a good opportunity for employment after graduation.
Matt Rice: Academic institutions seem to be continually chasing new GIS and web technology in order to update coursework and reflect changes in the GIS industry. What strategic skills and courses would you recommend to younger students to stay ahead of the technology curve? What skills are in highest demand at present (Spring 2018)?
Han Qin: Some of the hottest technology areas right now are artificial Intelligence (AI), machine learning, big data, cloud computing, high performance computing, and autonomous logistics and transportation systems. You hear these terms everywhere now, both inside the GIS industry, where they are making an impact, and outside of the GIS industry. Some knowledge of these areas is useful, and in depth knowledge of at least one or two is important. Solid programming skills and advanced data analysis capabilities are now a required element for mid- and high-level positions in the new GIS industry. Programming languages such as Python and Java are used widely in GIS industry, and I would recommend that younger students who decide to choose GIS as major to be proficient in at least one programming language. If one programming language is chosen, python may be the best choice right now. I would also recommending being familiar with advanced data processing methods, especially for earth science data.
Dr. Rice: How does CaGIS, as an organization, help you stay current? What value does CaGIS have for you? What CaGIS papers have you read recently that you can recommend?
Han Qin: I know CaGIS (the organization) primarily from the reputation of CaGIS (it’s journal), which has a history stretching back to the 1970s. It is a top-tier journal in GIS field and currently has a very high impact factor relative to other geospatial journals. I read and cite many papers from CaGIS in my research, and have also published papers with my colleagues and professors in CaGIS. Four CaGIS papers that I have cited recently, and recommend to others, are:
- Camponovo, M. E., & Freundschuh, S. M. (2014). Assessing uncertainty in VGI for emergency response. Cartography and Geographic Information Science, 41(5), 440–455. [10.1080/15230406.2014.950332]
- Kraak, M.-J. (2011). Is There a Need for Neo-Cartography? Cartography and Geographic Information Science, 38(2), 73–78. [10.1559/1523040638273]
- Thies, J., & Smith, V. (2011). Transitions in Digital Map Production: An Industry Perspective. Cartography and Geographic Information Science, 38(3), 310–312. [10.1559/15230406382310]
- Tsou, M.-H., Yang, J.-A., Lusher, D., Han, S., Spitzberg, B., Gawron, J. M., … An, L. (2013). Mapping social activities and concepts with social media (Twitter) and web search engines (Yahoo and Bing): a case study in 2012 US Presidential Election. Cartography and Geographic Information Science, 40(4), 337–348. [10.1080/15230406.2013.799738]
The Camponovo and Freundschuh paper is worthwhile due to it’s inspection of the quality of attribute data, which in my opinion is just as important as quality assessment of position, but it is much less commonly studied. The other papers look broadly at the neo- geospatial movements, and important technology trends.
CaGIS (the professional organization) also organizes conferences such as AutoCarto, which has a long history of GIS and GIScience research, and provides organizational support for the US National Committee for the International Cartographic Association (ICA). Many of my classmates and faculty at GMU participated in the recent ICC2017 meeting in Washington D.C., which was hosted and organized by CaGIS. While the role of professional organizations does seem to be changing, CaGIS still maintains a pretty high profile for research in cartography and GIScience and I believe that will continue. CaGIS will need to evolve to maintain relevance to young GIScience and cartography students and professionals, who as a group are less likely to join professional organizations.
Matt Rice: You began as a relatively unknown student in Wuhan, and now you are a well-known member of the geospatial community in the United States. What career advice can you pass along to young, developing students who aspire for success?
Han Qin: Although I have gained some standing in the geospatial community, I am still at the beginning of my professional career. One bit of advice I can pass along would be for students to have a good vision in terms of what he/she would like to do in the future so that many possibilities remain open. Industry and academia are two possible destinations for students, but there are other rewarding career options for students in the non-profit community, NGOs, think tanks, and government. Each possible career destination has specific skills that are required. Some skills are common to all, such as an ability to write and communicate well. Other career destinations, such as industry, prioritize technology and skills such as programming. If students have an idea which career area that want to pursue, they should talk to practitioners in that area about the specific skills that are needed, while maintaining as many options as possible.
I would also encourage students to be open to new ideas, and experiment and test those ideas as early in their career as possible. A good general approach for research is to try different ways, fail fast, and try to fix it faster. My last bit of advice is to be grateful for help from advisors and colleagues. You can’t imagine how much a good research team will elevate you in your research and professional career.
Matt Rice: Thank you for talking today! How can CaGIS members get in touch with you if they have questions or comments?
Han Qin: If they are in the DC region if would be easy to meet in person. Otherwise, we can arrange a phone call or skype. The best email address for me is firstname.lastname@example.org.