Archive for May, 2015

My takeaways from AAG 2015

May 21st, 2015

The 2015 Annual Meeting of the Association of American Geographers (AAG) in Chicago is long gone – time for a summary of key lessons and notable ideas taken home from three high-energy conference days.

Choosing which sessions to attend, was the first major challenge, as there were over ninety (90!) parallel sessions scheduled in many time slots. I put my program together based on presentations by Ryerson colleagues and students (https://gis.blog.ryerson.ca/2015/04/17/ryerson-geographers-at-aag-2015/) and those given by colleagues and students of the Geothink project (http://geothink.ca/american-associaton-of-geographers-aag-2015-annual-meeting-geothink-program-guide/), as well as by looking through the presenter list and finding sessions sponsored by select AAG specialty groups (notably GIScience and Cartography). Abstracts for the presentations mentioned in this blog can be found via the “preliminary” conference program at http://meridian.aag.org/callforpapers/program/index.cfm?mtgID=60.

Upon arrival, I was impressed by the size and wealth of the industrial and transportation infrastructure in Chicago as well as the volume of the central business district, as seen from the airport train and when walking around in the downtown core.

aag-photo1 aag-photo3

My conference started on Wednesday, 22 April 2015, with Session 2186 “Cartography in and out of the Classroom: Current Educational Practices“. In a diverse set of presentations, Pontus Hennerdal from Stockholm University presented an experiment with a golf-like computer game played on a Mercator-projected world map to help children understand map projections. Pontus also referred to the issue of “world map continuity” using an animated film that is available on his homepage at http://www.su.se/profiles/poer5337-1.188256. In the second presentation, Jeff Howarth from Middlebury College assessed the relationship between spatial thinking skills of students and their ability to learn GIS. This research was motivated by an anonymous student comment about a perceived split of GIS classes into those students who “get it” vs. those who don’t. Jeff notes that spatial thinking along with skills in orientation, visualization, and a sense of direction sets students up for success in STEM (science, technology, engineering, math) courses, including GIS. Next was Cindy Brewer, Head of the Department of Geography at Penn State University, with an overview of additions and changes to the 2nd edition of her Esri Press book “Designing Better Maps”. The fourth presentation was given by David Fairbairn of Newcastle, Chair of the Commission on Education and Training of the International Cartographic Association. David examined the accreditation of cartography-related programs of study globally, and somewhat surprisingly, reported his conclusion that cartography may not be considered a profession and accreditation would bring more disadvantages (incl. management, liability, barriers to progress) than benefits to the discipline. Finally, Kenneth Field of Esri took the stage to discuss perceptions and misconceptions of cartography and the cartographer. These include the rejection of the “map police” when trained cartographers dare to criticize the “exploratory playful” maps created by some of today’s map-makers (see my post at http://gis.blog.ryerson.ca/2015/04/04/about-quick-service-mapping-and-lines-in-the-sand/).

A large part of the remainder of Wednesday was spent in a series of sessions on “Looking Backwards and Forwards in Participatory GIS“. Of particular note the presentations by Renee Sieber, professor of many things at McGill and leader of the Geothink SSHRC Partnership Grant (http://www.geothink.ca), and Mike McCall, senior researcher at Universidad Nacional Autonoma de Mexico. Renee spoke thought-provokingly, as usual, about “frictionless civic participation”. She observes how ever easier-to-use crowdsourcing tools are reducing government-citizen interactions to customer relationships, and participation is becoming a product being delivered efficiently, rather than a democratic process that engages citizens in a meaningful way. Mike spoke about the development of Participatory GIS (PGIS) in times of volunteered geographic information (VGI) and crowdsourcing, arguing to operationalize VGI within PGIS. The session also included a brief discussion among members of the audience and presenters about the need for base maps or imagery as a backdrop for PGIS – an interesting question, as my students and I are arguing that “seed contents” will help generate meaningful discussion, thus going even beyond including just a base map. Finally, two thoughts brought forward by Muki Haklay of University College London: Given the “GIS chauffeurs” of early-day PGIS projects, he asked whether we continue to need such facilitators in times of Renee Sieber’s frictionless participation? And, he observed that the power of a printed map brought to a community development meeting is still uncontestable. Muki’s extensive raw notes from the AAG conference can be found on his blog at https://povesham.wordpress.com/.

In the afternoon, I dropped in to Session 2478, which celebrated David Huff’s contribution to applied geography and business. My colleague Tony Hernandez chaired and co-organized the session, in which Tony Lea, Senior VP Research of Toronto-based Environics Analytics and instructor in our Master of Spatial Analysis (MSA) program, and other business geographers paid tribute to the Huff model for predicting consumers’ spatial behaviour (such as the probability of patronizing specific store locations). Members of the Huff family were also present to remember the man behind the model, who passed away in Summer 2014. A written tribute by Tony Lea can be found at http://www.environicsanalytics.ca/footer/news/2014/09/04/a-tribute-to-david-huff-the-man-and-the-model.

Also on my agenda was a trip to the AAG vendor expo, where I was pleased to see my book – “Multicriteria Decision Analysis in Geographic Information Science” – in the Springer booth!

aag-springer-books

Thursday, 23 April 2015, began with an 8am session on “Spatial Big Data and Everyday Life“. In a mixed bag of presentations, Till Straube of Goethe University in Frankfurt asked “Where is Big Data?”; Birmingham’s Agnieszka Leszczynski argued that online users are more concerned with controlling their personal location data than with how they are ultimately used; Kentucky’s Matt Wilson showed select examples from half a century of animated maps that span the boundary between data visualization and art; Monica Stephens of the University at Buffalo discussed the rural exclusions of crowdsourced big data and characterized Wikipedia articles about rural towns in the US as Mad Libs based on Census information; and finally, Edinburgh’s Chris Speed conducted an IoT self test, in which he examined the impact of an Internet-connected toilet paper holder on family dynamics…

The remainder of Thursday was devoted to CyberGIS and new directions in mapping. The panel on “Frontiers in CyberGIS Education” was very interesting in that many of the challenges reported in teaching CyberGIS really are persistent challenges in teaching plain-old GIS. For example, panelists Tim Nyerges, Wenwen Li, Patricia Carbajalas, Dan Goldberg, and Britta Ricker noted the difficulty of getting undergraduate students to take more than one or two consecutive GIS courses; the challenge of teaching advanced GIS concepts such as enterprise GIS and CyberGIS (which I understand to mean GIS-as-a-service); and the nature of Geography as a “discovery major”, i.e. a program that attracts advanced students who are struggling in their original subjects. One of the concluding comments from the CyberGIS panel was a call to develop interdisciplinary, data-centred program – ASU’s GIScience program was named as an example.

Next, I caught the first of two panels on “New Directions in Mapping“, organized by Stamen’s Alan McConchie, Britta Ricker of U Washington at Tacoma, and Kentucky’s Matt Zook. A panel consisting of representative of what I call the “quick-service mapping” industry (Google, Mapbox, MapZen, Stamen) talked about job qualifications and their firms’ relation to academic teaching and research. We heard that “Geography” has an antiquated connotation and sounds old-fashioned, that the firms use “geo” to avoid the complexities of “geography”, and that geography is considered a “niche” field. My hunch is that geography is perhaps rather too broad (and “geo” even broader), but along with Peter Johnson’s (U Waterloo) comment from the audience, I must also admit that you don’t need to be a geographer to make maps, just like you don’t have to be a mathematician to do some calculations. Tips for students interested in working for the quick-service mapping industry included to develop a portfolio, practice their problem-solving and other soft skills, and know how to use platforms such as GitHub (before learning to program). A telltale tweet summarizing the panel discussion:

Thursday evening provided an opportunity to practice some burger cartography. It was time for the “Iron Sheep” hackathon organized by the FloatingSheep collective of academic geographers. Teams of five were given a wild dataset of geolocated tweets and a short 90-or-so minute time frame to produce some cool & funny map(s) and win a trophy for the best or worst or inbetween product. It was interesting to see how a group of strangers new to the competition and with no clue about how to get started, would end up producing a wonderful map such as this :-)

aag-sheep-map2

My last day at AAG 2015, Friday, April 24, took off with a half-day technical workshop on “Let’s Talk About Your Geostack”. The four active participants got a tremendous amount of attention from instructor-consultant @EricTheise. Basically, I went from zero to 100 in terms of having PostgreSQL, PostGIS, Python, NodeJS, and TileMill installed and running on my laptop – catching up within four hours with the tools that some of my students have been talking about, and using, in the last couple of years!

In the afternoon, attention turned to OpenStreetMap (OSM), with a series of sessions organized by Muki Haklay, who argues that OSM warrants its own branch of research, OpenStreetMap Studies. I caught the second session which started with Salzburg’s Martin Loidl showing an approach in development to detect and correct attribute (tag) inconsistencies in OSM based on information contained in the OSM data set (intrinsic approach). Geothink co-investigator Peter Johnson of UWaterloo presented preliminary results of his study of OSM adoption (or lack thereof) by municipal government staff. In eight interviews with Canadian city staff, Peter did not find a single official use of OSM. Extensive discussions followed the set of four presentations, making for a highly informative session. One of the fundamental questions raised was whether OSM is distinct enough from other VGI and citizen science projects that it merits its own research approach. While typically considered one of the largest crowdmapping projects, it was noted that participation is “shallow” (Muki Haklay) with only 10k active users among 2 million registered users. Martin Loidl had noted that OSM is focused on geometry data, yet with a flat structure and no standards other than those agreed-upon via the OSM wiki. Alan McConchie added the caution that OSM contributions only make it onto the map if they are included in the “style” files used to render OSM data. Other issues raised by Alan included the privacy of contributors and questions about authority. For example, contributors should be aware of the visualization and statistics tools developed by Pascal Neis at http://neis-one.org/! We were reminded that Muki Haklay has developed a code of engagement for researchers studying OSM (read the documentation, experience actively contributing, explore the data, talk to the OSM community, publish open access, commit to knowledge transfer). Muki summarized the debate by suggesting that academics should act as “critical friends” vis-à-vis the OSM community and project. To reconcile “OSM Studies” with VGI, citizen science, and the participatory Geoweb, I’d refer to the typology of user contributions developed by Rinner & Fast (2014). In that paper, we do in fact single out OSM (along with Wikimapia) as a “crowd-mapping” application, yet within a continuum of related Geoweb applications.

Notes from #NepalQuake Mapping Sessions @RyersonU Geography

May 4th, 2015

This is a brief account of two “Mapping for Nepal” sessions at Ryerson University’s Department of Geography and Environmental Studies. In an earlier post found at http://gis.blog.ryerson.ca/2015/04/27/notes-for-nepalquake-mapping-sessions-ryersonu-geography/, I collected information on humanitarian mapping for these same sessions.

Mapathon @RyersonU, Geography & Spatial on Monday, 27 April 2015, 10am-2pm. 1(+1) prof, 2 undergrads, 3 MSAs, 1 PhD, 1 alumnus came together two days after the devastating earthquake to put missing roads, buildings, and villages in Nepal on the map using the Humanitarian OpenStreetMap Team’s (HOT) task manager. Thank you to MSA alumnus Kamal Paudel for initiating and co-organizing this and the following meetings.

hotosm-for-nepal_msa-lab_27april2015

Mapathon @RyersonU, Geography & Spatial on Sunday, 3 May 2015, 4pm-8pm. Our second Nepal mapathon brought together a total of 15 volunteers, including undergraduate BA in Geographic Analysis and graduate Master of Spatial Analysis (MSA) students along with MSA alumni, profs, and members of the Toronto-area GIS community. On this Sunday afternoon we focused on completing and correcting the road/track/path network and adding missing buildings to the map of Nepal’s most affected disaster zones. Photos via our tweets:

 

My observations and thoughts from co-organizing and leading these sessions, and participating in the HOT/OSM editing:

  • In addition to supporting the #EqResponseNp in a small way, the situation provided an invaluable learning opportunity for everyone involved. Most participants of our sessions had never contributed to OSM, and some did not even know of its existence, despite being Geography students or GIS professionals. After creating OSM accounts and reading up on the available OSM and Nepal-specific documentation, participants got to map hundreds of points, lines, or polygons within just a couple of hours.
  • The flat OSM data model – conflating all geometries and all feature types in the same file – together with unclear or inconsistent tagging instructions for features such as roads, tracks, and paths challenged our prior experience with GIS and geographic data. Students in particular were concerned about the fact that their edits would go live without “someone checking”.
  • While the HOT task manager and general workflow of choosing, locking, editing, and saving an area was a bit confusing at first, the ID editor used by most participants was found to be intuitive and was praised by GIS industry staff as “slick”.
  • The most recent HOT tasks were marked as not suitable for beginners after discussions among the OSM community about poor-quality contributions, leaving few options for (self-identified) beginners. It was most interesting to skim over the preceding discussion on the HOT chat and mailing list, e.g. reading a question about “who we let in”. I am not sure how the proponent would define “we” in a crowd-mapping project such as OSM.
  • There was a related Twitter #geowebchat on humanitarian mapping for Nepal: “How can we make sure newbies contribute productively?”, on Tuesday, 5 May 2015 (see transcript at http://mappingmashups.net/2015/05/05/geowebchat-transcript-5-may-2015-how-can-newbies-contribute-productively-to-humanitarian-mapping/).
  • The HOT tasks designated for more experienced contributors allowed to add post-disaster imagery as a custom background. I was not able to discern whether buildings were destroyed or where helicopters could land to reach remote villages, but I noticed numerous buildings (roofs) that were not included in the standard Bing imagery and therefore missing from OSM.
  • The GIS professionals mentioned above included two analysts with a major GIS vendor, two GIS analysts with different regional conservation authorities, a GIS analyst with a major retail chain, and at least one GIS analyst with a municipal planning department (apologies for lack of exact job titles here). The fact that these, along with our Geography students, had mostly not been exposed to OSM is a concern, which however can be easily addressed by small changes in our curricula or extra-curricular initiatives. I am however a bit concerned as to whether the OSM community will be open to collaborating with the #GIStribe.
  • With reference to the #geowebchat, I’d posit that newbie != newbie. Geographers can contribute a host of expertise around interpreting features on the ground, even if they have “never mapped” (in the OSM sense of “mapping”). Trained GIS experts understand how feature on the ground translate into data items and cannot be considered newbies either. In addition, face-to-face instructions by, and discussion with, experienced OSM contributors would certainly help to achieve a higher efficiency and quality of OSM contributions. In this sense, I am hoping that we will have more crowd-mapping sessions @RyersonU Geography, for Nepal and beyond.

MSA Poster Day 2015

May 1st, 2015

composite-photo_msa-poster-day2015

On Wednesday, 29 April 2015, 22 Master of Spatial Analysis (MSA) students presented their research ideas/plans for their major research papers or theses to the Department. The students’ posters represented a mind-blowing diversity of research topics, with the important common denominator of using spatial analysis concepts and techniques. Techniques ranged from self-organizing maps to location-allocation, risk terrain modeling, cluster analysis, and various forms of regression analysis. Proposed tools included ArcGIS, MapInfo, QGIS/PySAL, and R. Finally, the fields of application span sustainable development to informal caregiving, housing deprivation, food access, regional transit planning, road salt usage, traffic injuries, ground-penetrating radar, urban heat island, and book retailing and mall tenant mix. Enjoy studying the following list of all poster titles!

msa-poster-day2015_working-titles

The MSA Best Poster Award 2015 went to Daniel Liadsky for the poster on “Neighbourhood Effects on Fruit and Vegetable Consumption in the Toronto CMA” shown below (click to enlarge). Congratulations, Daniel!!

daniels-winning-poster