Computers, Society, and Nature

Beginning reading with no prior awareness of cyberinfrastructures, Yang et al.’s article on Geospatial Cyberinfrastructures was pretty overwhelming. Yang et al. do such an incredible job of condensing huge amounts of information in a way that is fairly easy to follow (despite the multitude of acronyms) that admittedly, I don’t even know where to start in tackling it. What I found most interesting was the outline of the development of the semantic web and the data life cycle.

Throughout all readings in this course so far there has been mention of semantic differences in data, and the need to “facilitate the automatic identification, utilization, and integration of datasets into operational systems,” (272). With GCIs encompassing data from a huge array of different sources and different users (the Virtual Organisations are also really neat), the development of Web 3.0 is incredibly pressing in order to make sense of all this data and ensure interoperability.

I also really liked the section on Supporting the life cycle from data to knowledge. It is important to note that data is not information is not knowledge—it must be processed and synthesised in order to achieve a greater understanding of what the data represents.

Readings like Yang et al. really send home the point that this field is overarching and is growing at an incredible rate, and it’s really exciting to watch.

-sidewalk ballet

Chaowei Yang et al. very ambitiously discuss the development of geospatial cyberinfrastructure, including some of the challenges confronting this process. One of the aspects I found most interesting was the potential for increasing amounts of error being introduced as more data are generated by an ever growing number of users. The facilitation of a system, which can “collect, archive, share, analyze, visualize, and simulate data, information, and knowledge” increases the accessibility of data to a much wider array of people. While this is beneficial in terms of promoting research, it also, however, allows for a great deal of uncertainty to be introduced as there are are no clear standards for communicating this inherent component of data. Users not familiar with this notion – who are likely also those increasingly gaining access to this infrastructure – may further this problem.

Since the quality of this data may be questionable, ClimateNYC equates the development of GCIs to black boxes, and I think this has severe implications for the future of GIS. Madskiier_JWong, conversely, argues that scientists have much to gain from being able to easily share data with people in other fields, but I would be cautious with this. I am not questioning the notion that sharing data facilitates the production of knowledge. I am, however, concerned that if error and uncertainty are significantly present and not well-communicated, it can lead to severe divides and unnecessary arguments within fields of study. We all know how easily maps and data can be manipulated, for example, to convince a viewer of a point of view, so perhaps issues such as communicating error need to be better addressed as cyberinfrastructures are developed. From this, perhaps data will not only by more freely available, but it will also be more reliable.

– jeremy

I can relate to Andrew’s comment about Montrealers accepting a false north, which I find very interesting. In Vancouver, everyone knows that the local mountains are north, however, they aren’t actually. Despite this, it’s close enough for the purposes of traveling around the city and this is an important aspect of mental maps. By no means do I intend to flare up the ‘what is a mountain’ debate, but if a person’s mental map incorrectly associates a geographic feature with a compass reading in order to improve their navigational abilities, what does that say about the accuracy of their mental map?

Perhaps, as Tversky et al. illustrate, this notion reflects upon the nature of how we schematize, a process which accepts a loss of detail to “allow for efficient memory storage, rapid inference-making and integration from multiple media.” On the other hand, there may be more to this issue, as our ability to incorporate an individual’s cognitive map into a GIS is another problem that arises. How can we display and compare the landmarks, nodes, paths, etc. of cognitive maps when they are all, for example, represented at varying scales? To go back a step, how can we even be sure that the process of drawing a mental map isn’t completely fraught with error? These ideas relate to the varying ontologies that exist and trying to reconcile the differences between them, which – as we all know – is an extremely complicated task.

On a somewhat related note, everyone who is a map lover/artist (which I’m sure all of you are) should check out:

http://spacingmontreal.ca/2012/02/14/attention-all-map-lovers-spacings-creative-mapping-contest/

– jeremy

Tversky et al.’s article on spatial cognition categorizes our understanding of space into three inter-related categories: the outside, navigable world, the space around our bodies, and the space consisting of our bodies. At an individual level, it is axiomatically clear that we do not conceive of ourselves or our surroundings as a 2 dimensional space. It is also interesting to examine how detailed spatial cognition of our body can enable better movement physically.

                There are clear examples of the disconnect between our cognition of our body and representing it in a two dimensional form. Ask any person of the difficulty they had in trying to draw the hand of a person holding an object without a visual reference. This extends beyond the technical expertise of being able to draw the fingers in perspective, since it is difficult to simply imagine how the fingers position themselves in relation to each in two dimensions while remaining ‘realistic’ to our minds. It can be inferred that we cognize space of our bodies and our nearby surroundings in three dimensions (and thus rich in detail) because we have so much experience in these local matters. Does this raise an implication then of our 2-D conceptualizations of the ‘far outside world’ as being relatively poor in detail? By extension, do our two-dimensional maps reinforce a poorly performing cognition of space?

                On another point, I would argue that athletes have a heightened awareness of the space of their surroundings and of their body. The important functional relevance that the authors identified is likely to be stronger and more varied with athletes, since they have a more frequent and wider range of motion. Speaking from personal experience (I do martial arts), being able to mentally picture where my feet will land, where my arms must move, and how much space I require has helped me execute complex moves. Being able to orient yourself mentally is key in sports such as figure skating, and allows people to move better. Geography is in your body!

–  Madskiier_JWong

The Space of Navigation
Effects of the different elements such as frame of reference (the hierarchical encoding example) were all very interesting. What would be interesting to know though if certain of these elements are more dominant at certain scales. If we could identify that everyone thought at the provincial scale in the hierarchical example’s way, then we could develop a framework on how to communicate and teach people geographic information.

One problem I had with the article though, was that it was very much from a cog. sci. (or similar) approach, rather than a geography/geospatial view. What effect do the statements on our mental representations of space have on how we should do things in the future? Are the different biases observed in the different studies necessarily a bad thing? Should we shape our data output, such as local maps, to meet these methods of storing information in the brain, or should we stick to something that is as ‘accurate’ to being a representation of the ground surface as possible? At the very least, more knowledge on how humans perceive their environment should help us determine what we are doing wrong when presenting geographic information, an area that would have been interesting to see in this article.

-Peck

Sometimes I get the feeling that people view GCI as a single entity/unit. It isn’t. It is, too misquote Yang, a ‘system of systems’. This makes GCIs seem very flexible and able to do anything – as improvements in any of the systems will advance the GCI forward. However, the challenges are still immense. GCIs may be used as a way of sharing, analysing, storing data etc., but they are still limited by the rules we have, such as the semantic framework for sharing of data. However, this may make a GCI start to look a little cumbersome, at least when you are viewing it as a single entity. This is something I am not sure of – whether GCIs are adaptive to changing environments such as ontologies.

Future changes were also interesting. Virtual Organisations could start becoming more permanent as enabling technologies decrease physical limitations. GCIs though are still relatively closed environments though, and may benefit from more open sharing. This is what is expected to happen with the shift to ‘geospatial cloud computing’. However, the article doesn’t really define geospatial cloud computing – what’s the difference? Aren’t we already partly there?

-Peck

Cyberinfrastructures are a little outside of my comfort zone so I’m not sure I completely understand how they work, but I like how Yang places it in real world context.

It seems like the purpose of a cyberinfrastructure is to get more done, in less time, in more than one place. Great concept right? I agree that this sounds great, but I agree with some of the criticisms that Yang brings up. What if the type of research that you are performing is illegal in some countries, but not in others? The other day at work I heard some pretty well-known DJ’s talking about piracy. The first DJ was concerned that he would be fined for having pirated material on his computer. The other DJ brushed it off and stated that he would simply move his server to “somewhere like Thailand or Botswana” where it isn’t illegal. What happens if a computer that stores some of that information where it is illegal? Does the person responsible for the content get charged back home even a said law does not exist in that country, or does he have to be extradited first? Is the country where it isn’t illegal have to anything at all? I personally don’t think so, but it brings up some valid concerns. Will an international cyber-law enforcement come to fruition at some point soon? INTERINTERPOL (International Internet Criminal Police Organization) maybe.

On another note, but still somewhat related to international frontiers, I like how Yang continues the debate regarding ontologies. What becomes an official ontological language? How are we supposed to accommodate foreign languages in the web 3.0? Will many key words and ontological definitions from several languages all be linked to one parcel of data labelled “10010101” for example?

I don’t know much about programming or cyberinfrastructures but I get the impression that the issues that Yang brings to the surface are increasingly important. I don’t think that the solutions are impossible, I’m simply curious to know how the laws form around cloud servers and cyberinfrastructures.

Andrew

Last week I stepped out of the St. Joseph metro station at a stop I don’t normally exit. I gathered my bearings by assessing what surrounded me and began to walk to my final destination. After a second, I stopped to check the map on phone (a reaction I now have—my iPhone apparently rules my life) and realized I was walking in the wrong direction. “Thank iPhone!” I thought to myself, and plugged my headphones in and began the trek towards Saint Laurent. After 15 minutes I stopped and realized I had been walking in the wrong direction. I was right(or left??) all along!

I checked my phone and realized that with the new update, my phone now orients the Google maps in relation to where I’m pointing. Assuming that the map on my phone pointed north, when I was actually facing south, I ended up guiding myself in the wrong direction.

I really enjoyed the article about spatial cognition. It’s fascinating how we instinctively orient our world “North.” I believe, though, that this instinct is not actually instinct. I believe that it’s a product of our upbringing. Montrealer North is never actually north. Collectively, Montrealers accept this false north; we are however aware that our conception of North is in reality more north–west. The human brain is extremely elastic! It has the power to re-orientate itself after wearing “inverted goggles” (perceptual adaptation) and has the power to re-wire language and thought after a stroke (http://www.radiolab.org/2010/aug/09/). I imagine that no matter the convention, the brain has the ability to adapt to such changes.

If only it could turn off the compass on my iPhone…

Andrew

The concept of geospatial cyberinfrastructures seems to draw from all aspects of GIS: where it came from and where it is going.  The Yang et al article was a very thorough introduction to GCIs and their uses and limitations.  It also seemed to incorporate much advancement in GIS that we have read about over the last few weeks, and presented an opportunity to visualize how all these technologies may work together, their strengths, and their weaknesses.

This topic seems very current, as you hear more and more today of cloud computing, information being held and hosted on the world wide web, and so on.  It emphasized even more the need for shared knowledge and languages, good metadata, and fast processing.  The wealth of possibilities for GCI, as well as the inclusion of domains where it is already useful, was an interesting aspect of this article as well.

What I found to be the most important limitation, that seemed to run through not only most domains and uses mentioned in this article, but recalled as well many of the other tools we have discussed, was the difficulty in dealing with immense amounts of constantly flowing, real-time data.  This issue in itself seems to incorporate many of the needs mentioned above, and is really the crux of what, in my reading, GCIs are about: the ability to successfully, quickly, and knowledgeably share information, questions, and expertise, analyze and upload data, and more.  However, I agree with Madskiier in their suggestion that GCIs are very global by nature, and thus would presume that, through adequate cooperation, this could be a task undertaken by many, as opposed to just a few.

As a student, I found this prospect incredibly interesting, and it drew my mind to the countless hours spent searching for geospatial data for simple research projects.  While students perhaps a fewer connections than established scientists, we also have the power of McGill behind us–and yet finding (good) data is still tremendously time consuming and challenging in many cases.  The idea of a large infrastructure supporting free global geospatial data is quite appealing, and something I hope to see come to fruition.

–sah

Mapping as cartographer James Cook knew it is no more, but yet still fully present. Confused? Let me explain. In their paper entitled “Research Challenges in Geovisualization”, Maceachren and Kraak state that maps of the past were designed to be not only a visual aid to navigation, but also to be a database of spatial information (pg 3) such as place names, bays, coves, cities and related information such as their position (absolute and relative) and the distance between them and neighbouring features, to name a few.

Today, mapping is still very much a graphical aid to data visualization, but unlike in the past, maps are not just a static database of places and locations.  Today’s Geo-Web 2.0 and data visualizations platforms like Google Earth can do so much more than display local data, they have the whole internet as a database (pg 3) and can draw on information located in servers and on subjects all over the world with a single URL or script.

This means that the possibilities of today’s cartography are endless, we are not even limited to two, or even three dimensions any more.  Visual Earths (a 3D surface) can display 2D map data, 3D details such as buildings and topography and most importantly, changes over time with time sequenced raster playback.  In fact, display of change over time in Virtual Earths, rudimentary as it is, is still as good as, if not better than, many of the solutions proposed by GIScientists for use in traditional GIS analysis.

In conclusion, mapping today is just as useful as traditional maps, but more so.  We may not all be Cook, but we have access to a very powerful set of geo-visualization and analysis tools today that can only spell great things for our future and the future of GIS.

-rsmithlal

One strategy suggested by Elwood is that “enriching data with information will help the user assess heterogeneity” although to me this does not seem to assist with solving or managing the problem of data heterogeneity. It has been mentioned in class that data is not typically well documented in GIS and that one way to provide information about it is to create metadata. In the world of the internet where massive amounts of data now have spatial references, and in many cases change rapidly, it is not practical to try to provide more information about every piece of data to try to reduce heterogeneity and standardize the data. Since additional data about data would provide even more information to sift through, this also seems rather counterintuitive. While I recognize there is heterogeneity in data, I do not understand the use of assessing heterogeneity but see instead much more use in actually working with heterogeneous data and focusing more time and effort in promoting methods to do this such as in particular contexts as mentioned by Elwood.

-Outdoor Addict