Acquiring GIS Skills Through Formal Means
Dr. James Linders, Professor,
Department of Computing and Information Sciences,
University of Guelph, Guelph, Ontario, Canada.
| Abstract Of The Paper & The Profile of The Speaker | Speaker Index | Paper Title Index |
Introduction
The rapid evolution of GIS technology can, to a large degree, be directly attributed to the powerful tools created through advances in information technology over the past three decades. The power of GIS tools has closely followed the capabilities provided by information technology and have been particularly instrumental in augmenting the computational complexity of GIS processing. The development of GIS, in turn, has been motivated by the importance of landmass data to a large number of activities within society ranging from traditional mapping to complex modelling operations for earth science disciplines.
One of the many issues faced by users of GIS systems is concerned with how to acquire the necessary skill set to effectively exploit the full potential of a GIS. In order to properly address this issue, it is necessary to properly understand the difference between training within a discipline and comprehension arising from understanding the underlying concepts and notions within the discipline. At the same time it is very important to attempt to quantify exactly what GIS is as a discipline, if in fact it is truly a scientific discipline, as opposed to a methodology for handling land mass or georeferenced data. It is particularly important to differentiate between what constitutes GIS and what information technology specifically provides in support of GIS. Such issues will not be addressed in this paper, although notions dealing with what constitutes learning for GIS and specifically how this might be handled is addressed through a specific curriculum.
How We Learn
Learning theorists tell us that we learn through three main techniques, namely: a) through repetition (rote) whereby some mental and physical activities are repeated often enough that they essentially become second nature to the individual; b) through experience whereby we acquire knowledge through our sensory perceptions and last, but not least, by c) discovery.
Traditional pedagogy has built an educational system in which there is an attempt to convey basic concepts and notions to a student by having the student undertake a number of often repetitive exercises until they become ostensibly second nature. Unfortunately, this may not lead to a deep understanding of the ideas involved, although the student is able to repeat or demonstrate some skills to the satisfaction of an examiner. Clearly there has been some level of learning involved in this process, although the learner may be lacking in comprehension or fulfillment. Most people equate this level of learning with training, in that the prerequisite skills to function in some capacity have been instilled in the learner so that he is able to perform satisfactorily at a given level, especially where there is little or no need for innovative or creative thought.
It is often said that experience is one of the greatest teachers, primarily because of the impression that experience usually makes on our sensory organs and nervous system. If the experience has been painful or stimulating then the impression on the nervous system is sufficiently lasting that we make this part of our real world experience repertoire. This type of learning paradigm is now being sought by institutions of higher learning as they seek to create and maintain "learner-centred" environments in which students are challenged to learn by doing without the confines of a restrictive environment. One argument for this type of learning is that better students can advance quickly at their natural pace, while other students may be challenged in different ways to acquire the necessary skills for a given discipline. Experience may be a great teacher, however it unfortunately seems to work differently in all of us.
The third form of learning is through discovery which is really quite different from the other two forms of learning. Discovery accounts for a very small fraction of our learning but it is very important within a scientific and technical perspective. Clearly, abstraction and inference play a large part in this type of learning, although chance can also be a significant factor. Unfortunately one cannot easily teach methods of discovery, although it is possible to foster an environment where this is either developed or occurs.
It is interesting to note that computer scientists have been attempting to emulate learning by rote and experience through the use of powerful software tools such as neural networks [Hinton]. This is achieved by creating information structures which closely parallel the neural structures found in the brain of a human. Remarkable success has been reported for many applications [Swingler] using these techniques.
Motivation for a GIS Curriculum
Within many institutions of higher learning there have been ongoing discussions of what GIS is and how it should be taught. These issues have not always been easy to discuss without reference to a specific context. However, as GIS continues to mature and the potential of GIS is more demonstrable and better appreciated we are perhaps more able to provide convincing answers.
Clearly GIS is an ubiquitous technology which now impacts many different disciplines from the conventional to the unconventional. These range from traditional topographical mapping to the use of electronic charts in pleasure craft; from demographics to data mining techniques for trend analysis. Since GIS is found in many diverse disciplines, it is perceived by students and career professionals to be desirable for a resume. However, because GIS is so ubiquitous it can also act as an agent for integrating components from different disciplines. Because of the diversity of use for GIS, not unlike computing, GIS is seen to provide the means by which we can undertake tasks that in the past have been too daunting to even contemplate. As GIS matures and follows the role of defining itself through standards and models, it becomes an ideal candidate to be structured within a general engineering or science curriculum.
Defining a GIS Curriculum
There are many different views as to what constitutes a GIS. Most would provide a definition of a GIS which combines the notions of a collection of hardware and software which are designed specifically to deal with spatial, land related data problems. However, practitioners have perceptions of GIS ranging from hardware/software systems to methodologies or even a set of toolkits for working with georeferenced data. The perception of what GIS really is will to a large degree define how we approach GIS and the extent to which we can effectively use the technology. Hence when considering GIS, we should be thinking in terms of the following:
- nature of the problem domain - namely georeferenced data
- characteristic functionality which is unique to GIS
- conceptual models for handling georeferenced data
- the principles and techniques used for GIS thinking
- relationships to other science disciplines
The above are what uniquely define GIS to be what it is and also which distinguish it from just simply being an application of information technology. When the information technology components are taken out of our thinking about GIS, then we are left with that which really defines GIS.
The challenge, of course, is to structure a curriculum for GIS which is meaningful to the full spectrum of GIS users. In defining the GIS users we would have to include at least the following categories:
GIS professionals: those whose primary professional activities are associated with structuring, maintaining and exploiting GIS environments for solving real world georeferencing problems. Such professionals may have come from other disciplines including computing science, engineering, planning, economics etc. Their skills are essentially those of a problem solver, analyst, project manager etc.
GIS technologists: those whose primary role is to use GIS to achieve specific functions required within their job mandate. Their activities may range from production work to well defined projects. A GIS technologist has the skills to work without much guidance and is expected to perform or carry out traditional GIS activities.
casual users: anyone who has an interest in GIS and in some way or other finds practical applications of the technology for realizing some part of their mandate. Such users may simply browse georeferenced data or they could be interested in retrieving specific georeferenced knowledge.
Conspicuous by it's absence is any mention of GIS systems designers, system managers, GIS programmers etc. since these type of disciplines tend to fall into the category of computer system technology. This does not mean that they are not required to understand GIS, but rather that their direction for GIS development and support will likely come from GIS professionals.
Our real task is to define the curriculum needs for those requiring GIS skills. At the same time it is important to differentiate between curriculum, which is really a collection of courses needed to prepare students in the area of GIS, and syllabus which is an outline of the topics to be covered within a given course or curriculum. Regrettably we often use the terms interchangeably, especially when there is a real need to differentiate between curriculum and syllabus.
From the perspective of curriculum, it is clear that there are many notions which need to be mastered before embarking on the task of learning GIS. Some of these ideas are now covered in the newer texts on GIS [ ], however these are not intended to be exclusive. For those interested in developing a GIS curriculum, it is necessary to carefully consider how to structure a curriculum, as with most any discipline of science, in which a stream of courses builds up the required knowledge base in a logical and systematic manner. Since we may not have the luxury of structuring a full suite of courses to define a GIS curriculum, it is necessary to achieve the same end by carefully selecting topics for inclusion within a single or restricted set of GIS courses. Usually this involves taking specific topics from a GIS syllabus and structuring a complete course around specific notions in order to develop a good foundation for further study.
For the three categories of GIS users defined in this section, it is more than likely that their GIS knowledge base will be developed within another curriculum such as geography, geomatics, engineering etc. Hence there is a need for GIS courses which address GIS learning in terms of concepts and practical applications. Such courses are not intended to be a substitute for vendor training, but are expected to make vendor training as natural as learning another word processor, once you have mastered your first word processor.
It is the thesis of this author that a GIS course can and should be designed to facilitate the learning of the GIS concepts with the deep understanding of GIS coming through the use of properly laid out projects as lab assignments. At the same time there is a need for a basic set of skills that can usually be found within a traditional science curriculum, which should be sufficient to allow the student to progress rapidly and gain the confidence necessary for independent study or work in GIS. Essentially most disciplines do not have much room for more course work, hence any attempt to incorporate GIS must be well thought out and structured to optimize the effort involved.
Later on in the paper the ideas for a GIS course are illustrated by means of a practical example where these ideas are put into practice. The curriculum was actually developed for a specific group, namely environmental engineers and others who need to use GIS for research or practical projects.
Criteria for Assessing a Good Curriculum
With the introduction of the computer into the learning environment, the world of pedagogy is undergoing dramatic change. Even though we have to rethink many of our prized notions concerning teaching and learning, there is still much which remains invariant with changes in technology. The design of a good curriculum must in the last analysis be guided by sound principles.
Some principles for evaluation of a good curriculum are:
relevance - the relationship of the subject matter to current thinking and practices. All too often, our curricula lag behind the needs and practices of the worlds of industry, commerce and science, simply because of the difficulty by which feedback can take place from the workplace to the classroom. Ask any group of students about what they thought of a given course and all too often they will depreciate a course simply because they do not see its relevance in terms of career objectives or position within a structured curriculum.
comprehension - which simply states that the student must be able to grasp the basic concepts and notions of the subject matter within a natural and logical sequence. Comprehension normally requires a solid foundation of a prerequisite knowledge base, else further learning is impeded. However without comprehension there can be no satisfaction resulting from the learning process.
structured - A curriculum must be structured so that there is a logical progression of ideas to augment the learning process. In this way the student is challenged to learn the material and acquire new knowledge throughout the curriculum.
dynamic and adaptive - In our present day, where change is the order of the day, it is imperative that any curriculum be structured in such a manner that it is conducive to accepting new ideas and embedding them within a dynamic curriculum. This is perhaps most obvious in such areas as computing and information science where new concepts, ideas, techniques etc. are evolving at such as rapid pace that course content is never static.
presentation - The curriculum must lend itself to a natural form of presentation to facilitate comprehension. Perhaps the most difficult challenge to any course designer is how to organize and present the material so that it is both interesting and challenging. Some of this is part of the inspiration of teaching, however the success of a course can to a large degree be attributed to how well it is presented. This in turn means that there must be logic to the syllabus which is both attractive and meaningful to the student.
evaluation - In order to be able to judge the results delivered from a curriculum there must be effective and meaningful means of evaluation. This all to important aspect is usually omitted in curriculum design as we revert to the traditional and accepted means of testing such as class tests, examinations, projects etc. A proper evaluation mechanism is important to the student in that it establishes milestones for the student for measuring progress and establishing confidence in the discipline.
focus - Any curriculum worthy of consideration must have a clear focus with well defined objectives for both the student and the instructor. Perhaps this is even more important, for those disciplines such as GIS, which are heavily influenced by technology. It is far too easy to lose sight of one objectives and be consumed by the intricacies and fascination of a given technology.
The above criteria were very much in mind when considering the design of the practical GIS course for professionals at the university level.
GIS Course Development
The design of a single course which would meet the perceived needs for GIS by professionals was based on a number of considerations. The primary motivation was to provide a solid curriculum within a single course which would prepare the learner to use GIS effectively for real and significant applications. The sufficient exposure to the topics in the syllabus allows the student to seek more detailed knowledge if and when required, however the greater need was to develop sufficient confidence through practical examples, so that the student would be able to use the technology within any GIS environment with ease.
Some of the principles that were used in the design of the course included:
The syllabus consists of a number of important topics reflecting concepts within GIS. Concepts are introduced in class and reinforced with practical lab exercises synchronized with the classroom instruction.
The syllabus is structured so that topics are introduced in a logical sequence and build on previous material. The intent was to avoid the introduction of isolated topics.
Every attempt was made to encourage self directed learning and study. This was almost essential because of the amount of material that had to be covered. Those students who feel motivated to learn more are provided with the tools to pursue further study on any topic. However, the course provides sufficient coverage to permit a good level of comprehension of the concept or topic involved.
A major design criteria of the course was to require knowledge acquisition from outside sources using the Internet via the World Wide Web (WWW), FTP etc. At the same time, there was sufficient motivation to the student to undertake this type of learning activity because of the perceived relevance of using the WWW as a knowledge source for this and other courses. (For many students, surfing the Net was relatively new and sufficiently alluring that it did not take much encouragement).
Each assignment encourages a project perspective. This means that the assignments (projects) must be practical, relevant and interesting but at the same time challenging and demanding.
It is important to challenge motivated student through depth of study. Essentially this meant leaving sufficient room in each assignment for student enhancement of the project.
It was deemed essential to build as much variability into each assignment as possible so as to encourage individual work. For example, each student should be given some unique data sets to work with.
Projects were selected which represented practical real world scenarios and which students might wish to add to a portfolio, especially when entering a competitive job market. Students should feel encouraged to be proud of their achievements.
The course has a well defined schedule and there is an attempt to maintain the schedule. Each lab attempts to focus on a single concept. The labs must be completed within a normal lab cycle, i.e. before the next lab is scheduled.
Tutorial support is provided within a scheduled lab environment, however students are encouraged to use the system help files as much as possible. Students also have access to the labs during off peak periods.
The coverage of the syllabus is sufficient to fully appreciate the concept or notion involved. There is sufficient background to permit the motivated student to explore the topic further. Suggested further reading topics as well as questions raised in class tend to encourage this aspect, especially for those students who perceive a niche which they would like to develop for their marketable skills.
Every attempt is made to seek technology independence in course development. This means that little or no time is spent on the details of the system being used. This does however imply basic computer literacy skills operating within a windows environment.
Considerable time and thought was given to the design and development of the course, but it was always expected that there would need to be continual refinements to address shortcomings and new ideas.
Topics Covered in GIS Course
The syllabus of the course is based on a number of topics which are intended to prepare the student for extensive use of GIS as well as to encourage further independent study. The following major topics are addressed:
- introduction to GIS: land related data concepts, nomenclature, application domains, historical development of GIS
- georeferencing: recording and maintaining position on the earth's surface, projection systems, transformations, GPS technology
- computer mapping: digital representation of map data, data types and representation, map objects, symbology, map generalization
- spatial data processing: storage and management of spatial data structures, polygon data management, buffering, spatial algorithms, object libraries
- database processing: attribute and spatial data query processing
- digital imagery: remote sensing, scanner technology, image processing tools
- thematics and application areas: cadastral systems, resource data management, municipal information systems
- demographics and atlas preparation
- standards for the geomatics industry: data interchange, geospatial processing models
- data modelling for geomatics applications
- applications development at local project, enterprise (Intranet) and distributed (Internet) levels
- future directions of industry
Deployment of Course
The course has been in place for quite some time, however was somewhat constrained in that there were insufficient resources and system capabilities to provide the full range of topics that such a course should offer. Traditionally the course has been offered in the Engineering department for Environmental Engineering majors. In the past there were only a limited number of PC ArcInfo licenses and hence students had to work in groups. As with such pedagogical constructs, usually only a few students really end up doing the work and hence these are the students who gain the most from the course. It was for this reason that it was decided to revamp the course using new technology and the many other considerations identified earlier in the paper.
The course consisted of approximately 55 students, most of which were at the third year level in Environmental Engineering. The GIS package used was a comprehensive distributed GIS based on client-server architecture and objects, namely GeoMap( by GEOREF Systems Ltd. of Waterloo, ON Canada. The package is in use within a municipal environment (Cities of Cambridge and Guelph ON, Canada). Each student was required to do their own work, since the environment allowed all students to have their own workstation (133 MHz PC).
The course was originally scheduled to be given as 3 hour per week for a 12 week period, however it was deemed desirable to change this so that alternate weeks the students would use a 2 hour lecture period as a structured lab. This was necessary because of the difficulty in rescheduling the course as a 3 hour lecture and 2 hour lab, which is the ideal setup.
Although it was originally expected to give each student a separate dataset (e.g. map file), it was found that this was too difficult to administer because of the copyright requirements on the data. However, as much flexibility as possible was put into the assignments to encourage independent work.
Six rather extensive assignments were put into the course, each of which was as natural as possible and would represent a typical situation for GIS. In summary the nature of each assignment and objectives were as follows:
GIS data browsing and digital map display. The purpose of this assignment was to acquaint the student with the types of information that are found in a digital topographic database. The student was expected to explore the types of features depicted by the various graphics by picking the graphic entity and inquiring of its attributes, both spatial and textual. An area from a large seamless topographic database was selected and prepared for drawing on a large laser printer or plotter. Note that there was very little time spent in teaching students the details of the underlying GIS package, except for the essentials. Details about any specific function can be determined through the use of the online help function(s).
Cartographic editing . The purpose of this assignment was to demonstrate a working knowledge of cartographic editing according to specific rules. The assignment involved producing a large scale map of the campus, using the digital topographic database enhanced with whatever information the student wanted to add. Annotation and symbology were expected to be entered as may be appropriate for the map theme. Extraneous information was removed and in some cases new features were added by those students who wanted to extend their assignment. The output was a very high quality map prepared to publication standards using the standard Windows printing capabilities.
Polygon data management. This assignment was designed specifically to build knowledge in buffering, polygon overlay and spatial querying. An arbitrary waterbody (e.g. a pond) was created within the topographic database and the student was required to identify the number of soils polygons of various types (determined by attributes) which were (i) contained completely within the pond, (ii) contained at least partially within the pond, and (iii) intersected the edges (shoreline) of the pond. Computations, such as those required for areas and perimeters, were also required. There are many possibilities for variability, depending on the number of polygon classes available with the system. For this assignment only soils and a drainage data were considered although land use could have been another possibility.
Data query. The students were taught the rudiments of relational databases and SQL query in class. This type of assignment was expected to instill in them some understanding of demographics and simple database processing. Students learned the power of relational database queries when they are used with multiple tables and/or combined with spatial constraints.
Thematic shading. Thematic shading involved searching the database for specific spatial entities which fall within well defined classes and then shading (highlighting, colouring) the features on a map. Polygons in the database are shaded and depicted according to spatial or textual attributes. During this assignment, it was also possible to demonstrate spatial processing functions, such as the preparation of circulation notices to notify all party owners who are within a given distance of a property for which a zoning change is being considered.
Image processing. Since the topographic database had complete digital image coverage for over 2000 square kilometres acquired using MEIS [ ] technology, it was deemed desirable to have the students delineate features in the image to create a topographic database using heads-up digitizing. Originally it was expected to have them use image processing functions for the feature delineation, however this will have to wait until another offering when more time is scheduled for the course. This assignment was designed to illustrate the potential of using a totally digital platform for map production.
It should be noted that to function properly the student had to have a working knowledge of Windows 95 and some of the functionality of Microsoft OfficeProfessional. Many of the assignments required the use of a spreadsheet, word processing or a simple relational database, all of which were normally taken from Microsoft Office Professional.
When the course is offered again, there will be further assignments interspersed with the current assignments. These would have been included this year, but because of the inability to reschedule the course they could not be considered. Included in this subset are the following: using map projections, digital terrain modelling, polygon synthesis to create a cadastre from a set of cadastral points and boundaries, import/export functions between disparate systems, feature extraction from airborne remote sensing imagery, spatial reasoning problems such as viewsheds/watershed problems.
Each lab requires approximately 2 hours within a scheduled lab environment where tutors are available to assist in handling conceptual problems. As well the student is expected to spend up to another 3 hours during the week to complete the assignments.
It should be noted that no textbook was prescribed for this course, not because there was no suitable text but rather because of the need to train students to rely more on Web based resources to locate material. There are many good Web sites including ESRI [ ], University of Texas [ ], USGS [ ] etc.
Experiences Gained from the Course
The experiences from the perspective of the instructor were particularly interesting, because of the uncertainty of the audience. Perhaps the most difficult aspect of teaching this type of course is that there is always considerable diversity in student background. Because of the crowding of courses within the Engineering curriculum, it is no longer possible to expect rigid prerequisites in many undergraduate courses. Whereas one would optimally like to see more mathematics, some geography, more computing science etc., this is no longer practical. Hence it was for this reason that the course was designed to be as comprehensive as possible while at the same time trying to provide sufficient details as to be eminently useful to the student on graduation.
Since this was a required course for students in Environmental Engineering, the class was not necessarily full of students who were there because of interest. However, it soon became apparent that there was almost universal enthusiasm for the course, once the trauma of overcoming the fright of dealing with a complex computer based GIS was overcome. Students were really pleased with themselves when they could identify with the output produced which was of very high quality. Many students were proud of their assignments and placed them in their undergraduate portfolios, hopefully to be used when job searches begin.
The role of the GIS labs cannot be depreciated, in that it required the student to develop a fuller understanding of the subject matter and to demonstrate it in an independent manner.
Initially students were very apprehensive about what was happening in the course, because previous versions of the course did not live up to the expectations of their predecessors. The course was also very gratifying in other ways. For example, it forced the students to hone their skills in working with a current Windows based computing environment and it also provided good access to the WWW. Those students who were really motivated and challenged were able to glean sufficient knowledge from the course to be able to seek out further information on their own.
Based on the feedback of the students, it is proposed to put the course on a local Web site to allow students to operate within an "open-learning' environment. As well, more assignments/projects will be developed to further challenge those who have an insatiable appetite for more of this type of learning experience.
Summary
In order to master GIS, it is first of all necessary to have a sense or awareness about GIS as a discipline. This allows the user to operate within a conceptual environment which naturally exploits the full potential of GIS. GIS of itself is so intimately related to many other disciplines of science and engineering, including information science that a proper course in GIS must first of all develop the basic notions of the discipline and then build bridges to the associated disciplines and technologies. Hence any curriculum in GIS must seek to build a natural foundation based on solid principles which are invariant through time and technologies. With this type of approach the user is able to acquire the fundamentals and build the type of knowledge base that can endure through technological and pedagogical changes.
References:
Hinton G. E. Connectionist Learning Procedures - Artificial Intelligence Vol. 40 Number 1, 1989, pp 143-150
Swingler K. Applying Neural Networks, A Practical Guide - Academic Press, London 1996.
Following are the slides of the presentantion made by Dr. James Linders.
Motivation for Providing Formal GIS Education
|
What is GIS?Differing perspectives:
|
|
|
Conceptual Framework for GISWhen considering GIS, we should be thinking in terms of:
|
Problem Definition
|
Criteria for Assessing a Good Curriculum
|
Applications of GIS Curriculum
in reference to the perspective that GIS is about understanding and using land related data within a systematic context
|
Skill Sets Used for GIS
|
Some related Disciplines by Category
|
Structuring a GIS Curriculumconsiderations: 1. curriculum levels
2. audience
3. learning environment
4. content
5. other issues
|
Principles for Course Development
|
Topics Covered in GIS Course
|
Options for Structuring a Learning Environment
|
Examples of Classroom Exercises
|
Experiences with the Course
|
Instructor Experiences
|
Student Experiences
|
General Comments
|
Summary
|
| Abstract Of The Paper & The Profile of The Speaker | Speaker Index | Paper Title Index |
|
|
|
| |
