TECHNOTES
Vector Data Development Changes
New software provides a solution to the trade-off between cost and quality
By John Auble

Vector database development has always been considered the "necessary evil" of GIS projects. Unfortunately, database development can consume 50%-80% of the project's resources. The commercial GIS industry has attempted to address this problem by spawning a number of data provision companies. However, these off-the-shelf data often lack the required level of accuracy or the underlying spatial construction for the intended analysis. The fundamental dilemma for GIS professionals is the trade-off between the quality and the cost of the data. One solution to this dilemma is a tool that provides professionals the flexibility to create their own data, or edit third-party data inexpensively. Most importantly, the tool should be easy enough to use so professionals can quickly choose which option will optimize quality while minimizing cost.
    Shedding the overhead of trying to create a "single solution" software package, Clark Labs, located in Worcester, Massachusetts, has launched a new product, CartaLinx, that reduces the resources required for vector data development. This new program allows GIS professionals the opportunity to devote more of their precious resources to analysis and problem solving. Versatile, CartaLinx acts as a tool within the larger GIS toolkit. It enables users to capture data from disparate sources, process the data into an integrated topological database, then export the data to the user's desired GIS data format.

CartaLinx Prioritizes Compatibility with other GIS Tools
As a specialized tool for vector data development, CartaLinx is designed to maximize compatibility with the tools for data input as well as the tools that utilize the data product. For example, vector data can be derived from paper maps, digital imagery, COGO, spreadsheets, GPS surveys, other GIS databases, and many other sources. This represents the breadth of input options. However, the user requires significant depth to any of these specific input alternatives and CartaLinx delivers. One example of this depth is tablet digitizing created through a strategic relationship with Digitizer Technology Company in Redmonds, Washington, makers of the Virtual Tablet Interface driver that supports over 300 different digitizing tablets. The results have been staggering. Clark Labs' technical support staff reports that less than 1% of all CartaLinx users have needed advice on tablet configuration.
    Software compatibility is not just a practical consideration at Clark Labs, it represents a fundamental goal to participate in the development of a larger GIS community, based on the principle of inter-operability. This, of course, means adhering to standards. To this end, CartaLinx not only supports proprietary data formats such as Shape and MIF files, but also includes support for the USGS Spatial Data Transfer Standard and the British Ordnance Survey format. Similarly, CartaLinx supports real-time GPS input through the National Marine Electronics Association (NMEA) standard format.
    CartaLinx supports a wide variety of data formats and input/output devices to facilitate the flow of data through the editing process. Within the editing environment, there are dozens of functions neatly organized to enhance the user's ability to learn and use these functions.
    The first level of editing functions includes those operations that work on the entire coverage and will significantly impact the character of the coverage. These options are built into the pull-down menu structure and include operations such as Re-sample, Project, Clip, and Append. The pull-down menu also includes operations that can be applied to the entire coverage or to the features within the current display window. These include operations that create nodes at intersections, remove pseudo nodes, and automatically snap nodes.
    The second level of editing functions is accessed through a context-sensitive cursor pop-up menu. This innovation clears the workspace of floating toolbars and removes the need to learn the location of dozens of specialized editing functions. Each cursor pop-up menu contains a related set of functions. CartaLinx decides which menu to present based on where the user placed the mouse cursor before clicking, and in some cases, the last action performed by the user.
    The logic is simple. If a user clicks the right mouse button while the cursor is within the Tables Window, the cursor pop-up menu lists options related to attribute management. The same logic applies to working within the Map Display Window. If a user wants to break an arc at a specific location, he selects the arc at that location before clicking the right mouse button. CartaLinx then presents a cursor pop-up menu with options limited to the logical operations that can be performed on a selected arc. This innovation improves the user's workflow by avoiding the disruption of searching through menus and toolbars. Instead, at the click of a button, the user has the editing options he needs, when and where he needs them.
    Among those users who have benefited from these innovations is the Philippine Association for Intercultural Development (PAFID), which is currently engaged in a large mapping project to survey and plot the ancestral domain claims of tribal minorities. Lacking three significant resources: money, time, and data, PAFID has accomplished work using CartaLinx which, heretofore, might otherwise have only been possible with much larger budgets. Using this vector database development tool, this nonprofit group has been able to:
¥ Quickly create a regional base data from 1:50,000 scale maps to help triage the overwhelming number of requests for assistance.
¥ Integrate GPS survey data, traditional chain and compass survey data, and map sheet data into the production of the maps of ancestral domain claims.
    Reduce the time spent on training staff in digital data development by more than 50%. CartaLinx has helped PAFID deliver GIS products that meet national government standards for use in the high-stakes land claim arena. "What we really lack is GIS expertise," said Bruce Young, PAFID's mapping program officer. "I think this is the case for most groups with limited resources, which makes CartaLinx's ease of use even more valuable."

Efficient Data Production Can Include Topology
CartaLinx creates a vector data structure with full topological relationships. These relationships allow the user to know the area for each polygon, the contiguity (or neighbor relationship) between polygons, and the connectivity between lines. The topological attributes, along with user-defined attributes, are stored in a Microsoft Access database with three tables: Nodes, Arcs, and Polygons. Therefore, each coverage is represented by one file that carries the spatial data and a related file that carries the attribute information.
    The value of a topologically encoded data structure is its inherent flexibility. A robust data structure can support complex geographic analysis or be easily transformed to a data structure that better supports graphic representation. However, the reverse is not true. It is far more difficult to reverse engineer topological relationships from graphic data. Given the cost of data development, it is more efficient to create a database that can accommodate a number of different uses than to retool the data in the middle of a GIS project. Data with built-in flexibility carry greater value as a commodity.
    Dr. Gene Paull at the University of Texas - Brownsville, is currently developing a database with topology. "I am part of a mapping project with the Bureau of Economic Geology of the University of Texas. It is, in essence, the state mapping agency for Texas. My charge has been to digitize geologic maps of the U.S. - Mexico border region." Dr. Paull's group is one of several border universities responsible for developing the project's database. Their component of the project had stalled for 6-months before using CartaLinx. He said his group was amazed "at how fast we were able to work in CartaLinx - building and labeling polygons, then joining coverages."
    Choosing the right tool for the job is half the battle in any line of work. Many GIS software packages promoted as "single solutions", either do not include data input facilities, or provide difficult to use or limited capabilities. CartaLinx, as a specialized tool for vector data development, can be plugged into any GIS professional's toolkit to improve productivity immediately.

About the Author:
John Auble is a senior research associate at Clark Labs, specializing in vector, GIS. GPS, and GIS adoption. He can be reached at [email protected].

Back