In the realm of the XXI century information society, the significance of data acquisition, processing and transfer is increasing exponentially.
The given process has taken its toll on the spatial analysis and its specific, seeing how the spatial data structure and models seem to have undergone major changes over the past few years. In Spatial data structure and models, the fourth chapter of his book GIS basics, Shahab Fazal provides a succinct yet very informative account of the spatial data structure that is currently considered the most adequate in GIS.
The chapter starts, appropriately enough, with a definition of such concepts as “data” and “information.” As the author explains, these terms are often viewed as synonyms, yet, in fact, they are not. More to the point, according to Fazal, “information” is the result of data processing.
While data, as Fazal stresses, comes in four key types, i.e., linguistic, symbolic and mathematical expressions, as well as signals (Fazal, 2008, p. 84). Texts are traditionally defined as linguistic data, while symbolic data includes all types of signs, pictures and pictograms.
Mathematical expressions, in their turn, incorporate formulas and constants, while the concept of signals is deciphered as various acoustic phenomena. It should also be kept in mind that information is prone to aging; as a result, not all information can be classified as useful and suitable for keeping in data sets. Fazal specifies the following types of information as useless: irrelevant, unreliable, outdated, unintelligible, inconsistent, and causing difficulties in handling.
A logical continuation of the previous paragraph, the difference between geographic data and geographic information is defined by Fazal in the next part of the chapter. It is quite remarkable, though, that the author does not focus as much on drawing the line between these notions as much as he explains what data and information mean in geographical dimension.
According to Fazal, “geographic data” can be defined as the data that is “pertinent to features and resources of the Earth” (Fazal, 2008, p. 85), whereas “geographic information” is traditionally viewed as the geographic data that has been organized in accordance with the currently adopted principles, i.e., based on the “descriptive” and “graphical” elements of the geographic data (GD).
The aforementioned differentiation between descriptive and graphical elements is an integral part of the GIS information organization principles, as Fazal explains. Making it obvious that the information organization of the descriptive data is based on the distinction between the two, Fazal proceeds with the definition of key terms and the explanation of their relation to each other.
Fazal lists such elements as a data item, “an occurrence or instance of a particular characteristic” (Fazal, 2008, p. 86), which a data item represents, a stored field, which can be defined as a feature of a particular data item (Fazal, 208, p. 86), and permissible values, also known as a domain of values, which can be identified as the characteristics of a particular set. Apart from the aforementioned terms, Fazal also provides an overview of the concepts of a data file, a flat file, a filename, an ASCII file, and a binary file.
Most importantly, the concepts of a vector and a matrix are explained from the dimensional prospect. Finally, by splitting the exiting databases into database models, i.e., relational-data, network-data, hierarchal-data and object oriented-data ones, Fazal makes it possible to choose the appropriate database type as a reference for a particular type of information.
A next reasonable step in the description of the current system of data organization in GIS, the layout of graphical data allows for distinguishing between the basic concepts of GIS graphic data. While Fazal lists only the key concepts, i.e., a line, a point and a polygon, or an area, he still manages to provide a fairly decent account of the GIS graphic data functions and the ways in which these types of data can be used.
To be more exact, the fact that each of the items mentioned above represent a particular dimension, i.e., a 1-D, a 2-D and a 3-D universe, is represented to the readers. The following introduction of the concept of the vector data and the vector data model, in their turn, allows for creating a model of the object to represent its actual position in the real world; basically, it allows for the “object view of the real world” (Fazal, 2008, p. 92).
The raster method, however, should not be left forgotten, either; though being without the benefits that the vector model has to offer, it still has been relatively popular in GIS over the past few years. In fact, even nowadays, it is still used in bedrock geology, topography, land use, and many other fields of GIS as a superior method.
Choosing the levels of data abstraction as his next focus, Fazal explains that there is a huge difference between data models and database ones in that the former, including the vector and raster models, are supposed to represent the reality and, therefore, can be defined as methods of displaying the reality.
The database models, however, according to Fazal, include the four types of databases, or, to be more exact, the implementation of the existing types of data models. The latter, in their turn, are split into descriptive, graphical, and georelational data structures.
While the previous two have been expanded on in a rather detailed manner by Fazal, the third type of data structure is introduced for the first time in the given paragraph and is identified as a method of analyzing the data that helps coordinate between the spatial, or graphical, and non-spatial, or descriptive, type of data, bridging the gap between the two types of information and translating graphical data into descriptive and vice versa.
After Fazal tackles the tricky differentiation between the three data abstraction levels, he focuses on the relationship perspective of data organization. Thus, the author makes it possible for the audience to consider different elements of the GIS data organization as a system instead of viewing them as separate concepts.
According to Fazal, four scales of measuring the categorical relationship between data types are traditionally distinguished; namely, there are the nominal, the ordinal, the interval and the ratio scales serve as the basis for the GIS data measurement, nominal being qualitative, ordinal presupposing differentiation based on a certain order, interval being grounded in numerical ranking principle, and ratio relating the ranking process to the so-called “absolute datum” (Fazal, 2008, p. 96).
The spatial relationship, in its turn, presupposes that the data is arranged according to its spatial characteristics, i.e., a point, linear, or area position.
Finally, the issue regarding the operating system in general is touched upon. In other words, Fazal defines the process of data organization within the GIS system as a process coordinated within a particular operational system. The previous part of the chapter was devoted solely to defining the key types of databases, as well as the types of data and the methods of data arrangement within a particular data set, which is the reason why the next pat of the chapter concerns the operating system in which the given databases can be created and managed.
As Fazal explains, the root directory in the given operating system is called the topmost directory, the one that is linked to the main one is known as a sub-directory, and the one that is located below the latter is called the parent directory. The workspace concept as a directory structure is also mentioned among the key concepts that are crucial to understanding the operating system designed for the GIS data storage.
As soon as the operating systems, in which the data storage and processing stages are carried out, are identified, Fazal proceeds with expanding on the application architecture of the organization for GIS data. Stressing that most modern computers are constructed on the client-server principle, Fazal makes it clear that the given system architecture pretty much defines the algorithm of data structuring.
Traditionally, five ways of data organization are distinguished for client-server architecture; these are file (demanding a specific record), database (demanding a structured query language, or SQL, request), transaction (demanding a server-side transaction), web (demanding the use of the Hypertext Transfer Protocol, or HTTP) and groupware (allowing for the use of various types of media in messages) services.
Weirdly enough, even with all the information mentioned above in mind, one may stumble on a major obstacle when defining such a simple notion as data. To avoid the possible misconceptions and make the process of learning about the GIS data arrangement principles easier, Fazal provides his own definition of data.
According to Fazal’s own definition, “data are facts” (Fazal, 2008, p. 100). In other words, data are the elements that can be stored in a database. Fazal, however, does not restrict his endeavors to providing a shallow definition of data as pieces of information; on the contrary, he stresses that data may be defined as elements that are closely related to each other and that can be arranged in a specific hierarchy. Thus, Fazal justifies the existence of a concept of a GIS database.
As it has been stressed above, data is traditionally split into two major categories, i.e., the data related to space and the information that characterizes the objects in question from a different perspective. As Fazal explains, such elements as “location, shape, size, and orientation” (Fazal, 2008, p. 100) belong to the spatial data.
The non-spatial information, in its turn, embraces such characteristics as mass, temperature, etc.; in other words, non-spatial data is “independent of geometric considerations” (Fazal, 2008, p. 100), Fazal says. It should also be noted that both spatial and non-spatial data comes in a variety of forms, including text, pictures, graphs, tables, etc. As a rule, traditional database types are suitable for storing the above-mentioned types of information.
However, when it comes to such data type as texts, standard databases fail to perform their basic functions, which is why the standard database solutions are chosen typically to construct attribute tables only. It should also be kept in mind that in standard databases, the order of the elements is not considered the priority; as a result, the data elements often appeal to be jumbled. The given problem can be resolved with the help of either a total DBMS solution, or a mixed solution.
Another type of data storage, a repository can be considered a decent way out of the problem with retaining the required information. A repository offers carrying out such operations with the existing data sets as adding new information, securing the existing data and retrieving the necessary facts whenever the necessity appears. With a security mechanism incorporated into the principle of a repository operation, the given system of data storage allows for an enhanced security of the data in question.
It should be noted that the given method of data storage is considered superior, since it allows for reducing the redundancy of data and, therefore, for its better arrangement. Among the languages used in the DBMS queries, a data definition language, or DDL, a data manipulation language, or DML, a query and a query language should be listed.
In his book, Fazal often mentions the concept of data model, and, after contemplating on the spatial data models, he finally provides the definition of the given phenomenon.
According to Fazal, a data model is a “mathematical formalism” (Fazal, 2008, p. 102), which includes a data description notation and a range of operations with the help of which the data is arranged, distributed, analyzed and used. Based on what Fazal says, three levels of abstraction can be identified for a data model, i.e., a physical one, a conceptual one, and a one defined as a “view” (Fazal, 2008, p. 103).
Eventually, the issue of data modeling should be expanded on a bit more; Fazal specifies that there are two data modeling types, i.e., logical and physical. At this point, it is reasonable to recall the three levels of abstraction mentioned above, i.e., data model, data structure and file structure, seeing how Fazal relates each of the abstraction levels to the corresponding data modeling phase, i.e., conceptual, logical and physical data modeling.
Apart from the aforementioned types of data modeling, the fourth one should also be named; defined by Fazal as process modeling, it is called to determine the processes that the operation of the information system will include. The last, but definitely not the least, such tool as data flow must be mentioned, since it is constructed with the help of symbols representing “process, data stores, entities and data flow in a business function” (Fazal, 2008, p. 105).
Despite the fact that Fazal provides a relatively short account of the structure and models of spatial data, the chapter still offers a plethora of information regarding data acquisition, storage, processing and distribution.
More to the point, the author puts a very strong emphasis on the proper use of the acquired data, which makes the given chapter all the more informative and significant. A perfect introduction into the principles of data usage in GIS, the given excerpt is an integral part of the entire book and a crucial piece of information for a GIS expert.
Reference
Fazal, S. (2008). Spatial data structure and models. In S. Fazal (Ed.), GIS basics. New Delhi, IN: New Age Publications.