Survey on Spatial Database Systems



Abstract—In this survey, the term of “spatial database”, its data models, its data types with implementations, and its management techniques are described with providing at least spatial join methods. Also the storage and query processing algorithms for such databases are surveyed.

(Abstract section will be updated with final report)

1.Introduction

Various fields need various data types such as character, number, date, time, and image in data base management systems (DBMS). Some other fields need more specialized data types with geometric and geographical attributes. Those needs are satisfied by the spatial data. The Spatial Data is described as data related to time and space [12]. The most noticeable area for spatial data types is two dimensional abstraction of the earth surface [Figure 1]. Other examples are layouts of very large scale integration designs in electronics, 3D designs of biological issues like DNA, and the complicated models of the human brain systems.

Figure 1: Spatial Data. (Source: http://www.cubrid.org/blog/dev-platform/20-minutes-to-understanding-spatial-database/ [11])

Spatial database systems can be grouped as followings [1]:

1. Geographical Information Systems (GIS) deal with digitized maps displaying geographic or thematic information.

2. Automated Mapping/Facilities Management (AM/FM) systems which automate the management and maintenance of networks such as power grids or telephone lines.

3. Land Information Systems (LIS) manage information such as Image Processing systems which process remote sensing images acquired by aircraft and satellites. LIS also deal with the details of land parcel ownership.

Although the relational DBMS’s have been tried to manage those types of data, they did not meet the requirements properly [12]. Spatial database systems provide advantages in areas such as decision support, administration, transportation scheduling, resource management, environmental monitoring, real-time navigational systems, data quality and integrity enforcement, and impact assessment.

The remaining of this draft report is organized as follows: In Section 2, Modeling Spatial Database Systems is explained in detail. In Section 2.1, the characteristics of spatial data types are represented. The relations and related explanations are provided in Section 2.2. The querying and its techniques are represented in Section 2.3. At 2.4, indexing of spatial data is shown. Visualization of spatial data is explained in Section 2.5. Finally, concluding remarks are summarized in Section 3.

2. Modeling SPATIAL DATABASE SYSTEMS

Spatial database systems are either the new DBMS or additional features on Relational DBMS’s. It is a DBMS with additional capabilities for handling spatial data and Offers spatial data types in its data model and query language. For modeling such a database system, data types, relations, querying, indexing and visualization steps can be considered different parts of it.

2.1.Data TYPES

There are classical data types for all DBMS such as types of chars, types of numbers, date, and time. Spatial data shows the geometric and geographical variables such as point, line, region, polyline, and polygon. The presentation of those can be divided into two main groups [1].

2.1.1. Objects in space

It is a representation of spatial data types such as polygons, lines, polylines etc.

Point:

• As pairs of coordinates in lat/long or some other reference system
• A point feature is a zero-dimensional cartographic object.
• It specifies the geometric location and no other meaningful measurement
• The size of the point may vary, but the area of those symbols is meaningless
• Four types of points exist: entity point, label point, area point and node

Line:

• Ordered sequence of points connected by straight lines
• Line features are one dimensional features, despite occupying two-dimensional space.
• A line segment is the direct connection between two points
• A line feature is typically represented as a sequence of vectors
• An Arc is the location of points that are defined by a mathematical function to form a curve
• Link or edge is the connection between two nodes

Areas:

• As ordered rings of points connected by straight lines to form polygons
• Area is a two dimensional, bounded and continuous object
• Interior area is an area not including its boundary
• Simple polygon consists of an interior area and an outer ring. The boundary does not intersect itself
• Typically refers to vector polygons, but also relates to pixels and grid cells.

2.1.2.Space

It deals with Statement about every point in space such as partitions into states, counties, municipalities etc.

(This section will be detailed in final report)

2.2.REALATIONS

Spatial relationships are very important in the operations offered by spatial algebras. For instance, it is possible to ask for all objects in a given relationship with a query object, e.g. all objects within an object or intersection points. There are several classes [8, 3, 4]:

• Topological relationships

• Direction relationships

• Metric relationships

(This section will be detailed in final report)

2.3.QUERYING

Spatial data requires a graphical presentation of results. In addition, SDT values used in queries or graphical input of queries need graphical representations.

(This section will be detailed in final report)

2.3.1.Languages

Query languages for spatial databases can be used as candidates for the creation of a spatial language. Because of the extra semantic complexity added by spatial dimensions, it is desirable to have features in a spatial query language that go beyond those provided by currently available mainstream relational languages.

(This section will be detailed in final report)

2.3.2.Operators

There are several types of spatial operators [4] logical relationships, arithmetic, spatial metrics, position, orientation, area, volume, shape, extent, surface, disjunction, intersection, inclusion, neighborhood, and equality.

(This section will be detailed in final report)

2.4.INDEXING

For all DBMS, fast access to row data depends on the quality of indexing. Complex indexing methods can be used to rapidly locate single or multiple objects in the databases.

(This section will be detailed in final report)

2.4.1.Indexing Methods

For spatial databases, some indexing techniques such as quadtrees [6], R-Trees [2] are mostly used ones.

(This section will be detailed in final report)

2.5.vısualızatıon

The modern database management systems provide visualization tools to represent spatial data and queries about those data. Browsers, plotters and map displays can be considered as standard tools for spatial database systems. Although some researchers classify the spatial maps as maps showing qualitative, quantitative and composite change, and space-time ratios[9], some researchers made this classification like dance maps, chess maps and change maps to visualize time series data [7].

(This section will be detailed in final report)

3. Conclusıon

(This section will be detailed in final report)

4. ReferenceS

[1] Abel, D.J. What’s Special about Spatial?. Proc. of the 7th Australian Database Conference, Melbourne, Australia, 1996, 72-81.

[2] Guttman, A. R-trees: A Dynamic Index Structure for Spatial Searching. Proceedings of the ACM SIGMOD International Conference on Management of Data, 1984, 47-57.

[3] Egenhofer, M., A Formal Definition of Binary Topological Relationships. Proc. 3rd Intl. Conf. on Foundations of Data Organization and Algorithms, Paris, 1989, 457-472.

[4] Langran, G. Manipulation and Analysis of Temporal Geographic Information. Proc. of the Canadian Conference on GIS 93, Ottawa, Canada, 1993.

[6] Samet, H. The Design and Analysis of Spatial Data Structures. Addison-Wesley, 1990.

[7] Monmonier, M. Strategies for the Visualization of Geographic Time-Series Data, Cartographica, 1990, 30-45

[8] Pullar, D., and Egenhofer, M., Towards Formal Definitions of Topological Relations Among Spatial Objects. Proc. 3rd Intl. Symposium on Spatial Data Handling, Sydney, 1988, 225-242.

[9] Muehrcke, P.C.. Map Use, JP Publications, 1978.

[10] Worboys, M.F., A Generic Model for Planar Geographical Objects. Intl. Journal of Geographical Information Systems , 1992 , 353-372.

[11] 20 Minutes to Understanding Spatial Database. Retrieved October 20, 2014, from http://www.cubrid.org/blog/dev-platform/20-minutes-to-understanding-spatial-database/

[12] An introduction to spatial database systems. (1994). The International Journal on Very Large Data Bases, 3(4), 357-399.

(This section will be updated in final report)