An Urban GIS Database Model for Integrated Land and Building Property Management.

ABSTRACT

Since the introduction of Geographic Information Systems (GIS) in South Korea in the early 1980s, GIS have been widely used for effective facility and land resource management. Specifically, an extensive GIS implementation strategy was initiated by the National GIS committee with a two-stage plan: GIS database development (1996-2000) and a consumer-oriented database and application development (from 2001). The first stage can be characterized by the development of large-scale vector base maps along with the development of facility and utility maps. GIS implementation for the city of Seoul has been extensively researched since the early 1990s. Focusing on the facility management systems, 1:500 and 1:1,000-scale base-map databases have been developed following guidelines described by National Orders and Bylaws regarding digital mapping. Because the guidelines are based on a hierarchical structure of spatial features, it is necessary to provide semantic relationships among features for developing information-ri ch application databases. Using the land parcel and building themes, conceptual and logical database models were designed. Three entities and six relationships were identified. After analyzing their relations, a logical database model was designed by converting entities, relationships, and set attributes into objects.

INTRODUCTION

The growth of the city of Seoul, South Korea, has caused an increased demand for land resources and complexity of land uses. Under limited land resources, issues regarding effective land use planning and decision making have been major concerns to city governors, planners, administrators, and citizens. Inefficient and incomprehensive management of natural resources, ground and underground facilities, land cadastre data, and urban planning workflow may bring untimely decision-making thus making it difficult to manage the ever changing urban environment efficiently. This would decrease the quality of urban spaces and accrete unnecessary costs such as increasing traffic time and unwelcome public policies.

For the effective management of complex urban systems and the improvement of the quality of public services, geographic information systems (GIS) have been introduced beginning in the early 1980s by many local governments in South Korea such as Seoul, KwangJu, TaeKu, SeongNam, KwaChon, and InChon. Because of the complexity of urban facilities, most urban GIS implementations in South Korea have focused on the effective management of urban infrastructures such as power and communication cables, water, sewage, gas, roads, buildings, and land parcels at a very high accuracy level.

Many GIS projects, however, revealed several unexpected problems, specifically in creating very large-scale base-maps where urban facilities would be drawn. Even though there are land parcel maps at 1:600, 1:1,200, and 1:3,000 scales used in South Korea, many GIS projects created very large-scale base-maps (1:500 and 1:1,000 scales) from stereo plotting aerial photos because the projection used for the land parcel maps is obsolete and the land parcel maps keep ownership and planning-related features more accurately compared to less accurate topographic and cultural features. Even though some prototype systems were developed using 1:5,000-scale base maps, which are the largest-scale map among national topographic map series, a general consensus was to develop large-scale digital base-maps from scratch. This responsibility was assigned to the National GIS (NGIS) Initiative started in 1995. According to the initiative, most urban areas would be mapped at least to the scale of 1:1,000. Under the NGIS plan, large- scale base-maps have been developed at three scales, 1:1,000, 1:5,000, and 1:25,000 since 1996, and the base-map construction was scheduled to be finished by the year 2000.

With most GIS efforts occupied by the construction of large-scale base-maps, there has been little research on how to design application data models using the base-maps and other data sets. The construction of base-map databases does not mean that GIS database construction is finished; rather, it is a starting point for actual GIS application development. As the word "base" implies, the base-maps are the starting point of application database development. Base-map data are sometimes organized with poor semantic relationships. Consequently, application database modeling becomes very important for effective database structuring and efficient database querying. The base-maps constructed under the NGIS Initiative are organized by hierarchical entity definitions without semantic relationships among features at the same level. The hierarchical layer-based GIS databases may be good for producing maps, but they need to be restructured to have semantic relationships with other features for developing information-rich GIS applications. In this context, this research aims at introducing GIS implementation in South Korea and Seoul City and at investigating an application database logical modeling for urban land property management using base-map databases, land parcel maps, and land descriptions. This research will be helpful to the countries and local governments that are planning or currently developing an extensive GIS for managing urban facilities. After examining the experience of GIS implementation in South Korea and Seoul City, we will investigate an application database design.

GIS IN SOUTH KOREA AND SEOUL

National Geographic In formation Systems Initiative

In South Korea, GIS were introduced in the early 1980s and the need for GIS in mapping and facility management has been enthusiastically discussed since the late 1980s. In the early 1990s, some prototype GIS systems were developed by city governments such as TaeGu and KwangJu and by companies such as Korea Telecom and the Korea Electric Power Corporation. In most early systems, facility management was a major concern because many people are living in relatively limited urban space packed with various ground and underground facilities and because GIS have been proven to be an excellent vehicle for managing complex urban facilities. It was in 1995 that the NGIS was initiated to manage and control nationwide public GIS implementation plans. In the meantime, two gas explosions in metropolitan Seoul in December 1994 and in metropolitan TaeGu in April 1995 added significant weight to the NGIS activities. An NGIS master plan was substantiated in May 1995 and is currently coordinated by the Ministry of Construction and Transportation (MoCT), which manages various construction and road facilities and provides national topographic maps. The NGIS Initiative consists of two stages: GIS database development and application development.

The first stage (1996--2000) plan is characterized by the construction of nationwide geographic information databases. Because a database is critical to the success of subsequent development of applications, the NGIS approached database construction strategically under the consideration of various aspects including surveying the demand and supply of spatial data, designing a synoptic framework, building databases, and developing spatial data distribution systems. The second stage of the NGIS, beginning in 2001, is characterized by the use of data that are constructed during the first stage. Specific consumer-oriented database constructions are a major issue at this application stage, along with the development of spatial decision support systems and public administration support systems.

In relation to the NGIS, the role of the national government was identified to support practical issues such as,

* basic geographic information database construction,

* technology development and education,

* standardization of spatial information,

* government-wide GIS implementation and support,

* maximization of spatial information management and sharing,

* revision of regulations and laws,

* securance of resources from public and private sectors, and

* continuous evaluation and revision of the NGIS plans.

As a steering organization for effective implementation of the NGIS Initiative, five NGIS committees were organized to tackle those practical issues (see Figure 1). They consist of committees on geographic information, technology development, standardization, land information, and general administration. The main task of the geographic information committee is to develop digital maps at three scales: 1:1,000 for urban areas, 1:25,000 for mountainous areas, and 1:5,000 for the other areas. The 1:5,000 and 1:25,000 scale maps were digitized from the national topographic map series, but the 1:1,000 scale maps have been plotted from aerial photos because 1:5,000 is the highest map scale used for the national topographic map series. By late 1998, most parts of South Korea were digitized and plans were made for more than 95 percent of the digital maps to be constructed by the end of 1999. Also, GIS databases for thematic maps and seven underground facilities--water, sewage, gas, electricity, communication, heat, an d oil--are under development with the necessary revision and support of laws. The technology development committee focuses on the comprehensive and systematic development of basic GIS software technologies including mapping, fundamental GIS operation, database management, and system integration. The standardization committee develops various standards, including the content and representation method of digital maps, metadata standards, and data transfer standards. The land information committee on the other hand focuses on the management of land ownership and planning information including land cadastre and parcel maps. Under a plan to make a land information system operational by 1998, the committee developed a prototype system for the City of TaeJon in 1997, and the digitizing of land property maps was supposed to be finished by that time. Finally, the general administration committee coordinates the rest of the subcommittees, evaluating the progress of each subcommittee's activities, revising necessary reg ulations and laws, advertising the NGIS, and supporting various research projects. Also, the general administration committee intervenes with conflicts among institutions and resolves them through a central coordination committee. An NGIS law, which was prepared by the MoCT and submitted to the National Congress, is one of the outputs of the committee, and the law is anticipated to support and coordinate GIS-related sensitive issues such as administration, database security, organizational cooperation, auditing, and data distribution (MoCT 1999).

The NGIS Initiative and accompanying activities may be characterized by their comprehensiveness, importance of facility management, and information infrastructure. First, the NGIS Initiative is very comprehensive because it covers various themes with very high accuracies. GIS themes include topographic features, thematic maps, facilities, and land parcel maps. Also, the geographic extent covers the whole territory of South Korea. Second, the initiative is more likely focusing on the construction of facility management and land parcel management databases with high priorities. Synergy effects to physical environmental applications and mapping applications are considered at the application development stage. Considering various GIS applications, various base-map databases, for example, digital orthophoto database and scanned topographic map databases such as the Digital Raster Graphics from the United States Geological Survey, could be developed easily for environmental modeling and mapping, but the initiative more emphasizes vector-based large scale maps for using with land parcel and facility features. Third, the initiative focuses on the information infrastructure. Participants from administration offices represent all the departments using geographic data for management, planning, and decision making. Also, issues discussed at the committee cover not only GIS database construction but also system integration, software development, legal arrangement, security, system design, and distribution system.

GIS IN SEOUL

The applications of GIS for Seoul City have been investigated actively since the early 1990s. The first consideration was given by Chang (1992) regarding the strategic, mid-term, and long-term directions of urban GIS (UGIS) in Seoul. She researched UGIS applications implemented in other cities and discussed strategic GIS implementation plans for Seoul City government. Based on her research in 1992, Chang et al. (1993) extensively surveyed all the routine tasks and UGIS needs over all the departments and executive branches of Seoul City government. They also investigated institutional issues regarding the introduction of UGIS in an integrated and controlled manner. Technical issues such as database needs, software and hardware demands, networking strategies, application system priority, and cost/benefit analysis were analyzed too.

Based on the synoptic implementation design of UGIS for Seoul, practical UGIS implementation guidelines were researched by Kim et al. (1994). The research focused on twenty-one practical issues, such as the contents of databases, base map requirements, city-wide GIS computing environments, GIS education systems, and UGIS management systems. A prototype UGIS database was also developed in their research for Chung-Gu, which is located at the center of Seoul, an area of about 9.97 [km.sup.2]. Along with the UGIS implementation, a comprehensive urban information system design was discussed to couple the Seoul UGIS under a perspective of synoptic system integration (Chang et al. 1995).

Along with the NGIS initiatives started in 1995, various research projects have been performed on Seoul UGIS. The primary concern has been targeted to the management of complex urban facilities. Specifically, underground facilities have been regarded as very important after a gas explosion at AhyeunDong, Seoul, in 1994, which boosted the significance of UGIS for underground facility management. According to Kang et al. (1996), underground facilities including electricity, water, sewage, gas, subway, heating, and communication have been managed by several departments and public companies with little data sharing in Seoul. Specifically, they pointed to overlapping investments, unstandardized data management systems, little data sharing between departments, and insufficient communication among departments. As a solution to the problems, they suggested that the management of underground facilities be considered in connection to the NGIS. Also, they suggested the use of standardized base maps, an integrated datab ase construction, and a standard for underground facility database design. Regarding base maps, 1:1,000-scale and 1:500-scale maps were suggested for general purposes and road management systems respectively (Chang et al. 1996).

Recently, Kim et al. (1997) researched a specific application of UGIS for building safety management. Kang et al. (1998) also investigated the use of geometrically rectified digital aerial photos as timely UGIS base maps. In addition, an effective maintenance method of Seoul UGIS base maps was investigated by Kim et al. (1998), and the use of land parcel maps and planning maps in relation to topographic base maps was researched in depth by Kang et al. (1997). Because the land parcel and planning maps in Seoul have been managed with various-scale paper maps with the relative georeferencing method, which uses relative coordinates to nearby distinctive features instead of global or national coordinate systems, the use of the land parcel and planning maps in a digital environment has been a hot issue in Seoul, requiring political coordination and in depth technical consideration. Kang et al. (1997) tackled the problem and cast a strategy on the use of digital land parcel and planning maps in relation to the 1:1, 000-scale NGIS base maps.

In Seoul, the land cadastre maps mostly at 1:600 and 1:1,200 scales are the base-maps of urban planning maps. They have been compiled and managed separately from the topographic maps, which are mostly at 1:500, 1:1,000, 1:5,000, 1:25,000 and 1:50,000 scales. Specifically, the use of different projections--Transverse Mercator projection and Gauss-Kruger projection--and different survey bench marks, the deformation of paper maps, and the illegal use of public lands have been known to cause mismatches between topographic maps and land parcel maps. In their research, they examined three technical alternatives to match them: coordinate transformation, rubber sheeting, and manual moving methods, and they suggested the manual moving method to be the most practical strategy to match the land parcel maps to the topographic maps (Kang et al. 1997).

Based on guidelines specifically by Kim et al. (1994) and Chang et al. (1995), the Seoul City government initiated GIS implementation in two stages: GIS database and application development from 1996 to 2001 and integration to comprehensive urban information system from 2002 to 2010. During the first stage, 1:1,000-Oscale base maps were prepared by 1998 and facility GIS databases will be finished by the end of 2001. Also, GIS applications will be developed by the end of 2001 focusing on applications such as facility and utility management, road management, emergency management, planning, and policy development. The second stage is characterized by the integration of Seoul UGIS to a comprehensive urban information system that provides citizens with seamless spatial and aspatial (attribute) information through the World Wide Web.

The efforts by the NGIS and Seoul metropolitan government show the importance of effective coordination and management of UGIS plans in an integrated fashion. Extensive efforts have been made to find the best strategy for the implementation of Seoul UGIS emphasizing the importance of effective coordination and management in an integrated fashion. However, research has been mostly focused on the geometrical, technical, and organizational issues. Considering GIS may bring quadruple effects when the systems are used, not only for mapping but also for project development, planning, facility maintenance, coordination, general services, etc. (Bernhardsen 1999), it is very important to design GIS databases efficiently considering the maximization of their use.

In South Korea, GIS databases usually reference the Order 17 (MoCT 1995) and the Bylaw 71 (NGI 1995) for the construction of digital maps, which categorize map features in four-level entity codes. In the Order, the first-level layer codes contain nine categories: railroads, rivers, roads, buildings, land use, facilities, topography, boundaries, and annotations. The second-level layer codes are the subcategories of the nine first-level categories and so on. The categories, therefore, increase as the level deepens, so that there are 9, 28, 92, and 580 categories for the first, second, third, and fourth levels, respectively. A problem of using the guidelines is that the Order and Bylaw identify only the geographic entities so that they organize the entity types very well hierarchically, but they don't provide information on how to relate geographic objects to other objects. They only provide an implicit relationship in a hierarchical grouping so that there is no relationship between same level groups. Following just the Order and Bylaw may result in GIS databases with a layer-based structure which would be good for mapping features, but the lack of the relationships among objects that may limit the ability of GIS analysis and application development. As an example of a solution, the next section investigates a method to organize and model geographic objects using the analysis of entity-relationship and the relational database models.

GIS DATA MODELING FOR INTEGRATED LAND MANAGEMENT, SEOUL, KOREA

Database Design

A database is usually designed through three stages: conceptual database design, logical database design, and physical database design (Connolly et al. 1996). Based on the survey and analysis of database needs, necessary data types and relationships are identified at the conceptual database design stage. The conceptual schema developed at the conceptual database design stage is transformed into logical schema using a data model supported by a specific database management system (DBMS) at the logical database design stage. Finally, physical schemata are constructed to optimize the performance of DBMS by defining data storing methods and data access methods at the physical database design stage.

This research focused on the land parcel data because the land-related data are very important and directly connected to the land cadastre, building, taxation, and planning. The relationships among land cadastre, land parcel map, and building information were investigated in terms of the relational model, which is popular in practical GIS applications. Using the information from land and building data, this research investigated the conceptual and logical database designs with existing digital land parcel data for KangDong Cu, Seoul (see Figure 2). The attribute information of land parcels and the building records was carefully analyzed to find entities and relationships. From a survey, three major entities were identified in relation to land parcels: owner, land parcel, and building. The following describes the unique characteristic of each entity.

Owner Entity: The owner entity consists of name, owner identification, and address. Each owner has legal right to own and mortgage lands and buildings and also has exclusive right to the information on ownership change date, ownership change reason, mortgage change date, mortgage change reason, mortgage amount, mortgagor, and mortgagee.

Land Parcel Entity: Each land parcel has a unique identification and the location information composed of parcel number, sector, city, and province. Land use information such as land use type, zone, and district is also attached to each land and changes of the information are updated. Information such as size and planned use can be included to the land parcel entity, Also included are the general grade of the land, land grade change date, unit price, and attached buildings to the land parcel entity.

Building Entity: The building entity has a building identifier, usage, structure, and drawings. Building attributes also include current status, size, construction date, garage, sanitation facility, elevator, and attached buildings. Each attribute can even be subdivided in detail. For example, the garage contains attributes such as capacity, location and size, and the current status is composed of floor usage, related land parcel data, height, and floor sizes.

CONCEPTUAL MODELING

Conceptual models were designed based on the entities identified from the analysis of database needs and existing cadastre and building records. The main task in the conceptual database design stage is to define entity types and relationship types more specifically. During the definition of an entity type, the entity-specific attributes are defined and then relationship types are analyzed considering the relationships among entities.

Figures 3, 4, and 5 show the three entities with attributes attached. In the figures, rectangles mean entities, single circles indicate attributes, and double circles indicate set attributes. The set attribute is identified when a composite attribute is frequently changed. The underlined attribute means the attribute is used for a unique identifier. In the case of the owner entity, only name, address, and owner identification were used. The land parcel entity, however, was designed with more attributes. Specifically, five attributes were identified as very important: land parcel identifier, location of the land parcel, land use plan, land parcel representation, and land parcel grade. Most attributes were identified as composite attributes, except the land parcel identifier. Also, the land grade and land use plan attributes are continuously changed, therefore they were defined as set attributes. Building entity was identified to include ten unique attributes: building identifier, usage, structure, drawings, b uilding specifications, constructor attributes, construction date, parking data, sanitation facilities, and elevator. Among these, parking and elevator attributes were assigned as set attributes because their operation data are accumulated.

From the analysis of database needs, six relationships were identified among owner, land parcel, and building entities: land ownership, mortgage, lease, building, attached building, and building ownership (see Figure 6), The land ownership, lease, and mortgage relationships were defined as relationships between the land parcel entity and the owner entity. The building ownership relationship was identified between the owner entity and the building entity. Also, the building relationship was defined between the land parcel and building entities. The attached building relationship was also defined between building entities. In Figure 6, six relationships are shown in diamond shapes. The single line among entities, attributes, and relationships means the relation is not mandatory but optional; double lines mean the relation is mandatory. For example, a land parcel may have a building optionally, but a land parcel is mandatory to a building. The characters 'M' and 'N' indicate multiple mapping cardinality while '1' indicates single cardinality. Specifically, 'M' denotes the one-to-many cardinality between entity and relationship. 'N' denotes the many-to-one cardinality between entity and relationship. For example, multiple owners can share one land ownership making cardinality 'N', and a land parcel can have multiple land ownerships making cardinality 'M'. Therefore, the relationship between the owner and land parcel entities has N:M mapping cardinality. Similarly, the building and owner entities have M:N mapping cardinality, and the land parcel and building entities have M:N relationship. A building may have many attached-building relationships making cardinality 'N' and each attached bui lding has a one-to-one relationship to a main building making cardinality '1' so that the mapping cardinality between building and building is N:1.

LOGICAL MODELING

Logical modeling is a process to convert the components of a conceptual data model into logical schemata using a data model supported by a specific target DBMS. Entity types, composite attributes, set attributes, and relationship types are restructured to relations that are implemented in the attribute table structure.

During the logical modeling process, each object should have a unique identifier that is a primary key. Single attributes are transformed to relations attached to relevant entity types and relationship types. Also, composite attributes are decomposed and attached to entity or relationship types like the simple attributes because the composite attributes are just extensions of simple attributes which are not subject to change. However, set attributes are subject to change and must to be updated and managed more frequently; therefore, they can be organized into unique objects. Unique primary keys are to be used for making relations to relevant entity types. In the case of relationship objects, they should have two primary keys connecting the relationship objects to other entity objects. If the relationship has 1:N mapping cardinality, which means one is contained to another, the container object includes the other object.

Figure 7 shows the result of logical database modeling. Simple attributes such as name, identifier (ID), and address become relations attached to the owner entity using the ID as a primary key. In the case of the building object, it contains simple and composite attributes, and includes the relationship that indicates the container of the building object. Parking and elevator objects show they are set objects and are related to building identifiers. Also, the building ownership, mortgage, lease, building and land ownership relationships show they have two primary keys, occasionally with some single attributes contained. The logical models can be transformed easily to a relational database using a DBMS.

SUMMARY AND CONCLUSION

An extensive, nationwide GIS implementation initiative requires considerations from various aspects. Since the introduction of GIS in South Korea in the early 1980s, South Korea and Seoul City government have studied various strategies to implement GIS for their administration and public services. Specifically, two gas explosions in Seoul and TaeGu cities in 1994 and 1995, respectively, have expedited GIS implementation for urban facility management, including gas, sewage, road, power cable, and communication. Along with the National GIS initiative initiated in 1995, the use of GIS has been coordinated by the five NGIS committees with two five-year plans, 1996 through 2000 and 2001 through 2006. The main goal of the first five-year plan (phase) is to build national GIS databases at 1:1,000 scale for urban areas, 1:5,000 scale for the most other areas, and 1:25,000 scale for mountainous areas. Most databases are supposed to be constructed complying with MoCT's Order 17 on Digital Mapping. The main goal of the second phase is to develop applications using the databases. The NGIS Initiative can be characterized by its comprehensiveness, covering a variety of theme types and whole areas with a very high accuracy level. The initiative focuses more on vector-based databases targeting mainly facility management applications. Also, the initiative is touching extensive information infrastructure of the country, including database construction, system integration, software development, legal arrangement, security, and distribution systems.

Since the early 1990s, Seoul City has studied strategic plans, pilot systems, and sample application programs for GIS implementation. Using MoCT's Order 17, Seoul City has built GIS databases, and they are almost completed. MoCT's Order 17, however, groups geographic entities into four levels only with entity type Ids; therefore, it is virtually impossible to find relationships among features in the same grouping level. The layer-based cartographic GIS databases may be good for mapping applications; however, they could be more efficient and semantically rich in information when they have relationships among entities and attributes. This research investigated how the relationships can be implemented for more efficient urban land and building management using the entity-relationship model and relational database model.

Land parcel, building, and owner entities and their relationships were modeled using the entity-relationship model. Using the Seoul GIS database constructed upon MoCT's Order 17 on Digital Mapping, an example of conceptual and logical data models was constructed. For an effective land and building management, three entities were identified: Owner, Land, and Building. Also, six relationships were identified: Land Ownership, Mortgage, Lease, Building, Attached Building, and Building Ownership. Attributes were modeled into three types: Single Attribute, Composite Attribute, and Set Attribute. Using the entities, relationships, and attributes, this research investigated how to develop conceptual and logical GIS database models.

REFERENCES

BERNHARDSEN, T. 1999. Geographic information systems: An introduction (second edition). New York: John Wiley & Sons.

CHANG, Y. H. 1992. A study on geographic information systems for Seoul. Seoul Development Institute Research Report 92-P-01.

CHANG, Y. H., E. H. KIM, S.J. MOON, D. Y. LEE, S.0. YANG, AND Y. S. CH0, et al. 1993. A study on the implementation of geographic information systems for Seoul I. Seoul Development Institute Research Report 93-R-25.

CHANG, Y. H., Y. J. KIM, Y. O. KANG, D. Y. LEE, AND S. YANG. 1996. A technical proposal to GIS database construction for Seoul metropolitan government. Seoul Development Institute Research Report 96-R-35.

CHANG, Y. H., YU, J. Y., SEONG, J. C., AND HWANG, M. Y. 1995. Comprehensive public management information system plan for Seoul metropolitan government. Seoul Development Institute Research Report 95-R-35.

CONNOLLY, T., BEGG, C., AND STRACHAN, A. 1996. Database systems: a practical approach to design, implementation and management. New York: Addison-Wesley Longman.

KANG, Y. O., J. Y. AHN, AND T. Y. CHO. 1997. A technical guidelines for construction of cadastral and urban plan database in Seoul. Seoul Development Institute Research Report 97-R-35.

KANG, Y. O., J. Y. AHN, and E. M. KIM. 1998. A study on the creation and use of digital orthophoto for Seoul. Seoul Development Institute Research Report 98-R-30.

KANG, Y. O., T. H. CHO, M. Y. HWANG, J. Y. LEE, J. Y. YU, AND J. S. WON. 1996. Database development guidelines for the effective management of underground facilities in Seoul. Seoul Development Institute Research Report 96-R-36.

KIM, E. H., D. Y. LEE, S. J. MOON, Y. S. CHO, J. C. SEONG, AND M. Y. HWANG. 1994. A study on the implementation of geographic information systems for Seoul II. Seoul Development Institute Research Report 94-R-32.

KIM, Y. J., S. H. PARK, S. M. LEE, AND C. H. SONG. 1998. A study on the management and maintenance of digital base-maps in Seoul. Seoul Development Institute Research Report 98-R-29.

KIM, Y. J., C. H. SONG, and J. S. WON. 1997. A guideline for construction of urban disaster information system using GIS. Seoul Development Institute Research Report 97-R-36.

MOCT (MINISTRY OF CONSTRUCTION AND TRANSPORTATION, SOUTH KOREA). 1999. Proposed Act announcement on the construction and use of national geographic information systems. MoCT Announcement No. 1999-235.

-----. 1995. MoCT 17--Order about digital mapping works,

NGI (National Geography Institute). 1995. NGI Bylaw about Digital Mapping Works. NGI Bylaw 71.