The Geographical Ontology Page

3.1 General

In the geographic discipline, in both the physical and human sectors, there is a need for interoperability between researchers at different sites and different computer systems. The development of a definitive and authoritative nomenclature of the geospatial domain – an ontology of the geospatial domain is one way to achieve this state of interoperability. Ontology design is grounded on the idea that a knowledge base can be defined through the development of a set of unique, domain-specific concepts for objects and processes. A concept is an idea or notion that we apply to classify those things around us. For instance, if we were to list all those objects to which the concept of “being mortal” applies, we would form the set of all mortals. If we were to list all of those objects that relate to geospatial information, we would form the set of all things geographic. The following discussion of ontology design principles sets the groundwork for the development of just such a data structure – objects and concepts describing geospatial information. Because the final product has taken the form of both a hierarchical taxonomy (and ultimately an associated formal thesaurus), this data structure will be referred to as the Visual Objects Taxonomy/Thesaurus (VOTT).

The essence of any ontology development project is to form a set of link-node relationships between categories in category space and their corresponding concept nodes in concept space. Figure 1 illustrates the relationship between category space and concept space. In category space, there exists a unique set of categories within each different classification schema. In concept space, a set of domain specific concepts exists, as an interlinked network of nodes between and within domains. Based on existing equivalency between concept definitions and category meanings, each node in category space can be linked to its corresponding concept node in concept space. By explicitly defining those links between category space and concept space, a formal ontological data structure can be created.

In the case of geospatial information, category space can be thought of as the set of all heritage classification schemas. These heritage classification schemas take the form of a wide variety of existing map legends, word lists, dictionaries, thesauri, taxonomies, and ontologies. By reconciling all of these different forms of classification schema into one single, integrated, non-duplicitous ontological structure, a unified knowledge base of geospatial information can be achieved. Once mediated, this unified knowledge base would then represent all of those objects that are currently addressed in the broad expanse of geospatial information data handling. However, Patrick Hayes has estimated that this comprehensive list of both scientific and common sense objects would be in excess of 10,000 unique conceptualizations [Hayes, 1985]. Arriving at this list of 10,000 unique objects would require the systematic analysis of over 30,000 categories – all the accumulated categories from each of the available heritage classification schema. The systematic transformation of these 30,000 heritage categories, into a logical, hierarchical, non-duplicitous form, the actual mediation of this large data set, this is the subject of this research effort.

The Visual Objects Taxonomy/Thesaurus (VOTT) was created using existing classification schema as the basis for the hierarchical structure. Table 3, at the end of this document, lists those 173 heritage classification schemas used so far in this effort. The list includes a wide variety of classification schema to include systems for land cover/land use inventory, property management, urban planning, facility inventorying, wetlands mapping, mapping symbologies, industrial and occupational codes, standard data models, and ontologies of various types.

The primary intent in developing the VOTT has been to devise a system that allows the efficient and logical inventorying of natural and cultural features to such a level of detail that all those cultural and natural features seen while walking through a typical natural or cultural landscape could be inventoried. These features, for lack of a better term are tangible features because they have substance and occupy visible space on the landscape. A secondary intent has been to allow the exploitation of many other forms of readily available geospatial information generated outside the traditional mapping and charting areas of interest. The widespread use of Geographic Information System (GIS) technology has introduced a wealth of geospatial data into the public and private sectors. However, the GIS community is concerned with a more diverse set of objects than just those objects that are visible on or above the terrain. Routinely, GIS practitioners collect information on less tangible object than the mapping community does. These intangible features, i.e. events, situations, phenomena, and objects that are hidden from view, are an essential part of the overall human and physical geography discipline [Bitters, 2002]. The VOTT includes both tangible and intangible features. It was created by merging a wide variety of existing classification schema (Table 3) using a semantic integration process.

The VOTT, in its current state is an extensive list of concepts that have been, or are currently used in a broad range of disparate classification schema. The VOTT defines a geospatial nomenclature in the form of defined concepts of objects that span much of the physical and human geographic discipline. It contains more than 13,000 different entries. Approximately half of these concepts include explicit natural language definitions, associations, and relationships to other concepts within the data set.

Each VOTT concept not only has a unique textual name, but also has a unique short name in the form of an eight-digit, hierarchical, numeric short name. Figure 2 illustrates the generalized hierarchical structure of the VOTT and provides an example of the naming convention for the numeric short names. Each VOTT concept is referenced by a unique eight-digit, numeric code. The two left most digits indicate the top-level VOTT group designation. To the right, the next two digits indicate a class value. To the right of the class, the next two digits reference a subclass value and the right-most two digits identify the unit value. In the example above, the concept, “Gulch” would have a VOTT short name designation of 13020470.

The Visual Objects Taxonomy (VOT) Data Structure

The working version of the Visual Objects Taxonomy (VOT) is stored as a relational database using Microsoft Access^TM. Figure 3 shows the current structure of the VOT working database. The Heritage Description Document is a database of all heritage data structures that have been used in the development during this project. It contains bibliographic references to each heritage data structure and a unique identifier for each record. Heritage Master Databases have been created for each of the heritage taxonomic structure used in this project and each contains the original classification schema used to describe each heritage data set. For those heritage data structures that contained definitions, those heritage definitions have been have been stored in the VOT Definitions database and have been used as a starting point in developing VOTT compliant definitions. The same is true for heritage data that contained explicit association and relationship data. The final VOT hierarchical structure is stored in the Master VOT database with relational links to the VOT Definitions, VOT Associations, and VOT 3-D Models databases. Provisions have been made for the future addition of relational links to data models to support several GIS and modeling and simulation software packages – SEDRIS^TM, Terrain Experts (TerraVista^TM), MultiGen® (Creator Terrain Studio^TM) and ESRI (ARC/INFO and AarGIS^TM). Additionally, provisions have also been made to allow the future generation of Heritage Conversion Tables to capture the affiliation of each record in these heritage taxonomic structures to its counterpart record in the VOT. This will allow the generation of a near “lossless” data conversion capability for future implementations of the VOT data structure.

Output Formats

The distribution version of the VOTT contains only valid VOTT classes; their definitions, associations, relationships; associated 3-D models; and bibliographic references for the derivation of each concept definition. As a separate directory structure, the geometry and texture for the associated 3-D models from the VOTT 3-D model library are also available. This version of the VOTT is available in several different data formats shown in Table 1. The taxonomy is distributed as a single relational database in Microsoft Access^TM in standard .mdb format. It will be distributed with a unique user’s interface - the VOTT Browser that will allow searching the entire database for keywords and phrases. The taxonomy database will also be available in standard eXtensible Markup Language (XML). The VOTT thesaurus hierarchical and full file listings are available in ASCII text, XML, and HTML format. The 3-D model library is composed of a file directory containing model geometry files, each in standard OpenFlight^TM format [MultiGen-Paradigm, 2000] and a file directory containing model texture files in standard SGI image format.

Table 1. Formats Used to Distribute the VOTT Data.

Data Set	Format
TAXONOMY	MSAccess^TM Database in .mdb format
TAXONOMY	MSAccess^TM Database in .xml format (Not Yet Available)
THESAURUS	ASCII text format (Not Yet Available)
THESAURUS	.xml format (Not Yet Available)
THESAURUS	HTML format (Not Yet Available)
3-D MODELS	OpenFlight^TM format
MODEL TEXTURES	SGI image format (.rgb .rgba .int .inta)

VOTT Top-Level Groups

Table 2 identifies the 48 top-level groups used in this version of the VOTT. Of the 48 top-level groups in the taxonomy, the first 15 groups represent those broad top-level categories of cultural and natural features that have been traditionally used in mapping and charting classification schema – in particular in the Digital Feature Analysis Data (DFAD), Feature and Attribute Coding Catalogue (FACC), and most recently in the SEDRIS^TM EDCS classification schema. Groups 16 through 18 are an extension to these traditional groups and address the categorization of all forms of vehicles, human forms, and animal forms. The objects within the first 18 broad top-level groups represent the preponderance of feature classes for those tangible objects that would be encountered on the Earth’s surface. Groups 19 through 47 contain a set of classes for various forms of tangible and intangible feature data that can be encountered in the GIS community. Group 50 contains a set of standard units of measure and group 70 contains a set of non-feature concepts and their related definitions.

Table 2. The Visual Objects Taxonomy/Thesaurus (VOTT) Top-Level Groups.

ID	Top-Level Class	Status	Number of Valid Concepts	Number of 3-D Models
1	Industry	Complete	1867	59
2	Transportation	Complete	1343	334
3	Commercial	In-Review	851	124
4	Recreational	Complete	460	32
5	Residential	Complete	377	124
6	Agricultural	Complete	330	25
7	Communications	Complete	128	14
8	Governmental	Complete	234	30
9	Institutional	Complete	428	58
10	Military	In-Review	571	68
11	Storage	Complete	229	26
12	Hydrography	Complete	885	16
13	Physiography	Complete	592	3
14	Vegetation	Complete	233	37
15	Miscellaneous	In-Review	62	5
16	Vehicles	In-Work	1246	37
17	Human Forms	In-Work	34	0
18	Animal Forms (except Homo sapiens)	In-Work	105	1
19	Demarcation	In-Work	96	0
20	Map Symbology	In-Work	12	0

21	Crime & Law Enforcement	In-Work	241	0
22	Parks and Recreation	In-Work	200	160
23	Animal Studies	In-Work	10	0
24	Urban Studies	In-Work	26	0
25	Forestry	In-Work	20	0
26	Geology	In-Work	60	0
27	Atmosphere & Climate	In-Work	130	0
28	Real Estate	In-Work	10	0
29	Hazards and Hazmat	In-Work	27	20
30	Utilities	In-Work	207	0
31	Wetlands	Complete	301	0
32	Health and Disease	In-Work	44	0
33	Pollution	In-Work	31	0
34	FGDC Vegetation	Complete	704	0
35	FGDC Vegetation cont.	In-Work	10	0
36	Landmarks	In-Work	7	0
37	Landscaping	In-Work	5	0
38	Zoning	In-Work	4	0
39	Conservation	In-Work	18	0
40	Survey Control	In-Work	70	0
41	Archeology	In-Work	35	0
42	Parcels	In-Work	78	0
43	Land Use /Land Cover	In-Work	31	0
44	Soil Science	In-Work	75	0
45	Census	In-Work	67	0
46	Space	In-Work	52	0
50	Units of Measure	Complete	210	N/A
70	Non-Object Definitions	Complete	909	N/A

As an example of the specificity and granularity of this data structure, Figure 4 identifies the top-level classes within the Physiography Group. This group includes concepts concerning the types of natural material surfaces that may be encountered and the types of physiographic features (landforms), and topographic symbologies that are commonly used to portray the Earth’s surface.

As a demonstration of the increased granularity and specificity of the VOTT compared to other heritage classification systems, Table 4 compares the number of feature concepts for the Physiography group in the VOTT with the number of Physiography concepts in SEDRIS, FACC, DFAD, and SDTS. Notice that there is a 10 to 30-fold increase of unique Physiography concepts in the VOTT compared to any of the heritage systems; more precisely, an 831% increase over SEDRIS, a 1364% increase over FACC, a 2800% increase over DFAD and a 1716% increase over SDTS. In the VOTT data set, this increase in the clarity can be seen across the entire spectrum of traditional top-level groups. This increase is primarily because the VOTT was created by performing an intersection of all heritage categories and contains a superset of all physiographic categories used within all the different heritage classification schemes.

Table 4. Physiography concepts in the VOTT and Physiography categories in selected heritage classification systems.

Taxonomic Structure	Total No. of Categories	No. of Surface Categories	% of Whole	Percent Increase in Concepts
VOTT	12,500	532	4.2
SEDRIS	1225	64	5.2	831%
FACC	550	39	7.0	1364%
DFAD	309	19	6.1	2800%
SDTS	201	31	15.5	1716%

Concept Naming

Concept names have not been created in natural language text, but are a concatenated form without spaces and using initial uppercase characters. Although this is a difficult format for human understanding, it is readily machine recognizable. This convention was adopted to insure that naming was compatible with existing upper-level ontologies. To insure that this ontology could be merged with other existing ontologies, all concept names were compared to those used in the top-level and mid-level Suggested Upper Merged Ontology (SUMO). When VOTT concepts were found to correspond exactly to SUMO concept, SUMO names were adopted in the VOTT.

Concept Definitions

Developing explicit definitions has been a very challenging component of this dissertation project. One would think that when establishing any classification schema, the subject matter expert would have generated natural language definitions for each class. During the initial conceptualization of this project, we assumed that explicit definitions would be available for the preponderance of the categories within heritage schema. However, in many of the older heritage systems used here, the original authors depended on stand-alone class names to define their concepts and for this reason, many heritage systems contained no explicit definitions. Those subtle nuances in meaning that can be expressed in a natural language definitions, of course were missing. Therefore, deciphering exactly what was meant by each stand-alone class name has often been problematic. To say the least, this has made it exceedingly difficult to decipher the precise meaning of many heritage class concepts.

However, in the IT ontology environment it is commonplace to create very large ontological structures based purely on words devoid of definitions. In these situations, lexical differences and association differences are the only factors that can be used in the creation of a hierarchy of concepts. When available, definitions provided the following factors in the overall structure of the VOTT data set:

• Clarification of concept meaning,
• Clarification of natural language concept names,
• Linkage to a circular network of knowledge.

Throughout the semantic integration, our approach was to use definitions of categories, whenever they were available, as the primary determinant of class structure. Further, because definitions could provide concept-meaning clarification through a linkage to other terms, definitions have been considered a critical element in the overall VOTT design.

Figure 5 provides an example of a formal VOTT definition. Definitions are composed of two parts: the explicit natural language definition and as a suffix, an explicit natural language form of the concept name (highlighted in blue). An unconcatenated, natural language concept name is suffixed onto the definition to insure that there is a full understanding of the expanded concept name – expanded from the concatenated computer form. Embedded within each definition are XML tagged keywords (highlighted in red). The opening (<KW>) and closing (</KW>) XML keyword tags identify those words within the definition text that are further defined within the taxonomic data structure. This will allow for future automated identification of keywords that can be used to expand the definitions into a broad networked knowledge base.

Table 1. Heritage Taxonomic Data Structures Used in this Research Effort

The following 173 heritage taxonomic structures have been used to varying extents during the development of the VOTT data structure. This list includes formal ontological domain studies, taxonomies, thesauri, classification schemes, formal map legends, geospatial data models, and basic word lists developed for purely information technology purposes.

ID No.	Data Set Name	Short Name	Responsible Agency
1	Land Cover/Land-use Determinants, Impact	HUD_lulc	U.S. Department of Housing and Urban Development (HUD)
2	Land Cover Classification System	UNFAO	United Nations Food and Agriculture Organization
3	U.S. Department of Energy (USDOE): Site Development Planning Order	USDOE	U.S. Department of Energy (DOE)
4	Anderson Land-Use Land Cover Classification System: Levels I and II	Anderson	U.S. Geological Survey (USGS)
5	Atlanta Regional Commission, Georgia: Land Use Information System	Atlanta	Atlanta Regional Commission
6	Aurora, Colorado: Land-Use Map Categories	AuroraCO	Aurora Division of City Planning
7	Barton Aschman and Associates: Summary of Recommended Activity Coding System	Barton	Barton Aschman Associates
8	Cape Cod Commission: Summary of Land Uses	CodSLU	Cape Cod Commission
9	Cape Cod Commission: Regional Plan Atlas, Land-Use Map Color Codes	CodCC	Cape Cod Commission
10	Cape Cod Commission: Summary Description of Massachusetts Property Type Classification Codes	CodPTC	Cape Cod Commission
11	Chicago Fire Department: Occupancy Hazard Classification of Properties	Chicago_FH	Chicago Fire Department
12	Clark County, Nevada: Land-Use Categories for Use with	ClarkCO_LU	Clark County, Nevada Comp. Planning Dept.
13	Clark County, Nevada: Coding Scheme for Existing Land Uses	ClarkCO_EL	Clark County, Nevada Comp. Planning Dept.
14	Clark County, Nevada: Classification for Land-Use Compatibility in the Airport Overlay	ClarkCO_AC	Clark County, Nevada Comp. Planning Dept.
15	Cleveland, Ohio: Categories for the Future Land-Use Map	ClevelandCC	Cleveland Planning Department
17	International Association of Assessing Officers - Standard on Property Use Codes	IAAO_PU	International Association of Assessing Officers
18	Institute of Transportation Engineers: ITE Manual, 5th Edition	ITE_Man	Institute of Transportation Engineers
19	Cobb County, Georgia: Categories for Existing Land-Use	CobbCO	Cobb County, Georgia
20	CORINE Land Cover Nomenclature: Levels I, II, and III	Corine	European Commission
21	Cleveland, Ohio: Categories for the Future Land-Use Map	DechCC	Pratt Institute
22	Color Coding of Land Uses from DeChiara, Simplified Scheme	DechSCC	Pratt Institute
23	Land-Use Inventory Categories	Denver	Denver Regional COG
24	Dona Ana County, New Mexico: Assessors Department Land-Use Codes	DonaAC	Dona Ana County Community Dev. Dept
25	Dona Ana County, New Mexico: Land-Use Codes, Expanded	DonaEL	Dona Ana County Community Dev. Dept
26	Dona Ana County, New Mexico: Planning Department Land Uses	DonaPD	Dona Ana County Community Dev. Dept
27	DuPage County, Illinois: Land-Use Coding System - Combined Coding System	DuPageCO	DuPage County Development Dept.
28	Eagle Point Software: Graphic Database Structure Color Codes	EagleCC	Local Affairs Inc.
29	Eagle Point Software: Graphic Database Structure Hatch Style	EagleH	Local Affairs Inc.
30	Eau Claire: Land-Use Plan Categories	EauClaire	Eau Claire County Planning and Development Dept.
31	Fairfax County, Virginia: Coding Scheme for Conceptual & Area Plan Land Uses	FairfaxLU	Fairfax County Office of Comprehensive Planning
32	Fairfax County, Virginia: Coding Scheme for Existing Land Uses	FairfaxELU	Fairfax County Office of Comprehensive Planning
33	Fairfax County, Virginia: Coding Scheme for Planned Land Uses	FairfaxPLU	Fairfax County Office of Comprehensive Planning
34	FEMA: HAZUS, Airport System Classifications	HAZUS	U.S. Federal Emergency Management Agency (FEMA)
35	FEMA: Rapid Visual Screening of Buildings for Potential Seismic Hazards, Building Structure Codes	Screen	U.S. Federal Emergency Management Agency (FEMA)
36	Federal Geographic Data Committee Cadastral Data Content Standard for Parcel Type, Parcel Area Type, and Restriction Type	FGDC-P	U.S. Federal Geographic Data Committee (FGDC)
37	Federal Geographic Data Committee Types of Facilities (Informative)	FGDC_Fac	U.S. Federal Geographic Data Committee (FGDC)
38	Federal Geographic Data Committee Ground Transportation Network and Attributes	FGDC_G	U.S. Federal Geographic Data Committee (FGDC)
39	Federal Geographic Data Committee Utilities Feature Classes	FGDC_U	U.S. Federal Geographic Data Committee (FGDC)
40	Federal Geographic Data Committee National Vegetation Classification Standard	FGDC_V	U.S. Federal Geographic Data Committee (FGDC)
41	Guttenberg’s Multiple Land Use Classification System	Gut	University of Illinois - Urbana-Champaign
42	Coastal Change Analysis Program (C-CAP)	CCAP	U.S. National Oceanic and Atmospheric Agency
43	Land-Use Compatibility Zones Coding	AFCC	U.S. Department of the Air Force (USAF)
44	Hacienda & Rowland Heights Area Plan Classification	LAC_HAPC	County of Los Angeles
45	Santa Clarita Valley Area Plan Classification	LAC_SAPC	County of Los Angeles
46	Lincoln-Lancaster County, Nebraska: City County Planning Department Land-Use Inventory Coding Conventions	LincNE	Lincoln-Lancaster County Planning Dept.
47	MacConnell Land Cover Categories	MAC_LCC	University of Massachusetts - Amherst
48	Water Resources Administration: Wetlands Mapping	MDDNR	Maryland Dept. of Natural Resources
49	Massachusetts Executive Office of Environmental Affairs: MASSGIS	MassGIS	Massachusetts Dept. of Environmental Protection
50	Michigan Resource Information System	MichDNR	Michigan Department of Natural Resources
51	Missouri Land Cover Classification Scheme	MissLC	University of Missouri
52	Région dlle-de-France Categories for Land-Use Modes	RFR-LU	Institut dAméngagement dUbanisme
53	North American Industry Classification System (NAICS)	NAICS	U.S. Department of Labor (DOL)
54	Nancy, France: Urban Ecosystem Classification System	Nancy	France Ministry of the Environment
55	North Carolina Southeast Chapter of the A.I.P.: Color Code Scheme for "A Proposal for a Standardized Classification"	NCSAIP	APA - North Carolina Chapter
56	North Carolina State Standard for Land-Use Categories	NCLUC	North Carolina Dept. of Natural Resources
57	North Carolina DNR: Superconductor/Supercollider Project	NCDNR	North Carolina Dept. of Natural Resources
58	North Carolina State Standard for Land Cover Categories	NCLCC	North Carolina Dept. of Natural Resources
59	Northern Kentucky Area Planning Commission: Land-Use Codes	NKAPC	Northern Kentucky Area Planning Commission
60	Ohio DNR, Remote Sensing Program: Land Use/Land Cover Classification	OHDNR	Ohio Remote Sensing Program
61	Orange County, California: Coding Scheme for Existing Land Uses	OCCA	Orange County, California
62	Palm Beach County, Florida: Combined Property Appraiser Codes	PBFL-CPAC	Palm Beach County Property Appraiser
63	Palm Beach County, Florida: Land-Use Codes for the Existing Land-Use Database	PBFL-ELUC	Palm Beach County
64	SANDAG: 1968 Standard Land-Use Codes	Sandag-68	San Diego Association of Governments
65	SANDAG: Generalized Land Ownership Map Categories	Sandag-	San Diego Association of Governments
66	SANDAG: Property Use Codes	Sandag-PUC	San Diego Association of Governments
67	Southern California Assn. of Governments: 1990 Aerial Land Use Study, Land Use Level III/IV Classification	SCAG	Southern California Assn. of Governments
68	Standard Industrial Classification Manual (SIC)	SIC	U.S. Office of Management and Budget (OMB)
69	Standard Land-Use Coding Manual (SLUCM)	SLUCM	U.S. Department of Commerce- Bureau of Econ. Analysis
70	St. Louis, Missouri: Land Records Management System	StLoMO	St. Louis Community Development Agency
71	Classification System for Substandardness Criteria	UrbIL	University of Illinois - Urbana-Champaign
72	U.S. Air Force Center for Environmental Excellence: Suggested Land-Use Capability in Accident Potential Zones	USAF_Sugg	U.S. Department of the Air Force (USAF)
73	U.S. Air Force Land-Use Compatibility Zones Coding	USAFLUC	U.S. Department of the Air Force (USAF)
74	U.S. Air Force Land Use Planning Bulletin: Facilities List	USAFFac	U.S. Department of the Air Force (USAF)
75	U.S. Air Force Center for Environmental Excellence: People per Parking Space Land-Use Categories	USAFPark	U.S. Department of the Air Force (USAF)
76	U.S. Army: Guide to Army Real Property Codes	USARPC	U.S. Department of the Army
77	U.S. Department of Agriculture (USDA): Soil Classifications, Farmlands Classification	USDAFarm	U.S. Department of Agriculture - NRCS
78	U.S. Geological Survey Digital Line Graph - Enhanced	DLG	U.S. Geological Survey (USGS)
79	Washoe County, Nevada: Integrated Terrain Unit Mapping Classification System	Washoe	Washoe County, Nevada
80	Whittier College, Categories for the Environmental Justice Whittier Project	Whittier College
81	WISCLAND: Land Cover Classes	WISCLAND	Wisconsin Department of Natural Resources
82	SEDRIS: Environmental Data Coding System (EDCS)	EDCS_CLASS	www.SEDRIS.org
83	U.S. Defense Intelligence Agency Functional Category Codes (TDI Catcodes)	TDI_cat	U.S. Defense Intelligence Agency
84	Canada Council on Surveying and Mapping (CCSM)	SSCM_Clas	Canada Council on Surveying and Mapping
85	FACC	FACC_Code	DIGEST
86	MDA Railroads	MDA-R	MDA
87	National Wetland Inventory (NWI)	NWI	U.S. Fish and Wildlife Service (USFWS)
88	Digital Line Graph (DLG)	DLG	U.S. Geological Survey (USGS)
89	TigerLine Files	TIGER	US Census Bureau
90	National Vegetation Classification System	NATVEG	U.S. Federal Geographic Data Committee
91	Landuse Compatibility Codes	AFCC	U.S. Department of the Air Force (USAF)
92	Standard Occupational Classification System	SOC	U.S. Bureau of Labor Statistics (BLS)
93	Coastal Change Analysis Program	CCAP	U.S. National Osceanic and Atmopheric Administration
94	Industry and Occupational Codes	IOC	U.S. Census Bureau
95	EuroRegionalMap	ERM	EuroGraphics.org
96	Digital Feature Analysis Data (DFAD)	DFAD	National Geospatial Intelligence Agency (NGA)
97	U.S. Defense Information System	DDDS	U.S. Defense Information System
98	EuroGlobalMap	EGM	EuroGraphics.org
99	California Wetlands Classifications	CWC	University of California Santa Barbara
100	U.S. Federal Geographic Data Committee	FGDC	U.S. Federal Geographic Data Committee
101	Geosym	GEOSYM	National Geospatial Intelligence Agency (NGA)
102	Geographic Names Information System	GNIS	U.S. Geological Survey (USGS)
103	ISO-DIS-19110	ISO19110</