Sigcat Software Showcase 1993/Sigcat 93 Software Showcase dataDisc.ISO/research/library/docs/scchp9.asc

Text File | 1993-02-08 | 21.2 KB | 474 lines

                     CHAPTER NINE

        DEVELOPING AN SGML DATABASE FOR CD-ROM



           SGML PROOF-OF-CONCEPT PROTOTYPE
                DEFENSE MAPPING AGENCY
              DIGITAL SAILING DIRECTIONS

                     Walter Klaus
         Defense Mapping Agency Systems Center
                      Fairfax VA

                    Ronald Hawkins
    Science Applications International Corporation
                       McLean VA


     Abstract: The Defense Mapping Agency (DMA) is
     currently developing a Text Product Standard
     (TPS) to support the production, distribution
     and use of DMA publications in a digital
     environment [1]. Key features of the TPS
     include use of the Standard Generalized Markup
     Language (SGML) for text structuring and
     Compact Disc Read-Only Memory (CD-ROM) as the
     distribution media. This paper provides an
     overview of the effort, including the TPS
     proof-of-concept prototype on CD-ROM.


INTRODUCTION

The TPS is one leg of a standardization triad under
development by DMA for its digital data products. The
other two legs are the Vector Product Standard (VPS) and
the Raster Product Standard (RPS). VPS will be the
standard for DMA's digital vector data products and is in
the final stages of development. RPS will be the standard
for raster and gridded data products and is in the
preliminary stages of development. A short discussion
follows.
     Besides producing paper maps, charts, and supporting
publications for the Merchant Marine and the military
services, DMA already produces some digital data products
[2]. These products include vector, raster, and gridded
Mapping, Charting, and Geodesy (MC&G) data supporting
command, control, communications, and intelligence (C3I)
systems, precision-guided weapon systems, simulators,
training, scientific analyses and other applications.
     Examples of vector data products using the standard
Vector Product Format (VPF) include the new Digital Chart
of the World (DCW), a 1:1,000,000 scale worldwide
database of geographic features; World Vector Shoreline
(WVS), a 1:250,000 scale worldwide database of
shorelines, international boundaries, and country names
(now being converted to VPF); and, the Digital Nautical
Chart (DNC) for electronic navigation and chart display
currently under development.
     Examples of raster data include scanned color images
of paper maps and charts such as ARC Digitized Raster
Graphics (ADRG). Gridded data includes Digital Terrain
Elevation Data (DTED), a worldwide database of land
elevations and Digital Bathymetric Data (DBD), a
worldwide database of ocean depths.
     Of the products mentioned above, ADRG and DTED are
currently distributed on CD-ROM; DCW will be distributed
on CD-ROM shortly; a recent prototype CD-ROM for WVS is
now being converted to VPF. CD-ROM is an ideal
distribution media because of its low cost, high storage
density, and adherence to ISO standards which provide
computing platform independence. To facilitate the
production and interchange of these data products and the
interoperability of systems using them, DMA has embarked
on this program to standardize data structures and
formats.
     The TPS is the third major standard under
development by DMA and is at the forefront of a paper-to-
digital transition of DMA's publication development and
distribution environment. DMA produces a number of
publications related to mapping and charting. These
include the Sailing Directions and Fleet Guides, which
are voyage planning and navigation publications for
Merchant Marine and U.S. Navy ships; the Chart Update
Manual (CHUM), which provides listings, updates, and
corrections to DMA aeronautical charts; and the
Gazetteer, which is a publication containing place names
and related information worldwide. Efforts are underway
to produce digital versions for each of these textual
publications; they are further described later in this
paper. Developing a standard for digital text products
benefits both the producer and the users. The producer
benefits because the text and graphics data for each
publication is based on standard data structures,
allowing a common set of tools to be used for the entry
and editing of this data and permitting the transfer of
data among workstations at various stages in the
production process. In addition, since DMA exchanges
production data with other nations, adopting standard
data structures will facilitate international exchange
and support interoperability as well.
     The user benefits because development of the TPS
will allow DMA to distribute text products as self-
contained publications on CD-ROM. These "intelligent
publications" will include text and graphics data
integrated with retrieval software supporting access
techniques such as browsing, query, and hypertext. In
addition to providing an alternative to printed
publications, the intelligent publication concept will
support the integration of DMA's digital text products
with other computerized information systems. Standards
being considered and adopted under the TPS umbrella
include data structures and data access language
facilities which will support interoperability of
intelligent publications and text databases with other
standards-based digital data products.
     One of the key benefits of the DMA standardization
program is the eventual integration of vector, raster,
and text data in common systems. The capability to
reference and view DMA publication data by pointing and
clicking a mouse on a vector or raster map display offers
significant future potential for C3I, navigation,
training, and other automated systems.


EXISTING STANDARDS

DMA's formal hierarchy of standards, including existing
and planned ISO and ANSI standards, form the foundation
of the TPS; this functionally layered hierarchy is
reflected in the "TPS Meta-Model" (Figure 1). The DoD
CALS/CIM concepts influenced early TPS conception. The
extensive use of formal and defacto standards supports a
highly flexible interface architecture (Fig. 2) based on
the OSE/OSI standard models, provides independence from
storage (media) devices [3] and operating systems [4],
and facilitate COTS-based engineering. The TPS will
provide the capability to represent not only text, but
also embedded graphics in 'standard' vector and raster
formats. Standards which form the basis for TPS include:

Standard Generalized Markup Language (SGML) (ISO 8879)

     It provides a standard means for defining and
representing document structure using a set of 'tags'
placed within the document text to signify structural
components such as titles, chapters, paragraphs,
illustrations, etc.; SGML is the core standard for the
TPS [5].
     The development of common Document Type Declarations
(DTDs) (the formal expression of document structure in
SGML) will clearly define publication structure,
providing for controlled change as well as validation and
configuration management. Maintenance of change authority
and history logs ('audit trails') will also be supported,
an important consideration for many DMA publications.
     A key feature of SGML is its capability to reference
external entities from within the tagged text. This
capability will support the eventual integration of DMA's
textual products and vector/raster map data products. For
example, an SGML-structured digital Sailing Directions
could reference a DNC 'database'. The Sailing Directions
user could, by pointing and clicking with a 'mouse', call
up a section of DNC for his area of interest; he would
then be viewing the same chart he uses to navigate,
instead of a 'picture' of a chart, which is currently the
case. This type of capability would lead to greater
consistency between products, reduced chance for error,
and eventual development of integrated voyage planning
and navigation systems.
     SGML is also expected to provide a capability to
interface publication databases used for in-house
production to other DMA systems. An example is the
marriage of digital Sailing Directions to DMA's
Navigation Information Network (NAVINFONET), a special
online service to mariners providing remote access to
extensive maritime safety information. It can be queried
from anywhere in the world via modern communications
equipment on a 24-hour basis. It supports and supplements
a number of navigation publications, including the
Sailing Directions. SGML provides the capability to
reference processing instructions which download Sailing
Direction updates from the master database to the
NAVINFONET computers; such an arrangement would provide
faster and more accurate updates.
     An area in which SGML promises to be of great
benefit is that of defining standard text data structures
which can be used with intelligent retrieval systems.
Systems currently available convert input textual data
into a structure specific to the vendor's proprietary
retrieval software; the data is then indexed for direct
access and query capability. Some of these systems accept
SGML as an input structure, but converted data is not
SGML-structured. Retrieval systems could be developed
which index SGML-structured data directly, eliminating
the need to convert to vendor-specific structures. This
would promote interoperability because data supplied on
CD-ROM could be re-indexed for a different (or even
multiple) retrieval engine(s), if desired by the end-
user. The retrieval engine vendor's data preparation
software would be used to add the necessary index files,
but no data conversion would be required. Ultimately,
standard index structures for SGML-tagged data should be
developed to provide a completely open approach to text
retrieval [6].

Volume/File Structure of CD-ROM for Information
Interchange (ISO 9660)

It defines the physical organization of stored files and
related volume directories on CD-ROM. It also supports
the development of operating system-independent
interfaces to data stored on CD-ROM media. CD-ROM was
selected as a primary distribution media because of its
low cost, high storage density, platform independence,
and its potential for integrating other DMA digital
products [7].

Tag Image File Format (TIFF)

It is an industry (defacto) standard for raster images
using a flexible tag and directory scheme which is
extensible without sacrificing compatibility. TIFF will
be used for raster drawings and illustrations [8].

CCITT-Group/4 compression (FIPS PUB 150)

It provides a standard algorithm for the compression of
black-and-white raster images. It will be used in
conjunction with TIFF. providing standard compressed file
structures for black-and-white images [9].

Computer Graphics Metafile (CGM) (ISO 8632)

It is a formal standard for the representation of vector
graphics. It will be used for drawings and illustrations
which are created or maintained as vector data [10].

Data Access Languages

These standards (under development), attempt to define
protocols for accessing data on CDROM from retrieval
systems. Adoption of standard data access languages and
client-server architectures will separate user
applications ("front-ends") from text databases on CD-
ROM, allowing different standard protocol-compliant
applications to access the same database. This supports
interoperability and reduces the requirement to have a
unique user interface and means of accessing text data
for every CD-ROM database which is produced. The
situation is analogous to the use of Structured Query
Language for accessing traditional databases. Emerging
standards which are being considered include (among
others) the Compact Disc Read Only Data Exchange Standard
(CD-RDx) under development by the Intelligence Community
[11] and the Structured Full Text Query Language (SFQL)
under development by the Air Transport Association [12].


PROTOTYPE DEVELOPMENT

DMA is approaching the development of TPS with a
prototyping effort designed to validate: (1) the
standards and technologies and (2) the proposed
client/server architecture (Fig. 3) selected for the TPS.
It will also provide DMA publication users with a working
version of a TPS-based electronic publication for
evaluation and comment. The initial proof-of-concept
prototype, including an intelligent publication on CD-ROM
and a publication database consisting of SGML-tagged text
and graphics in 'standard' formats, is planned for
completion in May, 1992.
     The current, non-digital 47-volume series of Sailing
Directions uses an oceanic basin concept and provides,
for each basin, a planning guide with the oceanographic,
meteorological, route, and other information required for
an ocean passage. Three to seven enroute publications
accompany each oceanic planning guide to provide textual
and graphic information, including coastal views and
photographs required for inshore navigation and port
ingress.
The current non-digital, two-volume Fleet Guide, a sister
publication of the Sailing Directions, contains
information designed to acquaint incoming naval ships
with pertinent command, navigational, operational,
repair, and logistical information on frequently visited
ports in both the United States and foreign countries;
there is one volume for the Atlantic Fleet and one for
the Pacific Fleet. Much of the information contained in
the Fleet Guide is similar to that found in applicable
volumes of the Sailing Directions, but the Fleet Guide
emphasizes areas of special interest to U.S. Navy ships
such as command relationships, operational
responsibilities, and munitions support capabilities.
     The Digital Sailing Directions proof-of-concept
prototype data-set consists of the following publications
on CD-ROM: the Planning Guide for the North Pacific (PUB
152), the Enroute Guides for Japan, Volumes I and II
(PUBs 158 and 159) and Chapter 11 of the Pacific Fleet
Guide (PUB 941).
     The Digital Sailing Directions prototype
incorporates several key standards previously described,
including SGML (ISO-8879), CD-ROM (ISO-9660/10149), TIFF,
CCITT-Group/4 and CGM (ISO-8632). Technology areas which
are still emerging and are not demonstrated in this
prototype are direct indexing of SGML-compliant data for
full-text retrieval and use of a standard data access
language. Developments in these areas are being followed
with the goal of incorporating them into future Digital
Sailing Directions and other digital text data products
[13].
     Development of the Digital Sailing Directions
prototype began with the delivery of text in a tagged
format used by the Government Printing Office. Graphics
were provided in vendor-specific raster format. Analysis
of the provided data revealed that the text data could be
converted to the SGML structure relatively easily;
however, graphics data presented a problem for two
reasons: (1) the data structure was proprietary; and (2)
the quality of many of the digital graphics was poor. In
the end, a translator was found which could convert some
of the graphics to a TIFF structure, but many had to be
scanned from the original illustrations and photographs.
     Document Type Definitions (DTD) for each of the
publications (Planning, Enroute & Fleet Guides) were then
developed and completed as formal SGML DTDs. Sailing
Directions experts from DMA worked closely with the SGML
consultant during this phase of the project. In addition
to their experience in producing Sailing Directions
publications, most of these experts possess Merchant
Marine or U.S. Navy background, with 'hands on'
experience in using the Sailing Directions for voyage
planning. The participation of these experts was a key
factor in properly defining structures embodied in the
DTDs.
     Following development of the DTDs, processing
instructions were written to convert and add SGML tags to
the input text data. Manual proofing and editing was then
conducted to finalize conversion of the text data.
     Graphics were converted from the vendor-specific
format or scanned as necessary. Most graphics were
converted to TIFF with CCITT-Group/4 compression, but at
least one CGM vector graphic was included to demonstrate
the technology.
     The resulting database of text and graphics was then
integrated with a proprietary SGML-based retrieval
system. Processing instructions were developed to index
the SGML-tagged data for use with this retrieval system
and a customized user interface was designed for Sailing
Directions display.
     The Digital Sailing Directions prototype operates on
an IBM-compatible personal computer running the MS/DOS
operating system and the Microsoft Windows 3.0 graphical
user interface; both VGA and Super VGA displays are
supported. To allow smooth operation of the Windows
interface and relatively quick retrieval and display of
graphics, a minimum of a 80386/25 MHz microprocessor and
4 MB of system RAM is recommended. However, the prototype
has been successfully demonstrated with lesser
configurations.
     The Digital Sailing Directions prototype CD-ROM
includes two distinct components. The first consists of
the SGML compliant dataset described above; it is
included to demonstrate the capability for electronic
interchange of publications using SGML. The data can be
accessed using a SGML publishing system. The second
component is a complete, intelligent publication
consisting of the same prototype data-set; however, it
provides the additional capability to browse through the
publication and submit queries (e.g., word searches).
Hypertext capabilities are also provided; graphics may be
viewed by using a mouse to point and click on highlighted
references in the text; additionally, the user can move
between sections of the publications by pointing and
clicking on highlighted cross-references.
     The initial Digital Sailing Directions prototype is
planned to be released to the military services and other
designated evaluators in June, 1992. Concept planning is
underway for a product prototype which may include the
entire Sailing Directions (or a large subset) and
incorporate evaluation comments from the concept
prototype.
     In other related efforts, DMA recently completed
development of a Digital Gazetteer (DG) prototype on CD-
ROM. A second prototype currently being planned will
demonstrate the integration of vector map graphics (based
on the VPS) with the Digital Gazetteer text database.
Development of an Electronic Chart Update Manual (ECHUM)
prototype on CD-ROM is also being planned. Both the DG
and ECHUM products are potential candidates for TPS-
compliant text products and functional integration with
related digital products.


FUTURE DIRECTIONS

Although current information technology standards permit
development of a proof-of-concept prototype for TPS,
significant work must still be accomplished in several
areas; technology areas which require advancement to
eliminate 'dependencies' are: (1) direct indexing of
SGML-tagged data for use with retrieval systems; (2)
development of standard (abstract) index specifications
for text retrieval; (3) acceptance and implementation of
standard data access languages for text retrieval
applications (Fig. 4). However, the core standards for
the structuring and exchange of text and graphics,
including SGML, TIFF, CGM, and CD-ROM make the production
and distribution of intelligent publications on CD-ROM a
viable proposition in the near future.


REFERENCES

[1] Klaus, W., Memorandum - Proposed Text Product
Standard (TPS) for CD-ROM, Defense Mapping Agency, 8 May
1991.
[2] Defense Mapping Agency (DMA) - Digitizing the Future
(3rd Ed.), 1991.
[3] International Standard (ISO 10149), Information
Processing - Data Interchange on Read-Only 120mm Optical
Data Discs (CD-ROM), 1989.
[4] International Standard (ISO 9945), Information
Processing - Portable Operating System Interface (POSIx),
1990.
[5] International Standard (ISO 8879), Information
Processing - Standard Generalized Markup Language (SGML),
1986.
[6] United States Air Force (USAF), Computer Resource
Management Technology Program - (Draft) CD-ROM Index
Architecture Specification (CIAS), 1990.
[7] International Standard (ISO 9660), Information
Processing - Volume and File Structure of CD-ROM for
Information Interchange, 1988.
[8] The Microsoft and Aldus Corporations - Tag Image File
Format (TIFF) Specification (V 5.0), 1988.
[9] Facsimile Coding Schemes and Coding Control Functions
for CCITT-Group/4 Facsimile Apparatus, 1988.
[10] International Standard (ISO 8632-1), Information
Processing - Computer Graphics Metafile (CGM) for the
Storage and Transfer of Picture Description Information,
1987.
[11] Director Central Intelligence (DCI/IHC/ICS) -
(Proposed) CD-ROM Read-only Data EXchange (CD-RDx)
Standard (V 3.11), August 1991.
[12] Air Transport Association (ATA) Specification-100,
Manufacturers Technical Data, Digital Data Standards,
October 1990.
[13] Defense Mapping Agency (DMA) - Sailing Directions
Product Specifications (1st Ed.), 1977 & (Draft) Product
Specifications, 1990.


Related Graphics to this paper:

%g KLA01.pcx;
%g KLA02.pcx;
%g KLA03.pcx;
%g KLA04.pcx;
%g KLA05.pcx;
%g KLA06.pcx;
%g KLA07.pcx;
%g KLA08.pcx;
%g KLA09.pcx;
%g KLA10.pcx;
%g KLA11.pcx;
%g KLA12.pcx;