Where traditional document management ends, content management
begins. Document management handles information from a file perspective.
Each file is it's own entity, and it is indexed, stored, retrieved and
used at the file level. Content management breaks down a file into its
component parts, and these elements are indexed, stored, retrieved and
used at the content level.
A document file can be made up of multiple types of information,
such as charts, tables, headlines, captions, text and even sound or
video. Content management breaks out the content from the document. The
information can then be accessed individually and brought together in
different ways. As a result, you can "repurpose" information for
different mediums, including print, CD-ROM and the Web.
With the emergence of the Web, many businesses are discovering that
they are in the publishing business. According to analyst firm
International Data Corp. of Framingham, MA, "One of the biggest
challenges facing organizations deploying intranet applications is the
ongoing organization and management of the dynamic set of interrelated
documents [content], generally authored as a collaborative process, that
compose the content found on these corporate sites."
Document management systems have served many organizations well in
this regard, providing a way to share information internally and
externally across intranets. But what document management has been
lacking is the ability to use and reuse bits and pieces of information
for different purposes and with many different delivery options. This is
where content management comes in.
There is some overlap in the document management and content
management (as well as knowledge management) markets. "The process is
more important than the product", says Priscilla Emery, senior VP of
information products and services at AIIM International. "The product
must accomplish your process. With the growth of the Web in business,
everyone is becoming a publisher and content management deals with those
publishing issues. It is a publishing process that works in front of the
business process."
XML Provides a Foundation
XML (Extensible Markup Language) is a technology that is bringing
together the separate worlds of document management and content
delivery. Using XML, a document can be separated into its discrete
elements. For example, a 50-page report might contain a five-page
executive summary, three 15-page chapters, 20 charts, five tables, eight
photographs, 40 headlines and 30 captions. XML offers a way to identify
the different types of information presented and the relationships
between that information.
We are all familiar with HTML (Hypertext Markup Language), the
language most popularly used to display information on the Web. HTML
deals primarily with presentation of electronic documents and does not
offer much in the way of providing structure for the content. In the
50-page report example used above, HTML would not be able to provide
information on the difference or relationship between a headline and a
caption, with the possible exception that they might be presented in
bold type.
HTML was developed from the more complex SGML (Standard Generalized
Markup Language), which was developed in 1978 with the idea that a
standard "markup" could create structure separate from the appearance of
the document. Markup is simply providing information along with a
document that says something about its appearance or structure. SGML is
a language that is used to code nearly any type of document to describe
its structure. SGML is not as widely used as HTML because of its
complexity and the high cost of SGML editing tools.
XML was developed as a simpler version of SGML. XML documents are
composed of many entities better described as objects, such as the Word,
PowerPoint, TIF or Wave files that might make up a document. Each object
can contain one or more elements, which is a single bit of information
usually encapsulated as a string of text. Each of these elements can
have certain attributes or properties that will describe the way in
which the information is to be processed.
XML provides a way of describing the relationships between these
entities, elements and attributes. It tells the computer how it can
recognize the component parts of a document. It accomplishes this
through the use of metadata, which is simply coding information about
the information. Metadata is any information that accompanies an XML
document, including tags, DTDs, links and style sheets.
Elements in the document are marked with and identified by tags. The
tags are placed at the beginning and end of a string of text to describe
the attributes of the element. Tags do not describe the format ý whether
the text is bold or italic. Nor do they instruct the computer what to do
in a Web page when people click on it. Tags simply identify what the
element is. This is different from HTML, where the tags must perform all
these functions at once.
Separating the formatting from the element's identity makes it
easier to repurpose information for different mediums. With HTML you
have to write new tags each time you reuse an element for display on a
different medium. With XML a Document Type Definition (DTD) lets you
define the tags that were created and then use and reuse the elements.
A DTD is created once to describe the document type (i.e. expense
report, fiscal statement, press release), what the content elements are
and their structure. It does not define aspects of format, such as bold
or centered. Formatting is done later through XML style sheets using a
format not yet finalized called XSL (Extensible Style Language).
Once a file is created and submitted to the central repository, it
is validated against the DTD and parsed, meaning its content elements
are broken out. Validation is a key function of XML. It lets the
structure of data be checked before it is used in an application.
Once parsed, an XML document is manipulated through an object model
(or API). A standard DOM (Document Object Model) is being developed by
the World Wide Web Consortium (WC3), the international body that
develops standards for the Web.
The DOM lays out an XML document and its elements in a tree
structure.
Simply stated, XML provides a standard file format for presenting
data and a standard way to include information about the data (metadata)
to describe its own structure. XML Version 1.0 was finalized as a W3C
Recommendation in February of this year. Work is still underway on the
accompanying style language, called XSL (Extensible Style Language), and
a link language, called XLL (Extensible Linking Language).
The XML standard has been adopted by both Microsoft and Sun.
Microsoft has implemented support for XML within Internet Explorer 4.0
and will continue support with 5.0. XML will also be adopted in the next
release of Office and other Microsoft products. When these products
arrive, you will be able to "Save as XML" as simply as you do any other
format today. Standard DTDs are being developed for many other popular
applications, but as of now, users will need to design their own DTDs,
which is not an easy task.
How Content Management Works
The model for content management is to store and manage XML content
from a central repository, usually an object-oriented or relational
database or a hybrid of both. Content management systems offer many of
the same functions as document management systems, such as storage,
version control, check-in/check-out and text searches. What's added is
the ability to tag, index, search and reuse smaller elements of a
document.
Content management systems output content from the central
repository through any means of delivery available. Web delivery is
particularly efficient. Take as an example a company that manufactures
custom PCs that needs to provide customized manuals on demand for each
configuration ordered. Content management would let this manufacturer
break out and store content on individual PC parts as separate items
with their relationships to other parts.
The content on just those parts ordered by a specific customer could
be pulled together to create a manual. The manual could be printed,
output to CD-ROM or served up on a Web site at the customer's request.
The result is more efficient, dynamic and responsive use of content, and
this translates into cost savings for the manufacturer and better, more
personalized service for the customer.
Like many of today's document management systems, XML will take
advantage of three-tier architectures that support Web applications. In
this model, a browser front end interacts with a middle-tier Web server,
which in turn communicates with a back-end database server for central
storage. Information can be converted to XML on the middle tier,
offering new ways to access stored information from mainframes and
databases. This data can then be delivered to the Web and exchanged
online as easily as HTML pages using HTTP.
A Look At Three Products
The three content management systems described below are providing
innovative solutions for using information efficiently and for
automating the publication of information to the Web. The main features
that are common to all these products are a central repository based on
a database, XML content, HTML conversion and Web publishing. Many of
these vendors and products also have a history of supporting SGML. This
may offer those looking to move from SGML to XML some added expertise
and convenience.
DynaBase v3.0 from Inso (Boston, MA 617-753-6500 www.inso.com) is an
XML-based content management and web publishing system. XML content
management capabilities include import, parse and store XML components
with indexing and version control. Content for publishing can also be
served on the Web with XML tag-level scripting and XML tag-level search
and retrieval.
The content management components include a Data Server and a Web
Manager Client. The Data Server plugs into existing Web servers for
content management. It is a repository that includes a built-in
object-oriented database for storing any and all Web content. The
content can be XML, HTML, graphics, video, applets or scripts.
The Web Manager Client is the interface to the Data Server. From the
Web Manager Client, users can browse for files and check out content
from the Data Server or import new content using HTTP. The Web Manager
automatically launches authoring applications when you want to edit
content.
Inso's publishing system includes the Web Server Plug-In and the Web
Developer Client. The system lets publishers create Web sites and pages
on the fly.
Version 3.0 of DynaBase, which was released in September, includes
Inso's "Outside In" Web server-based file format converter, which
automatically converts popular file formats like Microsoft Word, Excel
and PowerPoint to HTML, GIF or JPEG. Also included are new workflow
capabilities that can be applied to any authoring, editing and
publishing process.
DynaBase supports Mac and Unix platforms through a cross-platform
Java client. There is also added plug-in support for Microsoft Internet
Information Server 4.0 running Windows NT and Netscape Enterprise Server
3.5.1 running Windows NT 4.0 and Sun Solaris 2.6.
Starting pricing for a DynaBase enterprise installation is about
$47,000. This includes ten client licenses, one data server that
supports unlimited users, three Web server plug-ins for management and
delivery to the Web and ten host or domain names for supporting ten
separate Web projects at once.
Blade Runner from Interleaf (Waltham, MA 781-768-1578
www.interleaf.com) is an enterprise content management system scheduled
for release this month. BladeRunner addresses the entire life cycle of
content in three layers: content creation, content repository and
content publish.
Content is created in Blade Runner using an XML creation toolbar
that is installed right into Microsoft Word. The XML toolbar lets you
use Word as an XML editor. Standard DTDs are implemented and used to
create Word templates that enforce certain rules or restraints on the
author. Users do have to know XML. They create Word files using a
template that matches the XML DTD. The author is guided with lists of
choices for valid content. The document is then validated against the
DTD and saved as an XML document. The document goes through a final
validation before check-in to the content repository.
A DTD creation tool, Microstar's Near and Far Designer, is included,
but any other DTD creation tools can be used. Authors are given lists of
choices for content structure that adhere to the XML DTD. The content is
validated against the DTD before it is checked into the content
repository.
Blade Runner's content repository is built on an object-oriented
database from Poet (San Mateo, CA 650-286-4640 www.poet.com). Once the
content is validated, the XML document is "burst" (i.e., broken down)
into its discrete elements. The reusable content elements can be brought
together and formatted for output using the menu-driven Composer/Style
Editor.
The database stores XML content along with the DTDs, XSL style
sheets and XLinks. The system provides versioning, check-in, check-out,
searching, navigational linking and referencing. Searches can be
performed on text as well as content, properties, structure and
metadata.
A tool is provided for creating XSL style sheets, which are applied
to the XML content to render a document. Users can preview the layout
and edit the final document.
The XML content is published using a two-part Batch Publishing
Engine that prepares content for print, CD or the Web. Since browsers
other than MS Internet Explorer have yet to natively support XML, the
publishing engine converts content to HTML for presentation on the Web.
The Assembly Engine uses Cascading Style Sheets for batch conversion to
HTML for the Web.
The XML toolbar can also be installed in the Interleaf 7 publishing
system. Interleaf supplies data adapters with Blade Runner that support
the integration of applications like SAP and Baan and the use of this
information by the content management system as XML content. Blade
Runner operates on Windows NT servers and Windows NT/9X clients.
Pricing had not been set yet but will be announced this month at the
release date and will be in the six-figure range for enterprise
applications, depending on how the solution is packaged.
Information Manager 2.0 from Texcel (Cambridge, MA 617-621-7004
www.texcel.no) puts a great deal of emphasis on collaboration. It is a
content and "process" management system that includes collaborative
tools for using content. Users can find, edit, review, reuse and
assemble information managed from an Information Manager (IM) database.
IM identifies all the elements of a document defined by XML or SGML
markup as separate objects. These can be searched, retrieved and
accessed for editing or viewing. IM tracks object versions in the
database and the software manages the Links between objects and
documents along with metadata.
The system integrates with SGML/XML editors such as Adobe's
FrameMaker+SGML and ArborText's Adept Editor, making it easy to update
XML and SGML data from the database in their native application. For
Microsoft Word files and other types of files the application can be
selected. Integrated ActiveX controls can be used to make different file
types like MS Word files accessible from the database.
Using a Windows Explorer-like interface, users can browse multiple
databases and click on an XML or SGML document. IM displays a view of a
document's contents in a small window. Search tools let users find any
content, link or metadata.
To support collaboration, an Electronic Review tool lets users
insert comments that can be passed along with the document. IM includes
a workflow system that lets you design and implement work processes.
IM's Document Assembly tool lets you gather content to create
documents or publications that can be changed or updated for different
purposes. A Document Assembly template can incorporate boilerplate text
and database queries.
Visual Basic can be used to add customizations. Open APIs are also
available to developers. IM supports Windows 95 and Windows NT clients.
It also can support Web browsers, but the separate IM Web Application is
needed. Windows NT servers and Unix servers are supported.
Information Manager costs $25,000 for the starter kit, which
includes the server, software, development, workflow and four concurrent
user licenses.
Doc. Managers Tackle Content
Many document management vendors have been slow to adopt XML. Those
that have adopted it offer useful systems for managing information at
the content level and delivering content to the Web as a front end to
their enterprise systems.
Without XML, you cannot achieve the same level of granular
information management. The 50-page report mentioned earlier could not
be broken down into its discrete parts unless those parts were saved as
separate files. To reuse these elements, you would have to cut and paste
content together in a manual, labor-intensive way.
This is the crucial difference between content management and
compound document management, which has long been supported by many
document management systems. With compound document management,
different objects, such as Word files, Excel files, images, graphics,
sound or video can be accessed from one system and pulled together for
delivery to the Web.
FileNet (Costa Mesa, CA 714-966-3400 www.filenet.com) is taking on
Web publishing and talking about "content" management with the
introduction of Panagon Web Publisher, which was introduced in November.
The product offers an automated solution for managing and incorporating
multiple types of information and the delivery of that information to
the Web.
In most organizations, information such as financial reports,
marketing materials and customer records are authored or created in
conventional applications. When the company wants to put that
information online, they send it to the Web master. The Web master is
the middle man responsible for managing the site, including the
structure, updates, revisions and purges.
Panagon Web Publisher distributes the job of the Web Master by
providing tools to automate the publication of information to the Web.
The system converts documents (or the discrete files of a compound
document) into HTML on the fly.
FileNet's Web Publisher is a toolkit for the Panagon IDM (Integrated
Document Management) system and can be used with imaging, COLD and
workflow. Features include the PWP Project Manager, which draws source
documents from Panagon libraries, organizes them into related sections
and subsections and automatically publishes linked Web sites and online
compound documents. Web sites can be published and managed from the
Panagon libraries within the document repository.
PWP Station uses translation templates that control the automatic
creation of HTML renditions of source documents, including format and
style. The PWP Station also generates hyperlinks, tables of contents,
keyword indexes and reference lists.
PWP Scheduler updates Web publications on a set schedule or whenever
new versions of existing materials become available. This automates the
process of revising Web publications to offer the very latest
information.
Related project files maintain an audit trail of what was published
at any given time, an important feature as the legal liabilities of Web
publishing emerge.
FileNet says it is waiting for enabling technologies, such as wider
Web browser support, before incorporating XML. When this happens, the
system will support the publishing of discrete content as well as
compound documents in XML as well as HTML pages. Panagon Web Publisher
is base priced at $19,500 for the server and three clients with Web
publishing stations.
Documentum (Pleasanton, CA 925-463-6800 www.documentum.com)
describes "content" as any business-critical document or information in
any format, such as word processing files, spreadsheets, graphics,
emails, HTML, Java, etc.
RightSite is the module included in Documentum's Enterprise Document
Management System (EDMS) that enables the assembly and delivery of
content-based information to Intranets. Using RightSite, EDMS users can
store, manage, retrieve and deliver mixed "content" from a common
repository. All objects are stored in the DocuBase object-relational
database and are shared with the RightSite Web server.
RightSite is not new, but support for XML has been added through
integration with other tools for enhanced content management
capabilities. Documentum promises to add native XML functionality in a
future release. Presently RightSite supports XML through their
partnerships with Abortext (Waltham, MA 781-529-1000 www.arbortest.com),
makers of Adept XML editor.
RightSite includes native support for SGML, which can be used along
with HTML to bring content to the Web. WebQL, a super set of the
Structured Query Language (SQL), is a published set of HTML directives
for dynamic page assembly. WebQL is used to tag items with attributes
and create live links to other items. RightSite automates the process of
link creation and deletion.
Saillant Consulting Group's (Denver, CO 303-708-8209) HTMLRender is
integrated with Documentum's RightSite to automate the conversion of
documents to HTML to simplify creating Web-ready documents.
Documentum says new APIs and SDKs will be offered in future releases
to add more flexibility to RightSite as well as added support for active
server pages and XML publishing. RightSite is included in EDMS 98 and is
not sold separately.
PC Docs (Burlington, MA 781-273-3800 www.pcdocs.com) offers what
they call a "compound document management and publishing module," called
Docs Binder, for its Docs Open document management system. Docs Binder
is geared toward collaborative environments where documents are made up
of multiple items and have multiple authors spread across the
enterprise.
Within a "binder," the separate components of varied file formats
are managed as individual documents. The documents can be edited
separately or simultaneously within the larger binder document.
Individual documents can be used in more than one binder, allowing for
the reuse of content. The content can be updated within all binders
automatically.
The contents of the binder are viewed in an Explorer-like tree
diagram. Documents can be dragged and dropped between and reordered
within binders. Binders are stored in XML format and contain tags that
identify links between documents.
For publishing, PC Docs integrates with third-party products such as
PDFfusion from Computerized Document Control (Monmouthshire, UK
888-240-1752 www.docctrl.com). PDFfusion will publish binders as PDF
documents on the fly. The software offers templates for layout. Other
third-part publishing systems can be integrated using the Docs Binder
toolkit.
