Enterprise Information Portal Revolution

Published in TDAN.com July 2001

Internet and web technologies have profoundly changed the way we view and interact with information. Information Portals, both on the Internet and in the Enterprise, have shifted our perception of
information from a document-centric model: in which static information is embedded in documents that require weighty client software—to a dynamic model: a model in which both structured and
unstructured information is presented in a web browser. In this role the browser acts as a ubiquitous window to data from an ever-widening variety of sources: legacy database management and
reporting systems, client-server networks, and increasingly, unrelated Internet and Intranet web sites and syndicated services.

Confronted with a vast number of information sources, our expectations of the quality of information we interact with have shifted as well. Information must be timely. It must be sorted, filtered,
and prioritized by relevance. Data must be presented in context, hyper-linked to glossaries and related knowledge, and offer meta-data that reveal its sources, authors and any assumptions that were
part of its creation.

In short, we have come to expect the world from our portals. We expect them to increase the ratio of signal to noise, and to perform dependably, accurately, and securely while doing so. We expect
them to be usable, and to coherently classify and make accessible the knowledge of our enterprise, creating discrete channels of information that users can identify with, and comprehend.

While this level of integration is technically challenging, it is—with commercial, off the shelf products—eminently feasible. The real challenge is found within the less tangible
aspects of portal building: capturing employee knowledge, recognizing relevance, identifying relationships between diverse information sources, and capitalizing on the human side of the human /
computer interface by clearly understanding the goals and expectations of the user community.

Enterprise Portal Building Blocks

While the defining characteristics of Enterprise Information Portals continue to evolve, a number of common functional components serve as essential building blocks. These include (but are
certainly not limited to): Content Management, Application Integration, Data Access and Reporting, Collaboration, Personalization, Search, and User Management and Security.

Content Management. Content management facilitates the processes of creating, using, storing, indexing, and retrieving primarily unstructured information, historically in the form
of office productivity documents (MS Word, Excel, etc.), images (logos, brochures, CAD drawings), email messages, project plans and other binary and text documents. These documents generally
possess no intrinsic meta data, and can be stored virtually anywhere within the organization – from the creator’s desktop PC, to network file servers and floating laptop computers.

Increasingly, content management concerns itself with the workflows employed to create documents, as well as document components: the granular elements of a document, separated from their structure
and presentation. The goal of current content management practices is to support a wide variety of access devices—standard web browsers, micro-browsers (found in cellular phones and palm-top
devices), server peers, and search engines—without additional device-specific formatting.

Content management systems often serve as the backbone of large web sites, provide version control and site management facilities, and may act as integration points for external sources of
unstructured data through syndicated news and information services.

In a similar vein, knowledge management systems attempt to capture the tacit information embedded in an organization, and make it explicit. Much of an organizations information capital is unwritten
and undocumented, and is instead filed away in the minds of its employees. Knowledge management employs tools intended to entice employees to document what they know, and then store this explicit
information in repositories that are highly searchable, and widely shared.

Business Intelligence Applications. While Content Management Systems serve unstructured data, Business Intelligence Applications (Data Marts, Data Mining, OLAP, Query and Reporting
products) aid organizations in discovering actionable business information in their structured data systems. On the whole these are large, complex and often expensive products, increasingly fronted
by web browsers which serve to increase the potential for information distribution.

Together, these product segments provide access to enterprise data, presentation services, business analytic services, multi-dimensional views and complex pattern analysis and trending
capabilities, all with increased user efficiency and less impact on production systems.

Business Intelligence Applications increasingly employ web-derived technologies (through various APIs, XML and other data “hooks”) in both their access and presentation service layers,
making them excellent targets for incorporation into the Enterprise Information Portal.

Collaboration. Collaboration among teams and workgroups is among the most critical of business practices, and yet it proves an elusive goal for organizations with the highest
operational capabilities. Much of the effort of today’s “knowledge worker” revolves around teams, and team-based activities: shared projects, discussions, calendars and ideas.

Collaboration is the root-stock of the Internet. Web-based collaborative tools exist in many forms: email, threaded discussion groups, chat, and instant messaging – each with its distinct
factors for persistence, search and archival capabilities and immediacy. The number and variety of these tools may lead to yet another issue: how to choose? A single worker may have a number of
collaborative tools at his disposal, and may be unaware of the both tools and the preferences of the other members of the team.

Hosting collaborative tools on the Enterprise Information Portal can solve the “which tool?” dilemma, while at the same time providing a means of ensuring that ideas and concepts are
not “lost” to applications that afford little persistence. Portal-based collaborative environments play a fundamental role in knowledge management, by capturing tacit information, and
making it both explicit and durable.

Personalization. Personalization comes in many forms—from simple name recognition, to checkbox-based profiling, to advanced collaborative filtering and dynamic rules engines.
Typically personal and preference data is saved to the individual user’s profile, where it can be referred to by dynamic pages and applications alike. In terms of functionality,
personalization can provide user-customizable interfaces and “themes” at a very basic level. More advanced personalization systems allow for content filtering and prioritization,
customized access to frequently-referenced data, even timely notice of relevant events, or notification that new materials and updated documents are available.

Personalization plays a significant role in the user’s experience, and can make or break a portal application in terms of its usability. As a rule, applications that are difficult to use are
not used, or are used only ineffectively.

Search. If there were but a single phrase to sum up the collective frustrations of our information age, it would probably read, ”I know it’s here…
somewhere.” While nearly every desktop productivity tool, and many of our enterprise systems incorporate index and search capabilities, tools that serve only specific data stores are
inadequate to the task.

The Enterprise Information Portal is uniquely suited to host unified search and taxonomy services that span multiple data stores, web sites, collaborative applications and other unstructured data
sources. Portal-based directories can supplement key-word and conceptual searches by offering a hierarchy of subjects that can be casually browsed, or used to narrow a search domain.

Search results can be made more relevant by matching application-specific meta data to user preferences, and by employing search agents that crawl Enterprise file systems and data stores,
automatically generating taxonomic classifications for the documents they discover.

User Management and Security. Security is a necessary component of the Enterprise Information Portal. It’s clearly important to safeguard access to sensitive and strategic
information assets. Equally important is the ability to share user credentials between multiple applications – even if the applications themselves aren’t integrated.

A policy store with delegated administration capabilities can provide single user sign-on capability, while maintaining flexible user rights / roles definitions. Users log in once to access
information on multiple applications, on multiple servers, across multiple platforms. The result is a higher quality user experience, personalized content, lower administrative costs, and
simplified password management for large groups of users.

In Summary

The transition from a document-centric to a dynamic information model is more revolution than evolution—it is a non-linear change. An Information Portal that is solidly aligned with the
organization’s business goals has the capacity not only to integrate, but to transform.


submit to reddit

About Doug Cadmus

Doug Cadmus is an Internet Strategist for CIBER, Inc.