Understanding The Semantic Value Proposition

By on

Part  1 – Understanding The Semantic Value Proposition

The term “Semantic Web” has developed some interesting yet confusing connotations since it was first introduced in the early 2000’s. Those misconceptions include but are not limited to:

  1. The notion that the associated technologies and practices are only applicable to web applications or the Internet in general.
  2. The assumption that Semantic Web technologies are not yet ready for prime time.
  3. The misconception that these technologies are not primarily focused upon data integration or focused on it all.
  4. The assumption that the Semantic Web is a long term goal rather than a set of near term opportunities.
  5. The notion that all semantic solutions must necessarily conform completely to W3C semantic web standards. (i.e. that legacy capability can’t play).

These misconceptions and the overall lack of direct focus thusfar within the Semantic Industry have led to a significant delay in the exploitation of semantic capabilities where they are needed most. The realization of where semantic technology is needed most is in fact the foundation of the true value proposition for Semantic  technology. 

Semantic technology is the next generation framework for Enterprise Integration and Cross Domain interoperability. It’s near-term value is related to cost savings, efficiencies and capability enhancements within every enterprise IT environment on the planet. As an added bonus, it also provides the foundation for exploitation of web-based resources and capabilities – something that really couldn’t even be considered as a comprehensive solution without semantic technology.  The reason why Semantics represents an immediate value proposition is this – it gives us for the first time ever the ability to conduct and manage dynamic integration & interoperability on multiple scales or levels, bypassing the current practices of predetermining all structures before deployment.

Part 2 – My Perspective

I’ve worked in Silicon Valley before, but the vast majority of my experience is not with start-ups or traditional internet related companies. I’ve followed the emergence and evolution of Semantic standards but I’ve never worked for a Semantic-focused company, at least not until I helped form my own two years ago. My focus over the past decade has been primarily directed at large scale Enterprise environments, often in the Federal sector. The problem space I work in is Enterprise Integration. This is a ubiquitous area to say the least and covers the following subcategories:

  1. Enterprise Architecture
  2. Enterprise Services Oriented Design
  3. Enterprise Data Integration
  4. Systems Engineering
  5. Business Process Analysis, Design & Management
  6. Enterprise Content Management
  7. Enterprise Learning Integration

The scope of Enterprise Integration often includes the full range of activities and applications available within an enterprise, and the goal of Enterprise Integration is to ensure the highest level of interoperability between them possible. Enterprise Integration is more than the development of specific data or application interfaces between systems – it is the development and maintenance of family of systems environments. 

Working in this problem space has given me a profound sense of what is working and what needs to occur in order to develop the holistic systems environments that organizations desire and need right now.  This is precisely what led me to view Semantic Technology as the basis for a new approach to systems integration at the enterprise level. But it is important to recognize that the existence of semantic technology or standards in themselves do not represent a solution – they are building blocks which must be applied in the proper manner in order to resolve the unique challenges of this problem space.

Part 3 – The Data Roadblock

By far, the greatest challenge facing the typical enterprise is data integration. The reason for this is based on the nature of current database technology. Relational databases and even older mainframe database technologies are predicated upon a philosophy which requires that the majority of data interactions have to be predetermined.  This was not as problematic 20, 30 or 40 years ago, when the number and complexity of the systems creating and presenting that data were relatively limited.  Unfortunately for those of us dependent on those technologies, the number of systems needing to pass data back and forth as well as the sheer quantity and complexity of data have both risen exponentially.

For those of us attempting to integrate dozens or hundreds of systems and /or their related data, the technology has created a capability barrier that can only be surpassed at great expense, usually with mediocre results. We’ve been stuck behind this roadblock for the past decade and it is only getting worse. New applications including data warehouses, master management and business intelligence tools have tried to address this – but often times these solutions aren’t much better than the underlying technologies that they try to cobble together.  The reason for that is similar to the primary problem facing database management systems – the nature of the integration has to be planned and configured in advance.  This often leads to multi-year efforts on data warehouse solutions, and long cycles to configure EAI, MDM or other message management solutions (ESBs).
The most ironic part of this data roadblock dilemma is the fact that when or if a project finally manages to get all of the interfaces and data models aligned in a complex system of systems enterprise, the situation has already changed and then a non-stop catch up effort begins – one that often costs as much as the initial integration. 

Part 4 – Exploiting the Opportunity

Every challenge provides us an opportunity for innovation – the innovation in this case is viewing Enterprise Systems Integration in a new light and correlating both the emerging set of semantic technologies with existing technologies and resources in order to support immediate solutions that redefine the industry.  In order to do this we need to begin with some basic assumptions:

  • That the problem space for enterprise integration cannot be solved with any one product, any one standard or any one technique.
  • That semantic technology can and must necessarily coexist with current legacy technologies and resources in order to support immediate opportunities for exploitation.
  • That semantic technology in itself is not a solution, any more than the legacy technology was – in both cases, specific processes or methodologies and sets of best practices need to be employed in order to properly exploit the technology in question. These are well understood for the current set of RDBMS related technologies but have yet to be fully defined for Semantic technology. 

In other words, the context for how to apply semantic technology to the enterprise is missing. This context can be referred to as Semantic Integration. Semantic Integration is both a practice methodology and a set of technical best practices related to the correlation and reconciliation of legacy and semantic IT capability. In many ways, Semantic Integration is an extension of Systems Integration with one critical difference – that difference is the deliberate coordination up front of all semantic entities within the enterprise. That may sound a bit academic but in fact it is a very practical distinction.

Part 5 – Practical Magic

As sophisticated as the typical IT enterprise has become, it has never been completely pragmatic.  What this means is that:

  • Databases and applications aren’t generally designed to interoperate – especially across heterogeneous sets.
  • Business processes rarely coordinate their terminology and process structures with either applications, services or data structures.
  • The presentation layers atop these other structures often don’t correlate to the underlying lexicons.
  • Definitions of assets or functions within the enterprise are usually not coordinated with other structures (this includes such things as network asset management, email server domains and other NETOPS capabilities). 
  • Often times, competing standards with their own syntax and lexicons are employed in the same enterprise.
  • Project management structures are seldom coordinated at the data level with the aspects of the enterprise they control.

This situation extends even further, but the bottom line is that every enterprise is built ad hoc, piece by piece, with no simple way to reconcile their many components within any sort of unified framework. We require a unifying medium to accomplish that – the unifying medium is the semantic layer. Everything that can be identified or can communicate can be resolved within this medium or fabric – allowing for a single foundation which can be managed dynamically rather than designed in an attempt at static perfection

My blog on the new Semantic Universe is dedicated to exploring and explaining the emerging realm of Semantic Integration (SI). We will discuss the best practices associated with SI, the vendor toolsets and combinations of tools that are providing real capability today and the emerging practice methodologies associated with Semantic Integration. 

My next article will present a taxonomy of Semantic Integration best practices and methodological components.  After that, I will begin exploring the business case for the new SI.

Copyright 2009, Stephen Lahanas



We use technologies such as cookies to understand how you use our site and to provide a better user experience. This includes personalizing content, using analytics and improving site operations. We may share your information about your use of our site with third parties in accordance with our Privacy Policy. You can change your cookie settings as described here at any time, but parts of our site may not function correctly without them. By continuing to use our site, you agree that we can save cookies on your device, unless you have disabled cookies.
I Accept