What Do I think of the Semantic Web - Somebody asked.

I hope we are soon in an era where I can find the best answers to your questions on the Semantic Web.
I feel strongly about the entire initiative considering the fact that it has brought in thoughts from a wide diaspora of industries and domains – a community effort to find a solution to a common problem of disparities in metadata and understanding.

The reach of the Semantic Web has turned out to be so widespread, to such an extent that the term “World Wide Web” itself might require rephrasing [unless the effort of migration is so humungous]. Wonder why? - Because it would not even matter if we are in the same world. [Hope that’s not too much to ask for]

Finally, we have come to the point where the Metadata has traversed up to the Root of the heirarchy, where entities have become self-descriptive, and no longer require external metadata for them to be defined.


I feel that the transition from human readable semantics, where the User knows that the content within an tag should be an author, to the machine readable semantics where the Document Processor also knows about this as a fact, is a success in the following respects.

1) The Definition of a Consumer has increased in breadth – Does not matter who the Consumer is – The Data is available at the disposal for consumption by any End System, as long as it treats it well.
2) Common View of the Data for Consumers irrespective of the Syntax.
3) Transparent and Adaptive Context Based Processing and Interpretation of Data for any Consumer of a Document.
4) Automatic Expansion of Data on discovery of matching Data Semantics.

While Programming languages shifted from their notion of a Procedural Paradigm to that of an Object Oriented Paradigm, so do have Applications moved from a Processing Paradigm to a Document Centric Paradigm – The realization that Processing is just about making decisions based on Data or just enable data flow, and the relevance of a Consistent and generic Data Model has been the most significant shift in the thoughts over the past decade. Both these systems have felt the significance of Data Sharing.

One common misconception that has evolved around the Semantic Web – I have noticed it, is the myth that
a) Semantic Web requires a Browser and a Search Engine
b) Semantic Web is all about, and only about Semantic matches for data. Of course, discoverability is one of the principal concepts that Semantic Web holds, but it is unfair to restrict its definition to such a small subset.

But I definitely foresee the Web evolving into a virtual database with infinite tables, rows and columns, and inspite of this, providing easy access to the relevant information that Applications and People really want. In any case, semantically, a Document Query closely resembles a database query in many respects.
a) Both are aimed at extracting relevant information from a data source
b) Both use a linguistic query syntax for retrieving data
c) Both access data that exists in a predefined structural format.

It’s interesting to note the similarities in the notion of semantic equivalence in Documents and terminologies, and Objects in programming languages.
Object Oriented Programming had crude notions of semantic associations between Objects in the form of “is a”, “has a” relationships, domains, ranges that enable storage of data within the Object that it considers correct and valid, and help build networks of Objects that can work together.

This eventually led to the natural evolution of Web Services that allowed loosely coupled interaction between Objects [well, at the rock bottom, it boils down to Objects, right?].
There’s a plenty of scope, and plenty of research already happening – in the area of Semantic discovery of Services – based on the publicly visible behaviour, assisting discovery and selection of Partners based on Semantic Annotations on the Service Definitions through variety of mechanisms.

Well, as soon as the industry realized that Web Services alone are not going to solve all the business problems in the world – they needed to talk to the Organizations’ internal Business Processeses – for creating a Business Process Orchestration that would police the interaction between the Services, Semantic Integration has come onto the forefront – where communities started mail threads talking on Process Ontologies, and automatic orientation of the Organizational direction to changing Business Needs – capturing metrics of partners and processes in ontologies and optimizing the Business Processes based on the metadata.

Maybe I feel I should stop here – for this answer could run into pages, weeks and months, and by the time I finish it, I would be able to see my PC with an inbuilt intelligent inference engine giving me instructions on optimized work-time productivity and suggest times when I should continue writing this article – based on my Project Deadline metadata, my personal metadata and information from external data sources.

Well, in Summary, I feel that the Semantic Web Initiative holds a bright future by
1) Allowing people and Applications to understand that they are talking about the same thing or a related fact.
2) Providing space for Resilient Knowledge Management by moving forward from structural and linguistic constraints on definition of data, to a space that allows for change in the Organization’s data representation without impacting Business and avoiding downtimes.
3) Spreading awareness among organizations that a generic and extensible data model will solve problems that would otherwise impact the Business, if not at the immediate present.
4) Involvement and endorsal of a huge intellectual community under the umbrella of people like TBL, who can dream of the impossible and make them possible.