Dr. David J. Roberts, iBASEt, Inc.
Imagine this: You go online to book the perfect vacation. Instead of having to search through two hundred flights, forty hotel rooms, and eight car rental companies, an Internet application willyour personal ontology, consisting of contextual descriptors you’ve decided to share, and give you a short, customized list of options. No more having to manually enter in everything from the price of the flight to the type of car – you’ll get your very own view of products and information on the Internet, based on your preferences. This is Web 3.0.
As the second evolution of the Internet, Web 3.0 jumps into the middle of the information exchange problem with a semantic view of the information highway. Much like how humans use spoken language to communicate, Web 3.0 uses a linguistic base for exchanging information on the Internet, as opposed to the functional and process-driven base used by Web 2.0.
Web 2.0 can be seen as another level of information integration and services built around the performance of tasks. In a Web 2.0 environment, applications are connected in useful processes. In a Web 2.0 integration, all information exchange is negotiated and understood, and processes are locked into a well-defined sequence flow. Web 3.0 unlocks those processes.
Presently, there are two main problems associated with Web 3.0:
- Machines have a terrible time resolving ambiguity. Humans are good at it, because we have learned to use contextual clues and pattern interpolation. Just one look from an annoyed spouse, along with the comment, "You’re in my way," are enough to get us to move. A computer might have a difficult time resolving the spatial and temporal context for the indirect request for movement.
- Core pieces of the semantic "language" of Web 3.0 are missing, such as domain-based languages, ontology builders, and classification and authentication services. Some of us already know what those pieces are, and where they go.
Web 2.0 applications can be useful for focused and well-defined tasks. When information is integrated from many sources, in varying degrees of completeness and meaning, the problem set ceases to be imperative (rule and sequence driven), and is more functional in nature. Think of the world as one giant spreadsheet with thousands of simple rules linking different areas of interest. Change one variable, and whole parts of the spreadsheet can be impacted. It is in this functional model of the world that Web 3.0 provides a useful paradigm for exchanging information.
Web 3.0 takes a page from nature in constructing useful patterns into clusters of correlated concepts and associations. Some of these patterns can be declared, others derived, but all are based on shared contextual attributes. In applications in which sequences or contexts are not well specified, such as loosely associated news articles, using a semantic approach offers pattern detection based on self-describing pieces of information and associations. Web 3.0 offers a method for arbitrating disparate information sources based on published definitions.
But the big plus is that information can be combined in new and creative ways. Web 2.0 assumes you know how, why and when information is associated. Web 3.0 only assumes the context of information has been declared to some level of specificity, and you have some mechanism to make sense out of the pieces.
What about skipping Web 2.0 and going directly to Web 3.0? Aside from the problems mentioned above, some shifts in thinking, along with a little set theory and logic, will be necessary to make Web 3.0 applications work: What are contextual constants? What are the associations between objects?
However, there are viable Web 3.0 solution sets that can be immediately exploited. News feed integration, core languages in publications, chemistry, medicine, and other well-defined information domains offer easy entries and even easier solutions. Well-described and published information can go directly to Web 3.0 applications; we see whole information communities doing just that. For example, in 2006 iBASEt produced a technology assessment service based on Oracle 10g R2 that semantically integrated over 400 RSS newsfeed, multiple databases, and even library documents.
Web 3.0 currently offers viable solutions to some difficult problem sets, but it is also the next personal Internet that understands your tasks and context. In the future, you will use certification services to ensure the proper description of information, and classification engines to build your personal profiles. The Internet will be your personal viewport of the world, and performing tedious tasks such as searching through thousands of options for that perfect vacation will be a thing of the past.
Dr. David J Roberts is currently the Chief Scientist at the Cross-domain Information Exchange Framework (CIEF) project at iBASEt, Inc., San Diego, CA.
