Monday, October 1, 2001

Deeper into managing corporate aliteracy

Active knowledge portals provide users with a window-to-the-world through which they access and author information. While some would state that this is suffice, the non-reflexive nature of information inhibits the automatic cross-semantic associative binding that can take place in active-reflexive knowledge repositories. Furthermore, through concepts like inductive association, reflexive knowledge can reach beyond direct associative relationships into levels of indirection, beyond the immediate semantic fabric into second, third, and more semantic horizons.

Such inductive semantic reaching facilitates knowledge churn that can rekindle knowledge associations, such that new information acts as a knowledge catalyst, setting off a chain reaction of associative knowledge building and rebinding. As a result, near-surface but remote and far semantic horizon inductive knowledge associations nodes can surface when apparently unrelated information gets added to the knowledge repository or used accordingly, setting up an associative knowledge building and rebinding chain much like a ripple effect. Knowledge portals that don't support active-reflexive data technology cannot harvest to the same depth the underlying knowledge associations that lurk below the surface as killer detail.

Implementing an active-reflexive knowledge repository in Domino

Domino provides a natural object repository through its DOM/XML (eXtensible Markup Language) document architecture. New versions of Domino support XML directly through DXL and XML Parsers in LotusScript and Java. Domino also supports a Java server for running Java servlets. These can be moved outside the Domino environment into TomCat and J2EE (Java 2 Enterprise Edition) Java servers. Communications from the LotusScript environment can be provided transparently to an underlying J2EE Java Servlet through an XML SOAP (Simple Object Access Protocol) Knowledge Management transaction.

The SOAP transaction metaphor can be realized within Lotus Notes using a Notes document as an object container, with fields wrapped in appropriate XML Tags accordingly. RFT fields provide the requisite source document and knowledge object transport mechanism, passing requests to Notes Knowledge Management agents from external arbitrary XML applications that drive underlying J2EE Java Sevlets through a Knowledge Management API (Application Programming Interface) library. The Knowledge Management API Library in turn passes requests through a SOAP XML knowledge transaction to an underlying J2EE Servlet Knowledge Server application, which processes the results, retrieves and contextualizes a response, and returns requested knowledge objects in the SOAP transaction envelope's request RTF field to the calling LotusScript Agent. This can then pass it back to the calling program or process it with subsequent LotusScript Agents.

Furthermore, SOAP transaction results can be parsed and placed in Notes databases where fulltext indexing can retrieve kThread profiles and kSig document signatures based on attribution, acting as shadow indexes for associated documents and for users through their profile signatures. Powerful knowledge portal applications can drive corporate productivity to the next plateau, facilitating automatic reflexive knowledge churn to activate knowledge publishing and collaboration and automatic portal page personalization that brings relevant context to users based not only on the semantic containment of discovery context from documents, assertions or copy-pasted text snippets from germane documents, but also on automatically updated user profiles that learn from what they author, read, and collaborate.