Not only are search engines early-stage technology that get left behind in a sea of InfoGlut, they aren't really that accurate because they suffer from the inherent limitations imposed by the browser. Good semantic search containment requires an order of magnitude more descriptive constraints than is possible using standard query strings played through a browser.

2G-KM achieves this accuracy through automatic parallel query transformations of associative signature patterns that map onto flat full text information bulk indexing. This automatically prepares and executes 30 to 40 search strings in parallel on behalf of the user that define concept hierarchies extracted from discovery documents and document viewpoint profiles. This approach delivers a manageably small set of highly accurate, componetized results organized around the user's needs and their learning profiles as articulated in their knowledge discovery search submission. As such, search engines can serve a useful role in knowledge management as addressable raw information resource pools that can be leveraged by 2G-KM technology as fodder for knowledge assimilation.

If we are to turn the tables on InfoGlut, we must take a different approach to solving the knowledge assimilation problem by harnessing how we learn, understand, and leverage reusable patterns for profit.

We don't need more information, we need better information automatically reorganized and filtered to our needs, presented to our personal learning style, knowledge experience, and problem context.

We need decision-ready data that permits us to cut-to-the-chase, make well-informed decisions quickly. The data must be well-prepared and organized to our problem concepts automatically, without need for expensive manual intervention at significant compute-expense and time delay, so we can get on with making good decisions early to capture critical problems before they take root and significant opportunities before they disappear.

We need to harness reusable corporate knowledge and memory trapped in documents scattered around our networks that describe meaningful business concepts, their relationships, and their values. No longer can we afford the expense, time, and effort required to rebuild business concepts in documents time and time again.

We must recognize, capture, preserve, and leverage the knowledge concepts and relationships found in documents so we can stand on the shoulders of our business experts, not stomp on them.

A better way
Users have not discounted the value that corporate knowledge experts have articulated in business documents. In fact, they're frustrated at not being able to get at them. But until now they haven't had a way to recognize, capture, preserve, leverage, and track the value in a fully automated fashion. Semi-automated methods require too much effort to stay in tune. Once they fall into discord, they become a business expense with little return and little trust. They get fast-tracked on their way to the knowledge management application graveyard as solutions that are long on promise but short on delivery.