The Solution - Thematix™

Thematix, Microlanguage's intelligent language processing technology, hurdles the shortcomings of conventional search technology by (a.) delivering levels of precision and recall that far exceed what are possible with Boolean and Bayesian technologies, and (b.) extracting concepts from unstructured and structured text.

The figure below illustrates what affect precision and recall performance can have on the task of finding relevant information, particularly within large knowledge bases.

Conventional Search Technology	Thematix Technology

For each side of the figure, the outer circle A represents all of the information assets available. The inner circle B represents all of the relevant information that the knowledge worker is after. The circle C represents the information returned by a search. The relationship between C and D is where performance is measured. The standard metrics to measure the difference between what is relevant (B) and what is retrieved (C) are recall and precision.

• Recall measures how much of the relevant information was returned (D) as against how much is available (B); it is measured by D/B.
• Precision measures how much of the returned information is relevant (D) as against how much total information was returned (C); it is measured by D/C.

Conventional search technologies miss much of the relevant information, and return lots of irrelevant (i.e. spurious) information. With such results, the knowledge worker must spend time wading through all of the information returned in search of the relevant information. Moreover, a new search must be initiated in an attempt to gather the missed information, or worse yet, the knowledge worker has no idea how much relevant information was missed.

The illustration also promotes an understanding of the impact precision and recall have on knowledge worker productivity, but Thematix goes one step further and allows the knowledge worker to search on concepts, not just on keywords. Using patent pending technology, Thematix recognizes and retrieves target concepts or "themes" (selected by the customer).

The results can be both dramatic and profound:

• Knowledge workers can focus on all available, and only relevant data, and avoid sifting through, scanning or reading irrelevant data.
• Searches will retrieve information that is relevant to the search concept, not just the information that matches the search terms.
• Companies and agencies can avoid the loss of productivity resulting from fruitless or unnecessarily lengthy searches, and capture the strategic and financial benefits of making timely, informed decisions.
• Thematix can reduce knowledge worker productivity losses by up to 80%.

These factors make Thematix technology a vital component of knowledge management, data visualization, voice recognition, and customer relationship management (CRM) systems, content filters, and database systems.

How does Thematix accomplish these remarkable results?

Next: Technology