 
|
|
the
information
bottleneck
challenge 
|
Today, the intelligence community faces an information bottleneck of unprecedented proportions: large volumes and types of unstructured data that analysts must quickly, thoroughly and precisely scan for relevance to a large number of threats, route to the appropriate subject matter analyst for analysis and assessment, and then distribute precise and actionable information to appropriate military war fighters or civilian first responders. As tough a challenge as that represents, it is only getting worse:
- Ever-expanding
number of subjects - Political insurgencies and new terror groups are constantly arising, broadening the scope of real, implied or hoax threats.
- Ever-expanding
input - The sources and amounts of unstructured digital data on threats (real, implied or hoax) are increasing geometrically, enabled by new and disparate communication systems that include e-mail, chat, instant messaging, cellular and satcom telephony, print, voice to text communications, and commercial news feeds. This means that there is more information about more threats than ever before.
- Conventional
search technology - Conventional search technologies (e.g. Boolean, Bayesian) which underlie knowledge management and information retrieval applications are inadequate since they can not distinguish between relevant and irrelevant information. Conventional technologies are not intelligent enough for the job.
Commercial
knowledge workers spend almost one third of their time searching
for relevant information .
Microlanguage has demonstrated
that conventional search technologies waste up to 80% of that
time because the information they return is irrelevant. In
the intelligence community, where information is the primary
commodity, even more time is being lost looking for relevant
information and turning it into actionable
information.
If there
were a way to program the knowledge of the intelligence analyst
into a knowledge management application, the application could
distinguish between relevant and irrelevant information, and
get the right information to the right analyst. The information
bottleneck would be overcome, and analysts could spend more
time analyzing relevant information and acting upon it. What
is needed is a next-generation technology that intelligently
distinguishes between relevant and irrelevant information
and can convert simple data to actionable
information.
|
|
|
thematix™
-
the
information
bottleneck
solution
|
Microlanguage's
next-generation, intelligent language processing technology,
Thematix, is aimed at applications that demand highly precise
filtering of unstructured digital information. Thematix features
two components:
- The
parsing engine breaks incoming information into its constituent
parts, such as words and punctuation, so they can each be
examined individually.
- The
grammarbase™ is a repository of knowledge that gives
further instructions to the parsing engine. The analyst
knowledge of what information is relevant and how to use
it is programmed into a grammarbase.
Thematix
recognizes relevant, actionable
information, marks it as such, and enables the knowledge management
application with which it is integrated to distribute it to
the appropriate analyst in 'analyst ready' form: nothing less
than the right information to the right person, in an actionable
form.
As a component
software technology, Thematix integrates with third party
knowledge management applications, giving them much higher
search acuity, while interfacing seamlessly with their document
management tools.
|
|
|
the
benefits
|
Thematix
technology enables intelligence analysts to focus on knowledge,
not on sorting through thousands of bits of information that
may or may not be relevant to their specific interest. Thematix
enriched knowledge management applications offer several important
benefits to the intelligence community:
- Reduced
cost of information sorting
- Reduced
operator fatigue
- Conserved
processing time
- Closed
gap between data and actionable
information
Footnotes:
1 - SIGNAL Magazine, August 2001. Electronics
Transform the Army
2 - Infoworld Magazine, January 4,
2002 issue
3 - Microlanguage Whitepaper, Conventional
Search Technology vs. Intelligent Language Processing
Technology
return
to top
|
|
|
