NarraText is managed
and directed by its principle co-founder, Dean
is an abbreviation of "Narrative
Mission is to "Tell the
Business Story" that is inherent
somewhere in all unstructured text.
aims to "engineer the best
solutions" based on the underpinning
"disciplines" and "theories" of Linguistics-based
Text Analytics, including Domain Pragmatics, Discourse
Structure, Semantics, Genre', Syntax, and Morphology.
to collaborate and share insights, theory and methods with
academic and technical colleages who are interested in
working and sharing in a similar Text Analytics Paradigm and
- NarraText researches and develops
prototype Text Analytic software systems
in investors, who want to "commercialize" the
NarraText Text Analytics Paradigm.
flagship software, InfoMator(tm), is based on a code-generating
prototype architecture (and Work Flow Process), which is
(1) a high-performing, full-function,
fully-transparent NLP application prototype, and (2) a
complete software specification.
NarraText targets all text
domains, disciplines, and markets, and regards all text content as coherent
streams or bursts of utterances manifest as
clauses and sentence fragments ,
which map to event structures, discourse structures, and ultimately to
story roles that articulate the "Busness Story"
Lake City, UT USA
- NarraText was formed in 2011
with a focus on insurance claim notes processing and other types of
domain-specific and nuanced unstructured text.
- NarraText deploys
a proprietary Text Analytics Work
Flow Process-based software prototype, aka InfoMator(tm),
which leverages lexicons, taxonomies, and context vectors, using text
markers to that capture patterns (extractions in the language levels of
pragmatics,discourse, semantic, genre', syntax, and
engineers text via theories, methods, and
insights of Linguistics in its Text
Analytics approach to communication and language.
- Narratext parses
unstructured text to recover the
aspects of Pragmatics, Discourse Analysis, Narrative Analysis,
Genre' Analysis, Semantics, Syntax, and Morphology that are
inherent in all text.
- Narratext uses a
full-function prototype and
code-generator for building Text Analytics workflow processes
based on the CRISP-DM.
emphasizes use of Open-Source software
- Humans make sense of the world and form memories
based on "stories".
- Stories are captured and articulated via human
- Human language codes messages based on a multi-layer
hierarchy of production rules that are tacitly known
and used by an author/reader.
- Once the message is explicated -- via an verbal
utterance or written text sequence -- those organizational
rules can be employed by a hearer/reader to reconstruct the
- Engineered strategies for Natural Language Processing
(NLP) can explicitly identify, extract, label, and vectorize
the semantic primitives, syntactic roles, thematic
roles, constituent types and patterns, story roles,
and the story's actor/event-state event (chronological)
structure and discourse structure.
- The extracted roles and patterns of utterances or
sentences can be represented by data vectors that are amenable to
formal qualitative and quantitative analytical techniques and machine
learning to reveal rich new informational patterns for making
sense of the organization's environment, operations, and history.
- All Business Analysts act as narrators that
"Tell the Business Story"