:: A NATURAL LANGUAGE PROCESSING (NLP) SOLUTION:: PARSEON

HOME

ABOUT PARSEON

SOLUTION

TECHNOLOGY

PROFILE

Scientific text is an acquired electronically and fed into the NLP engine

Words are tagged as part of speech and each sentence is broken down into its component phrases. The relationships are recognized and stored, as are the source sentences, all are indexed to the source text.

The OnSolver DataStream^TM feeds extracted knowledge to the OnSolver DataRepository^TM on an ongoing basis, - as soon as text becomes available online, it can be processed and added to the database, eliminating the 3 to 6 month time lag inherent in manually curated databases.

The extracted knowledge is simplified into binary relationships to enable it to be displayed graphically, but the complex relationships and their indexing are stored also. Key words mentions are stored as well.

Scientific text

Pars Engine

Onsolver data stream

Onsolver data Reposition

Periodic updates

Database

Over 10 million relationships between over 1.4 million terms are currently available in the OnSolver DataRepository^TM can easily be visualized in a variety of Graphical User Interfaces. Relationships can be displayed in context hyperlinked to abstracts and sources.

Graphical User Interface

Parseon data can be stored in a unique proprietary graphical database, developed by one of Parseon’s partners. The use of this database enables extremely fast searching of Parseon’s over 1 terabyte of data, as well as ontological storage and retrieval of data.

HOME | ABOUT PARSEON | SOLUTION | TECHNOLOGY | PROFILE