Back to results
US 11,636,275 B1Granted
Natural language processing system and method using a synapper model unit
OwnerMin Ku Kiminventor
Life of this patent
Estimating the term…
Abstract
Disclosed herein are a natural language processing system and method, more particularly a natural language processing system and method using a synapper model unit.
Claims19
A natural language processing method using a synapper model to provide a correct structure to a sentence comprising:
- a preprocessing step in which text written or pronounced in a first natural language is received and converted into a first sentence comprised of a plurality of words in the first natural language;
- a first conversion step in which the first sentence is converted into a plurality of neural concept codes through a first code converter (WNC), wherein each of the plurality of neural concept codes provides sufficient information to determine a part of speech and a meaning for a corresponding word of each of the plurality of words;
- a step of creating synapper model data in a computer-readable storage medium for the plurality of neural concept codes, wherein the synapper model data is in the form of main loop data and data for one or more branches;
- wherein the main loop data is in the form of data for one of six sentence word orders of:
- (a) subject-object-verb, (b) subject-verb-object, (c) object-verb-subject, (d) verb-object-subject, (e) verb-subject-object, and (f) object-subject-verb, wherein the subject, object and verb are connected in a circular shape, and wherein the sentence word orders (a), (c), and (e) move in a clockwise direction, and the sentence word orders (b), (d), and (f) move in a counterclockwise direction;
- wherein the step of creating the synapper model data includes determining whether each of the plurality of neural concept codes is a node or a branch, wherein a node is one of either a subject, an object or a verb of the main loop data;
- and
- wherein the data for the one or more branches includes data for one or more supplementary words related to a particular node of the main loop data.
5 more elements
A natural language processing system using a synapper model to provide a correct structure to a sentence, comprising:
- an input unit to which a first natural language is input;
- an analog-to-digital converter (ADC) converting analog data input to the input unit into digital data;
- a first code converter (WNC) converting the digital data into a plurality of neural concept codes;
- a language processing unit configured to receive the neutral concept codes from the first code converter (WNC) and configured to create synapper model data for the plurality of neural concept codes, wherein the synapper model data is in the form of main loop data and data for one or more branches;
- wherein the main loop data is in the form of data for one of six sentence word orders of:
- (a) subject-object-verb, (b) subject-verb-object, (c) object-verb-subject, (d) verb-object-subject, (e) verb-subject-object, and (f) object-subject-verb, wherein the subject, object and verb are connected in a circular shape, and wherein the sentence word orders (a), (c), and (e) move in a clockwise direction, and the sentence word orders (b), (d), and (f) move in a counterclockwise direction;
- wherein the language processing unit is configured to create the synapper model data, at least in part, by determining whether each of the plurality of neural concept codes is a node or a branch, wherein a node is one of either a subject, an object or a verb of the main loop data;
- and
- wherein the data for the one or more branches includes data for one or more supplementary words related to a particular node of the main loop data.
6 more elements
Dependencies are inferred from the claim text.