US6604094B1
Expired / Lapsed
US
⏱️ 0.0 years remaining
Simulating human intelligence in computers using natural language dialog
Assignee: SUMBIONAUTICS Corp
•
Filing: 2000-08-09
•
Granted: 2003-08-05
•
Expires: 2020-08-09
Abstract
Abstract A method and apparatus for simulating human intelligence and natural language dialog capability is disclosed. The present invention contains a cognitive model of human intelligence ( 20 ), a mathematical model of information abstraction, synthetic dialog interaction ( 202 ), a method of language-independent computer learning through training ( 201 ), interaction and document reading ( 203 ) and a method of efficient computer implementation ( 200 ) of all preceding parts. The cognitive model ( 20 ) is the theoretical basis of the entire invention, describes the way humans learn and interact in general terms, provides a mathematical basis for natural language ( 40 ) learning and interaction and establishes a basis for detailed computer implementation of the theory.
Inventors
- 👤 Henry M. Harris
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-24 16:22
CPC Classifications
G; G06; G06F; G06F40/00; G06F40/40; G06F40/55; G06F40/56; G06N; G06N3/00; G06N3/02; G10; G10L; G10L13/00; G10L15/00; G10L15/26
Claims Analysis
Total Claims: 23
Claim 1
Independent Claim
102 words
An apparatus comprising: a processing means ( 20 ) for simulating human capability to understand and to converse; said processing means ( 20 ) including automated, interactive dialog and adaptive functions; said processing means ( 20 ) using as an input a stream of natural language ( 40 ); said processing means ( 20 ) having a cognitive model ( 10 ) of human intelligence; said cognitive model ( 10 ) of human intelligence including pattern recognition, association and sensory memory capabilities; and said processing means ( 20 ) simulating a plurality of human intelligence characteristics in order to perform useful tasks.
Claim 2
Independent Claim
102 words
An apparatus comprising: a processing means ( 20 ) for simulating human capability to understand and to converse; said processing means ( 20 ) including automated, interactive dialog and adaptive functions; said processing means ( 20 ) using as an input a stream of natural language ( 40 ); said processing means ( 20 ) having a cognitive model ( 10 ) of human intelligence; said cognitive model ( 10 ) of human intelligence including pattern recognition, association and sensory memory capabilities; and said processing means ( 20 ) simulating a plurality of human intelligence characteristics in order to perform useful tasks.
Claim 3
Dependent Claim
263 words
The apparatus as claimed in claim 1 in which said processing means ( 20 ) includes: a conversion means ( 41 ) for converting a stream of said natural language ( 40 ) in electronic format ( 50 ) into sentences (w); said sentences (w) containing declarative information ( 12 ), and a stimulus ( 16 ); said stimulus ( 16 ) being a said sentence (w) which is a response and a said sentence (w) which requests a response ( 18 ); associative memory means ( 14 ) for storing said natural language ( 50 ); said associative memory means ( 14 ) storing said electronic format ( 50 ) as an abstract Concept ( 42 ); said abstract Concept ( 42 ) including a pattern (p) of said natural language ( 40 ); said associative memory means ( 14 ) being searched for a same said pattern (p) of said natural language ( 40 ); said same pattern (p) being previously stored as a pattern of response (p) to a prior stimulus ( 16 ); said Concept ( 42 ) being abstracted from said associative memory ( 14 ) when said pattern (p) matches a pattern of response (p) to said prior stimulus ( 16 ); a matching knowledge content ( 30 , 32 ) from said associative memory ( 14 ) being attached to said pattern of response (p); and said pattern (p) and said knowledge content ( 30 , 32 ) being converted to an electronic equivalent ( 28 ) which can be visually displayed, aurally displayed and printed.
Claim 4
Dependent Claim
36 words
The apparatus as claimed in claim 2 in which said associative memory means ( 14 ) includes a conceptual database ( 30 ), a contextual database ( 32 ) and a pattern buffer ( 34 ).
Claim 5
Dependent Claim
60 words
The apparatus as claimed in claim 3 in which said sentences (w) containing declarative information ( 12 ), and said sentences (w) containing said stimulus ( 16 ) are processed and stored ( 219 ) in a working memory ( 242 ), converted to a said pattern (p) and appended to said pattern buffer ( 34 ) for long-term storage.
Claim 6
Dependent Claim
53 words
The apparatus as claimed in claim 2 in which said abstract Concept ( 42 ) is created by de-constructing a sentence ( 120 ) into a pattern (p) and eigen words; said eigen word being a word that can be replaced in a same sentence without changing the conceptual basis of said sentence.
Claim 7
Dependent Claim
52 words
The apparatus as claimed in claim 2 in which said knowledge content ( 30 , 32 ) is derived from structured, interactive training ( 201 ); said structured, interactive training ( 201 ) including inputting sentence (w) pairs which are analyzed by said cognitive model ( 10 ) to produce abstract information.
Claim 8
Dependent Claim
34 words
The apparatus as claimed in claim 2 in which said cognitive model ( 10 ) of human intelligence includes an ability to transform said sentence (w) into a pattern (p) represented by substantive examples.
Claim 9
Dependent Claim
85 words
The apparatus as claimed in claim 2 in which said Concept ( 42 ) is represented mathematically by a set of Functors ( 122 ); said Functor ( 122 ) being a construct having two components, namely a pattern (p) and an argument list (e n , e m . . . ); said argument list (e n , e m . . . ) including sets of all words in said associative memory means ( 14 ) associated with said pattern ( 14 ).
Claim 10
Dependent Claim
22 words
The apparatus as claimed in claim 1 in which said stream of natural language ( 40 ) is speech ( 40 ).
Claim 11
Dependent Claim
25 words
The apparatus as claimed in claim 1 in which said stream of natural language ( 40 ) emanates from a text document ( 203 ).
Claim 12
Dependent Claim
32 words
The apparatus as claimed in claim 1 in which said stream of said natural language ( 40 ) is independent of characteristics common to a particular nation, tribe and other speech community.
Claim 13
Independent Claim
93 words
A method of simulating human intelligence comprising the steps of: providing a processing means ( 20 ) for simulating human capability to understand and to converse with a cognitive model of human intelligence including automated, interactive dialog and adaptive functions; providing pattern recognition and sensory memory association capabilities in said cognitive model of human intelligence; using natural language ( 50 ) as an input ( 40 ) to said processing means ( 20 ); and simulating a plurality of human intelligence characteristics with said processing means ( 20 ) to perform useful tasks.
Claim 14
Independent Claim
93 words
A method of simulating human intelligence comprising the steps of: providing a processing means ( 20 ) for simulating human capability to understand and to converse with a cognitive model of human intelligence including automated, interactive dialog and adaptive functions; providing pattern recognition and sensory memory association capabilities in said cognitive model of human intelligence; using natural language ( 50 ) as an input ( 40 ) to said processing means ( 20 ); and simulating a plurality of human intelligence characteristics with said processing means ( 20 ) to perform useful tasks.
Claim 15
Dependent Claim
291 words
The method as claimed in claim 12 in which the step of simulating human intelligence with a processing means ( 20 ) includes the steps of: converting a stream of said natural language ( 40 ) into sentences (w) in an electronic format ( 50 ) with a conversion means ( 41 ); said sentences (w) containing declarative information ( 12 ), and a stimulus ( 16 ); said stimulus ( 16 ) being a said sentence (w) which is a response and a said sentence (w) which requests a response ( 18 ); storing said sentences (w) in electronic format ( 50 ) as an abstract Concept ( 42 ) in an associative memory means ( 14 ); including a pattern (p) of said natural language ( 40 ) in said abstract Concept ( 42 ); storing a pattern (p) of said natural language ( 40 ) as a pattern (p) in response to a stimulus ( 16 ); storing a knowledge content ( 30 , 32 ) associated with said pattern (p) in said associative memory means; searching said associative memory means ( 14 ) for a same said pattern (p) previously stored as a pattern (p) in response to an earlier applied stimulus ( 16 ); abstracting said Concept ( 42 ) from said associative memory ( 14 ) when said pattern (p) matches said pattern (p) in response to said earlier applied stimulus ( 16 ); attaching said matching knowledge content ( 30 , 32 ) from said associative memory ( 14 ) to said pattern (p); and converting said pattern (p) and said knowledge content ( 30 , 32 ) to an electronic equivalent ( 28 ) which can be visually displayed, aurally displayed and printed.
Claim 16
Dependent Claim
48 words
The method as claimed in claim 13 in which the step of converting a stream of said natural language ( 40 ) into sentences (w) in an electronic format ( 50 ) with a conversion means ( 41 ) includes converting a stream of speech ( 40 ).
Claim 17
Dependent Claim
49 words
The method as claimed in claim 13 in which the step of converting a stream of said natural language ( 40 ) into sentences (w) in an electronic format ( 50 ) with a conversion means ( 41 ) includes converting text emanating from a document ( 203 ).
Claim 18
Dependent Claim
59 words
The method as claimed in claim 13 in which the step of searching said associative memory means ( 14 ) for a same said pattern (p) of said natural language ( 40 ) includes the steps of searching a conceptual database ( 30 ), searching a contextual database ( 32 ) and searching a pattern buffer ( 34 ).
Claim 19
Dependent Claim
89 words
The method as claimed in claim 13 in which the step of abstracting said Concept ( 42 ) from said associative memory ( 14 ) when said pattern (p) matches a pattern of response (p) to said prior stimulus ( 16 ), includes the step of: creating said abstract Concept ( 42 ) by de-constructing a sentence ( 120 ) into a pattern (p) and eigen words; said eigen word being a word that can be replaced in a same sentence without changing the conceptual basis of said sentence.
Claim 20
Dependent Claim
78 words
The apparatus as claimed in claim 13 in which the step of storing a knowledge content ( 30 , 32 ) associated with said pattern (p) in said associative memory means includes the steps of: deriving said knowledge content ( 30 , 32 ) from structured, interactive training ( 201 ); and inputting sentence (w) pairs in said structured, interactive training ( 201 ), which are analyzed by said cognitive model ( 10 ) to produce abstract information.
Claim 21
Dependent Claim
69 words
The method as claimed in claim 13 in which the step of converting a stream of said natural language ( 40 ) into sentences (w) in an electronic format ( 50 ) with a conversion means ( 41 ) includes the step of: converting said stream of said natural language ( 40 ) sentences (w) being independent of characteristics common to a particular nation, tribe and other speech community.
Claim 22
Independent Claim
141 words
A method of simulating human intelligence comprising the steps of: storing patterns of language ( 40 ) and specific knowledge electronically in an associative memory ( 14 ) database ( 30 , 32 , 34 ) and creating therein patterns (p) of natural language ( 40 ) having specific knowledge content; comparing a new pattern (p) of natural language to ( 40 ) said patterns of language electronically in said associative memory ( 14 ) database ( 30 , 32 , 34 ); selecting a previously stored, same said pattern (p) of language from said data base ( 30 , 32 , 34 ); extracting electronically a set of words associated with said pattern (p) and said specific knowledge content from said data base ( 30 , 32 , 34 ); and converting said set of words into human comprehensible form.
Claim 23
Independent Claim
141 words
A method of simulating human intelligence comprising the steps of: storing patterns of language ( 40 ) and specific knowledge electronically in an associative memory ( 14 ) database ( 30 , 32 , 34 ) and creating therein patterns (p) of natural language ( 40 ) having specific knowledge content; comparing a new pattern (p) of natural language to ( 40 ) said patterns of language electronically in said associative memory ( 14 ) database ( 30 , 32 , 34 ); selecting a previously stored, same said pattern (p) of language from said data base ( 30 , 32 , 34 ); extracting electronically a set of words associated with said pattern (p) and said specific knowledge content from said data base ( 30 , 32 , 34 ); and converting said set of words into human comprehensible form.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2006-08-04 | 2007-02-03 | 2007-08-05 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2010-08-04 | 2011-02-03 | 2011-08-05 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2014-08-04 | 2015-02-03 | 2015-08-05 | $7700.00 | Window Closed / Lapsed |