Cognitive operating system

• US 20030049589A1
• Filed: 03/21/2002
• Published: 03/13/2003
• Est. Priority Date: 09/10/2001
• Status: Active Grant

• Current Assignee: Michael J. Feldhake
• Original Assignee: Michael J. Feldhake
• Inventors: Feldhake, Michael J.
• Granted Patent: US 7,577,631 B2
• Time in Patent Office: Days
• US Class Current: 434/238
• CPC Class Codes: G09B 19/00 Teaching not covered by other main groups of this subclass teaching or practice apparatus for gun-aiming or gun-laying F41G3/26

Abstract

An operating system configured to support cognitive capable environments is described. The system comprises a memory structure and an I/O process configured to update inputs and outputs. The operating system further includes a process to determine changes to symbols within the symbol space due to the stimulus and create STM images of the inputs. Additional processes are configured to pass stimulations between connecting symbols, measure temporal and spatial properties of images, filter the STM images based upon the properties, and propagate stimuli through hereditary structures. Further processes analyze propagated symbol groups of STM images for novel distinctions, create new symbols, connect novel stimuli together for symbol groupings, update existing symbols to include connections to the new symbols, and assign weights to the connections. Additional processes form stimulus-response pairs from received images, provide time-based erosion of connection weights of symbols and adapt learned responses to long term memories.

Claims

1. An operating system configured to support cognitive capable environments, said operating system comprising:

a memory structure, said memory structure comprising a symbol allocation table, a symbol space, and a short term memory (STM) image space;

at least one input/output (I/O) process configured to update inputs and outputs;

at least one short term memory load process configured to determine changes to symbols within said symbol space due to the stimulus and create STM images of the inputs;

at least one symbolic associative stimulation (SAS) process configured to pass stimulations (propagate) between connecting symbols within said symbol space;

at least one focus process configured to measure temporal and spatial properties of STM images and filter the STM images based upon the properties;

at least one process configured for propagation of stimuli through hereditary structures;

at least one process configured to analyze propagated symbol groups of STM images for novel distinctions;

at least one process configured to create new symbols, connect novel stimuli together for symbol groupings, update existing symbols to include connections to the new symbols, and assign weights to the connections;

at least one adaptation process configured to form stimulus-response pairs from received STM images;

at least one habituation process configured to provide time-based erosion of connection weights of symbols; and

at least one migration process to adapt learned responses to long term memories (LTM).

2. An operating system according to claim 1 wherein said STM image space is configured with at least one of an energy level, temporal attributes, spatial attributes, and a pointer reference into said symbol space.

3. An operating system according to claim 1 wherein said STM image space is configured to allow single-stimuli pattern matching of symbols.

4. An operating system according to claim 1 wherein said symbol space is configured with symbols having variable length alphanumeric codes (UACs).

5. An operating system according to claim 1 wherein said symbol space is configured with connection attributes having at least one of temporal, spatial and weight information.

6. An operating system according to claim 1 further configured to attach bi-polar weights to symbol connections within said symbol space.

7. An operating system according to claim 1 wherein said at least one habituation process is configured to purge STM image space based on a last conscious state of the image.

8. An operating system according to claim 1 wherein to analyze STM images for novel distinctions, said process is configured to base novelty upon an absence of conscious connected symbol image.

9. An operating system according to claim 1 wherein said adaptation process is configured to adapt connection weights of response connections for reinforcement learning.

10. An operating system according to claim 1 wherein said assimilation process comprises:

a distinctive definition process configured to connect novel stimuli together for symbol groupings; and

an association process configured to assimilate symbols based upon STM image temporal and spatial contiguity.

11. An operating system according to claim 10 wherein said focus process is driven by dynamic boundaries which set temporal and spatial settings for STM image filtering.

12. An operating system according to claim 1 wherein said memory structure is configured to store STM images based on classifications within STM channels.

13. An operating system according to claim 12 wherein said short term memory (STM) image space comprises at least one STM image channel for each of output responses, for input stimuli, for hereditary stimuli, for focal stimuli, and for abstracts.

14. An operating system according to claim 1 wherein said at least one process configured for propagation of stimuli through hereditary structures comprises a process that protects the hereditary structures from the learning processes by virtue of their classification.

15. An operating system according to claim 14 wherein said at least one process configured for propagation of stimuli using carrier attributes to determine which SAS node to assist in propagation of the stimuli.

16. An operating system according to claim 1 wherein to filter the STM images, said process is configured with filter elements comprising one or more temporal center elements, temporal bounds elements, spatial position elements, and spatial bounds elements thereby making the filter elements dynamic.

17. A method for building a short term memory (STM) structure within an operating system for processing, said memory structures being stored in a STM image and including elements having attributes in time, space, and stimulus, said method comprising:

initializing filter elements;

determining a setting for each filter element;

applying the filter to each STM image element;

determining if errors within the STM image elements are within bounds; and

setting a focal point attribute for each STM image based on whether the errors are within the bounds.

18. A method according to claim 17 wherein initializing filter elements comprises initializing at least one of temporal center elements, temporal bounds elements, spatial position elements, and spatial bounds elements.

19. A method according to claim 17 wherein applying a filter to each STM image element comprises calculating filter errors by a difference between STM image attributes and the filter elements.

20. A method according to claim 17 wherein determining if errors within the STM image elements are within bounds comprises:

ensuring that temporal errors are less than temporal bounds; and

ensuring that spatial errors are less than spatial bounds.

21. A computer program for the propagation of signals through a connection system, the signals being stored as symbolic associative stimulation (SAS) nodes configured for stimulation, carrier, and target information, the connection system configured to store symbol structures in memory, the symbol structures including stimulus, temporal information, spatial information, and connection data for connectivity to other symbols, said program configured to:

delay each SAS node for proper timing;

set carrier attributes of SAS nodes for propagation control;

propagate a summed symbolic signal excitation by creating new SAS nodes for each connection;

set the SAS node attributes based on connection information; and

discard SAS nodes after processing of the nodes is completed.

22. A computer program according to claim 21 wherein SAS nodes are stored as a stack and all non-zero latencies are retained for one iteration.

23. A computer program according to claim 21 wherein SAS nodes carry spatial information and are added to the symbol during propagation.

24. A computer program according to claim 21 wherein connection data is filtered based on connection attributes comprising at least one of temporal elements and spatial elements elements.

25. A computer program according to claim 21 wherein to create a symbolic associative stimulation nodes for processing said program is configured to create a node for at least one of further processing of abstract connections and processing in a memory area designated for protected hereditary structures.

26. A computer program according to claim 21 wherein connections between symbols comprise at least one of an experience type connection or an introspective type connection, a latency and spatial error between connections.

27. A method for propagating signals through a connection system, the signals being stored as symbolic associative stimulation (SAS) nodes configured for stimulation, carrier, and target information, the connection system configured to store symbol structures in memory, the symbol structures including stimulus, temporal information, spatial information, and connection data for connectivity to other symbols, said method comprising:

(a) delaying each SAS node for proper latency;

(b) setting a current symbol address for a symbol structure specified with target information in the SAS node;

(c) checking the SAS nodes for carrier information;

(d) adding stimulation to SAS nodes with a target information of stimulus;

(e) repeating steps (a) through (d) for all SAS nodes;

(f) resetting to the first SAS node;

(g) finding carrier SAS nodes by checking for matches between carrier information and target information within the SAS node;

(h) for matching information, calculating a target excitation;

(h1) determining whether excitation levels of current symbols is above a threshold;

(h2) creating a symbolic associative stimulation (SAS) node for processing of each connection system;

(h3) setting SAS node attributes for propogated stimulus based on connection data; and

(h4) repeating (h1) through (h3) for all other connections between symbols;

(i) for non-matching carrier information, discarding the SAS node; and

(j) returning to step (g) until last SAS node is processed.

28. A memory structure for storing short term memory (STM) images, the images being representations of symbolic or neural firing, the memory structure configured for the processing of the images, said memory structure comprising:

at least one input channel;

at least one output channel; and

at least one abstract channel.

29. A memory structure according to claim 28 further comprising:

at least one channel for hereditary traits, visual inputs, auditory inputs, and output feedback.