Topological methods to organize semantic network data flows for conversational applications
First Claim
1. A method executed by a data processor system for quantifying clarity in a natural language processing system by measuring abstractness deviation for sets of inheritance sibling nodes within a semantic inheritance network, comprising the steps of:
- a) identifying a set of inheritance sibling nodes;
b) computing an average abstractness for the set of inheritance sibling nodes;
c) computing an abstractness deviation from the average abstractness, for each inheritance sibling node in the set of inheritance sibling nodes;
d) summing the abstractness deviations of each inheritance sibling node in the set of inheritance sibling nodes;
e) comparing the summed abstractness deviation of a set of sibling nodes to a summed abstractness deviation of an alternative topology for the set of inheritance nodes; and
f) optimizing the semantic network inheritance link topology by selecting the alternative topology with less abstractness deviation.
6 Assignments
0 Petitions
Accused Products
Abstract
A system and methods for enforcing uniform branching of node-to-node inheritance links within semantic networks, to control data flows to and from such networks in conversational applications. Enforced uniform branching criteria converge the population of directly connected nodes of each node toward a small system-wide constant, and converge each sibling inheritance node to a similar level of abstractness, and are also used to select the best candidate tree from a set of competing representation trees within the semantic network. Uniform branching criteria are applied to competing trees for speech recognition, for object recognition in vision systems, for concept recognition in text scanning systems, and for algorithm definition. For speech recognition, phonemes are identified and matched to dictionary nodes in the semantic network. For visual object recognition, object features are identified and matched. For text scanning, words are identified and matched. For speech, visual and text the sets of competing representation trees are formed from alternative combinations of matched dictionary nodes.
367 Citations
42 Claims
-
1. A method executed by a data processor system for quantifying clarity in a natural language processing system by measuring abstractness deviation for sets of inheritance sibling nodes within a semantic inheritance network, comprising the steps of:
-
a) identifying a set of inheritance sibling nodes;
b) computing an average abstractness for the set of inheritance sibling nodes;
c) computing an abstractness deviation from the average abstractness, for each inheritance sibling node in the set of inheritance sibling nodes;
d) summing the abstractness deviations of each inheritance sibling node in the set of inheritance sibling nodes;
e) comparing the summed abstractness deviation of a set of sibling nodes to a summed abstractness deviation of an alternative topology for the set of inheritance nodes; and
f) optimizing the semantic network inheritance link topology by selecting the alternative topology with less abstractness deviation. - View Dependent Claims (2, 3, 4, 5, 6, 11)
-
-
7. An apparatus in a data processing system for quantifying clarity in a natural language processing system by measuring abstractness deviation for sets of inheritance sibling nodes within a semantic inheritance network, comprising:
-
a) a set of inheritance sibling nodes;
b) a stored computed abstractness for each inheritance sibling node;
c) a stored computed average abstractness for the set of inheritance sibling nodes;
d) a stored measured abstractness deviation for each inheritance sibling node, from a difference of the abstractness of the sibling node and the average abstractness of the set of inheritance sibling nodes;
e) a stored measured abstractness deviation, from a sum over each abstractness deviation;
f) a measured abstractness deviation for a set of inheritance sibling nodes and a measured abstractness deviation for an alternative topology for the set of inheritance sibling nodes; and
g) an optimized semantic network inheritance link topology selected from an alternative topology with less abstractness deviation. - View Dependent Claims (8, 9, 10, 12)
-
-
13. A method executed by a data processor system for quantifying clarity in a semantic inheritance network of a natural language processing system by measuring a deviation in branching from an optimal number of inheritance branches for a set of nodes, comprising the steps of:
-
a) identifying a set of nodes having inheritance links;
b) counting a population count of inheritance links for each of the identified nodes;
c) computing a deviation in branching for each nodes in the identified set of nodes, from a difference between the optimal number of inheritance branches and a population count of inheritance links;
d) comparing the deviation in branching of a set of sibling nodes to deviation in branching of an alternative topology for the set of inheritance nodes; and
e) optimizing the semantic network inheritance link topology by selecting the alternative topology with less deviation in branching. - View Dependent Claims (14, 15, 16)
a) identifying a set of inheritance sibling nodes;
b) computing an average abstractness for the set of inheritance sibling nodes;
c) computing an abstractness deviation from the average abstractness, for each inheritance sibling node in the set of inheritance sibling nodes, d) summing the abstractness deviations of each inheritance sibling node in the set of inheritance sibling nodes; and
e) calculating the overall deviation from optimal topology by combining via multiplication, the summed abstractness deviation with the deviation in branching.
-
-
15. The method of claim 13, further comprising a set of nodes representing elements of a visual image, further comprising a step of identifying elements of the visual image to represent as a set of nodes.
-
16. The method of claim 13, further comprising a set of nodes representing elements of a sequence of sounds, further comprising a step of identifying elements of the sequence of sounds to represent as a set of nodes.
-
17. An apparatus in a data processing system for quantifying clarity in a semantic inheritance network of a natural language processing system by measuring a deviation in branching from an optimal number of inheritance branches for a set of nodes, comprising:
-
a) a set of nodes having inheritance links;
b) a stored population count of inheritance links for each node having inheritance links;
c) a stored measured deviation in branching for each node in the identified set of nodes having inheritance links, from a difference in magnitude between the optimal number of branches and a population count of inheritance links;
d) a measured deviation in branching for a set of inheritance sibling nodes and a measured deviation in branching for an alternative topology for the set of inheritance sibling nodes; and
e) an optimized semantic network inheritance link topology selected from an alternative topology with less deviation in branching. - View Dependent Claims (18, 19, 20)
a) a set of inheritance sibling nodes, b) for each inheritance sibling node, a computed abstractness, c) for the set of inheritance sibling nodes, a computed average abstractness, d) for each inheritance sibling node, a computed deviation, computed from a difference of the computed abstractness and the average abstractness, e) a summed abstractness deviation, summed over each computed deviation, and f) an overall deviation from optimal topology computed by combining via multiplication, the summed abstractness deviation with the deviation in branching.
-
-
19. The apparatus of claim 17, further comprising a set of nodes representing elements of a visual image.
-
20. The apparatus of claim 17, further comprising a set of nodes representing elements of a sequence of sounds.
-
21. A method executed by a data processor system for identifying optimal paths of meaning from a set of shortest inheritance paths between a first node and a second node stored in a semantic inheritance network, comprising the steps of:
-
a) identifying a first set of inherited nodes of the first node;
b) identifying a second set of inherited nodes of the second node;
c) identifying a set of common nodes including inheritor nodes corresponding to the first set which are also inheritor nodes corresponding to the second set; and
d) identifying the optimal paths of meaning by identifying a set of shortest inheritance paths traversing the first node to a node in the first set, to a node in the set of common nodes, to a node in the second set, to the second node. - View Dependent Claims (22, 23, 24, 25, 26)
-
-
27. An apparatus in a data processing system to identify optimal paths of meaning from a set of shortest inheritance paths connecting a first node and a second node stored in a semantic inheritance network, comprising:
-
a) a first set of inherited nodes of the first node;
b) a second set of inherited nodes of the second node;
c) a set of common nodes including inheritor nodes corresponding to the first set which are also inheritor nodes corresponding to the second set; and
d) a set of optimal paths of meaning from a set of shortest inheritance paths traversing from the first node to a node in the first set, to a node in the set of common nodes, to a node in the second set, to the second node. - View Dependent Claims (28, 29, 30, 31, 32)
-
-
33. A method executed by a data processor system for quantifying clarity of a concept stored in a natural language processing system semantic inheritance network, by measuring a conceptual extent deviation from a difference between an optimal number of nodes and a population count of nodes linked by inheritance links from a concept node to a set of nodes representing the concept, comprising the steps of:
-
a) identifying a set of nodes representing the concept;
b) counting a population count of nodes linked by inheritance links from the concept node to a set of nodes representing the concept;
c) computing a conceptual extent deviation from a difference in magnitude between the optimal number of nodes and the population count of nodes;
d) comparing the conceptual extent deviation of a set of nodes representing a concept to conceptual extent deviation of an alternative topology for the set of inheritance nodes; and
e) optimizing the semantic network inheritance link topology by selecting the alternative topology with less conceptual extent deviation.
-
-
34. An apparatus in a data processing system for quantifying clarity of a concept stored in a natural language processing system semantic inheritance network, by measuring a conceptual extent deviation from the difference between an optimal number of nodes and a population count of nodes linked via inheritance links from a concept node to a set of nodes representing the concept, comprising:
-
a) a set of inheritance nodes representing the concept;
b) a stored population count of nodes linked via inheritance links from the concept node to a set of nodes representing the concept;
c) a stored measured conceptual extent deviation computed from a difference in magnitude between the optimal number of nodes and the population count of nodes;
d) a measured conceptual extent deviation for a set of nodes representing a concept and a measured conceptual extent deviation for an alternative topology for the set of inheritance sibling nodes; and
e) an optimized semantic network inheritance link topology selected from an alternative topology with less conceptual extent deviation.
-
-
35. A method executed by a data processor system for quantifying average abstractness of a concept stored in a natural language processing system semantic inheritance network, by measuring the average abstractness of nodes representing the concept, comprising the steps of:
-
a) identifying a set of inheritance nodes representing the concept;
b) counting a population count of inheritor nodes for each node representing the concept;
c) computing a node abstractness number from a population count of inheritor nodes for each node representing the concept;
d) computing the average abstractness by averaging the node abstractness numbers;
e) computing the average abstractness over inheritance sibling nodes of each node representing the concept;
f) comparing the average abstractness of a set of nodes representing a concept to average abstractness of an alternative topology for the set of inheritance nodes representing a concept; and
g) optimizing the semantic network inheritance link topology by selecting the alternative topology with average abstractness closest to the average abstractness over inheritance sibling nodes of each node representing the concept.
-
-
36. An apparatus in a data processing system for quantifying average abstractness of a concept stored in a natural language processing system semantic inheritance network, by measuring the average abstractness of nodes representing the concept, comprising:
-
a) a set of inheritance nodes representing the concept;
b) a stored population count of inheritor nodes for each node representing the concept;
c) a stored node abstractness number, measured from a population count of inheritor nodes for each node representing the concept;
d) an average abstractness computed by averaging the node abstractness numbers;
e) a measured average abstractness over inheritance sibling nodes of each node representing the concept;
f) a measured average abstractness for a set of nodes representing a concept and a measured average abstractness for an alternative topology for the set of inheritance sibling nodes; and
g) an optimized semantic network inheritance link topology selected from an alternative topology with average abstractness closest to the measured average abstractness over inheritance sibling nodes of each node representing the concept.
-
-
37. A method executed by a data processor system for quantifying an emotional status of a conversation stored in a natural language semantic inheritance network, processing conversation through a set of processes and representing a conversation as a set of nodes which are mapped to a normalized measure of conversational abstractness, a normalized measure of conversational extent, a normalized measure of conversational clarity, and a normalized measure of achieved conversational clarity which are mapped to quadrants of a multi-dimensional space having an origin point representing an absence of a strong emotional status and having points of the multi-dimensional space far from the origin point representing the presence of strong emotional status, comprising the steps of:
-
a) mapping the normalized measure of conversational abstractness, by determining from the topology of the set of nodes, an average abstractness of the set of nodes representing the conversation, and comparing the average abstractness to an ideal average abstractness or to an average abstractness over nodes in the semantic inheritance network;
b) mapping the normalized measure of conversational extent, by determining from the topology of the set of nodes, a population count of nodes linking nodes sufficient to link each node from the set of nodes to some other node from the set of nodes via semantic inheritance links, and comparing the population count of nodes to an ideal population count or to an average population count over conversations represented by the semantic inheritance network;
c) mapping the normalized measure of conversational clarity, by determining from the topology of the set of nodes, an amount of inconsistency in the number of inheritance links connected to sets of sibling nodes within the set of nodes, and comparing the amount of inconsistency to an ideal amount of inconsistency or to the average amount of inconsistency over conversations represented by the semantic inheritance network;
d) mapping the normalized measure of achieved conversational clarity, by comparing a change in the normalized measure of conversational clarity during the set of processes to an ideal amount of change or to the average amount of change stored from a previous set of processes;
e) mapping the normalized measure of conversational abstractness, the normalized measure of conversational extent, the normalized measure of conversational clarity, and the normalized measure of achieved conversational clarity each to separate dimensions of the multi-dimensional space;
f) mapping from quadrants of the multi-dimensional space to a state of emotion; and
g) confirming from user feedback that the mapping to state of emotion is correct.
-
-
38. An apparatus in a data processing system for quantifying an emotional status of a conversation stored in a natural language semantic inheritance network, processing conversation through a set of processes and representing a conversation as a set of nodes which are mapped to a normalized measure of conversational abstractness, a normalized measure of conversational extent, a normalized measure of conversational clarity, and a normalized measure of achieved conversational clarity which are mapped to quadrants of a multi-dimensional space having an origin point representing an absence of a strong emotional status and having points of the multi-dimensional space far from the origin point representing the presence of strong emotional status, comprising:
-
a) a stored measured conversational abstractness, calculated from a topology of the set of nodes, calculated from an average abstractness of the set of nodes representing the conversation compared to an ideal average abstractness or to an average abstractness over nodes in the semantic inheritance network;
b) a stored measured conversational extent, calculated from the topology of the set of nodes, calculated from a population count of nodes linking nodes sufficient to link each node from the set of nodes to some other node from the set of nodes via semantic inheritance links, compared to an ideal population count or to an average population count over conversations represented by the semantic inheritance network;
c) a stored measured conversational clarity from the topology of the set of nodes, by calculating from an amount of inconsistency in the number of inheritance links connected to sets of sibling nodes within the set of nodes, compared to an ideal amount of inconsistency or to the average amount of inconsistency over conversations represented by the semantic inheritance network;
d) a stored measured achieved conversational clarity, by calculated from a change in the normalized measure of conversational clarity during the set of processes compared to an ideal amount of charge in the normalized measure of conversational clarity or to the average amount of change in the normalized measure of conversational clarity stored from a previous set of processes;
e) a multi-dimensional space which maps to separate dimensions the normalized measure of conversational abstractness, the normalized measure of conversational extent, the normalized measure of conversational clarity, and the normalized measure of achieved conversational clarity;
f) a mapping from quadrants of the multi-dimensional space to a state of emotion; and
g) a user feedback confirming that the mapping to state of emotion is correct.
-
-
39. A method executed by a data processor system for quantifying empathy between conversations stored in a natural language semantic inheritance network, processing conversational symbols using a set of processes and a set of nodes and emotional path mappings through a multi-dimensional space representing degrees of emotional intensity, storing emotional path mappings for at least a first conversation and a second conversation, comprising the steps of:
-
a) mapping the emotional path for the first conversation as a sequence of points in the multi-dimensional space by mapping emotional shifts for the set of processes and the set of nodes of the first conversation;
b) mapping the emotional path for the second conversation as a sequence of points in the multi-dimensional space by mapping emotional shifts for the set of processes and the set of nodes of the second conversation;
c) calculating the empathy between the first conversation and the second conversation by calculating the distance through the multi-dimensional space between the emotional path of the first conversation and the emotional path of the second conversation; and
d) confirming from user feedback that the calculated empathy is correct.
-
-
40. An apparatus in a data processing system for quantifying empathy between conversations stored in a natural language semantic inheritance network, processing conversational symbols using a set of processes and a set of nodes, and emotional path mappings though a multi-dimensional space representing degrees of emotional intensity, storing emotional path mappings for at least a first conversation and a second conversation, comprising:
-
a) a stored map of an emotional path for the first conversation, the map having point sequences in the multi-dimensional space which map emotional shifts for the set of processes and the set of nodes of the first conversation;
b) a stored map of an emotional path for the second conversation, the map having point sequences in the multi-dimensional space which map emotional shifts for the set of processes and the set of nodes of the second conversation;
c) a stored measured empathy between the first conversation and the second conversation, measured from the distance through the multi-dimensional space between the emotional path of the first conversation and the emotional path of the second conversation; and
d) a user feedback confirming that the measured empathy is correct.
-
-
41. A method executed by a data processor system for verifying successful propagation of meaning from a first conversation between distributed semantic network servers to a second conversation between distributed semantic inheritance network servers by comparing a distance between an emotional path stored for the first conversation to an emotional path stored for the second conversation, comprising the steps of:
-
a) mapping the emotional path for the first conversation as a sequence of points in the multi-dimensional space by mapping emotional shifts for the set of processes and the set of nodes of the first conversation;
b) mapping the emotional path for the second conversation as a sequence of points in the multi-dimensional space by mapping emotional shifts for the set of processes and the set of nodes of the second conversation;
c) calculating the distance through the multi-dimensional space between the emotional path mapping of the first conversation and the emotional path mapping of the second conversation; and
d) verifying that the distance is not more than a maximum acceptable distance between emotional paths or an average amount of distance for conversations where meaning was successfully propagated.
-
-
42. An apparatus in a data processing system for verifying successful propagation of meaning from a first conversation between distributed semantic network servers to a second conversation between distributed semantic inheritance network servers by comparing a distance between an emotional path stored for the first conversation to an emotional path stored for the second conversation, comprising:
-
a) a stored map of an emotional path for the first conversation, the map having a sequence of points in the multi-dimensional space mapping emotional shifts for the set of processes and the set of nodes of the first conversation;
b) a stored map of an emotional path for the second conversation, the map having a sequence of points in the multi-dimensional space mapping emotional shifts for the set of processes and the set of nodes of the second conversation;
c) a stored measured distance through the multi-dimensional space between the emotional path mapping of the first conversation and the emotional path mapping of the second conversation; and
d) a generated confirmation of successful propagation of meaning, if the stored measured distance is not more than a maximum acceptable distance between emotional paths or an average amount of distance for conversations where meaning was successfully propagated.
-
Specification