Granted Patent B2

US 12,475,889 · App. 18/523,447 · Granted Nov 18, 2025

Inventor: Jodi Kodish-Wachs (Alburquerque, NM)

Assignee: Cerner Innovation, Inc.

G10L15/22G10L15/1815G10L21/028G10L2015/223

Abstract

Claims (69)

1 . One or more non-transitory computer-readable media comprising computer-executable instructions that, when executed by one or more processors, cause performance of operations, comprising:

determining a first candidate concept corresponding to a first dialogue segment, and a first weight indicating a first relevancy of the first candidate concept to a knowledge domain;

determining a second candidate concept corresponding to the first dialogue segment, and a second weight indicating a second relevancy of the second candidate concept to a knowledge domain;

determining that the first weight is higher than the second weight;

responsive to determining that the first weight is higher than the second weight: selecting the first candidate concept, from a group comprising the first candidate concept and the second candidate concept, as a first target concept;

determining a first dialogue goal for the first target concept; and

generating a first structured link associating the first dialogue goal to the first target concept.

2 . The one or more non-transitory computer-readable media of claim 1 , wherein the operations further comprise:

excluding the second candidate concept from qualification as a target concept based at least on the second weight falling below a qualification threshold.

3 . The one or more non-transitory computer-readable media of claim 1 , wherein the operations further comprise:

determining a third candidate concept corresponding to a second dialogue segment, wherein the second dialogue segment differs from the first dialogue segment;

selecting the third candidate concept as a second target concept;

determining a second dialogue goal for the second target concept; and

generating a second structured link associating the second dialogue goal to the second target concept.

4 . The one or more non-transitory computer-readable media of claim 3 ,

wherein the first dialogue segment corresponds to a first entity and the second dialogue segment corresponds to a second entity, wherein the second entity differs from the first entity.

5 . The one or more non-transitory computer-readable media of claim 3 , wherein the operations further comprise:

determining a third weight indicating a third relevancy of the third candidate concept to the knowledge domain; and

selecting the third candidate concept as the second target concept based at least on the third weight meeting a qualification threshold.

6 . The one or more non-transitory computer-readable media of claim 5 , wherein the operations further comprise:

determining the second dialogue goal based on the first target concept and the second target concept.

7 . The one or more non-transitory computer-readable media of claim 5 , wherein the operations further comprise:

determining an extracted concept based on the second target concept; and

determining the second dialogue goal based on the extracted concept, wherein the second structured link associates the second dialogue goal to the extracted concept.

8 . The one or more non-transitory computer-readable media of claim 5 , wherein the operations further comprise:

storing the first structured link and the second structured link in a structured dataset;

wherein the structured dataset is queried based on at least one dialogue goal, and

wherein a first query of the structured dataset corresponds to the first dialogue goal,

wherein the first query returns a first query result comprises the first target concept and the second target concept.

9 . The one or more non-transitory computer-readable media of claim 1 , wherein the first dialogue goal comprises a purpose for a communication between a plurality of entities.

10 . The one or more non-transitory computer-readable media of claim 1 ,

wherein the first dialogue segment comprises a first set of one or more utterances and a second set of one or more utterances,

wherein the first candidate concept corresponds to the first set of one or more utterances, and

wherein the second candidate concept corresponds to the second set of one or more utterances.

11 . The one or more non-transitory computer-readable media of claim 1 , wherein the operations further comprise:

parsing a second dialogue segment to obtain an utterance; and

based on the utterance, modifying the first candidate concept.

12 . The one or more non-transitory computer-readable media of claim 11 ,

wherein modifying comprises at least one of: confirming, validating, negating, denying, qualifying, or quantifying.

13 . The one or more non-transitory computer-readable media of claim 1 , wherein the operations further comprise:

determining an extracted concept, wherein the extracted concept is correlated to the first target concept based on a semantic scheme,

wherein the first structured link associates the first dialogue goal to the extracted concept.

14 . The one or more non-transitory computer-readable media of claim 13 ,

wherein the extracted concept comprises at least one of:

a confirmation of the first target concept, a validation of the first target concept, a negation of the first target concept, a denial of the first target concept, a qualification of the first target concept, or a quantification of the first target concept.

15 . The one or more non-transitory computer-readable media of claim 1 , wherein the first weight is determined based on a weighting scheme, wherein the weighting scheme is configured for use in determining whether to select a candidate concept as a target concept.

16 . The one or more non-transitory computer-readable media of claim 15 ,

wherein the weighting scheme comprises at least one of: a rule, a logical criterion, a condition, a prediction, a pattern inference algorithm, or a machine learned model.

17 . The one or more non-transitory computer-readable media of claim 1 , wherein selecting the first candidate concept as the first target concept comprises:

determining, based on a weighting scheme, that the first candidate concept qualifies for inclusion in a set of one or more target concepts.

18 . The one or more non-transitory computer-readable media of claim 17 ,

wherein the weighting scheme comprises at least one of: a fuzzy logic algorithm, a neural network, a finite state machine, a support vector machine, a logistic regression, a clustering algorithm, or a machine learning algorithm.

19 . A method, comprising:

determining a first candidate concept corresponding to a first dialogue segment, and a first weight indicating a first relevancy of the first candidate concept to a knowledge domain;

determining a second candidate concept corresponding to the first dialogue segment, and a second weight indicating a second relevancy of the second candidate concept to a knowledge domain;

determining that the first weight is higher than the second weight;

responsive to determining that the first weight is higher than the second weight: selecting the first candidate concept, from a group comprising the first candidate concept and the second candidate concept, as a first target concept;

determining a first dialogue goal for the first target concept; and

generating a first structured link associating the first dialogue goal to the first target concept;

wherein the method is performed by at least one device including a hardware processor.

20 . A system, comprising:

at least one hardware processor;

the system being configured to execute operations, using the at least one hardware processor, the operations comprising:

determining a first candidate concept corresponding to a first dialogue segment, and a first weight indicating a first relevancy of the first candidate concept to a knowledge domain;

determining a second candidate concept corresponding to the first dialogue segment, and a second weight indicating a second relevancy of the second candidate concept to a knowledge domain;

determining that the first weight is higher than the second weight;

responsive to determining that the first weight is higher than the second weight: selecting the first candidate concept, from a group comprising the first candidate concept and the second candidate concept, as a first target concept;

determining a first dialogue goal for the first target concept; and

generating a first structured link associating the first dialogue goal to the first target concept.

Assignments (1)

Continuity (3)

References Cited (85)