Granted Patent B2

US 11,809,825 · App. 17/235,329 · Granted Nov 7, 2023

Inventor: Boris Galitsky (San Jose, CA)

Assignee: Oracle International Corporation

G06F40/295G06F16/00G06F16/9014G06F40/211G06F40/253G06F40/289G06F40/35G06N20/00

Abstract

Claims (120)

1. A computer-implemented method for navigating text using discourse trees, the method comprising:

creating a discourse tree from text comprising fragments, wherein the discourse tree comprises a root node, nonterminal nodes, and terminal nodes, each nonterminal node representing a rhetorical relationship between two of the fragments, and each terminal node is associated with one of the fragments, wherein the fragments are elementary discourse units;

determining, from the discourse tree, a subset of terminal nodes that are (i) associated with a nonterminal node representing a rhetorical relationship of type elaboration and (ii) associated with a nucleus elementary discourse unit of the elaboration;

calculating, for each node of the subset of terminal nodes, a respective path length from the root node;

identifying, from the subset of terminal nodes, a topic node having a path length that is a smallest path length of the path lengths;

determining, from the topic node, a topic of the discourse tree by extracting a noun phrase from the nucleus elementary discourse unit associated with the topic node;

determining, from an utterance received from a user device, a user intent, wherein determining the user intent comprises:

providing the utterance to a machine-learning model that is trained to classify utterances; and

receiving, from the machine-learning model, an indication that the user intent relates to the topic; and

providing, based on the user intent, text associated with the topic.

2. The method of claim 1 , further comprising:

identifying, in the discourse tree, a satellite elementary discourse unit associated with the topic node via the rhetorical relationship of type elaboration; and

providing text associated with the satellite elementary discourse unit to the user device.

3. The method of claim 1 , further comprising:

identifying, in the discourse tree, an additional terminal node from the subset of terminal nodes, the additional terminal node having an additional path length equal to or greater than the path length;

extracting additional text from an additional nucleus elementary discourse unit associated with the additional terminal node; and

providing the additional text to the user device.

4. The method of claim 1 , further comprising:

navigating, in the discourse tree, to an additional non-terminal node that is connected to the topic node and represents an additional rhetorical relation;

determining, from the additional non-terminal node, an additional topic by extracting an additional noun phrase from an additional nucleus elementary discourse unit associated with additional non-terminal node; and

responsive to determining that the additional rhetorical relation is one of type elaboration, condition, or temporal sequence:

extracting additional text from the additional nucleus elementary discourse unit; and

providing the additional text to the user device.

5. The method of claim 1 , further comprising:

determining that the utterance indicates familiarity with an additional topic;

identifying, in the discourse tree, an additional non-terminal node that comprises text representing the additional topic and represents an additional rhetorical relation; and

responsive to determining that the additional rhetorical relation is one of type elaboration, condition, or contrast:

extracting additional text from an elementary discourse unit associated with a terminal node that is connected to the additional terminal node; and

providing the additional text to the user device.

6. The method of claim 1 , further comprising:

determining that the utterance comprises a confirmation of an interest in an additional topic;

identifying, in the discourse tree, an additional non-terminal node that comprises text representing the additional topic and represents an additional rhetorical relation;

extracting additional text from an elementary discourse unit associated with a terminal node that is connected to the additional non-terminal node; and

providing the additional text to the user device.

7. The method of claim 1 , further comprising:

determining that the utterance comprises a request for additional details related to the topic; and

responsive to the determining:

identifying, in the discourse tree, a parent node directly connected to the topic node;

identifying, in the discourse tree, an additional terminal node that is related to the parent node via a rhetorical relation of type elaboration;

extracting additional text from an elementary discourse unit associated with the additional terminal node; and

providing the additional text to the user device.

8. The method of claim 1 , further comprising:

determining that the utterance comprises a question that is unrelated to the topic; and

responsive to the determining:

identifying one or more keywords from the utterance;

identifying an additional text by providing a query comprising the keywords to a search engine; and

providing text from one or more elementary discourse units of the additional text to the user device.

9. The method of claim 1 , further comprising:

forming, from the discourse tree, a communicative discourse tree that represents the text by matching each fragment that has a verb to a respective verb signature;

identifying, from the communicative discourse tree, a verb signature that corresponds to the topic node;

extracting a subject from the verb signature; and

providing the subject to the user device.

10. A system comprising:

a non-transitory computer-readable medium storing computer-executable program instructions; and

a processing device communicatively coupled to the non-transitory computer-readable medium for executing the computer-executable program instructions, wherein executing the computer-executable program instructions configures the processing device to perform operations comprising:

creating a discourse tree from text comprising fragments, wherein the discourse tree comprises a root node, nonterminal nodes, and terminal nodes, each nonterminal node representing a rhetorical relationship between two of the fragments, and each terminal node is associated with one of the fragments, wherein the fragments are elementary discourse units;

determining, from the discourse tree, a subset of terminal nodes that are (i) associated with a nonterminal node representing a rhetorical relationship of type elaboration and (ii) associated with a nucleus elementary discourse unit of the elaboration;

calculating, for each node of the subset of terminal nodes, a respective path length from the root node;

identifying, from the subset of terminal nodes, a topic node having a path length that is a smallest path length of the path lengths;

determining, from the topic node, a topic of the discourse tree by extracting a noun phrase from the nucleus elementary discourse unit associated with the topic node;

determining, from an utterance received from a user device, a user intent, wherein determining the user intent comprises:

providing the utterance to a machine-learning model that is trained to classify utterances; and

receiving, from the machine-learning model, an indication that the user intent relates to the topic; and

providing, based on the user intent and to the user device, text associated with the topic.

11. The system of claim 10 , wherein the operations further comprise:

identifying, in the discourse tree, a satellite elementary discourse unit associated with the topic node via the rhetorical relationship of type elaboration; and

providing text associated with the satellite elementary discourse unit to the user device.

12. The system of claim 10 , wherein the operations further comprise:

identifying, in the discourse tree, an additional terminal node from the subset of terminal nodes, the additional terminal node having an additional path length equal to or greater than the path length;

extracting additional text from an additional nucleus elementary discourse unit associated with the additional terminal node; and

providing the additional text to the user device.

13. The system of claim 10 , wherein the operations further comprise:

navigating, in the discourse tree, to an additional non-terminal node that is connected to the topic node and represents an additional rhetorical relation;

determining, from the additional non-terminal node, an additional topic by extracting an additional noun phrase from a nucleus elementary discourse unit associated with additional non-terminal node; and

responsive to determining that the additional rhetorical relation is one of type elaboration, condition, or temporal sequence:

extracting additional text from the additional nucleus elementary discourse unit; and

providing the additional text to the user device.

14. The system of claim 10 , wherein the operations further comprise:

determining that the utterance indicates familiarity with an additional topic;

identifying, in the discourse tree, an additional non-terminal node that comprises text representing the additional topic and represents an additional rhetorical relation; and

responsive to determining that the additional rhetorical relation is one of type elaboration, condition, or contrast:

extracting additional text from an elementary discourse unit associated with a terminal node that is connected to the additional non-terminal node; and

providing the additional text to the user device.

15. The system of claim 10 , wherein the operations further comprise:

determining that the utterance comprises a confirmation of an interest in an additional topic;

identifying, in the discourse tree, an additional non-terminal node that comprises text representing the additional topic and represents an additional rhetorical relation;

extracting additional text from an elementary discourse unit associated with a terminal node that is connected to the additional non-terminal node; and

providing the additional text to the user device.

16. The system of claim 10 , wherein the operations further comprise:

forming, from the discourse tree, a communicative discourse tree that represents the text by matching each fragment that has a verb to a respective verb signature;

identifying, from the communicative discourse tree, a verb signature that corresponds to the topic node;

extracting a subject from the verb signature; and

providing the subject to the user device.

17. A non-transitory computer-readable medium storing computer-executable program instructions that when executed by a processor, perform operations comprising:

creating a discourse tree from text comprising fragments, wherein the discourse tree comprises a root node, nonterminal nodes, and terminal nodes, each nonterminal node representing a rhetorical relationship between two of the fragments, and each terminal node is associated with one of the fragments, wherein the fragments are elementary discourse units;

determining, from the discourse tree, a subset of terminal nodes that are (i) associated with a nonterminal node representing a rhetorical relationship of type elaboration and (ii) associated with a nucleus elementary discourse unit of the elaboration;

calculating, for each node of the subset of terminal nodes, a respective path length from the root node;

identifying, from the subset of terminal nodes, a topic node having a path length that is a smallest path length of the path lengths;

determining, from the topic node, a topic of the discourse tree by extracting a noun phrase from the nucleus elementary discourse unit associated with the topic node;

determining, from an utterance received from a user device, a user intent, wherein determining the user intent comprises:

providing the utterance to a machine-learning model that is trained to classify utterances; and

receiving, from the machine-learning model, an indication that the user intent relates to the topic; and

providing, based on the user intent and to the user device, text associated with the topic.

18. The non-transitory computer-readable medium of claim 17 , wherein the operations further comprise:

navigating, in the discourse tree, to an additional non-terminal node that is connected to the topic node and represents an additional rhetorical relation;

determining, from the additional non-terminal node, an additional topic by extracting an additional noun phrase from a nucleus elementary discourse unit associated with additional non-terminal node; and

responsive to determining that the additional rhetorical relation is one of type elaboration, condition, or temporal sequence:

extracting additional text from the additional nucleus elementary discourse unit; and

providing the additional text to the user device.

19. The non-transitory computer-readable medium of claim 17 , wherein the operations further comprise:

determining that the utterance indicates familiarity with an additional topic;

identifying, in the discourse tree, an additional non-terminal node that comprises text representing the additional topic and represents an additional rhetorical relation;

responsive to determining that the additional rhetorical relation is one of type elaboration, condition, or contrast:

extracting additional text from an elementary discourse unit associated with a terminal node that is connected to the additional non-terminal node; and

providing the additional text to the user device.

20. The non-transitory computer-readable medium of claim 17 , wherein the operations further comprise:

determining that the utterance comprises a confirmation of an interest in an additional topic:

identifying, in the discourse tree, an additional non-terminal node that comprises text representing the additional topic and represents an additional rhetorical relation:

extracting additional text from an elementary discourse unit associated with a terminal node that is connected to the additional non-terminal node: and

providing the additional text to the user device.

Assignments (1)

Continuity (13)

Cited By (1)

US 12,530,531