Using communicative discourse trees to detect distributed incompetence

Patent metadata

Field	Value
Publication number	US-12001804-B2
Application number	US-202217749019-A
Country	US
Kind code	B2
Filing date	May 19, 2022
Priority date	May 10, 2017
Publication date	Jun 4, 2024
Grant date	Jun 4, 2024

Abstract

Techniques are disclosed for detecting distributed incompetence in text of a conversation using communicative discourse trees and then inserting an automatic response from an autonomous agent (chatbot) or other entity. For example, a computing system generates a communicative discourse tree from utterances from multiple agents to a user. The computing system obtains a prediction of whether the text includes distributed incompetence by applying a trained predictive model to the communicative discourse tree. Based on the detection, the computing system generates an updated response to a user device.

First claim

What is claimed is:

1. A computer-implemented method for determining a presence of distributed incompetence by analyzing a communicative discourse tree, the method comprising:
generating a communicative discourse tree from fragments of text, wherein the communicative discourse tree represents rhetorical relationships between the fragments of text, and wherein generating the communicative discourse tree comprises:
creating, from the fragments of text, a discourse tree that comprises nodes, wherein each nonterminal node of the nodes represents a rhetorical relationship between two of the fragments of text, and each terminal node of the nodes is associated with one of the fragments of text; and
matching each fragment that has a verb to a verb signature that comprises the verb of a respective fragment and one or more thematic roles that describe a relationship between the verb and related words in the fragment;

detecting a presence of distributed incompetence in the fragments of text by applying, to the communicative discourse tree, a machine-learning model that is trained to detect distributed incompetence;
generating a response based on the detection of the presence of distributed incompetence; and
presenting the generated response to a user device.

2. The method of claim 1 , further comprising:
identifying, via the machine-learning model and in the communicative discourse tree, a first communicative action that identifies a first entity as a first actor and a second entity as a first recipient of the first communicative action; and
identifying, via the machine-learning model and in the communicative discourse tree, a second communicative action that identifies the second entity as a second actor and the first entity as a second recipient of the second communicative action; and
presenting information comprising the first actor and the second actor to the user device.

3. The method of claim 1 , further comprising:
presenting a message to a user device;
receiving, from the user device, a response to the message; and
extracting the fragments of text from the message and the response.

4. The method of claim 1 , further comprising:
identifying, via the machine-learning model and in the communicative discourse tree, a first communicative action that is of class “deny” and identifies a first actor; and
presenting information comprising the first actor to the user device.

5. The method of claim 1 , wherein the matching comprises:
selecting a particular verb signature, comprising a verb of a respective fragment, from a plurality of verb signatures based on the particular verb signature having one or more roles of one or more words in the respective fragment; and
associating the particular verb signature with the fragment.

6. The method of claim 5 , wherein the associating further comprises:
identifying each of a plurality of thematic roles in the particular verb signature; and
matching, for each of the plurality of thematic roles in the particular verb signature, a corresponding word in the respective fragment to the thematic role.

7. The method of claim 1 , wherein the verb is a communicative verb.

8. A system for determining a presence of distributed incompetence by analyzing a communicative discourse tree comprising:
one or more processors configured to perform operations comprising:
generating a communicative discourse tree from fragments of text, wherein the communicative discourse tree represents rhetorical relationships between the fragments of text, and wherein generating the communicative discourse tree comprises:
creating, from the fragments of text, a discourse tree that comprises nodes, wherein each nonterminal node of the nodes represents a rhetorical relationship between two of the fragments of text, and each terminal node of the nodes is associated with one of the fragments of text; and
matching each fragment that has a verb to a verb signature that comprises the verb of a respective fragment and one or more thematic roles that describe a relationship between the verb and related words in the fragment;
detecting a presence of distributed incompetence in the fragments of text by applying, to the communicative discourse tree, a machine-learning model that is trained to detect distributed incompetence;
generating a response based on the detection of the presence of distributed incompetence; and
presenting the generated response to an user device.

9. The system of claim 8 , wherein the operations further comprise:
identifying, via the machine-learning model and in the communicative discourse tree, a first communicative action that identifies a first entity as a first actor and a second entity as a first recipient of the first communicative action; and
identifying, via the machine-learning model and in the communicative discourse tree, a second communicative action that identifies the second entity as a second actor and the first entity as a second recipient of the second communicative action; and
presenting information comprising the first actor and the second actor to the user device.

10. The system of claim 8 , wherein the operations further comprise:
presenting a message to a user device;
receiving, from the user device, a response to the message; and
extracting the fragments of text from the message and the response.

11. The system of claim 8 , wherein the operations further comprise:
identifying, via the machine-learning model and in the communicative discourse tree, a first communicative action that is of class deny and identifies a first actor; and
presenting information comprising the first actor to the user device.

12. The system of claim 8 , wherein the matching comprises:
selecting a particular verb signature, comprising a verb of a respective fragment, from a plurality of verb signatures based on the particular verb signature having one or more roles of one or more words in the respective fragment; and
associating the particular verb signature with the fragment.

13. The system of claim 12 , wherein the associating further comprises:
identifying each of a plurality of thematic roles in the particular verb signature; and
matching, for each of the plurality of thematic roles in the particular verb signature, a corresponding word in the respective fragment to the thematic role.

14. The system of claim 8 , wherein each verb signature of the verb signatures comprises one of (i) an adverb, (ii) a noun phrase, or (iii) a noun.

15. A non-transitory computer-readable medium storing a set of instructions for determining a presence of distributed incompetence by analyzing a communicative discourse tree, the set of instructions comprising:
one or more instructions that, when executed by one or more processors of a device, cause the device to perform operations comprising:
generating a communicative discourse tree from fragments of text, wherein the communicative discourse tree represents rhetorical relationships between the fragments of text, and wherein generating the communicative discourse tree comprises:
creating, from the fragments of text, a discourse tree that comprises nodes, wherein each nonterminal node of the nodes represents a rhetorical relationship between two of the fragments of text, and each terminal node of the nodes is associated with one of the fragments of text; and
matching each fragment that has a verb to a verb signature that comprises the verb of a respective fragment and one or more thematic roles that describe a relationship between the verb and related words in the fragment;

detecting a presence of distributed incompetence in

Assignees

• Oracle Int Corp

Inventors

• Galitsky Boris

Classifications

• G06F40/35 Primary

  Discourse or dialogue representation · CPC title

• G06F40/253

  Grammatical analysis; Style critique · CPC title

• G06F40/289

  Phrasal analysis, e.g. finite state techniques or chunking · CPC title