Relying on discourse trees to build ontologies

Patent metadata

Field	Value
Publication number	US-12153889-B2
Application number	US-202318540615-A
Country	US
Kind code	B2
Filing date	Dec 14, 2023
Priority date	Jan 7, 2021
Publication date	Nov 26, 2024
Grant date	Nov 26, 2024

Abstract

Systems, devices, and methods of the present invention involve discourse trees. In an example, a method involves generating a discourse tree. The method includes identifying, from the discourse tree, a central entity that is associated with a rhetorical relation of type elaboration and corresponds to a topic node that identifies a central entity of the text. The method includes determining a subset of elementary discourse units of the discourse tree that are associated with the central entity. The method includes forming generalized phrases from the subset of elementary discourse units. The method includes forming tuples from the generalized phrases, where a tuple is an ordered set of words in normal form. The method involves responsive to successfully converting an elementary discourse unit associated with an identified tuple into a logical representation, updating the ontology with an entity from the identified tuple.

First claim

What is claimed is:

1. A method of extending an ontology, the method comprising:
generating, from text, a discourse tree that represents rhetorical relationships between elementary discourse units generated from the text, wherein the discourse tree comprises a plurality of nodes, each nonterminal node representing a rhetorical relationship between two of the elementary discourse units, and each terminal node of the nodes of the discourse tree being associated with one of the elementary discourse units;
identifying, from the discourse tree, a node of the plurality of nodes that identifies a central entity of the text, the node being identified based at least in part on identifying an association between the node and a rhetorical relationship type corresponding to an elaboration type;
determining, from the discourse tree, a subset of elementary discourse units that are associated with the central entity, wherein determining the subset of elementary discourse units comprises identifying nucleus elementary discourse units that are associated with a corresponding rhetorical relation type that differs from the elaboration type or a joint type;
forming generalized phrases by identifying, in corresponding text associated with the subset of elementary discourse units, one or more elements common to two or more elementary discourse units in the subset of elementary discourse units;
forming at least one tuple from the generalized phrases by applying one or more syntactic or semantic templates to a respective generalized phrase, wherein the tuple comprises an ordered set of words in normal form;
converting, based at least in part on a type identified for the tuple, an elementary discourse unit associated with the tuple into a logical representation comprising a predicate and an argument; and
based at least in part on successfully converting the tuple, updating the ontology with an entity identified from the tuple.

2. The method of claim 1 , further comprising in response to receiving a query from a user device, locating the entity within the ontology and providing data corresponding to the entity to the user device.

3. The method of claim 2 , further comprising identifying an entity class by:
encoding the tuple as a vector representation;
providing the vector representation to a machine learning model; and
receiving, from the machine learning model, the entity class, wherein providing the data corresponding to the entity to the user device comprises providing the entity class to the user device.

4. The method of claim 1 , wherein identifying the central entity comprises:
locating a root node in the discourse tree;
determining, from the discourse tree, a subset of terminal nodes that are (i) associated with a corresponding nonterminal node representing a corresponding rhetorical relationship type corresponding to the elaboration type, and that (ii) represent a corresponding nucleus elementary discourse unit;
calculating, for each node of the subset of terminal nodes, a respective path length from the root node; and
identifying, from the subset of terminal nodes, a particular topic node having a path length that is a smallest path length of the path lengths.

5. The method of claim 1 , wherein converting the elementary discourse unit associated with the tuple into the logical representation comprises:
identifying the tuple as corresponding to one of: (i) a noun phrase type, (ii) a verb phrase type, (iii) an adjective phrase type, or (iv) a prepositional phrase type;
identifying that the tuple corresponds to the noun phrase type or the prepositional phrase type;
extracting, from the tuple, at least one of a head noun or a last noun for the predicate of the logical representation; and
extracting one or more other words as the argument of the logical representation.

6. The method of claim 1 , wherein converting the elementary discourse unit associated with the tuple into the logical representation comprises:
identifying the tuple as corresponding to one of: (i) a noun phrase type, (ii) a verb phrase type, (iii) an adjective phrase type, or (iv) a prepositional phrase type;
identifying that the tuple corresponds to the verb phrase type; and
extracting a verb of the tuple for the predicate of the logical representation and one or more other words for the argument of the logical representation.

7. The method of claim 1 , the tuple comprises one or more of: a verb, a subject, and an object.

8. The method of claim 1 , further comprising:
identifying an entity class of the tuple corresponding to the generalized phrases, wherein the entity class represents a category of the entity, wherein the updating comprises updating the ontology with the entity class.

9. A system, comprising:
a non-transitory computer-readable medium storing computer-executable program instructions associated with extending an ontology; 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 causes the processing device to perform operations comprising:
generate, from text, a discourse tree that represents rhetorical relationships between elementary discourse units generated from the text, wherein the discourse tree comprises a plurality of nodes, each nonterminal node representing a rhetorical relationship between two of the elementary discourse units, and each terminal node of the nodes of the discourse tree being associated with one of the elementary discourse units;
identify, from the discourse tree, a node of the plurality of nodes that identifies a central entity of the text, the node being identified based at least in part on identifying an association between the node and a rhetorical relationship type corresponding to an elaboration type;
determine, from the discourse tree, a subset of elementary discourse units that are associated with the central entity, wherein determining the subset of elementary discourse units comprises identifying nucleus elementary discourse units that are associated with a corresponding rhetorical relation type that differs from the elaboration type or a joint type;
form generalized phrases by identifying, in corresponding text associated with the subset of elementary discourse units, one or more elements common to two or more elementary discourse units in the subset of elementary discourse units;
form at least one tuple from the generalized phrases by applying one or more syntactic or semantic templates to a respective generalized phrase, wherein the tuple comprises an ordered set of words in normal form;
convert, based at least in part on a type identified for the tuple, an elementary discourse unit associated with the tuple into a logical representation comprising a predicate and an argument; and
based at least in part on successfully converting the tuple, update the ontology with an entity identified from the tuple.

10. The system of claim 9 , wherein executing the computer-executable program instructions further causes the processing device to, in response to receiving a query from a user device, locate the entity within the ontology and providing data corresponding to the entity to the user device.

11. The system of claim 10 , wherein executing the computer-executable program instructions further causes the processing device to identify an entity class by:
encoding the tuple as a vector representation;
providing the vector representation to a machine learning model; and
receiving, from the machine learning model, the entity class, wherein providing the data corresponding to the entity to the user device compri

Assignees

• Oracle Int Corp

Inventors

• Galitsky Boris

Classifications

• G06N5/04

  Inference or reasoning models · CPC title

• G06F40/35 Primary

  Discourse or dialogue representation · CPC title

• G06F16/3329

  Natural language query formulation · CPC title

• G06F40/295 Primary

  Named entity recognition · CPC title