Generating question templates in a knowledge-graph based question and answer system

Patent metadata

Field	Value
Publication number	US-11983640-B2
Application number	US-201916730083-A
Country	US
Kind code	B2
Filing date	Dec 30, 2019
Priority date	Dec 30, 2019
Publication date	May 14, 2024
Grant date	May 14, 2024

Abstract

Techniques for generating a natural language question template for an artificial intelligence question and answer (QA) system are disclosed. A graph database query relating to a QA system is parsed using a predefined schema. The parsing includes extracting a first plurality of values from the graph database query relating to a where clause in the graph database query, extracting a second plurality of values from the graph database query relating to a return clause in the graph database query, identifying a QA template rule relating to the graph database query, based on a match clause in the graph database query. A natural language question template is generated based on the first plurality of values, the second plurality of values, and the identified QA template rule. The natural language question template is suitable for use by the QA system as part of generating a response to a natural language question.

First claim

What is claimed is:

1. A method, comprising:
parsing a graph database query relating to an automated artificial intelligence question and answer (QA) system using a predefined schema, using a parser engine, the parsing comprising:
processing a where clause in the graph database query to generate a where clause processor output, comprising:
extracting a first plurality of values, comprising a node, an operation, a function, and a condition, from the graph database query;

processing a return clause in the graph database query to generate a return clause processor output, comprising:
extracting a second plurality of values, comprising a return context and a return function, from the graph database query; and

identifying a QA template rule relating to the graph database query, based on a match clause in the graph database query, comprising:
selecting the QA template rule from a collection of QA template rules based on the match clause; and

generating a natural language question template based on the where clause processor output, the return clause processor output, and the identified QA template rule, using a template generator; and
providing the natural language question template to the automated QA system so that the QA system can generate a response to a natural language question using the natural language question template.

2. The method of claim 1 , wherein the graph database query comprises a Cypher query.

3. The method of claim 1 , wherein the generated natural language question template identifies a type of natural language question.

4. The method of claim 1 , wherein the first plurality of values comprises the node relating to the where clause and a first one or more attributes relating to the node.

5. The method of claim 4 , wherein the second plurality of values comprises a second node relating to the return clause and a second one or more attributes relating to the second node.

6. The method of claim 4 , wherein the first plurality of values further comprises the operation relating to the where clause.

7. The method of claim 6 , wherein the first plurality of values further comprises at least one of a function relating to the where clause or a condition relating to the where clause.

8. The method of claim 1 , wherein the predefined schema comprises a predefined knowledge graph schema, and wherein the predefine schema comprises:
a first portion relating to one or more nodes and comprising one or more attributes relating to question type; and
a second portion relating to one or more relations between nodes and comprising a forward phrase attribute relating to a first question template and a backward phrase attribute relating to a second question template.

9. The method of claim 1 , wherein identifying the QA template rule relating to the graph database query comprises retrieving the QA template rule from a storage location.

10. The method of claim 1 , wherein generating the natural language question template based on the where clause processor output, the return clause processor output, and the identified QA template rule using the template generator further comprises:
generating a phrase for the natural language question template based on a sentence structure by using the predefined schema to identify at least one of: (i) a forward phrase relating to a first question template, or (ii) a backward phrase relating to a second question template.

11. A system, comprising:
a processor; and
a memory containing a program that, when executed on the processor, performs an operation, the operation comprising:
parsing a graph database query relating to an automated artificial intelligence question and answer (QA) system using a predefined schema, using a parser engine, the parsing comprising:
processing a where clause in the graph database query to generate a where clause processor output, comprising:
extracting a first plurality of values, comprising a node, an operation, a function, and a condition from the graph database query;
processing a return clause in the graph database query to generate a return clause processor output, comprising:
 extracting a second plurality of values, comprising a return context and a return function, from the graph database query; and
identifying a QA template rule relating to the graph database query, based on a match clause in the graph database query, comprising:
 selecting the QA template rule from a collection of QA template rules based on the match clause; and

generating a natural language question template based on the where clause processor output, the return clause processor output, and the identified QA template rule, using a template generator; and
providing the natural language question template to the automated QA system so that the QA system can generate a response to a natural language question using the natural language question template.

12. The system of claim 11 , wherein the graph database query comprises a Cypher query and wherein the generated natural language question template identifies a type of natural language question.

13. The system of claim 11 , wherein the first plurality of values comprises the node relating to the where clause and a first one or more attributes relating to the node and wherein the second plurality of values comprises a second node relating to the return clause and a second one or more attributes relating to the second node.

14. The system of claim 11 , wherein the predefined schema comprises a predefined knowledge graph schema, and wherein the predefine schema comprises:
a first portion relating to one or more nodes and comprising one or more attributes relating to question type; and
a second portion relating to one or more relations between nodes and comprising a forward phrase attribute relating to a first question template and a backward phrase attribute relating to a second question template.

15. The system of claim 11 , wherein generating the natural language question template based on the where clause processor output, the return clause processor output, and the identified QA template rule using the template generator further comprises:
generating a phrase for the natural language question template based on a sentence structure by using the predefined schema to identify at least one of: (i) a forward phrase relating to a first question template, or (ii) a backward phrase relating to a second question template.

16. A non-transitory computer program product comprising:
a computer-readable storage medium having computer-readable program code embodied therewith, the computer-readable program code executable by one or more computer processors to perform an operation, the operation comprising:
parsing a graph database query relating to an automated artificial intelligence question and answer (QA) system using a predefined schema, using a parser engine, the parsing comprising:
processing a where clause in the graph database query to generate a where clause processor output, comprising:
extracting a first plurality of values, comprising a node, an operation, a function, and a condition, from the graph database query;

processing a return clause in the graph database query to generate a return clause processor output, comprising:
extracting a second plurality of values, comprising a return context and a return function, from the graph database query; and

identifying a QA template rule relating to the graph database query, based on a match clause in the graph database query, comprising:
selecting the QA template rule from a collection of QA template rules based on the match clau

Assignees

• IBM

Inventors

• Huang Zi Ming
• Wang Jian
• Li Jing
• Jiang Jian Min
• Wang Ke
• Ni Xin

Classifications

• G06N5/04 Primary

  Inference or reasoning models · CPC title

• G06F16/2458

  Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries · CPC title

• G06F16/3329 Primary

  Natural language query formulation · CPC title

• G06F16/9024

  Graphs; Linked lists (G06F16/9027 takes precedence) · CPC title

• G06F40/186

  Templates · CPC title