Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Target domain knowledge base generation method and device and question answering method and device

A technology for target fields and problems, applied in digital data information retrieval, special data processing applications, instruments, etc., can solve problems such as unable to generate answers

Active Publication Date: 2020-08-14
HUAWEI TECH CO LTD
View PDF6 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] This application provides a method and device for generating a knowledge base in the target field and answering questions, which is used to solve the problem that the question answering system in the prior art cannot generate an answer if the content of the user's question does not exist in the original knowledge base

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Target domain knowledge base generation method and device and question answering method and device
  • Target domain knowledge base generation method and device and question answering method and device
  • Target domain knowledge base generation method and device and question answering method and device

Examples

Experimental program
Comparison scheme
Effect test

example 1

[0161] The embodiment of the present application provides a knowledge representation method in the field of financial business accounting rules. The field of business accounting rules mainly defines accounting rules or accounting calibers for the company's main operating indicators. The hardware platform used in the implementation of this application can be a general-purpose PC computer, and the software can be implemented using Python3.5. The steps of building the target domain knowledge base can include:

[0162] Determine three types of knowledge representation: concept words, events, and predicate logic formulas. Concept words include concept instances or concepts, which are the abstraction of objects in the target domain knowledge and express static knowledge. Events describe the behavior, action or state of objects, expressing dynamic knowledge. The predicate logic formula is an abstract description of the rules in the target domain knowledge (for example, the norms and...

example 1

[0179] Specifically, the service cost rules include: the expenditure incurred by activities or personnel directly related to the delivery of the service contract shall be included in the service cost, including the expenditure incurred directly in the service and the direct management and support of these personnel engaged in the service.

[0180] In the service cost rules, identifiable events include: generation of expenditure events, contract delivery events, service cost confirmation events, and identifiable concept instances include: activities and personnel. Therefore, the service cost rule can be expressed as: ЭxЭy (incurring expenditure (X, Y) Λ contract delivery (X) Λ (activity expenditure (Y) ∨ personnel expenditure (Y))) -> service cost (Y)

[0181] According to the generation rule of the predicate logic, it is determined that the rule header included in the service cost rule is a service cost confirmation event. The rule body includes: generating expenditure events,...

example 2

[0191] Labor cost rules, including mainly referring to the company's own and outsourced labor expenditures engaged in delivery activities, which can include wages, bonuses, social insurance, water and electricity fees, communication fees, travel expenses, etc.

[0192] In the labor cost rules, identifiable events include: expenditure events, contract delivery events, labor cost confirmation events, identifiable concepts include: company-owned manpower, outsourced manpower, expenditure content: wages, bonuses, social insurance, Water and electricity costs, communication costs, travel expenses, etc. Therefore, the predicate logic formula corresponding to the labor cost rule can be expressed as: ЭxЭy (expenditure (X, Y) Λ contract delivery (X) Λ (company personnel expenditure (Y) ∨ outsourcing personnel expenditure (Y))) -> labor cost (Y)

[0193] In the labor cost rule, the rule header is the service cost confirmation event. The rule body includes: generating expenditure event...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a target domain knowledge base generation and question answering method and device, and the method comprises the steps: determining a concept graph, a reason graph and a predicate logic formula of target domain knowledge according to the knowledge type of the target domain knowledge; and generating a target domain knowledge base, wherein the concept map is used for representing a static relationship among the concept words, the factorial map is used for representing the sequence between the events and the factorial relationship between the events, and the predicate logic formula is used for representing business rules in the target domain knowledge. The question answering method comprises the steps that M events triggered by N word segmentation phrases of a questionare determined from an affair graph, and slot position values of slot positions of the M events are determined according to K word segmentation phrases, matched with a concept graph, in the N word segmentation phrases; a predicate logic formula corresponding to the consulting object of the question is calculated according to the M events and the slot values of the slots of the M events; an answerto the question is determined.

Description

technical field [0001] This application relates to the field of computer technology, mainly to natural language processing technology in artificial intelligence, and in particular to a method and device for generating a knowledge base in a target field and answering questions. Background technique [0002] The core step of the question answering system is to search relevant knowledge from the existing reserve knowledge base, and then generate answers. In the existing technology, knowledge catalogs can be classified according to the scope of the problem, and a knowledge map can be established. Or, directly facing the question-and-answer examples of the application, a Frequently Asked Questions (FAQ) library can be established, and the knowledge graph can also be combined with the FAQ library, and the nodes of the knowledge graph can be linked to content-related FQAs. However, no matter whether it is retrieval in the form of FAQ or the fusion of FAQ and ontology database, the...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/36G06F16/332G06F40/289G06F40/284
CPCG06F16/367G06F16/3329
Inventor 胡康兴段戎张明仕郭定平
Owner HUAWEI TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products