Entity question answering method and device based on neural network and terminal

A neural network and entity technology, applied in the computer field, can solve problems such as optimization, limited overall effect, heavy system calculation, etc., and achieve the effects of reducing explicit calculation and cumulative errors, improving timeliness, and improving positioning accuracy

Inactive Publication Date: 2018-12-07
BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF5 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The disadvantages of the traditional entity question answering system mainly include: (1) the above-mentioned functional modules involve a large number of key technologies such as lexical analysis, syntactic analysis, semantic analysis, and knowledge engineering, which makes the system calculation very heavy; (2) the overall effect of the system is limited by each The individual effects of functional modules have cumulative errors and are not conducive to sustainable effect optimization

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
  • Entity question answering method and device based on neural network and terminal
  • Entity question answering method and device based on neural network and terminal
  • Entity question answering method and device based on neural network and terminal

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0050] In a specific embodiment, a neural network-based entity question answering method is provided, such as figure 1 with figure 2 shown, including:

[0051]Step S100: convert the words contained in the question and candidate documents into word vectors, and generate corresponding question word vector sequences and candidate document word vector sequences.

[0052] In the presentation layer, the words in question q and document p are respectively vectorized. Specifically, initially, each word in the question and document is initialized as a random floating-point vector of fixed dimension, and the question word vectors are arranged to form a question word vector sequence q emb , the candidate document word vectors are arranged to form a candidate document word vector sequence p emb , these can be used as initial representations of words for questions and documents. The representations of questions and documents are then continuously refined during system training. In on...

Embodiment 2

[0073] In another specific embodiment, a neural network-based entity question answering device is provided, such as image 3 shown, including:

[0074] The vector conversion module 10 is used to convert the words contained in the question and the candidate document into word vectors respectively, and generate corresponding question word vector sequences and candidate document word vector sequences;

[0075] Sequence encoding module 20, for inputting the question word vector sequence and the candidate document word vector sequence respectively into the long short-term memory network model, outputting the word encoding sequence of the question and the word encoding sequence of the candidate document;

[0076] Question and document matching module 30, for matching the word coding sequence of the question and the word coding sequence of the candidate document, generate a candidate document representation based on matching information, and the candidate document representation incl...

Embodiment 3

[0090] The embodiment of the present invention provides a neural network-based entity question answering terminal, such as Image 6 shown, including:

[0091] A memory 400 and a processor 500 , the memory 400 stores computer programs that can run on the processor 500 . When the processor 500 executes the computer program, the neural network-based entity question answering method in the foregoing embodiments is implemented. The number of memory 400 and processor 500 may be one or more.

[0092]The communication interface 600 is used for the memory 400 and the processor 500 to communicate with the outside.

[0093] The memory 400 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one magnetic disk memory.

[0094] If the memory 400, the processor 500, and the communication interface 600 are implemented independently, the memory 400, the processor 500, and the communication interface 600 may be connected to e...

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 provides an entity question answering method and device based on a neural network and a terminal, wherein the method comprises converting a word contained in a question and a candidate document into a word vector, and generating a corresponding question word vector sequence and a candidate document word vector sequence; inputting the problem word vector sequence and the candidate document word vector sequence into the long-short-term memory network model, and outputting the word coding sequence of the problem and the candidate document; matching the word encoding sequence of theproblem and the word encoding sequence of the candidate document to generate a candidate document representation based on matching information, wherein the candidate document representation comprisesa plurality of word representations; selecting start and end words from all word representations and generating entity answers based on the start and end words. The method reduces explicit computationand cumulative error, effectively utilizes semantic representation between questions and documents, and improves positioning accuracy of entity answers.

Description

technical field [0001] The present invention relates to the field of computers, in particular to a neural network-based entity question answering method, a neural network-based entity question answering device, and a neural network-based entity question answering terminal. Background technique [0002] Based on the relevant documents of the given question and answer, the traditional entity question answering system needs to calculate multiple functional modules such as question type analysis, entity recognition, entity type matching, and context matching. The explicit calculation of these functional modules often makes The entity question answering system becomes heavy, and the final system effect is limited by the error accumulation of all modules. The disadvantages of the traditional entity question answering system mainly include: (1) the above-mentioned functional modules involve a large number of key technologies such as lexical analysis, syntactic analysis, semantic an...

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): G06F17/30G06F17/27G06N3/04G06N3/08
CPCG06N3/08G06F40/295G06F40/30G06N3/044
Inventor 韦豪杰
Owner BEIJING BAIDU NETCOM SCI & TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap