Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Method for realizing interpretable neural network of text matching

A neural network and text technology, applied in the field of interpretable neural networks, which can solve problems such as performance dependence

Active Publication Date: 2019-11-19
TIANJIN UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although these methods have achieved some success, the performance improvement relies heavily on parameter tuning

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
  • Method for realizing interpretable neural network of text matching
  • Method for realizing interpretable neural network of text matching
  • Method for realizing interpretable neural network of text matching

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] The technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings, but the protection scope of the present invention is not limited to the following description. Specific steps are as follows:

[0027] Such as figure 1 Shown, the present invention provides a kind of method that realizes the interpretable neural network of text matching, comprises the following steps,

[0028] S1. Establish a tensor network model through the relevant basic information of the convolutional network; in the deep network, the convolutional network is often used in information retrieval and text matching networks. In order to analyze the relationship between the number of channels and data features in a convolutional network, the convolutional network is first represented by a tensor network.

[0029] S2. Use the minimum cut method to treat the tensor network as a graph and obtain short-range related classes and long-rang...

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 method for realizing an interpretable neural network of text matching. The method comprises the following steps: mapping a deep convolutional network into a tensor network; adopting the tensor network as a graph, and analyzing the relation between language characteristics (long-distance correlation and short-distance correlation) and the number of channels of the deep convolutional network through the minimum cut theory; for a data set (such as a QA task) of a matching task, quantifying the language characteristics of sentence pairs by utilizing quantum entanglement entropy, wherein the larger the entanglement entropy is, the farther the sentence pairs belong to long-distance correlation, and the smaller the entanglement entropy is, the closer the sentence pairs belong to short-distance correlation. Therefore, the data set is divided into a long-distance related sub-data set and a short-distance related sub-data set. The network architecture, namely the channel number of each layer of the convolutional network, is dynamically adjusted according to different sub-data sets.

Description

technical field [0001] The invention relates to the technical field of natural language matching tasks, in particular to an interpretable neural network method for language matching through a tensor network. Background technique [0002] Neural network architectures are widely used in language modeling and matching, such as question answering, information retrieval, and semantic analysis. While these methods have achieved some success, performance improvements largely rely on parameter tuning. To mitigate this, researchers tend to look at neural networks from a different angle to gain new intuitions and insights. In particular, the basic connection between neural networks and quantum mechanics was established. For example, neural networks are used to solve the quantum many-body problem, and quantum mechanics is used to explain the expressive power of neural networks. This connection could help us study neural networks from the esoteric mathematics of quantum theory. Rece...

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
IPC IPC(8): G06F16/33G06F16/332G06F16/35G06F17/27G06N3/04G06N3/08
CPCG06F16/3329G06F16/3344G06F16/35G06N3/084G06N3/045
Inventor 毛晓柳张鹏
Owner TIANJIN UNIV
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
Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com