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

Sentence pair semantic matching method and device for intelligent interaction

A technology of semantic matching and intelligent interaction, applied in semantic analysis, character and pattern recognition, natural language data processing, etc. It can solve the problems of ignoring interaction and not considering re-encoding semantic features, so as to achieve accurate matching representation tensors and semantic features. Rich, Accurate Effects

Active Publication Date: 2022-05-24
南方电网互联网服务有限公司
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The interaction-based method generally captures the interaction features between two sentences through the attention mechanism, and then aggregates the matching results through a certain structure to obtain the final semantic representation; the advantage of this method is that it can better capture the interaction between sentences. Interaction dependencies among them, grasping the semantic focus, and reasonably modeling the importance of context; however, most of the current work focuses on word-level interactions; Although the work considers two granularities, they still ignore the interaction between the two granularities, and do not consider re-encoding after interaction to extract deeper semantic features

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
  • Sentence pair semantic matching method and device for intelligent interaction
  • Sentence pair semantic matching method and device for intelligent interaction
  • Sentence pair semantic matching method and device for intelligent interaction

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0117] as attached Figure 9 As shown, the main frame structure of the present invention includes a multi-granularity embedding module, a multi-level alignment matching network module, a feature enhancement module and a label prediction module. Among them, the multi-granularity embedding module performs embedding operations on the input sentences at word granularity and word granularity respectively, and transmits the results to the multi-level alignment matching network module of the model. Multi-level alignment matching network modules such as Figure 8 As shown, there are four alignment matching modules with unified structure, namely the basic module. The structure of the basic module is as Figure 7 As shown, the basic module takes tensor 1 and tensor 2 as input and calculates the attention scores of the two tensors, and then multiplies the input tensor 1 and the attention score to perform the alignment operation to obtain the alignment result 2, and then uses The align...

Embodiment 2

[0123] as attached figure 1 As shown, the intelligent interaction-oriented sentence pair semantic matching method of the present invention, the specific steps are as follows:

[0124] S1. Build a sentence pair semantic matching knowledge base, as attached figure 2 shown, the specific steps are as follows:

[0125] S101. Download a dataset on the Internet to obtain original data: download a dataset of semantic matching of sentence pairs or an artificially constructed dataset that has been published on the Internet, and use it as the original data for constructing a knowledge base of semantic matching of sentence pairs.

[0126] Example: There are many public datasets of sentence pair semantic matching on the web. The present invention collects and downloads these data, thereby obtaining the original data for constructing the semantic matching knowledge base of sentence pairs. For example, an example from the LCQMC dataset is as follows:

[0127] Sentence 1 Whic...

Embodiment 3

[0237] as attached Image 6 As shown, based on the intelligent interaction-oriented sentence pair semantic matching device of Embodiment 2, the device includes,

[0238] The sentence pair semantic matching knowledge base construction unit is used to obtain a large amount of sentence pair data, and then preprocess it to obtain the sentence pair semantic matching knowledge base that meets the training requirements; the sentence pair semantic matching knowledge base construction unit includes,

[0239] The sentence pair data acquisition unit is responsible for downloading the sentence pair semantic matching dataset or artificially constructed dataset that has been published on the Internet, and using it as the original data for constructing the sentence pair semantic matching knowledge base;

[0240] The raw data hyphenation / word segmentation preprocessing unit is responsible for preprocessing the raw data used to construct the sentence-pair semantic matching knowledge base, and ...

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 sentence-to-semantic matching method and device for intelligent interaction, belonging to the fields of artificial intelligence and natural language processing. The technical problem to be solved by the present invention is how to capture semantic features and sentence interaction information to realize intelligent semantic matching of human-computer interaction sentence pairs. The technical solution adopted is to construct and train a sentence-pair semantic matching model composed of a multi-granularity embedding module, a multi-level alignment matching network module, a feature enhancement module, and a label prediction module to realize the multi-level alignment and matching representation of sentence information. Dimensional maximum pooling and interactive generation of matching tensors of sentence pairs to determine the matching degree of sentence pairs to achieve the goal of intelligent matching of sentence pairs. The device includes a sentence-pair semantic matching knowledge base construction unit, a training data set generation unit, a sentence-pair semantic matching model construction unit and a sentence-pair semantic matching model training unit.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence and natural language processing, and in particular to a method and device for semantic matching of sentences pair oriented to intelligent interaction. Background technique [0002] With the development of artificial intelligence technology, more and more intelligent systems are widely used in people's daily life, such as Ali intelligent customer service robot, Apple Siri intelligent voice assistant, etc. How to effectively interact with these intelligent systems is a key factor in determining the user experience. Currently, most intelligent systems can be interactively controlled through graphical user interfaces, keyboards, mice, languages, and gestures. Among them, language interaction is the most important way of human-computer interaction; for humans, this way is the most convenient. Through language interaction, the user's instructions are directly transmitted to 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 Patents(China)
IPC IPC(8): G06F40/30G06F40/194G06F40/284G06K9/62
CPCG06F40/30G06F40/194G06F40/284G06F18/22G06F18/214G06F18/2411G06F18/253
Inventor 鹿文鹏左有慧张旭阚保硕赵鹏宇
Owner 南方电网互联网服务有限公司
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

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