Semantic model optimization method, device, intelligent device, and storage medium

A semantic model and optimization method technology, applied in semantic analysis, semantic tool creation, natural language data processing, etc., can solve the problems of wide coverage of semantic matching, low precision of semantic matching, high difficulty of semantic matching, etc., and achieve enhanced semantic matching ability, reduce the amount of acquisition, and improve the effect of accuracy

Active Publication Date: 2021-07-27
TENCENT TECH (SHENZHEN) CO LTD
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The general semantic model can handle semantic matching similarity calculations in various fields, and there are factors such as wide coverage of semantic matching, high difficulty of semantic matching, and low accuracy of semantic matching

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
  • Semantic model optimization method, device, intelligent device, and storage medium
  • Semantic model optimization method, device, intelligent device, and storage medium
  • Semantic model optimization method, device, intelligent device, and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0033] In the embodiment of the present invention, data such as unlabeled data, query (query) click data, machine translation data, and dialogue flow data can be used as the initial value and characteristics of the semantic model trained by manual labeling data, and the general semantic model and the domain semantic model can be implemented. Training optimization. In the embodiment of the present invention, the domain semantic model can be obtained by training based on manually labeled data, and then the domain semantic model can be trained and optimized based on a large amount of unlabeled data and natural labeled data to obtain the final domain semantic model. Wherein, in the embodiment of the present invention, the domain semantic model is the first model, and the general semantic model is the second model.

[0034] In this embodiment of the present invention, the training of the second model may be multi-task, the second model includes at least two sub-models, and there is...

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 embodiment of the present invention discloses a semantic model optimization method, device, and smart device, wherein the method includes: acquiring the first sentence and the second sentence used to train the first model; Coding the sentence and the second sentence to obtain a first coding result; using the coding rules configured for the first model to code the first sentence and the second sentence to obtain a second coding result; The first coding result and the second coding result are subjected to similarity training processing, and the first model is optimized according to the training processing results. By adopting the embodiment of the present invention, the semantic matching ability of the domain semantic model can be significantly improved, and the acquisition amount of training sentences can be reduced to a certain extent, so that the optimization speed of the domain semantic model is faster.

Description

technical field [0001] The present invention relates to the field of computer application technology, in particular to a semantic model optimization method, device, intelligent device, and storage medium. Background technique [0002] In scenarios such as retrieval-based dialogue robots or search engines, semantic matching is the core technology, and the performance of the constructed semantic model directly affects the performance of dialogue robots and search engines. [0003] When constructing semantic models, different models can be constructed for different domains. For example, domain semantic models in the fields of e-commerce and communication services can be constructed. Among them, based on the domain semantic models in the e-commerce field, dialogue robots in e-commerce can be realized, and some interactions with customers, such as Analyze customer statements for price replies, product introductions, and other functions. [0004] A general semantic model can als...

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): G06F16/36G06F40/194G06F40/30
CPCG06F16/36G06F40/194G06F40/30
Inventor 鲁亚楠林芬
Owner TENCENT TECH (SHENZHEN) 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