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

Cross-modal sequence-to-sequence generation method based on topic awareness

A sequence generation, cross-modal technology, applied in the field of data processing, can solve problems such as topic consistency and numerical coding problems

Active Publication Date: 2021-03-16
STATE GRID TIANJIN ELECTRIC POWER +1
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem of topic consistency and numerical encoding existing in existing data-to-text generation tasks, the present invention provides a cross-modal sequence-to-sequence generation method based on topic awareness

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
  • Cross-modal sequence-to-sequence generation method based on topic awareness
  • Cross-modal sequence-to-sequence generation method based on topic awareness
  • Cross-modal sequence-to-sequence generation method based on topic awareness

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0066] The subject-aware-based cross-modal sequence-to-sequence generation method of the present invention will be described in detail below with reference to the accompanying drawings.

[0067] The present invention mainly adopts deep learning technology and natural language processing related theoretical methods to realize the generation of data to text, and ensure the subject consistency between data and text. In order to ensure the normal operation of the system, in the specific implementation, the computer platform used is required to be equipped with no less than 8G memory, no less than 4 CPU cores and no less than 2.6GHz main frequency, GPU environment, Linux operating system, and Install Python 3.6 and above, pytorch0.4 and above and other necessary software environments.

[0068] Such as figure 1 As shown, the topic-aware based cross-modal sequence-to-sequence generation method provided by the present invention specifically includes the following steps executed in or...

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 cross-modal sequence-to-sequence generation method based on topic awareness. The method comprises the following steps: 1, data record encoding: learning context semantic representation of each record in a data table by utilizing a bidirectional long-short-term memory network; 2, learning word distribution corresponding to each topic according to the title of the data table and the text corresponding to the data table, and performing weighted summation on vector representation of words to obtain vector representation of the topics; 3, based on the hidden vector representation sequence of the data record obtained by the coding layer in the step 1 and the topic representation obtained in the step 2, generating an analytical text by using an attention mechanism-basedLSTM structure as a decoder; 4, model training: constructing a loss function to optimize the model parameters in the step 13; and 5, text generation: for a given data table in an inference process, approximately obtaining an optimal text generation result by utilizing cluster search. According to the method, the theme consistency of the data table and the generated text can be enhanced, and the quality of the generated text is improved.

Description

technical field [0001] The invention relates to the field of data processing, in particular to a topic-aware cross-modal sequence-to-sequence generation method. Background technique [0002] With the advent of the big data era, all walks of life have gradually accumulated massive amounts of industry data. These data are closely related to the production management of human society, and are the main objects of analysis and research in various fields. Among these industry data, structured data has become the most common form of data because of its simple format and easy recording and storage, such as company financial statements, equipment sensor records, etc. However, structured data is usually highly domain-specific, and it is difficult for people who lack industry knowledge to understand the meaning behind its values ​​and indicators. Therefore, how to accurately and efficiently convey the semantic information contained in structured data is an important cross-modal gener...

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): G06F16/33G06F40/30G06N3/04G06N3/08G06N5/04
CPCG06F16/3344G06F40/30G06N3/084G06N5/04G06N3/045G06N3/044
Inventor 王旭强张旭郑阳杨青
Owner STATE GRID TIANJIN ELECTRIC POWER
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