Unlock instant, AI-driven research and patent intelligence for your innovation.

Machine language generation method and device

A technology of machine language and language generation, applied in natural language translation, instruments, computing, etc., can solve the problems of huge time and manpower consumption, poor user experience, limited coverage, etc., and achieve the effect of saving manpower and material resources and good user experience

Active Publication Date: 2017-10-24
ZHEJIANG TMALL TECH CO LTD
View PDF8 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, based on the template filling scheme, it is necessary to manually construct the template first.
Manually constructing templates is time-consuming and labor-intensive, and the coverage is very limited
And based on the template filling scheme, the constructed template is a sentence lacking core words, and the core words are filled in when generating, resulting in obvious traces of the template in the returned sentence, and the user experience is poor

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
  • Machine language generation method and device
  • Machine language generation method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] The technical solution of the present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments, and the following embodiments do not constitute a limitation of the present invention.

[0037] Such as figure 1 As shown, the machine language generation method of this embodiment includes:

[0038] Step S1, receiving an input sentence, and obtaining a frame list corresponding to the input sentence according to a preset frame model.

[0039] The method of this embodiment is preset with a framework model and a language generation model.

[0040] The framework model and language generation model include but are not limited to statistics-based ngram language model, deep learning-based neural network language model, or encoder-decoder-based neural network model, etc., through a large number of training samples, such as a large number of question-answer corpus, Or other corpus used for training to learn the required frame...

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 machine language generation method and device. The machine language generation method comprises the steps of receiving an input statement; obtaining a framework list corresponding to the input statement according to a preset framework model; inputting the input statement into a preset language generation model; and generating a reply statement consistent with any one framework in the framework list and outputting the reply statement. The machine language generation device provided by the invention comprises a framework generation module and a statement generation module. According to the method and the device provided by the invention, when the reply statement is generated, a plurality of candidate following parts of a text are gradually generated according to the preceding part of the text, and the candidate following part of the text with the highest probability is selected, so the coherence of the generated reply statement is ensured; and moreover, through combination of the framework generated by the framework model, the reply statement is guaranteed to comprise a certain fixed structure, on the premise that the reply statement is guaranteed to comprise the framework, the reply statement is very coherent and is relatively consistent with emantics and a human natural language, and the relatively good experience is provided.

Description

technical field [0001] The invention belongs to the technical field of artificial intelligence, and in particular relates to a machine language generation method and device. Background technique [0002] With the development of science and technology, artificial intelligence has gradually begun to enter the fields of industry, commerce and life. Artificial voice assistants are widely used at present, such as Baidu's Dumi, Microsoft's Xiaobing and Cortana, etc., which can provide voice interaction and help users arrange their schedules and answer questions according to their preferences and habits. The core of these products is based on the idea of ​​a language model. According to the above text input by the user, the following text is predicted and output. For example, the user asks a question through voice, and the artificial voice assistant predicts the corresponding answer to answer. [0003] For the given above information as a condition to predict the probability of th...

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): G06F17/28
CPCG06F40/42
Inventor 黎槟华江会星孙健初敏
Owner ZHEJIANG TMALL TECH CO LTD