Auxiliary control method and system based on post-validation

An auxiliary control and verification result technology, applied in the field of semantic recognition, can solve problems such as semantic loss, huge keyword list, heavy workload and system resources, and achieve the effect of eliminating ambiguity

Active Publication Date: 2019-03-22
珠海希音医疗科技有限公司
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] 1. It is difficult to balance between recognition synchronization and resource occupation: as things are updated and developed, the keyword list will become very large
Maintaining corresponding keyword lists of different grammars will lead to huge workload and system resources, thereby reducing recognition performance; if the maintenance of keyword lists is not synchronized, new keywords will not be recognized; if fuzzy retrieval is added, it will further Leading to resource occupation and user experience degradation
[0005] 2. The standardization of grammatical expression: users must strictly follow the system's preset method for grammatical expression, and it is difficult to clearly identify the correct semantics of colloquial input, inversion, postposition or emphasis sentence patterns
Makes the level of knowledge expression to the user higher in the actual recognition process, which limits the popularization of speech technology
[0006] 3. There is no follow-up intelligent processing when there is ambiguity or no matching result: When the same grammar is used in multiple different fields, it is easy to conflict due to ambiguity in the understanding of keywords
[0007] To sum up, although the existing recognition technology can identify the general intention of the user through grammar to a certain extent, it is difficult to accurately identify the precise parameters based on the intention. The result of the recognition is to directly judge the content containing the meaning of the parameter as ordinary characters, so that the semantics of user input are lost

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
  • Auxiliary control method and system based on post-validation
  • Auxiliary control method and system based on post-validation
  • Auxiliary control method and system based on post-validation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0064] Such as figure 2 As shown, the present invention has the implementation of the auxiliary control method of post-verification, which specifically includes the following steps:

[0065] 1) In the process of performing grammar template matching on the original input in this embodiment, domain information of at least one matching grammar template is recorded; for example, when the original input is: I want to listen to you tell a story. but:

[0066] A. Using "I want to listen|" as a grammar template for matching, you can register to get the result with the audio player as domain information; or

[0067] B. Using "listen to you||story" as the grammatical template for matching, you can directly register to get the result of using the audio player to play the story collection as the domain information.

[0068] 2) Then match the operation parameter part obtained after registration with the entity knowledge base corresponding to the above domain information, and ...

Embodiment 2

[0076] In some more complex situations, the semantics of the original input includes multiple types and a large number. This example is based on image 3 A method for generating auxiliary control instructions is given, and the specific semantics of each part in the original input is further judged on the basis of Example 1.

[0077] For example, the original input is: I will have dinner with Xiao Wang in Xujiahui tomorrow morning from 10 am to 12 am, remind me 15 minutes in advance.

[0078] 1) The field information obtained by matching the syntax template "|reminder|" is an alarm clock or a reminder program, and the parameter of the corresponding field information obtained according to the regular expression is "15 minutes in advance", and the sub-parameter body is "tomorrow Have dinner with Xiao Wang in Xujiahui from ten to twelve o'clock in the morning"

[0079] 2) For the sub-parameter body "eating with Xiao Wang in Xujiahui tomorrow morning from 10 am to 12 am"...

Embodiment 3

[0083] In some specific cases, the original input will further contain adverbs of degree, such as "very", "certainly", "possibly", etc., which do not have specific behavior or entity description, but have priority or Classification adjustments.

[0084] Due to the limited number of degree description semantics, this embodiment sets corresponding degree knowledge bases for the entity knowledge base and the behavior knowledge base. When the entity knowledge base or the behavior knowledge base cannot match the result, the degree knowledge base is used for matching. , and further assign fixed values ​​to each matching result in the degree knowledge base, as the weighted result of matching results in the behavior knowledge base or entity knowledge base due to post-verification calculation.

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 provides an auxiliary control method and system based on post verification. The method comprises the steps that through the pre-processing of original command signals, obtaining auxiliary control information, conducting grammatical template matching of the auxiliary control information based on fields to obtain command operator parts and operation parameter parts, using the post verification method corresponding to the fields to analyze the operation parameter parts, treating the result which is obtained from the analysis and with the highest matching level as the operation parameters of the command operator parts, generating auxiliary control commands and sending the control commands to a central processing unit for operation, thus achieving the auxiliary control. The invention hugely enhances the precision of semantic identification based on current grammatical retrieval. By simulating the manner of human understanding of language, all the semantic meanings which can be extracted from the colloquial expression content inputted by the user are identified, and the identification rate and the user's experience are enhanced.

Description

technical field [0001] The invention relates to a technology in the field of semantic recognition, in particular to an auxiliary control method and system based on post-verification. Background technique [0002] Artificial intelligence can better understand human beings through semantic recognition, and finally adjust the working status of hardware is the main problem in the field of industrial automation control. Existing semantic recognition is mainly based on natural language understanding, and usually uses a regular expression retrieval method of grammar + keyword list. For example: the grammar is: I want to eat <slot1>; while the keywords, namely <slot1>For specific food names, restaurant names, etc. The general recognition method matches all the characters after the keyword "eat" with the keyword list one by one under the condition of conforming to the grammar, and only performs subsequent operations when there is a matching result. [0003] Although...

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): G06F17/27G06F16/332
Inventor 不公告发明人
Owner 珠海希音医疗科技有限公司
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