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

Semantic recognition method and device, electronic equipment and storage medium

A technology of semantic recognition and recognition, applied in the field of semantic recognition, can solve the problem of high cost of error correction data, achieve the effect of improving accuracy and reducing cost

Pending Publication Date: 2021-08-31
BEIJING ORION STAR TECH CO LTD
View PDF0 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] However, the error correction models in the above error correction methods all require a large amount of high-quality training data to ensure the effect of the model, resulting in the high cost of error correction data accumulation. To update the error correction lexicon, it is necessary to retrain the model or add additional module

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 recognition method and device, electronic equipment and storage medium
  • Semantic recognition method and device, electronic equipment and storage medium
  • Semantic recognition method and device, electronic equipment and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0063] Embodiment 1: slot error correction processing

[0064] figure 2 It is a schematic flowchart of a semantic recognition method provided in the embodiment of the present application, the method may include:

[0065] Step S210, acquiring text to be recognized.

[0066] In the embodiment of the present application, the text to be recognized is user input information, such as query text, and the user input information may be one or more combinations of voice information, text information, and picture information. When the user input information is voice information, an existing voice recognition method can be used to obtain the text to be recognized corresponding to the voice information. When the information input by the user is picture information, the text to be recognized contained in the picture information can be obtained by using the existing image recognition technology.

[0067] Step S220, using the semantic recognition module to perform semantic recognition pro...

Embodiment 2

[0123] Embodiment 2: slot expansion processing

[0124] In the embodiment of the present application, as another possible implementation, the slot value of the slot identified in the text to be recognized is a configured entity word in the thesaurus data, that is, the slot value is the value of a known entity word. In some cases, because the entity words involved in some fields may have homonym entity words, in order to recall the user’s real slot value, avoid false recall, and improve the accuracy of semantic recognition, the grammar rule recognition module and / or NLU module have When outputting, the slot identified by the grammar rule identification module and / or the NLU module may be expanded to obtain an accurate slot value of the corresponding slot.

[0125] In a possible implementation, if the grammar rule recognition module or the NLU module recognizes that the domain corresponding to the text to be recognized is the target domain, and there is at least one slot value c...

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 semantic recognition method and device, electronic equipment and a storage medium. The method comprises steps of after a to-be-recognized text is obtained, adopting a semantic recognition module to perform semantic recognition processing on the to-be-recognized text; if the semantic recognition module recognizes that the field corresponding to the to-be-recognized text is a target field, and at least one first slot position of which the slot position value is not matched with any configured entity word exists in slot positions corresponding to the to-be-recognized text, performing error correction processing on the slot position value of the first slot position; and determining a target recognition result of the to-be-recognized text according to the slot position value after error correction of the first slot position so as to improve the semantic recognition accuracy. Meanwhile, according to the method, the accuracy of semantic recognition is improved by correcting the slot position value of the slot position corresponding to the to-be-recognized text, and compared with a trained error correction model in the prior art, the semantic recognition cost is reduced.

Description

technical field [0001] The present application relates to the technical field of semantic recognition, in particular to a semantic recognition method, device, electronic equipment and storage medium. Background technique [0002] With the popularity of deep learning, major breakthroughs have been made in computer vision, speech recognition, and natural language processing. Compared with other human-computer interaction methods, voice interaction is more in line with people's daily habits and more efficient, so voice recognition technology will be widely used in various fields such as smart home, industrial production, communication, medical treatment, and automatic driving. [0003] In an actual voice interaction scenario, that is, a task-based voice interaction scenario, if the intention of the user query is to reach the target store, but the machine cannot truly interpret the user's intention because the user's pronunciation is not standard or the store name is misremember...

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 Applications(China)
IPC IPC(8): G06F40/30G06F40/284
CPCG06F40/30G06F40/284
Inventor 王艳蒋卓
Owner BEIJING ORION STAR TECH CO LTD