Search

CN-121979973-A - Domain-falling processing method, device and equipment of query statement and readable storage medium

CN121979973ACN 121979973 ACN121979973 ACN 121979973ACN-121979973-A

Abstract

The disclosure discloses a domain-falling processing method, device and equipment for query sentences and a readable storage medium, and relates to the technical fields of voice processing, intelligent voice interaction and the like. The method comprises the steps of obtaining a query statement input by a user, and carrying out at least one-level domain falling processing on the query statement based on at least one of a preset instruction domain falling library, a preset rule domain falling library, a preset intervention domain falling rule library and a pre-trained domain falling classification model to obtain a target domain identifier matched with the query statement.

Inventors

  • LI LIN

Assignees

  • 北京百度网讯科技有限公司

Dates

Publication Date
20260505
Application Date
20251216

Claims (13)

  1. 1. A domain-falling processing method of query sentences is applied to a vehicle-mounted voice interaction system and comprises the following steps of; Acquiring a query sentence input by a user; And carrying out at least one level domain falling processing on the query statement based on at least one of a preset instruction domain falling library, a preset rule domain falling library, a preset intervention domain falling rule library and a pre-trained domain falling classification model to obtain a target domain identifier matched with the query statement.
  2. 2. The method of claim 1, wherein performing at least one level of domain processing on the query statement based on at least one of a preset instruction domain library, a preset rule domain library, a preset intervention domain rule library, and a pre-trained domain classification model to obtain a target domain identifier for matching the query statement, comprises: And carrying out domain falling processing on the query statement based on the instruction domain falling library to determine the target domain identifier matched with the query statement, wherein the instruction domain falling library comprises a plurality of instructions acquired in advance and domain identifiers configured for the instructions.
  3. 3. The method of claim 2, wherein the query statement is subjected to at least one level of domain-falling processing based on at least one of a preset instruction domain-falling library, a preset rule domain-falling library, a preset intervention domain-falling rule library, and a pre-trained domain-falling classification model to obtain a target domain identifier for matching the query statement, the method further comprising: If domain falling processing cannot be performed on the query statement based on the instruction domain falling library, domain falling processing is performed on the query statement based on the preset rule domain falling library, and the target domain identifier matched with the query statement is determined, wherein the preset rule domain library comprises a plurality of preset rules with definite intention and domain identifiers configured for the rules.
  4. 4. The method of claim 3, wherein the query statement is subjected to at least one level of domain-falling processing based on at least one of a preset instruction domain-falling library, a preset rule domain-falling library, a preset intervention domain-falling rule library, and a pre-trained domain-falling classification model to obtain a target domain identifier for matching the query statement, the method further comprising: If domain falling processing cannot be performed on the query statement based on the preset rule domain falling library, domain falling processing is performed on the query statement based on the intervention domain falling rule library, and the target domain identifier matched with the query statement is determined, wherein the intervention domain falling rule library comprises a preset entity name library and a preset domain identifier matched with the entity name library, and is used for matching the domain identifier with the query statement with ambiguous intention.
  5. 5. The method of claim 4, wherein if domain dropping processing cannot be performed on the query statement based on the preset rule domain dropping library, determining the target domain identifier matched by the query statement based on the intervention domain dropping rule library comprises: Determining that the intent of the query statement is ambiguous; detecting whether the query statement comprises entity names in a preset entity name library or not by adopting the intervention domain rule library; and when the entity name library is included, acquiring the domain identifier matched with the entity name library from the intervention domain rule library as the target domain identifier.
  6. 6. The method of claim 5, wherein the domain identifier matched by the entity name library is obtained from the intervention domain rule base, and after being the target domain identifier, the method further comprises: Guiding the user to talk based on the query statement by adopting a pre-trained dialogue guiding model, and acquiring the supplementary information of the user; and generating a rewrite query sentence with clear intention based on the query sentence and the supplementary information by adopting a pre-trained rewrite model.
  7. 7. The method of claim 4, wherein the query statement is subjected to at least one level of domain-falling processing based on at least one of a preset instruction domain-falling library, a preset rule domain-falling library, a preset intervention domain-falling rule library, and a pre-trained domain-falling classification model to obtain a target domain identifier for matching the query statement, the method further comprising: if the domain-falling processing cannot be performed on the query statement based on the intervention domain-falling rule base, domain-falling processing is performed on the query statement based on the domain-falling classification model, and the target domain identifier matched with the query statement is obtained.
  8. 8. The method of claim 7, wherein prior to domain-dropping the query statement based on the domain-dropping classification model, the method further comprises: Acquiring historical context information of the query statement, wherein the historical context information of the query statement comprises dialogue contents between the user and a system in a preset time period before the query statement, and the dialogue turns do not exceed preset turns; Performing domain-falling processing on the query statement based on the domain-falling classification model to obtain the target domain identifier matched with the query statement, including: Based on the query statement and the historical context information, adopting the domain classification model to identify confidence scores of the query statement in each of a plurality of preset domains; And determining at least one target domain identifier matched with the query statement based on the confidence scores of the query statement in each of a plurality of domains and a preset threshold.
  9. 9. The method according to any one of claims 1-8, wherein, based on at least one of a preset instruction domain library, a preset rule domain library, a preset intervention domain rule library, and a pre-trained domain classification model, performing at least one level of domain processing on the query statement, and after obtaining the target domain identifier matched by the query statement, the method further comprises: and based on the target domain identifier, requesting a business service module of the downstream matching domain to respond to the query statement.
  10. 10. A domain-falling processing device of a query sentence is applied to a vehicle-mounted voice interaction system and comprises; The acquisition module is used for acquiring a query statement input by a user; The domain falling processing module is used for carrying out domain falling processing of at least one level on the query statement based on at least one of a preset instruction domain falling library, a preset rule domain falling library, a preset intervention domain falling rule library and a pre-trained domain falling classification model to obtain a target domain identifier matched with the query statement.
  11. 11. An electronic device, comprising: at least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1-9.
  12. 12. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-9.
  13. 13. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any of claims 1-9.

Description

Domain-falling processing method, device and equipment of query statement and readable storage medium Technical Field The disclosure relates to the technical field of computers, in particular to the technical fields of voice processing, intelligent voice interaction and the like, and particularly relates to a domain-falling processing method, device and equipment of query sentences and a readable storage medium. Background In the existing intelligent question-answering processing system, domain classification is carried out on query sentences, and a pre-trained classification model is mainly used for determining classification of the query sentences. For example, the classification model may be an end-to-end model obtained by fine-tuning a pre-trained large language model. The classification model is adopted to classify the query sentences, so that deep semantics can be effectively captured, the precision is high, but the model is huge, the calculation cost is high, the reasoning speed is low, and the model is a black box and is difficult to debug and intervene. Disclosure of Invention The disclosure provides a domain-falling processing method, device and equipment for query sentences and a readable storage medium. According to one aspect of the present disclosure, a domain-falling processing method of a query sentence is provided, which is applied to a vehicle-mounted voice interaction system, and includes; Acquiring a query sentence input by a user; And carrying out at least one level domain falling processing on the query statement based on at least one of a preset instruction domain falling library, a preset rule domain falling library, a preset intervention domain falling rule library and a pre-trained domain falling classification model to obtain a target domain identifier matched with the query statement. According to another aspect of the present disclosure, there is provided a domain-falling processing apparatus of a query sentence, applied in a vehicle-mounted voice interaction system, including; The acquisition module is used for acquiring a query statement input by a user; The domain falling processing module is used for carrying out domain falling processing of at least one level on the query statement based on at least one of a preset instruction domain falling library, a preset rule domain falling library, a preset intervention domain falling rule library and a pre-trained domain falling classification model to obtain a target domain identifier matched with the query statement. According to still another aspect of the present disclosure, there is provided an electronic apparatus including: at least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the aspects and methods of any one of the possible implementations described above. According to yet another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the method of the aspects and any possible implementation described above. According to yet another aspect of the present disclosure, there is provided a computer program product comprising a computer program which, when executed by a processor, implements the method of the aspects and any one of the possible implementations described above. According to the technology disclosed by the invention, the domain falling processing efficiency of the query statement can be effectively improved. It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the disclosure, nor is it intended to be used to limit the scope of the disclosure. Other features of the present disclosure will become apparent from the following specification. Drawings The drawings are for a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein: FIG. 1 is a schematic diagram according to a first embodiment of the present disclosure; FIG. 2 is a schematic diagram according to a second embodiment of the present disclosure; FIG. 3 is a schematic diagram according to a third embodiment of the present disclosure; FIG. 4 is a schematic diagram according to a fourth embodiment of the present disclosure; Fig. 5 is a block diagram of an electronic device used to implement the methods of embodiments of the present disclosure. Detailed Description Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and should be considered as merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments descri