CN112837678B - Private cloud recognition training method and device - Google Patents
Private cloud recognition training method and device Download PDFInfo
- Publication number
- CN112837678B CN112837678B CN202011617714.9A CN202011617714A CN112837678B CN 112837678 B CN112837678 B CN 112837678B CN 202011617714 A CN202011617714 A CN 202011617714A CN 112837678 B CN112837678 B CN 112837678B
- Authority
- CN
- China
- Prior art keywords
- training
- service
- developer
- private cloud
- recognition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012549 training Methods 0.000 title claims abstract description 115
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000012545 processing Methods 0.000 claims abstract description 15
- 238000005516 engineering process Methods 0.000 claims description 27
- 238000003860 storage Methods 0.000 claims description 14
- 230000004044 response Effects 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 5
- 230000010354 integration Effects 0.000 abstract description 3
- 238000007726 management method Methods 0.000 description 34
- 238000010586 diagram Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000011161 development Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000010295 mobile communication Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 241001672694 Citrus reticulata Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013550 semantic technology Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/18—Speech classification or search using natural language modelling
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Acoustics & Sound (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a private cloud identification training method and a private cloud identification training device, which are used for a knowledge base management platform, wherein the private cloud identification training method comprises the following steps: responding to different business scene requirements selected by a developer at the knowledge base management platform, and collecting corresponding business data to the developer based on the business scene requirements; and training and subsequent processing corresponding to the business scene are carried out based on the collected business data. The scheme meets different requirements of different scenes of a customer by providing the private cloud identification training method and the private cloud identification training device, the customer can be used after opening the box during integration, and the operation is simple.
Description
Technical Field
The invention belongs to the technical field of internet, and particularly relates to a private cloud identification training method and device.
Background
The current society develops rapidly, and the fields such as social chat, on-vehicle system, intelligent hardware, recreation have been a part in people's life, and people can communicate with machine pronunciation, and the machine receives the input signal and can give corresponding answer after handling, and this is speech recognition technology, converts input speech such as mandarin, sichuan chinese, cantonese into characters output.
In business application scenarios such as industry solutions, knowledge question answering, knowledge maps, etc., how to automatically recognize customized semantic statements into texts by using a speech recognition technology, and then request the speech texts for semantics to acquire natural language generation. Because the voice recognition technology is a single technology and can come from different manufacturers, how to automatically combine the technology and the custom semantic expression is difficult to realize, the development difficulty is high, in addition, the customers pay attention to data confidentiality, many customers can select private cloud deployment, the customers can customize the system, and how to apply the custom system expression to the voice recognition technology is a challenge.
Disclosure of Invention
The embodiment of the invention provides a private cloud identification training method and device and a private cloud identification application method, which are used for solving at least one of the technical problems.
In a first aspect, an embodiment of the present invention provides a private cloud identification training method, which is used for a knowledge base management platform, and includes: responding to different business scene requirements selected by a developer at the knowledge base management platform, and collecting corresponding business data to the developer based on the business scene requirements; and training and subsequent processing corresponding to the business scene are carried out based on the collected business data.
In a second aspect, an embodiment of the present invention provides a private cloud identification application method, including: receiving a voice request of a user, and acquiring a product ID transmitted by a recognition service based on the voice request; and acquiring the bound information or resource based on the product ID, and acquiring the voice recognition text corresponding to the voice request based on the bound information or resource.
In a third aspect, an embodiment of the present invention provides a private cloud identification training apparatus, used for a client, including: the service selection program module is configured to respond to different service scene requirements selected by a developer on the knowledge base management platform and collect corresponding service data from the developer based on the service scene requirements; and the training processing program module is configured to perform training and subsequent processing corresponding to the business scene based on the acquired business data.
In a fourth aspect, an embodiment of the present invention provides a private cloud identification application apparatus, configured for a client, including: the parameter transmission program module is configured to receive a voice request of a user and acquire a product ID transmitted by the identification service based on the voice request; and the identification text acquisition module is configured to acquire the bound information or resource based on the product ID and acquire the voice identification text corresponding to the voice request based on the bound information or resource.
In a fifth aspect, an embodiment of the present invention further provides an electronic device, which includes: 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 steps of the method of the first aspect.
In a sixth aspect, an embodiment of the present invention further provides a storage medium, which includes: which when executed by a processor performs the steps of the method of the first aspect
According to the scheme, different requirements of different scenes of a client are met by providing the private cloud identification training method and the private cloud identification application method, the client can be used after opening the box during integration, and the operation is simple.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the description below are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a flowchart of a private cloud identification training method according to an embodiment of the present invention;
fig. 2 is a flowchart of a private cloud identification application method according to an embodiment of the present invention;
fig. 3 is a flowchart of private cloud identification training in a specific embodiment of the private cloud identification training scheme according to an embodiment of the present invention;
fig. 4 is another flowchart of private cloud identification training according to a specific embodiment of the private cloud identification training scheme according to the embodiment of the present invention;
fig. 5 is a flowchart of a private cloud identification application according to a specific embodiment of the private cloud identification application scheme of the embodiment of the present invention;
fig. 6 is another flowchart of private cloud identification training according to a specific embodiment of the private cloud identification training scheme according to the embodiment of the present invention;
fig. 7 is a block diagram of a private cloud identification training apparatus according to an embodiment of the present invention;
fig. 8 is a block diagram of a private cloud identification application apparatus according to an embodiment of the present invention;
fig. 9 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a flowchart of an embodiment of a private cloud identification training method of the present invention is shown.
As shown in fig. 1, in step 101, in response to a developer selecting different business scenario requirements at the knowledge base management platform, collecting corresponding business data from the developer based on the business scenario requirements;
in step 102, training and subsequent processing corresponding to the business scenario are performed based on the collected business data.
In this embodiment, for step 101, the knowledge base management platform responds to a developer selecting different service scenario requirements at the knowledge base management platform, and then collects corresponding service data from the developer based on the service scenario requirements, where the different service scenarios include a group link, a single technology QA and recognition training tool, and a single QA window deployment, and each service scenario correspondingly requires corresponding service data to perform a subsequent service process.
Thereafter, for step 102, the knowledge base management platform performs training and subsequent processing corresponding to the business scenario based on the collected business data. For example, when a service scenario is deployed in a single technology QA window, based on service data required by the service data in the scenario, the knowledge base management platform acquires a corresponding xbnf file, and then establishes a connection between the knowledge base management platform and a client according to service requirements, and performs data transmission and other processes, which are not described herein again.
According to the scheme, the private cloud identification training method is provided, different requirements of customers in different scenes are met, the customers can use the cloud identification training method after opening the box during integration, and the operation is simple.
In some optional embodiments, the service scenario includes full link automatic identification, and the method includes: acquiring a recognition engine selected by the developer based on the full link automatic recognition; a training service is called to upload question statements from the knowledge base management platform and train the questions, and a file synchronization service is called to store trained resources; and responding to an asynchronous callback instruction of the knowledge base management platform model training ID after the training service is successfully trained, and storing the model training ID into a configuration list of the product ID automatically identified by the full link. For example, a developer accesses a knowledge base management platform, selects an identification engine, the knowledge base management platform calls a training service (emzt) to perform uploading problem description training, stores resources to a synchronous File service (File Share) after the training is successful, then stores an lmId (lmId) trained by an asynchronous callback language model of the knowledge base management platform into a configuration list (config) of a product ID (product ID), and details are not repeated here.
In some optional embodiments, the business scenario includes a single technology QA and an identification training tool, the method comprising: training a question statement file in response to a developer exporting the question statement file from the knowledge base management platform and training instructions of the developer; calling a file synchronization service to synchronously store the trained resources; and in response to successful training, binding the single technology QA and the product ID of the identification training tool to the redis after successful training. For example, a developer derives a question explanation txt File from a knowledge base management platform, enters a private cloud platform selection language model → creation language → uploading language material for training, file Share synchronously stores the trained resources after the training is successful, and meanwhile, the private cloud platform binds the redis bound with the product Id after the training is successful, which is not described again.
In some optional embodiments, the business scenario comprises a single technology QA window deployment, the method comprising: in response to the developer successfully publishing a question-and-answer library at the knowledge base management platform, generating an xbnf file based on the questions in the published question-and-answer library; and establishing websoket connection with a client, and sending the xbnf file to the client. For example, after the user successfully enters the knowledge base management platform to issue the question-answer base, the knowledge base management platform generates an xbnf file based on all the questions, then establishes a websoket connection with the client, sends the address of the xbnf file to the client, and the client receives the address of the xbnf file, which is not described herein again.
Referring to fig. 2, a flowchart of an embodiment of a private cloud identification application method of the present invention is shown.
As shown in fig. 2, in step 201, the client receives a voice request of a user, and acquires a product ID delivered by the recognition service based on the voice request. For example, a user sends a voice request through a client, and the client invokes a recognition service (CASR) to obtain an lmId of a product binding and associate a productId, which is not described herein again.
Then, in step 202, the client obtains the bound information or resource based on the product ID, and obtains the voice recognition text corresponding to the voice request based on the bound information or resource. For example, the client calls the CASR to obtain the lmId of the product binding from the redis; meanwhile, dds calls CASR to transmit the voice request and the multi-path lmId; the CASR service obtains the speech recognition text from the shared file decoding lmId, which is not described in detail herein.
In some optional embodiments, the bound information or resource includes a configuration sheet or a redis, where the lmId may be obtained through the configuration sheet (config) or the redis, which is not described herein again.
It should be noted that, although the above embodiments adopt numbers with definite precedence order such as step 101 and step 102 to define the precedence order of the steps, in an actual application scenario, some steps may be executed in parallel, and the precedence order of some steps is also not defined by the numbers, and this application is not limited herein and is not described herein again.
The following description is given to a specific example describing some problems encountered by the inventor in implementing the present invention and a final solution so as to enable those skilled in the art to better understand the solution of the present application.
The inventors discovered the defects of these similar techniques in the process of implementing the present invention:
for example, a customer may customize his or her semantic utterance and then add it to a product, and when the product is consumed by the device, the customer's voice requests the customized utterance, hopefully to hit speech recognition. However, in the present stage, when both the speech recognition technology (ASR) and the semantic service scene are provided, a client who wants to realize the function needs to make the statement of the service scene into an input format required by ASR service and then transfers a recognition service language model for training; when consuming, developers are required to access the ASR recognition and semantic technology and to be chained, which is high in development cost and difficult to understand for clients.
The inventors have found in the course of carrying out the invention why the reason is not easily imaginable:
in order to solve the defects, the speaking method of a customized system is usually derived in the market at present, and then the language model is made into a format required by ASR recognition service training for language model training; customer developers access ASR and Natural Language Understanding (NLU) technologies themselves at the time of consumption, and as noted above, the difficulty of development and difficulty of understanding can be significant.
Secondly, the system can be customized when the private cloud of the client is deployed, if the problem of the input of the knowledge base management platform does not exist in the recognition field, the recognition rate is inaccurate and the experience is not good when the user inputs the voice statement of the knowledge base management platform.
In addition, other platforms support a wide variety of types of recognition for privatized environments, such as: (1) the full link automatic identification is supported, and the speech of the statement input by the customer customized system can be automatically identified during consumption; (2) supporting a single technical question-answering and language model recognition training tool, leading out txt text from a customized utterance by a user, and leading the text into the recognition training tool; (3) the full-link automatic recognition and recognition training tool is supported, and an ASR automatic recognition technology is combined with the recognition training tool, so that customized utterance training of customers is fully met; (4) and window localization deployment is supported, and recognition voice interaction is realized by using xbnf (ebnf grammar format) and offline recognition training. Supporting the different needs of the different scenarios of the customers as described above, and the customers can be used out of box when integrating, and the operation is simple, which is certainly a big challenge to meet these and easy to use.
The invention has the technical innovation points that:
in order to support the different scenes, and the client can use the method after opening the box, the scheme provides a development toolkit sdk, combines the speech recognition technology ASR and the semantics, and for the client, the inside is a black box which can be directly used by the client.
First, a customer needs to customize a product ID (productId), configure information of the productId, and the productId can run through the whole process of recognition training, recognition consumption and semantic consumption.
Aiming at different scenes, the implementation is different, for example, full link automatic identification is carried out, because public cloud industry solutions and private cloud customized systems are involved, dds service is introduced, a producer can automatically call a training service (ezmt) to carry out identification training when producing data, then training information is stored in a product configuration sheet, dds can read identification information from the configuration sheet when consuming, and then an ASR identification service is called to obtain a voice text; a single technology QA and a language model recognition training tool are simple in type, and do not relate to dds service, a single technology QA exports a statement, the recognition training tool is accessed, a productId and the statement are imported into a training language model, a productId and language model lmId relation is established, and a client requests the recognition training tool to acquire a voice text during consumption.
Referring to fig. 3, a private cloud identification training flowchart of a private cloud identification training scheme according to an embodiment of the present invention is shown. The figure is mainly directed to a full link automatic identification flow chart.
As shown in fig. 3, the producer side: entering a knowledge base management platform, identifying and training when the skill is released, selecting an identification engine, and calling ezmt service to upload question explanation training; after the ezmt service is successfully trained, storing a resource into a synchronous File service (File Share), and then asynchronously calling back a 'knowledge base management platform' lmId; the knowledge base management platform stores the lmId to the config configuration ticket for the productId.
The consumer end: a client voice requests dds, wherein the dds reads a config configuration list according to the productId to obtain a multipath lmId of the ASR; and the dds calls a recognition service (CASR), transmits the voice request and one path, two paths, three paths, four paths and QA lmId, and acquires the voice text.
Please refer to fig. 4, which shows another private cloud identification training flowchart of a specific embodiment of the private cloud identification training scheme according to the embodiment of the present invention. The diagram is primarily directed to a single technique QA and recognition training tool flow diagram.
As shown in fig. 4, the producer side: the user derives a question and explanation txt file from a knowledge base management platform; a user enters a 'thinking and relaxation marking training integrated platform', selects a language model → creates a language → uploads a corpus, then trains and synchronizes resources; and binding the productId ID by redis after the language model of the 'thinking must navigation and marking training integrated platform' is successfully trained.
The consumer side: the client receives the voice request and calls the CASR service to transmit the parameter productId; the CASR service obtains the lmId from redis and then decodes the lmId from the shared file to obtain the voice text.
Referring to fig. 5, a flowchart of private cloud identification application according to a second specific embodiment of the scheme of the private cloud identification application according to the embodiment of the present invention is shown. The figure is primarily directed to a full link auto-id and id training tool flow diagram.
As shown in fig. 5, the producer side: the production mode of 'full link automatic identification' + 'single technology QA + identification training tool' is combined.
The consumer end: the client voice requests dds, and the dds reads a config configuration list according to the productId to obtain the lmId of asr; acquiring a label training lmId bound with a product from redis; the dds calls the CASR service to transmit the parameter voice request and the multipath lmId; the CASR service obtains the voice text from the shared file decoding lmId.
Please refer to fig. 6, which shows another private cloud identification training flowchart of a specific embodiment of the private cloud identification training scheme according to the embodiment of the present invention. The figure is mainly directed to a single technology QA window deployment flow chart.
As shown in fig. 6, the producer side: after a user enters a knowledge base management platform to issue a question-answer base successfully, an xbnf file is generated based on all questions; establishing websocket connection between the knowledge base management platform and the client, and sending an xbnf file address to the client; and the client websocket receives the file address sent by the server.
The consumer end: and the client receives the voice request, decodes the voice by using the xbnf sent by the server and acquires the text.
Beta version formed by the inventor in the process of implementing the invention:
the method is characterized in that xbnf recognition is used for QA window deployment of a single technology, interaction with a client is designed at first and is not websocket connection data transmission, the client is informed of updating if a client customizes a speaking method, and the client takes local latest xbnf decoding when consuming. The initial design is that when a client issues QA in a management background, the client records xbnf and a version number, when the client consumes, the client requests the client, the client calls the QA management background in real time to acquire the latest version number of the xbnf, and if the acquired version number is the same as the version number locally recorded by the client, the client decodes the local xbnf; if not, calling the QA management background to acquire the latest xbnf file, and then decoding by using the latest xbnf. However, if the speaking quantity of the client is large, the file trained by the xbnf is also large, the client responds to the client result after the three steps of acquiring, storing and decoding are completed, time consumption is long, user experience is not friendly, and therefore the client is notified to store in real time when the brainstorming storm is changed into updating.
The inventor finds that deeper effects are achieved in the process of implementing the invention:
by the aid of the solution, different requirements of different scenes of a client can be met, the user can use the product industry solution and the QA question-answer library, customized statements are automatically recognized and trained, and the user can use the product industry solution and the QA question-answer library after opening a box without developing access recognition; if a customized utterance needs to be accessed into a product, a matched recognition training tool can be used, training can be performed only by inputting the product and importing the utterance, the trained language model is bound with the product, the client calls a self-training platform recognition service during consumption, the recognition service can give a recognition voice result, and a client does not need to pay attention to a self-training platform training logic and can use the self-training platform recognition service only as a tool.
The system provides a complete speech recognition end-to-end experience combined autonomous optimization system scheme; based on the scheme, the standard productization of the voice recognition optimization capability can be realized, and the commercial value is generated through the output of public cloud and privatization.
Referring to fig. 7, a block diagram of a private cloud identification training apparatus according to an embodiment of the present invention is shown.
As shown in fig. 7, the private cloud recognition training apparatus 700 includes a business selection program module 710 and a training processing program module 720.
The service selection program module 710 is configured to respond to a developer selecting different service scenario requirements at the knowledge base management platform, and collect corresponding service data from the developer based on the service scenario requirements; and the training processing program module 720 is configured to perform training and subsequent processing corresponding to the business scenario based on the collected business data.
It should be understood that the modules recited in fig. 7 correspond to various steps in the method described with reference to fig. 1. Thus, the operations and features described above for the method and the corresponding technical effects are also applicable to the modules in fig. 7, and are not described again here.
It should be noted that the modules in the embodiments of the present application are not used to limit the solution of the present application, for example, a business selection program module configured to respond to a developer selecting different business scenario requirements at the knowledge base management platform, and collect corresponding business data from the developer based on the business scenario requirements; in addition, the related function module may also be implemented by a hardware processor, for example, the service selection program module may be implemented by a processor, which is not described herein again.
Referring to fig. 8, a block diagram of a private cloud identification application apparatus according to an embodiment of the present invention is shown.
As shown in fig. 8, the private cloud recognition application device 800 includes a parameter transferring module 810 and a recognition text acquiring module 820.
Wherein, the parameter transmission program module 810 is configured to receive a voice request of a user, and obtain a product ID for identifying service delivery based on the voice request; an identification text obtaining module 820 configured to obtain the bound information or resource based on the product ID, and obtain the voice identification text corresponding to the voice request based on the bound information or resource.
It should be understood that the modules recited in fig. 8 correspond to various steps in the method described with reference to fig. 2. Thus, the operations and features described above for the method and the corresponding technical effects are also applicable to the modules in fig. 8, and are not described again here.
It should be noted that the modules in the embodiments of the present application are not used to limit the solution of the present application, such as a reference program module configured to receive a voice request of a user, and obtain a product ID delivered by a recognition service based on the voice request; in addition, the related functional modules may also be implemented by a hardware processor, for example, the reference program module may be implemented by a processor, which is not described herein again.
In other embodiments, an embodiment of the present invention further provides a non-volatile computer storage medium, where computer-executable instructions are stored in the computer storage medium, and the computer-executable instructions may execute the private cloud identification training method in any of the above method embodiments;
as one embodiment, a non-volatile computer storage medium of the present invention stores computer-executable instructions configured to:
responding to different business scene requirements selected by a developer at the knowledge base management platform, and collecting corresponding business data to the developer based on the business scene requirements;
and training and subsequent processing corresponding to the business scene are carried out based on the collected business data.
The non-volatile computer-readable storage medium may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created from use of the private cloud recognition training apparatus, and the like. Further, the non-volatile computer-readable storage medium may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, the non-transitory computer-readable storage medium optionally includes memory remotely located from the processor, which may be connected to the private cloud identification training apparatus over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
Embodiments of the present invention further provide a computer program product, where the computer program product includes a computer program stored on a non-volatile computer-readable storage medium, and the computer program includes program instructions, and when the program instructions are executed by a computer, the computer executes any one of the private cloud identification training methods described above.
Fig. 9 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 9, the electronic device includes: one or more processors 910 and memory 920, one processor 910 being exemplified in fig. 9. The apparatus for the private cloud recognition training method may further include: an input device 930 and an output device 940. The processor 910, the memory 920, the input device 930, and the output device 940 may be connected by a bus or other means, and fig. 9 illustrates an example of a connection by a bus. The memory 920 is a non-volatile computer-readable storage medium as described above. The processor 910 executes various functional applications and data processing of the server by running the nonvolatile software programs, instructions and modules stored in the memory 920, that is, implementing the above method embodiments for the private cloud identification training apparatus method. The input device 930 may receive input numeric or character information and generate key signal inputs related to user settings and function control for the private cloud recognition training device. The output device 940 may include a display device such as a display screen.
The product can execute the method provided by the embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.
As an embodiment, the electronic device is applied to a private cloud identification training apparatus, and includes:
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 cause the at least one processor to:
responding to different business scene requirements selected by a developer at the knowledge base management platform, and collecting corresponding business data to the developer based on the business scene requirements;
and training and subsequent processing corresponding to the business scene are carried out based on the collected business data.
The electronic device of the embodiments of the present application exists in various forms, including but not limited to:
(1) A mobile communication device: such devices are characterized by mobile communications capabilities and are primarily targeted at providing voice, data communications. Such terminals include smart phones, multimedia phones, functional phones, and low-end phones, among others.
(2) Ultra mobile personal computer device: the equipment belongs to the category of personal computers, has calculation and processing functions and generally has the characteristic of mobile internet access. Such terminals include: PDA, MID, and UMPC devices, etc.
(3) A portable entertainment device: such devices may display and play multimedia content. The devices comprise audio and video players, handheld game consoles, electronic books, intelligent toys and portable vehicle-mounted navigation devices.
(4) The server is similar to a general computer architecture, but has higher requirements on processing capacity, stability, reliability, safety, expandability, manageability and the like because of the need of providing high-reliability service.
(5) And other electronic devices with data interaction functions.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A private cloud identification training method is used for a knowledge base management platform and comprises the following steps:
responding to different business scene requirements selected by a developer at the knowledge base management platform, and acquiring corresponding business data from the developer based on the business scene requirements, wherein the business scene requirements comprise full link automatic identification, single technology QA, identification training tools and/or single technology QA window deployment;
and training and subsequent processing corresponding to the business scene are carried out based on the collected business data, wherein the subsequent processing comprises binding the successfully trained resources to the product ID of the developer.
2. The method of claim 1, wherein the traffic scenario is full link automatic identification, the method comprising:
acquiring a recognition engine selected by the developer based on the full link automatic recognition;
a training service is called to upload question statements from the knowledge base management platform and train the questions, and a file synchronization service is called to store trained resources;
and responding to an asynchronous callback instruction of the knowledge base management platform model training ID after the training service is successfully trained, and storing the model training ID into a configuration list of the product ID automatically identified by the full link.
3. The method of claim 1 or 2, wherein the business scenario is a single technology QA and recognition training tool, the method comprising:
training a question statement file in response to a developer exporting the question statement file from the knowledge base management platform and training instructions of the developer;
calling a file synchronization service to synchronously store the trained resources;
and in response to successful training, binding the single technology QA and the product ID for identifying the training tool with the redis after successful training.
4. The method of claim 1, wherein the business scenario is a single technology QA window deployment, the method comprising:
responding to the success of the developer in publishing a question and answer library in the knowledge base management platform, and generating an xbnf file based on the questions in the published question and answer library;
and establishing a websoc key connection with a client, and sending the xbnf file to the client.
5. A private cloud identification application method is used for a client and comprises the following steps:
receiving a voice request of a user, and acquiring a product ID delivered by the recognition service trained according to the method in claim 1 based on the voice request;
and acquiring the bound information or resource based on the product ID, and acquiring the voice recognition text corresponding to the voice request based on the bound information or resource.
6. The method of claim 5, wherein the bound information or resource comprises a configuration sheet or a redis.
7. A private cloud recognition training apparatus for a knowledge base management platform, comprising:
the service selection program module is configured to respond to different service scene requirements selected by a developer on the knowledge base management platform and collect corresponding service data from the developer based on the service scene requirements, wherein the service scene requirements comprise full link automatic identification, single technology QA, an identification training tool and/or single technology QA window deployment;
and the training processing program module is configured to perform training and subsequent processing corresponding to the business scene based on the collected business data, wherein the subsequent processing comprises binding the successfully trained resources to the product ID of the developer.
8. A private cloud identification application for a client, comprising:
a parameter transmission program module configured to receive a voice request of a user, and obtain a product ID transmitted by the recognition service trained according to the method of claim 1 based on the voice request;
and the identification text acquisition module is configured to acquire the bound information or resource based on the product ID and acquire the voice identification text corresponding to the voice request based on the bound information or resource.
9. 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 steps of the method of any of claims 1 to 6.
10. A storage medium having a computer program stored thereon, wherein the program, when executed by a processor, performs the steps of the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011617714.9A CN112837678B (en) | 2020-12-31 | 2020-12-31 | Private cloud recognition training method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011617714.9A CN112837678B (en) | 2020-12-31 | 2020-12-31 | Private cloud recognition training method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112837678A CN112837678A (en) | 2021-05-25 |
| CN112837678B true CN112837678B (en) | 2022-12-02 |
Family
ID=75923939
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011617714.9A Active CN112837678B (en) | 2020-12-31 | 2020-12-31 | Private cloud recognition training method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112837678B (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107247726B (en) * | 2017-04-28 | 2018-06-08 | 北京神州泰岳软件股份有限公司 | Suitable for the implementation method and device of the intelligent robot of multi-service scene |
| CN109410922A (en) * | 2018-10-09 | 2019-03-01 | 苏州思必驰信息科技有限公司 | Resource preprocess method and system for voice dialogue platform |
| CN111049996B (en) * | 2019-12-26 | 2021-06-15 | 思必驰科技股份有限公司 | Multi-scene voice recognition method and device and intelligent customer service system applying same |
| CN111508479B (en) * | 2020-04-16 | 2022-11-22 | 重庆农村商业银行股份有限公司 | Voice recognition method, device, equipment and storage medium |
| CN111753075A (en) * | 2020-08-12 | 2020-10-09 | 腾讯科技（深圳）有限公司 | Method and device for creating question and answer data of customer service robot and computer equipment |
-
2020
- 2020-12-31 CN CN202011617714.9A patent/CN112837678B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN112837678A (en) | 2021-05-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3095113B1 (en) | Digital personal assistant interaction with impersonations and rich multimedia in responses | |
| EP3633947B1 (en) | Electronic device and control method therefor | |
| CN107463700B (en) | Method, device and equipment for acquiring information | |
| CN102819969A (en) | Implementation method for multimedia education platform and multimedia education platform system | |
| CN111090533B (en) | Bulletin information sending method, bulletin information sending device, electronic equipment and storage medium | |
| CN112364144B (en) | Interaction method, device, equipment and computer readable medium | |
| CN108491188A (en) | The exploitative management method and device of voice dialogue product | |
| CN110727782A (en) | Question and answer corpus generation method and system | |
| EP2747464A1 (en) | Sent message playing method, system and related device | |
| CN111679811B (en) | Web service construction method and device | |
| CN112185362A (en) | Voice processing method and device for user personalized service | |
| JP7375089B2 (en) | Method, device, computer readable storage medium and computer program for determining voice response speed | |
| CN113672748A (en) | Multimedia information playing method and device | |
| CN110418181B (en) | Service processing method and device for smart television, smart device and storage medium | |
| CN107888989A (en) | A kind of interactive system and method live based on internet | |
| CN111933118B (en) | Method and device for optimizing voice recognition and intelligent voice dialogue system applying same | |
| KR101351264B1 (en) | System and method for message translation based on voice recognition | |
| CN110442698B (en) | Dialog content generation method and system | |
| CN112837678B (en) | Private cloud recognition training method and device | |
| CN111726284A (en) | WeChat sending method and device for vehicle-mounted intelligent sound box | |
| CN114422468A (en) | Message processing method, device, terminal and storage medium | |
| CN110600021A (en) | Outdoor intelligent voice interaction method, device and system | |
| CN112802478B (en) | Hybrid cloud fusion scheduling method and device | |
| CN111966803B (en) | Dialogue simulation method and device, storage medium and electronic equipment | |
| CN108810105A (en) | For pushed information, the method and apparatus for showing information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |