Natural language processing arbitration-based central control method

Abstract

The application discloses a natural language processing dispatching method based on vertical arbitration, which comprises the following steps: a vertical arbitration service generates a vertical arbitration result according to input data, wherein the vertical arbitration result comprises a skill field to which the data belongs; the vertical arbitration result and the input data are sent to a scheduling service; a mapping table of the vertical arbitration result and a natural language processing service is read; and the scheduling service forwards the input data to a natural language processing vendor according to the skill field to which the data belongs and the mapping table. The application can help users to select the most suitable natural language processing vendor according to the vertical arbitration result, improve the skill hit rate and the recognition rate, and improve the overall response speed of the service without being limited by the use conditions of a single natural language processing vendor.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a natural language processing central control method based on vertical arbitration. Background Technology

[0002] In natural language processing (NLP) development scenarios, technologies such as automatic speech recognition can convert user voice input data into a data format that NLP algorithms can recognize. Before inputting the data into the NLP algorithm, it is usually first determined to belong to a specific skill domain through vertical arbitration, and then the NLP service is determined based on the arbitration result to improve the accuracy and recognition rate of skill matching.

[0003] Currently, among mainstream natural language processing (NLP) products, central control systems based on vertical arbitration typically target only a single NLP vendor. This means the product only interfaces with that vendor's own NLP services and doesn't offer the ability to interface with other NLP vendors. Therefore, the scenarios handled by the central control service are simpler; it only needs to consider the implicit relationship between the arbitration result and the NLP service, or it can forward the data directly without relying on the arbitration service. Furthermore, the central control service and the NLP service usually belong to the same service chain and cannot run concurrently, making the overall service response susceptible to the influence of the central control system.

[0004] Existing natural language processing central control scenarios include: 1) Connecting to a single vendor, without vertical arbitration, directly forwarding results to the deployed natural language processing service; 2) Connecting to a single vendor, with vertical sanctions defining skill domains, and forwarding results based on the mapping relationship between the domain and the deployed natural language processing service.

[0005] The above two scenarios mainly have the following drawbacks: 1) Only a single vendor is connected, and users cannot choose different natural language processing service providers according to their needs; 2) The skill areas handled by a single vendor are limited, which affects the user experience; 3) Users cannot customize the selection rules when connecting to multiple natural language processing providers according to their actual needs; vertical arbitration, central control services, and natural language processing services can only be called in a linear order, which makes the overall service call latency affected by the central control services. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a method for separating a central control service and maintaining the mapping relationship between the vertical arbitration results and the natural language processing service, based on the vertical arbitration results and user configuration, in scenarios with multiple natural language processing providers, in order to address the above-mentioned deficiencies of the prior art.

[0007] The technical solution adopted by this invention to solve its technical problem is: constructing a natural language processing central control method based on vertical category arbitration, including the following steps:

[0008] The vertical arbitration service generates vertical arbitration results based on input data, and the vertical arbitration results include the skill domain to which the data belongs;

[0009] The vertical arbitration result and the input data are sent to the scheduling service;

[0010] Read the mapping table between vertical arbitration results and natural language processing services;

[0011] The scheduling service forwards the input data to the natural language processing vendor based on the skill domain to which the data belongs and the mapping table.

[0012] 7. In the natural language processing central control method based on vertical arbitration described in this invention, the step of the scheduling service forwarding the input data to the natural language processing vendor according to the skill domain to which the data belongs and the mapping table includes:

[0013] The scheduling service queries the mapping table based on the skill domain to which the data belongs;

[0014] Determine if there is a separate configuration corresponding to the skill domain to which the data belongs. If there is, return the service address; otherwise, select the default vendor's service address.

[0015] The scheduling service forwards the input data to the natural language processing vendor based on the service address.

[0016] In the natural language processing control method based on vertical arbitration described in this invention, the mapping table is stored in the real-time hotspot service in the form of key-value pairs.

[0017] In the natural language processing central control method based on vertical arbitration described in this invention, the key-value pair includes the following configuration items: when vertical arbitration is not enabled, select the default natural language processing vendor; when vertical arbitration is enabled, if the vertical arbitration result is not mapped to the currently supported skill domain, transfer the processing to the default natural language processing vendor; when vertical arbitration is enabled, select the natural language processing vendor specified for a single skill domain; information on the connected natural language processing vendors and their service addresses.

[0018] The natural language processing control method based on vertical arbitration described in this invention also includes maintaining the mapping table in real time according to user selection.

[0019] The natural language processing central control method based on vertical arbitration described in this invention also includes maintaining the mapping table in real time according to the health status of the natural language processing service.

[0020] The present invention also discloses a natural language processing central control system based on vertical arbitration, including at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform the natural language processing central control method based on vertical arbitration as described above.

[0021] The present invention also discloses a non-transitory computer storage medium storing computer-executable instructions that can execute the above-described vertical arbitration-based natural language processing control method.

[0022] The vertical arbitration-based natural language processing central control method of the present invention has the following beneficial effects: In this invention, after the vertical arbitration service returns the result, it sends the result to the scheduling service, without being processed by the central control service; the scheduling service reads the mapping table between the vertical arbitration results and the natural language processing services cached in the real-time hot service, and forwards the result to the corresponding natural language processing service according to the mapping relationship of the domain. Therefore, it can avoid adjusting business requirements due to the limitations of a single natural language processing provider, reducing user experience, improving service response speed, and ensuring that the service can automatically select the most suitable natural language processing provider based on the vertical arbitration result. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort:

[0024] Figure 1 The diagram shows a flowchart of a natural language processing central control method based on vertical arbitration provided in an embodiment of the present invention;

[0025] Figure 2 The diagram illustrates how the scheduling service determines the forwarding address using a mapping table.

[0026] Figure 3 This diagram illustrates a specific structural diagram of a natural language processing control system based on vertical arbitration, according to an embodiment of the present invention. Detailed Implementation

[0027] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Typical embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0029] The definitions are as follows:

[0030] Vertical category arbitration: refers to classifying user statements into corresponding business skills;

[0031] Natural Language Processing: The study of various theories and methods that enable effective communication between humans and computers using natural language;

[0032] Automatic speech recognition technology refers to the process of converting the lexical content of human speech into computer-readable input.

[0033] Skill domain: Used to distinguish the applicable scenarios for different natural language processing algorithms, such as weather, vehicle control, etc.

[0034] A service is a set of definitions, procedures, and protocols that provide a common set of functions and enable data sharing across various platforms.

[0035] Figure 1 The diagram shown is a flowchart of a natural language processing central control method based on vertical category arbitration provided by an embodiment of the present invention; as follows: Figure 1 As shown, the natural language processing central control method based on vertical category arbitration provided by this invention includes:

[0036] Step S1: The vertical arbitration service generates a vertical arbitration result based on the input data, wherein the vertical arbitration result includes the skill domain to which the data belongs;

[0037] Step S2: Send the vertical arbitration result and the input data to the scheduling service;

[0038] Step S3: Read the mapping table between vertical arbitration results and natural language processing services;

[0039] Step S4: The scheduling service forwards the input data to the natural language processing vendor according to the skill domain to which the data belongs and the mapping table.

[0040] In this invention, after the vertical arbitration service returns its result, it sends the result to the scheduling service, bypassing the central control service. The scheduling service reads the mapping table between the vertical arbitration results and natural language processing services cached in the real-time hotspot service, and forwards the result to the corresponding natural language processing service according to the mapping relationship of that domain. This avoids adjusting business requirements due to limitations imposed by a single natural language processing provider, reducing user experience, improving service response speed, and ensuring that the service can automatically select the most suitable natural language processing provider based on the vertical arbitration result.

[0041] Specifically, in one embodiment of the present invention, such as Figure 2 As shown, the scheduling service determines the forwarding address through the mapping table, specifically including:

[0042] The scheduling service queries the mapping table based on the skill domain to which the data belongs;

[0043] Determine if there is a separate configuration corresponding to the skill domain to which the data belongs. If there is, return the service address; otherwise, select the default vendor's service address.

[0044] The scheduling service forwards the input data to the natural language processing vendor based on the service address.

[0045] Specifically, in one embodiment of the present invention, a real-time hotspot service is built using ETCD or other key-value databases. The central control service stores the mapping table between vertical arbitration results and natural language processing services in the real-time hotspot service in the form of key-value pairs. The key-value pair access method is as follows:

[0046] Example: Key Value

[0047] Here, key is a keyword, in string format, referring to the configuration name; value is the value, referring to the value of the configuration item.

[0048] Furthermore, the key-value pairs mainly include four configuration items: a. The default natural language processing vendor when vertical arbitration is not enabled; b. The default natural language processing vendor to be transferred for processing when vertical arbitration is enabled and the vertical arbitration result is not mapped to the currently supported skill domain; c. The natural language processing vendor specified for a single skill after vertical arbitration is enabled; d. Information on the connected natural language processing providers and their service addresses.

[0049] Table 1 below is a specific example of a key-value pair:

[0050]

[0051]

[0052] Furthermore, in one embodiment of the present invention, the mapping table is maintained in real time according to user selection. Users can adjust the above four configuration items at any time through the configuration interface provided by the central control service. Specifically, this includes:

[0053] When a user updates vendor information or connects to a new vendor, the central control service queries the vendor's service information; then the central control service modifies or updates the vendor's natural language processing service information.

[0054] Configure the default vendor when vertical arbitration is not enabled: When vertical arbitration is not enabled, the central control service queries the registered natural language processing service information of the vendor by allowing the user to select the default vendor; updates the default service configuration based on the query results; and updates the mapping relationship of all skill areas.

[0055] After enabling vertical arbitration, configure the default vendor: the user selects the default vendor; the central control service queries the registered natural language processing service information of that vendor; based on the query results, update the default service configuration; update the mapping relationship of skill areas for which no vendor has been selected separately;

[0056] Configure vendor selection for a single skill domain: The user selects a vendor for a single skill domain; the central control service queries the registered natural language processing service information of that vendor; and updates the mapping relationship for that skill domain.

[0057] Furthermore, in one embodiment of the present invention, the mapping table is maintained in real time based on the health status of the natural language processing service. Specifically, the central control service monitors the health status of the natural language processing service, such as whether it can be called normally and whether the return result speed is too slow; if a natural language processing service goes offline, the central control service will push a notification to the user and switch the mapping relationship using that natural language processing service in the configuration items to other selectable vendors.

[0058] This invention utilizes an independent central control service to maintain a mapping table between vertical arbitration results and natural language processing (NLP) services in real time, based on user selection, skill domain, and service health. The scheduling service then selects the appropriate NLP provider using this mapping table. This allows users to avoid being limited by the availability of a single NLP provider, enabling them to choose the most suitable provider based on the vertical arbitration results, thereby improving skill hit rate and recognition accuracy, and enhancing overall service response speed.

[0059] Figure 3 This is a schematic diagram of the hardware structure of the natural language processing central control system based on vertical arbitration provided in an embodiment of the present invention, as shown below. Figure 3 As shown, the device includes one or more processors 31 and a memory 32. Figure 3 Taking a processor 31 as an example, the processor 31 and the memory 32 can be connected via a bus or other means. Figure 3 Taking the example of a connection between China and Israel via a bus.

[0060] Processor 31 can be a Central Processing Unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations thereof. The general-purpose processor can be a microprocessor or any conventional processor.

[0061] The memory 32, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the method for improving the gradient effect of LED lamps in the embodiments of the present invention. The processor 31 executes various functional applications and data processing of the server by running the non-transitory software programs, instructions, and modules stored in the memory 32, that is, it implements the natural language processing central control method based on vertical arbitration in the above embodiments.

[0062] The memory 32 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the LED driving pulse modulation device. Furthermore, the memory 32 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 32 may optionally include memory remotely located relative to the processor 31, and these remote memories can be connected to the LED driving pulse modulation device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0063] The one or more modules are stored in the memory 32, and when executed by the one or more processors 31, they perform the natural language processing central control method based on vertical arbitration as described above.

[0064] The above-described product can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in [reference 1]. Figure 1 The relevant descriptions in the illustrated embodiments.

[0065] This invention also provides a non-transitory computer storage medium storing computer-executable instructions that can execute the method described above for improving the gradient effect of LED lighting fixtures. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM) > random access memory (RAM) > flash memory > hard disk drive (HDD) or solid-state drive (SSD), etc.; the storage medium may also include combinations of the above types of memory.

[0066] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0067] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the scope defined by the appended claims.

Claims

1. A natural language processing central control method based on vertical category arbitration, characterized in that, Includes the following steps: The vertical arbitration service generates vertical arbitration results based on input data, and the vertical arbitration results include the skill domain to which the data belongs; The vertical arbitration result and the input data are sent to the scheduling service; Read the mapping table between vertical arbitration results and natural language processing services, wherein the natural language processing services are provided by multiple different natural language processing service providers; The scheduling service forwards the input data to the natural language processing service provider according to the skill domain to which the data belongs and the mapping table, wherein the mapping table is dynamically updated based on user configuration or the health status of the natural language processing service.

2. The natural language processing central control method based on vertical category arbitration according to claim 1, characterized in that, The step of the scheduling service forwarding the input data to the natural language processing service provider according to the skill domain to which the data belongs and the mapping table includes: The scheduling service queries the mapping table based on the skill domain to which the data belongs; Determine if there is a separate configuration corresponding to the skill domain to which the data belongs. If there is, return the service address; otherwise, select the default vendor's service address. The scheduling service forwards the input data to the natural language processing service provider based on the service address.

3. The natural language processing central control method based on vertical category arbitration according to claim 1, characterized in that, The mapping table is stored in the real-time hotspot service in the form of key-value pairs.

4. The natural language processing central control method based on vertical category arbitration according to claim 3, characterized in that, The key-value pairs include the following configuration items: When vertical arbitration is not enabled, select the default natural language processing service provider; When vertical arbitration is enabled, if the vertical arbitration result is not mapped to a currently supported skill area, the default natural language processing service provider will be used to transfer the processing. After enabling vertical arbitration, select the natural language processing service provider specified for a single skill area; Information on connected natural language processing providers and their service addresses.

5. A natural language processing central control system based on vertical category arbitration, characterized in that, It includes at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform the natural language processing central control method based on vertical arbitration as described in any one of claims 1 to 3.

6. A non-transitory computer storage medium, characterized in that, The computer storage medium stores computer-executable instructions that can execute the natural language processing central control method based on vertical arbitration as described in any one of claims 1 to 3.

Images