Task strategy recommendation and recommendation model training method for component change

Abstract

The embodiment of the invention provides a task strategy recommendation method for component change and a recommendation model training method, and the task strategy recommendation method for component change comprises the steps: obtaining a component change request which comprises the associated information of a target change component; the associated information is input into a pre-trained recommendation model, a target task strategy for the target change component is obtained, the recommendation model is obtained through training based on sample associated information of the multiple sample change components and a label task strategy, and the sample associated information comprises at least one sample associated data item; the label task strategy is determined based on the matching relationship between the task strategy keyword in the sample text and each sample associated data item, and the sample text is a text related to the sample change task. The tag task strategy is obtained by matching the task strategy keyword in the sample text with the at least one sample associated data item, and the recommendation model is trained by using the tag task strategy, so that the accuracy of an output result of the recommendation model is ensured.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular to a method for training a task strategy recommendation and a recommendation model for component changes. Background Art

[0002] The recommendation system is an information retrieval tool in the Internet era. It originated in the 1990s. After more than 20 years of accumulation and precipitation, the recommendation system has gradually become an independent discipline and has achieved many results in academic research and business applications.

[0003] Conventional recommendation systems use binary classification labels for model training. However, binary classification labels may have inaccurate expression problems in some fields, resulting in low accuracy of the trained recommendation system. In addition, current recommendation systems are mostly used for advertising recommendations, and there is no recommendation method adapted to component change scenarios. It is impossible to intelligently recommend strategies for component change tasks. Therefore, there is an urgent need for a method to accurately recommend task strategies for component change scenarios. Summary of the invention

[0004] In view of this, an embodiment of the present application provides a method for recommending a task strategy for component changes. One or more embodiments of the present application also involve a method for recommending a task strategy for vehicle part changes, a method for training a recommendation model, a device for recommending a task strategy for component changes, a device for recommending a task strategy for vehicle part changes, a device for training a recommendation model, a computing device, a computer-readable storage medium, and a computer program to solve the technical defects in the prior art.

[0005] According to a first aspect of an embodiment of the present application, a method for recommending a task strategy for component changes is provided, comprising:

[0006] Obtaining a component change request, wherein the component change request includes associated information of a target change component;

[0007] The association information is input into a pre-trained recommendation model to obtain a target task strategy for the target change component, wherein the recommendation model is trained based on sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between task strategy keywords in a sample text and each sample association data item, and the sample text is text related to the sample change task.

[0008] According to a second aspect of an embodiment of the present application, a method for recommending a task strategy for vehicle part changes is provided, comprising:

[0009] Acquire a vehicle part change request, wherein the vehicle part change request includes change-related information of a target vehicle part;

[0010] Inputting the change association information into a pre-trained recommendation model to obtain a target task strategy for the target vehicle part, wherein the recommendation model is trained based on sample association information and label task strategy of multiple sample vehicle parts, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task;

[0011] Based on the target task strategy, it is recommended to execute the target vehicle part change task.

[0012] According to a third aspect of an embodiment of the present application, a training method for a recommendation model is provided, comprising:

[0013] Acquire a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task;

[0014] Extracting sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of the multiple sample change components, and the first label task strategy is the label task strategy of the first sample change component;

[0015] Inputting the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy;

[0016] Based on the prediction task strategy and the first label task strategy, the model parameters of the recommendation model are adjusted, and the step of extracting the sample association information of the first sample change component and the first label task strategy from the sample set is returned to execute until the training stop condition is reached to obtain a recommendation model that has completed training.

[0017] According to a fourth aspect of an embodiment of the present application, a task strategy recommendation device for component change is provided, comprising:

[0018] A first request acquisition module is configured to acquire a component change request, wherein the component change request includes associated information of a target change component;

[0019] The first acquisition module is configured to input the association information into a pre-trained recommendation model to obtain a target task strategy for the target change component, wherein the recommendation model is trained based on sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between task strategy keywords in a sample text and each sample association data item, and the sample text is text related to the sample change task.

[0020] According to a fifth aspect of an embodiment of the present application, a task strategy recommendation device for vehicle parts change is provided, comprising:

[0021] A second request acquisition module is configured to acquire a vehicle part change request, wherein the vehicle part change request includes change association information of a target vehicle part;

[0022] a second obtaining module, configured to input the change association information into a pre-trained recommendation model to obtain a target task strategy for the target vehicle part, wherein the recommendation model is trained based on sample association information and label task strategy of multiple sample vehicle parts, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task;

[0023] The recommendation module is configured to recommend executing a target vehicle part change task based on the target task strategy.

[0024] According to a sixth aspect of an embodiment of the present application, a training device for a recommendation model is provided, comprising:

[0025] A sample set acquisition module is configured to acquire a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task;

[0026] an extraction module, configured to extract sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of the multiple sample change components, and the first label task strategy is the label task strategy of the first sample change component;

[0027] An input module, configured to input the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy;

[0028] The adjustment module is configured to adjust the model parameters of the recommendation model based on the prediction task strategy and the first label task strategy, and return to execute the step of extracting the sample association information of the first sample change component and the first label task strategy from the sample set until the training stop condition is reached to obtain a recommendation model that has completed training.

[0029] According to a seventh aspect of an embodiment of the present application, there is provided a computing device, including:

[0030] Memory and processor;

[0031] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model are implemented.

[0032] According to an eighth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores computer-executable instructions, which, when executed by a processor, implement the steps of the above-mentioned task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model.

[0033] According to the ninth aspect of an embodiment of the present application, a computer program is provided, wherein when the computer program is executed in a computer, the computer is caused to execute the steps of the above-mentioned task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model.

[0034] An embodiment of the present application obtains a component change request, wherein the component change request includes association information of the target change component; the association information is input into a pre-trained recommendation model to obtain a target task strategy for the target change component, wherein the recommendation model is trained based on sample association information and label task strategy of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on the matching relationship between the task strategy keywords in the sample text and each sample association data item, and the sample text is the text related to the sample change task. The recommendation model is obtained by training the sample association information and label task strategy of multiple sample change components, the label task strategy is obtained by matching the task strategy keywords in the sample text with at least one sample association data item, so that the label task strategy corresponds to the task strategy keywords matched with at least one association data item, and the corresponding accuracy of the label task strategy to the sample association information of the sample change component is achieved, thereby ensuring the accuracy of the output result of the recommendation model, that is, by inputting the association information of the target change component in the component change request into the recommendation model with high accuracy, an accurate target task strategy for the target change component is obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a schematic diagram of an interaction process under a task strategy recommendation system architecture for component changes provided by an embodiment of the present application;

[0036] Figure 2 It is a framework diagram of a task strategy recommendation system for component changes provided by an embodiment of the present application;

[0037] Figure 3 is a flowchart of a task strategy recommendation method for component changes provided by an embodiment of the present application;

[0038] Figure 4 is a flow chart of a method for recommending a task strategy for vehicle parts changes provided by an embodiment of the present application;

[0039] Figure 5 is a flow chart of a training method for a recommendation model provided by an embodiment of the present application;

[0040] Figure 6 It is a process flow chart of a task strategy recommendation method for component change provided by an embodiment of the present application;

[0041] Figure 7 It is a structural diagram of a task strategy recommendation device for component change provided by an embodiment of the present application;

[0042] Figure 8It is a structural schematic diagram of a task strategy recommendation device for vehicle parts change provided by an embodiment of the present application;

[0043] Fig. 9 It is a structural schematic diagram of a training device for a recommendation model provided by an embodiment of the present application;

[0044] Fig.10 It is a structural block diagram of a computing device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0045] Many specific details are described in the following description to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present application, so the present application is not limited by the specific implementation disclosed below.

[0046] The terms used in one or more embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of the present application. The singular forms of "a", "said" and "the" used in one or more embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of the present application refers to and includes any or all possible combinations of one or more associated listed items.

[0047] It should be understood that, although the terms first, second, etc. may be used to describe various information in one or more embodiments of the present application, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of the present application, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0048] In addition, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0049] First, the terms involved in one or more embodiments of the present application are explained.

[0050] The recommendation system is an information retrieval tool in the Internet era. It originated in the 1990s. After more than 20 years of accumulation and precipitation, the recommendation system has gradually become an independent discipline and has achieved many results in academic research and business applications.

[0051] Conventional recommendation systems use binary classification labels for model training, but binary classification labels may have inaccurate expression problems in some fields, resulting in low accuracy of the recommendation system obtained through training. For example, in the industrial production part change scenario, a part change task usually includes at least one associated data item in the associated information, and each associated data item requires a corresponding strategy. However, binary classification labels cannot set corresponding strategy labels for each associated data item. In addition, current recommendation systems are mostly used for advertising recommendations, and there is no recommendation method adapted to component change scenarios. It is impossible to intelligently recommend strategies for component change tasks. Currently, for component change scenarios, component change strategies are usually determined based on expert experience in the component change field and the inventory of components in the factory. When expert experience is outdated or factory inventory information is inaccurate, it often causes difficulties in component changes. Therefore, there is an urgent need for a method to accurately recommend task strategies for component change scenarios.

[0052] In the present application, a task strategy recommendation method for component changes is provided. The present application also relates to a task strategy recommendation method for vehicle part changes, a training method for a recommendation model, a task strategy recommendation device for component changes, a task strategy recommendation device for vehicle part changes, a training device for a recommendation model, a computing device, and a computer-readable storage medium, which are described in detail one by one in the following embodiments.

[0053] See also Figure 1 , Figure 1 FIG. 1 shows a schematic diagram of an interaction process under a task strategy recommendation system architecture for component changes provided by an embodiment of the present application, such as Figure 1 As shown, the system includes a server 100 and a client 200 .

[0054] Client 200: used to send a component change request to the server 100;

[0055] The server 100 is used to obtain a component change request, wherein the component change request includes association information of a target change component; input the association information into a pre-trained recommendation model to obtain a target task strategy for the target change component, wherein the recommendation model is trained based on sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task;

[0056] The client 200 is also used to receive a target task policy for a target change component, and recommend an execution component change task based on the target task policy.

[0057] By applying the scheme of the embodiment of the present application, a recommendation model is obtained by training sample association information and label task strategies of multiple sample change components. The label task strategy is obtained by matching task strategy keywords in the sample text with at least one sample association data item, so that the label task strategy corresponds to the task strategy keywords that match the at least one association data item, thereby achieving the accuracy of the corresponding label task strategy to the sample association information of the sample change component, thereby ensuring the accuracy of the output results of the recommendation model, that is, by inputting the association information of the target change component in the component change request into the recommendation model with high accuracy, an accurate target task strategy for the target change component is obtained.

[0058] See also Figure 2 , Figure 2 The framework diagram of a task strategy recommendation system for component changes provided by an embodiment of the present application is shown, and the system may include a server 100 and multiple clients 200. Multiple clients 200 may establish a communication connection through the server 100. In the task strategy recommendation scenario for component changes, the server 100 is used to provide task strategy recommendation services for component changes between multiple clients 200. Multiple clients 200 may serve as senders or receivers respectively, and realize communication through the server 100.

[0059] The user can interact with the server 100 through the client 200 to receive data sent by other clients 200, or send data to other clients 200, etc. In the task strategy recommendation scenario for component changes, the user can issue a task strategy recommendation request for component changes to the server 100 through the client 200, and the server 100 generates a task strategy recommendation result for component changes according to the task strategy recommendation request for component changes, and pushes the task strategy recommendation result for component changes to other clients 200 that have established communication.

[0060] The client 200 and the server 100 are connected via a network. The network provides a medium for a communication link between the client 200 and the server 100. The network may include various connection types, such as wired or wireless communication links or optical fiber cables, etc. The data transmitted by the client 200 may need to be encoded, transcoded, compressed, etc. before being released to the server 100.

[0061] The client 200 can be a browser, an application (APP, Application), or a web application such as Hypertext Markup Language Version 5 (H5, Hyper Text Markup Language5) application, or a light application (also known as a mini-program, a lightweight application) or a cloud application, etc. The client 200 can be based on the software development kit (SDK, Software Development Kit) of the corresponding service provided by the server, such as based on the real-time communication (RTC, Real Time Communication) SDK development and acquisition. The client 200 can be deployed in an electronic device, and needs to rely on the device to run or some APPs in the device to run. For example, the electronic device can have a display screen and support information browsing, such as a personal mobile terminal such as a mobile phone, a tablet computer, a personal computer, etc. Various other types of applications can also be configured in the electronic device, such as human-computer dialogue applications, model training applications, text processing applications, web browser applications, shopping applications, search applications, instant messaging tools, mailbox clients, social platform software, etc.

[0062] The server 100 may include servers that provide various services, such as servers that provide communication services to multiple clients, servers for background training that provide support for models used on clients, and servers that process data sent by clients. It should be noted that the server 100 can be implemented as a distributed server cluster consisting of multiple servers, or as a single server. The server can also be a server of a distributed system, or a server combined with a blockchain. The server can also be a cloud server for basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content distribution networks (CDN, Content Delivery Network), and big data and artificial intelligence platforms, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology.

[0063] It is worth noting that the task strategy recommendation method for component changes provided in the embodiment of the present application is generally executed by the server 100. However, in other embodiments of the present application, the client 200 may also have similar functions as the server, thereby executing the task strategy recommendation method for component changes provided in the embodiment of the present application. In other embodiments, the task strategy recommendation method for component changes provided in the embodiment of the present application may also be jointly executed by the client 200 and the server 100.

[0064] See also Figure 3 , Figure 3 A flowchart of a method for recommending a task strategy for component changes provided by an embodiment of the present application is shown, which specifically includes the following steps.

[0065] Step 302: Obtain a component change request, wherein the component change request includes associated information of a target change component.

[0066] The embodiment of the present application is applied to a client and / or server to which a task strategy recommendation method for component changes is applied, and the following description is given using the server as an example.

[0067] When there is a need to recommend a task strategy for component change, the server obtains the component change request to obtain the target task strategy of the target change component using the component change request and a pre-trained recommendation model, so that the component change task corresponding to the component change request is executed using the target task strategy.

[0068] Specifically, a component change request refers to a request to change or replace a component. The component change request includes associated information of the changed component. The associated information refers to project information required to change the component. The associated information includes at least one associated data item. The associated information may be presented in the form of a data item. A component refers to a part that makes up an entity. For example, if part 1 on vehicle A needs to be replaced, then based on the need to determine the part model, associated factory, associated vehicle model, process leader, planned time, actual time and other associated information, the component is part 1. By integrating these example information, a component change request can be generated.

[0069] Optionally, the component change request also includes a part text description, and the associated information may only include at least one data item entry and does not contain the data corresponding to the actual data item, so that the recommendation system can output a recommended strategy corresponding to the target change component that meets the data requirements based on the part text description, wherein the data requirement is an integrated data item requirement composed of the association of various data items.

[0070] The method for obtaining the component change request can be that the user generates the component change request through the front-end operation, and the server obtains the component change request from the client. Alternatively, after the client generates the component change request, the client sends the component change request to the designated location where the server stores the request, and the server obtains the component change request from the designated location.

[0071] The implementation method of generating component change requests through user operations on the front end can be to generate component change requests through immediate operations of the user on the front end; or it can be to generate component change requests according to the request generation time interval pre-set by the user on the front end and the associated information of the target change component.

[0072] Step 304: Input the association information into a pre-trained recommendation model to obtain a target task strategy for the target change component, wherein the recommendation model is trained based on sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on the matching relationship between the task strategy keywords in the sample text and each sample association data item, and the sample text is text related to the sample change task.

[0073] Specifically, the recommendation model refers to an algorithm that processes the associated information to obtain the target task strategy for the target change component. The recommendation model can be a machine learning model, for example, the recommendation model is a gradient boosting decision tree (GBDT). The recommendation model can also be a deep learning model, for example, the recommendation model is a deep neural network (DNN). The target task strategy refers to the execution method of component change. The component can be changed according to the target task strategy. The target task strategy corresponds to the associated information of the target change component in the component change request. The target task strategy records the task sub-strategies corresponding to each associated data item in the associated information.

[0074] The label task strategy refers to the benchmark task strategy corresponding to the sample association information. The recommendation model can be trained through the label task strategy and sample association information to obtain a trained recommendation model. The label task strategy is determined by the matching relationship between the task strategy keywords in the sample text and each sample association data item. The label task strategy records the matching relationship between the task strategy keywords in the sample text and each sample association data item, so that each sample association data item has its own matching task strategy keyword, and the sample text is the text related to the sample change task. The sample change task refers to a task with a task strategy that already has a change component. The sample change task is usually a historical task that includes a benchmark answer. The sample change task usually contains sample association information of the sample change component.

[0075] Inputting the associated information into a pre-trained recommendation model to obtain a target task strategy for the target change component can be inputting the associated information into a pre-trained recommendation model so that the recommendation model performs corresponding calculations according to the model parameters obtained in pre-training to obtain the target task strategy for the target change component, wherein the target task strategy includes sub-strategies corresponding to each associated data item.

[0076] The recommendation model is trained based on sample association information and label task strategy of multiple sample change components. The sample association information includes at least one sample association data item. The label task strategy is determined based on the matching relationship between the task strategy keywords in the sample text and each sample association data item. The sample text is text related to the sample change task. Specifically, the label task strategy is constructed based on the sample text. The recommendation model is trained specifically by using the sample association information of multiple sample change components and the text related to the sample change task to construct and train the label task strategy, thereby realizing the simultaneous use of text and data.

[0077] In an optional embodiment of the present application, before the above step of inputting the associated information into the pre-trained recommendation model to obtain the target task strategy for the target change component, the following steps are also included:

[0078] Acquire a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components;

[0079] Extracting sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of the multiple sample change components, and the first label task strategy is the label task strategy of the first sample change component;

[0080] Inputting the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy;

[0081] Based on the prediction task strategy and the first label task strategy, the model parameters of the recommendation model are adjusted, and the step of extracting the sample association information of the first sample change component and the first label task strategy from the sample set is returned to execute until the training stop condition is reached to obtain a recommendation model that has completed training.

[0082] Specifically, a sample set refers to a set that provides a data basis for model training. A sample set usually includes multiple pairs of samples, and a sample pair includes an input and a benchmark output corresponding to the input. For example, a sample set includes sample association information and label task strategies of multiple sample change components. The prediction task strategy refers to the result obtained by the recommendation model based on the sample association information of the input sample change component.

[0083] The sample set may be obtained by receiving a model training request and obtaining the sample set based on the model training request.

[0084] The implementation method of obtaining the sample set based on the model training request can be to parse the model training request to obtain the sample set; it can also be to obtain the sample set from the storage based on the storage location carried in the model training request, where the storage location stores the sample set generated by sample association information based on multiple sample change components and sample text related to multiple sample change tasks.

[0085] The sample association information of the first sample change component is input into the recommendation model to obtain the prediction task strategy. The implementation method can be that the sample association information of the first sample change component is input into the recommendation model, and the recommendation model obtains the prediction task strategy according to a preset algorithm, wherein the prediction task strategy corresponds to the sample association information.

[0086] An implementation method for adjusting the model parameters of a recommendation model based on a prediction task strategy and a first label task strategy, calculating a loss value based on the prediction task strategy and the first label task strategy, and adjusting the model parameters of the recommendation model based on the loss value, wherein adjusting the model parameters of the recommendation model may be adjusting adjustable parameters in the recommendation model, wherein the adjustable parameters may include weight items, bias items, etc.

[0087] Optionally, the training stop condition may include the loss value reaching a loss value threshold, the number of cycles reaching a number threshold, etc.

[0088] By obtaining a sample set including sample association information and label task strategies of multiple sample change components, the parameters of the recommendation model are adjusted to obtain a trained recommendation model, so that the recommendation model can be used to obtain recommendations for target task strategies for target change components.

[0089] In an optional embodiment of the present application, the training process of the recommendation model is specifically carried out using sample association information of multiple sample change components and a label task strategy. The label task strategy is constructed based on the sample text. Therefore, the sample text and the sample association information of multiple sample change components are used as the data basis to train the recommendation model to obtain a trained recommendation model, that is, the above steps of obtaining a sample set include the following steps:

[0090] Acquire sample association information of a plurality of sample change components and sample texts related to a plurality of sample change tasks, wherein the sample association information includes at least one sample association data item;

[0091] Perform keyword extraction on the sample text related to the first sample change task to obtain at least one task strategy keyword corresponding to the first sample change task, wherein the first sample change task is any one of the multiple sample change tasks;

[0092] Matching the sample-related data items of the first sample change component with at least one task strategy keyword corresponding to each sample change task, respectively, to determine the label task strategy of the first sample change component, wherein the first sample change component is any one of the multiple sample change components;

[0093] According to the sample association information and label task strategy of each sample change component, a sample set is obtained by combining.

[0094] The method for obtaining sample association information of multiple sample change components and sample texts related to multiple sample change tasks may be to obtain sample association information of multiple sample change components and sample texts related to multiple sample change tasks from a storage location where pre-stored information is stored.

[0095] The method of performing keyword extraction on the sample text related to the first sample change task to obtain at least one task strategy keyword corresponding to the first sample change task may be to analyze each word in the sample text related to the first sample change task to obtain at least one task strategy keyword corresponding to the first sample change task; or it may be to input the sample text related to the first sample change task into a keyword extraction model to obtain at least one task strategy keyword corresponding to the first sample change task output by the keyword extraction model.

[0096] The sample-associated data items of the first sample change component are matched with at least one task strategy keyword corresponding to each sample change task respectively, and the implementation method of determining the label task strategy of the first sample change component can be to calculate the similarity between the sample-associated data items of the first sample change component and at least one task strategy keyword corresponding to each sample change task respectively, and determine the label task strategy of the first sample change component based on the similarity result.

[0097] According to the sample association information and label task strategy of each sample change component, the implementation method of combining to obtain the sample set can be to establish a correspondence between the sample association information and the label task strategy of each sample change component, and based on the correspondence, combine the sample association information and the label task strategy of each sample change component to obtain the sample set.

[0098] By applying the solution of the embodiment of the present application, a label task strategy is constructed using sample text, so that the recommendation model is trained using the expression task strategy and the sample association information of multiple sample change components as the data basis. The sample text is related to the sample change task, so that the determined label task strategy is more adapted to the sample association information of multiple sample change components, thereby ensuring the accuracy of the recommendation model application.

[0099] In an optional embodiment of the present application, the above step extracts keywords from the sample text related to the first sample change task to obtain at least one task strategy keyword corresponding to the first sample change task, including the following steps:

[0100] Extracting word features of each word in the sample text related to the first sample change task;

[0101] The word features of each word are analyzed to determine at least one task strategy keyword corresponding to the first sample change task.

[0102] Specifically, word features represent the key information of a word and do not contain useless information of the word.

[0103] The implementation method of extracting the word features of each word in the sample text related to the first sample change task can be to segment the sample text related to the first sample change task into words, obtain multiple words, and extract the word features of each word; or it can be to input the sample text related to the first sample change task into a word feature processing model to obtain the word features of each word output by the word feature processing model.

[0104] The implementation method of analyzing the word features of each word and determining at least one task strategy keyword corresponding to the first sample change task can be to cluster the words based on the word features of each word, and determine at least one task strategy keyword corresponding to the first sample change task based on the clustering results; or it can be to analyze the word features of each word to obtain the analysis results of each word feature, judge the association between the words based on the analysis results, and determine the words with the association as the at least one task strategy keyword corresponding to the first sample change task; or it can be to analyze the word features of each word to obtain the analysis results of each word feature, determine the importance of each word in the sample text based on the analysis results, and determine the words whose importance meets the degree threshold as the at least one task strategy keyword corresponding to the first sample change task.

[0105] By applying the solution of the embodiment of the present application, word features of each word in the sample text related to the first sample change task are extracted, so that at least one task strategy keyword corresponding to the first sample change task can be determined from each word through the key information of each word, thereby improving the efficiency of determining at least one task strategy keyword corresponding to the first sample change task, thereby improving the efficiency of constructing the label task strategy and the efficiency of using the label task strategy for model training.

[0106] In an optional embodiment of the present application, the above steps analyze the word features of each word to determine at least one task strategy keyword corresponding to the first sample change task, including the following steps:

[0107] Clustering the words based on the word features of the words to obtain at least one cluster corresponding to the first sample change task;

[0108] At least one task strategy keyword corresponding to the first sample change task is determined based on the at least one cluster.

[0109] There are many ways to cluster the words based on the word features of each word to obtain at least one cluster corresponding to the first sample change task. The specific implementation method is determined according to the actual situation and is not limited in this application.

[0110] In a possible implementation of the present application, based on the word features of each word, each word is clustered to obtain at least one cluster corresponding to the first sample change task. A standard distance may be set, the word features of each word are traversed, each word is divided into at least one class, and at least one cluster corresponding to the first sample change task is obtained.

[0111] On the basis of a possible implementation method, based on at least one clustering cluster, an implementation method of at least one task strategy keyword corresponding to the first sample change task is determined, a class center is determined from at least one clustering cluster, and the class center is determined as at least one task strategy keyword corresponding to the first sample change task; it can also be that a benchmark keyword related to each associated data item is determined from at least one cluster, and the benchmark keyword is determined as at least one task strategy keyword corresponding to a sample change task.

[0112] In another possible implementation of the present application, based on the word features of each word, each word is clustered to obtain at least one cluster corresponding to the first sample change task. It can also be a specified distance threshold and cluster center. Based on the word features of each word, each word is divided into at least one cluster, and then the center point of at least one cluster is recalculated, and then the next word is iteratively assigned until the stopping division condition is reached, and at least one cluster corresponding to the first sample change task is obtained, wherein the stopping division condition can be that the change of the center point of each cluster reaches a threshold, reaches a specified number of iterations, etc.

[0113] On the basis of another possible implementation method, based on at least one clustering cluster, an implementation method of at least one task strategy keyword corresponding to the first sample change task is determined, a class center is determined from at least one clustering cluster, and the class center is determined as at least one task strategy keyword corresponding to the first sample change task.

[0114] By applying the scheme of the embodiment of the present application, each word is clustered based on the word features of each word to obtain at least one clustering cluster corresponding to the first sample change task, and at least one task strategy keyword corresponding to the first sample change task is determined based on the at least one clustering cluster, so that the at least one task strategy corresponding to the first sample change task is obtained by clustering the words and obtaining them from the clustering clusters, that is, by clustering the words and then obtaining at least one task strategy keyword corresponding to the first sample change task from the clustering clusters, the efficiency of determining the task strategy keywords is improved, thereby improving the efficiency of training the model.

[0115] In an optional embodiment of the present application, the above steps match the sample-related data items of the first sample change component with at least one task strategy keyword corresponding to each sample change task, respectively, to determine the label task strategy of the first sample change component, including the following steps:

[0116] Determine data item features of sample-related data items of the first sample change component and word features of at least one task strategy keyword corresponding to each sample change task;

[0117] Calculating a first similarity between the first sample change component and each sample change task according to a data item feature of the sample-related data item of the first sample change component and a word feature of at least one task strategy keyword corresponding to each sample change task;

[0118] Based on at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than a first preset threshold, a label task strategy of the first sample change component is determined.

[0119] An implementation method for determining data item features of sample-associated data items of the first sample change component and word features of at least one task strategy keyword corresponding to each sample change task may be to extract data item features of sample-associated data items of the first sample change component and extract word features of at least one task strategy keyword corresponding to each sample change task.

[0120] Optionally, if the sample-related data items are discrete data items, feature extraction can be performed directly on the discrete data items to obtain data item features. For example, if the sample-related data items are part numbers, car attributes, etc., one-hot encoding can be used directly on the part numbers and car attributes to obtain data item features corresponding to the part numbers and car attributes.

[0121] Optionally, if the sample-associated data item is a continuous data item, the continuous data item is first discretized, and then feature extraction is performed to obtain data item features. For example, if the sample-associated data is time, the time data item is first discretized, and then the discretized data is one-hot encoded to obtain data item features.

[0122] An implementation method for calculating the first similarity between the first sample change component and each sample change task based on data item features of the sample-associated data items of the first sample change component and word features of at least one task strategy keyword corresponding to each sample change task may be to respectively calculate the data item features of the sample-associated data items of the first sample change component and word features of at least one task strategy keyword corresponding to each sample change task to obtain the first similarity between the first sample change component and each sample change task.

[0123] The implementation method of calculating the data item feature of the sample association item of the first sample change component and the word feature of at least one task strategy keyword corresponding to any sample change task to obtain the first similarity between the first sample change component and each sample change task may be to respectively calculate the similarity of the first data item feature of the sample association item of the first sample change component and the word feature of at least one task strategy keyword corresponding to any sample change task, obtain at least one result corresponding to the first data item, select the largest one from among them as the similarity corresponding to the first data item, respectively calculate the similarity of the second data item feature of the sample association item of the first sample change component and the word feature of at least one task strategy keyword corresponding to any sample change task, obtain at least one result corresponding to any data item, select the largest one from them as the similarity of the second association item, the second data item feature is the data item feature of the task data item in the data item other than the first data item in the sample association item, and so on, to obtain at least one similarity for the first sample change component and any sample change task.

[0124] An implementation method for determining the label task strategy of the first sample change component based on at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than the first preset threshold may be to determine at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than the first preset threshold, align at least one task strategy keyword and the sample-related data item of the first sample change component, assemble the at least one aligned task strategy keyword, and generate the label task strategy of the first sample change component.

[0125] Among them, the implementation method of assembling at least one task strategy keyword after alignment can be to assemble the task strategy keywords corresponding to each sample-related data item according to the alignment relationship between each sample-related data item and at least one task strategy keyword in the first sample change component and the data execution logic of each sample-related data item. The assembly can be a simple word splicing, or it can utilize the characteristics of language expression to obtain the assembly result by constructing sentences based on the task strategy keywords.

[0126] By applying the scheme of the embodiment of the present application, based on the data item features of the sample-associated data items of the first sample change component and the word features of at least one task strategy keyword corresponding to each sample change task, a first similarity corresponding to the first sample change component and each sample change task is calculated, and based on the comparison between the first similarity and a first preset threshold, a target sample change task is selected from each sample change task, and based on at least one task strategy word corresponding to the target change task, a label task strategy of the first sample change component is determined, so that the selected label task strategy is selected from at least one task strategy word corresponding to the target change task whose similarity meets the requirements, thereby ensuring the accuracy of the label task strategy selection, and further ensuring the accuracy of the model obtained by model training based on the label task strategy in the application stage.

[0127] In an optional embodiment of the present application, the above step determines the label task strategy of the first sample change component based on at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than the first preset threshold, including the following steps:

[0128] Determine at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than a first preset threshold;

[0129] Performing alignment processing on the at least one task strategy keyword and the sample associated data item of the first sample change component;

[0130] Based on the alignment processing result, a label task strategy of the first sample change component is generated.

[0131] The implementation method for determining at least one task strategy keyword corresponding to a target sample change task whose first similarity is greater than a first preset threshold value may be to compare each first similarity with the first preset threshold value, determine a target similarity whose first similarity is greater than the first preset threshold value, determine a target sample change task corresponding to the target similarity, and determine at least one task strategy keyword corresponding to the target sample change task.

[0132] An implementation method for aligning at least one task strategy keyword and a sample-associated data item of a first sample change component may be to establish a correspondence between at least one task strategy keyword and a sample-associated data item of the first sample change component, and align at least one task strategy keyword and a sample-associated data item of the first sample change component based on the correspondence.

[0133] An implementation method for establishing a correspondence between at least one task strategy keyword and a sample-associated data item of a first sample change component may be to calculate a second similarity between each task strategy keyword and each sample-associated data item based on word features of each task strategy keyword and data item features of each sample-associated data item, and establish a correspondence between a task strategy keyword whose second similarity is greater than a second preset threshold and the data item features of each sample-associated data item.

[0134] Based on the correspondence, an implementation method of aligning at least one task strategy keyword and the sample-related data item of the first sample change component may be to align the task strategy keyword and the sample-related data item whose second similarity is greater than a second preset threshold.

[0135] The implementation method of generating the label task strategy of the first sample change component based on the alignment processing result can be to obtain the label task strategy of the first sample change component based on the alignment processing result and the task strategy keywords corresponding to each associated data item in the first sample change component.

[0136] By applying the scheme of the embodiment of the present application, at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than a first preset threshold is determined; the at least one task strategy keyword and the sample-related data item of the first sample change component are aligned; based on the alignment result, a label task strategy of the first sample change component is generated, and through the correspondence between the first sample change component and the target sample change task, the label task strategy corresponding to the sample-related data item is determined from at least one task strategy keyword of the target sample change task, thereby ensuring the accuracy of the label task strategy determination.

[0137] In an optional embodiment of the present application, the above step aligns at least one task strategy keyword and the sample associated data item of the first sample change component, including the following steps:

[0138] Calculate the second similarity between each task strategy keyword and each sample-related data item according to the word feature of each task strategy keyword and the data item feature of each sample-related data item;

[0139] Align the task strategy keywords and the sample-related data items whose second similarity is greater than a second preset threshold.

[0140] The implementation method of calculating the second similarity between each task strategy keyword and each sample-associated data item based on the word features of each task strategy keyword and the data item features of each sample-associated data item may be to calculate, for any task strategy keyword, the second similarity between the word features of any task strategy keyword and the data item features of each sample-associated data item, respectively, to obtain multiple second similarities for any strategy keyword.

[0141] The implementation method for aligning the task strategy keywords and sample-associated data items whose second similarity is greater than the second preset threshold value can be to determine, for any of the task strategy keywords, the target sample-associated data items whose second similarity is greater than the second preset threshold value, and align each task strategy keyword with the corresponding target sample-associated data item.

[0142] Among them, the implementation method of aligning each task strategy keyword with the corresponding target sample associated data item can be to identify the word feature dimension of each task strategy keyword and the dimension of the data item feature of the target sample associated data item corresponding to each task strategy keyword, and align the word feature dimension and the data item feature dimension.

[0143] The alignment of word feature dimensions and data item feature dimensions can be achieved by setting a baseline feature dimension and converting both the word feature dimension and the data item feature dimension to the baseline feature dimension; or by taking the word feature dimension as a baseline and converting the data item feature dimension to the word feature dimension; or by taking the data item feature dimension as a baseline and converting the word feature dimension to the data item feature dimension.

[0144] By applying the scheme of the embodiment of the present application, the second similarity between each task strategy keyword and each sample-associated data item is calculated based on the word features of each task strategy keyword and the data item features of each sample-associated data item; the task strategy keywords and sample-associated data items whose second similarity is greater than the second preset threshold are aligned, so that the correspondence between the aligned task strategy keywords and each sample-associated data item is obtained through two similarity calculations and two preset threshold screenings, thereby ensuring the accurate correspondence between the aligned task strategy keywords and the sample-associated data items, ensuring the accuracy of the label task strategy of the first sample change component, and further ensuring the accuracy of the model application obtained by model training using the label task strategy.

[0145] Using one or more embodiments of the present application, experiments were conducted on the recommendation model, and the experimental results obtained showed that the average accuracy rate reached 96%, and the average recall rate reached 97%.

[0146] See also Figure 4 , Figure 4 A flowchart of a method for recommending a task strategy for vehicle parts changes provided by an embodiment of the present application is shown, which specifically includes the following steps.

[0147] Step 402: Obtain a vehicle part change request, wherein the vehicle part change request includes change-related information of a target vehicle part.

[0148] The embodiment of the present application is applied to the client and / or server to which the task strategy recommendation method with vehicle parts change belongs. The following description takes the server as an example.

[0149] Specifically, a vehicle part change request refers to a request to change a part in a vehicle. For example, if a user replaces part 1 of vehicle A and wants to obtain a task strategy for the part, a vehicle part change request is generated based on information such as the part model, associated factory, associated vehicle model, process manager, planned time, and actual node time corresponding to the part replacement, so that the server obtains the vehicle part change request.

[0150] Step 404: Input the change association information into a pre-trained recommendation model to obtain a target task strategy for the target vehicle part, wherein the recommendation model is trained based on sample association information and label task strategy of multiple sample vehicle parts, the sample association information includes at least one sample association data item, the label task strategy is determined based on the matching relationship between the task strategy keywords in the sample text and each sample association data item, and the sample text is text related to the sample change task.

[0151] Exemplarily, the change-related information includes part model, associated factory, associated vehicle model, process manager and planned time. The change-related information is input into a pre-trained recommendation model to obtain a target task strategy for the target vehicle part. It can be that for the part with part model F on the associated vehicle model E, according to the planned time G, go to the associated factory H and find the corresponding process manager I to replace the part.

[0152] Step 406: Based on the target task strategy, recommend executing the target vehicle part change task.

[0153] Specifically, based on the target task strategy, the implementation method of recommending the execution of the target vehicle parts change task can be to determine the target task sub-strategy corresponding to each associated data item in the associated information based on the target task strategy, and recommend the execution of the target vehicle parts change task according to the association relationship between each associated data item and the target task sub-strategy.

[0154] The technical solution of the task strategy recommendation method corresponding to the vehicle parts change in steps 402 to 404 is the same as the above Figure 3 The technical solutions corresponding to the method for recommending task strategies for component changes belong to the same concept. For details not described in detail in the technical solutions of the method for recommending task strategies for vehicle parts changes, please refer to the description of the technical solutions of the method for recommending task strategies for component changes mentioned above.

[0155] See also Figure 5 , Figure 5 A flowchart of a training method for a recommendation model provided by an embodiment of the present application is shown, which specifically includes the following steps.

[0156] Step 502: Obtain a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on the matching relationship between the task strategy keywords in the sample text and each sample association data item, and the sample text is text related to the sample change task.

[0157] Step 504: extracting sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of a plurality of sample change components, and the first label task strategy is the label task strategy of the first sample change component.

[0158] Step 506: Input the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy.

[0159] Step 508: Based on the prediction task strategy and the first label task strategy, adjust the model parameters of the recommendation model, and return to execute the step of extracting the sample association information of the first sample change component and the first label task strategy from the sample set until the training stop condition is reached to obtain a recommendation model that has completed training.

[0160] The technical solution of the training method of the recommendation model corresponding to the above steps 502 to 508 is the same as the above Figure 3 The technical solution for training the recommendation model in the above-mentioned Figure 3 Description of the technical solution for the task strategy recommendation method for component changes.

[0161] The following combination Figure 6 Taking the application of the task strategy recommendation method for component changes provided by the present application in vehicle parts changes as an example, the task strategy recommendation method for component changes is further described. Figure 6A process flow chart of a task strategy recommendation method for component change provided by an embodiment of the present application is shown, which specifically includes the following steps.

[0162] Sample data items are acquired through step 602, and sample text is acquired through step 306. The specific process is as follows:

[0163] Step 602: Obtaining sample data items: Obtain sample association information of multiple historical sample change components, wherein the sample association information includes at least one sample data association item, such as the sample data association item is part model, associated factory, associated vehicle model, process manager, planned time and actual node time, etc.

[0164] Step 604: extract features from sample data items: extract features from each sample data associated item in the sample associated information to obtain data item features of each sample data item.

[0165] Step 606: Sample text acquisition: acquiring sample texts related to multiple sample change tasks.

[0166] The two branch steps corresponding to the following steps 608 and 616 are respectively executed for the sample text.

[0167] Step 608: extract keywords: extract at least one word in the sample text related to the first sample change task, wherein the first sample change task is any one of the multiple sample change tasks.

[0168] Step 610: Extract word features: extract word features of each keyword in the sample text related to the first sample change task.

[0169] Step 612: Clustering: for at least one sample change task, cluster each keyword based on the word features of each keyword in each sample change task to obtain at least one cluster corresponding to each sample change task.

[0170] Step 614: Generate a tag candidate set: take each cluster as a tag candidate set, and take the center of each cluster as a recommendation candidate word.

[0171] Step 616: Extracting sample text features: extracting text features corresponding to the sample text.

[0172] For the tag candidate set and sample text features generated in steps 614 and 616 , the following step 618 is performed.

[0173] Step 618: Determine the task strategy keywords: Calculate the similarity between the text features and the word features of each keyword respectively, and determine the task strategy keywords corresponding to the sample texts of each sample change task based on the similarity calculation results.

[0174] Step 620: Data and text alignment: The first step is to determine the target change task of each sample change component; the second step is to align the sample-related data items of each sample change component with at least one task strategy keyword corresponding to the corresponding target change task.

[0175] The first step specifically includes: determining the data item features of the sample-related data items of the first sample change component and the word features of at least one task strategy keyword corresponding to each sample change task; calculating the first similarity between the first sample change component and each sample change task based on the data item features of the sample-related data items of the first sample change component and the word features of at least one task strategy keyword corresponding to each sample change task, and determining the target sample change task whose first similarity is greater than a first preset threshold. The second step specifically includes: aligning at least one task strategy keyword and the sample-related data items of the first sample change component; generating a label task strategy for the first sample change component based on the alignment processing result.

[0176] Step 622: Model training: The recommendation model is trained using the sample association information of multiple sample change components and the label task strategy of each sample change component to obtain a trained recommendation model.

[0177] By applying the scheme of the embodiment of the present application, a recommendation model is obtained by training sample association information and label task strategies of multiple sample change components. The label task strategy is obtained by matching task strategy keywords in the sample text with at least one sample association data item, so that the label task strategy corresponds to the task strategy keywords that match the at least one association data item, thereby achieving the accuracy of the corresponding label task strategy to the sample association information of the sample change component, thereby ensuring the accuracy of the output results of the recommendation model, that is, by inputting the association information of the target change component in the component change request into the recommendation model with high accuracy, an accurate target task strategy for the target change component is obtained.

[0178] Corresponding to the above method embodiment, the present application also provides a task strategy recommendation device embodiment for component changes, Figure 7 A schematic diagram of the structure of a task strategy recommendation device for component changes provided by an embodiment of the present application is shown.

[0179] like Figure 7 As shown, the device comprises:

[0180] A first request acquisition module 702 is configured to acquire a component change request, wherein the component change request includes associated information of a target change component;

[0181] The first acquisition module 704 is configured to input the association information into a pre-trained recommendation model to obtain a target task strategy for the target change component, wherein the recommendation model is trained based on sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on the matching relationship between the task strategy keywords in the sample text and each sample association data item, and the sample text is text related to the sample change task.

[0182] Optionally, the task strategy recommendation device for component change also includes a training module, which is configured to obtain a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components; extract sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of the multiple sample change components, and the first label task strategy is the label task strategy of the first sample change component; input the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy; based on the prediction task strategy and the first label task strategy, adjust the model parameters of the recommendation model, and return to execute the step of extracting the sample association information and the first label task strategy of the first sample change component from the sample set until the training stop condition is reached to obtain a recommendation model that has completed training.

[0183] Optionally, the training module is further configured to obtain sample association information of multiple sample change components and sample texts related to multiple sample change tasks, wherein the sample association information includes at least one sample association data item; perform keyword extraction on the sample text related to the first sample change task to obtain at least one task strategy keyword corresponding to the first sample change task, wherein the first sample change task is any one of the multiple sample change tasks; match the sample association data items of the first sample change component with at least one task strategy keyword corresponding to each sample change task, respectively, to determine the label task strategy of the first sample change component, wherein the first sample change component is any one of the multiple sample change components; and combine the sample association information and label task strategy of each sample change component to obtain a sample set.

[0184] Optionally, the training module is further configured to extract word features of each word in the sample text related to the first sample change task; analyze the word features of each word to determine at least one task strategy keyword corresponding to the first sample change task.

[0185] Optionally, the training module is further configured to cluster each word based on word features of each word to obtain at least one cluster corresponding to the first sample change task; based on at least one cluster, determine at least one task strategy keyword corresponding to the first sample change task.

[0186] Optionally, the training module is further configured to determine data item features of the sample-associated data items of the first sample change component and word features of at least one task strategy keyword corresponding to each sample change task; calculate a first similarity between the first sample change component and each sample change task based on the data item features of the sample-associated data items of the first sample change component and the word features of at least one task strategy keyword corresponding to each sample change task; and determine a label task strategy for the first sample change component based on at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than a first preset threshold.

[0187] Optionally, the training module is further configured to determine at least one task strategy keyword corresponding to a target sample change task whose first similarity is greater than a first preset threshold; align at least one task strategy keyword and a sample-associated data item of a first sample change component; and generate a label task strategy for the first sample change component based on the alignment result.

[0188] Optionally, the training module is further configured to calculate the second similarity between each task strategy keyword and each sample-associated data item based on the word features of each task strategy keyword and the data item features of each sample-associated data item; and align the task strategy keywords and sample-associated data items whose second similarity is greater than a second preset threshold.

[0189] By applying the scheme of the embodiment of the present application, a recommendation model is obtained by training sample association information and label task strategies of multiple sample change components. The label task strategy is obtained by matching task strategy keywords in the sample text with at least one sample association data item, so that the label task strategy corresponds to the task strategy keywords that match the at least one association data item, thereby achieving the accuracy of the corresponding label task strategy to the sample association information of the sample change component, thereby ensuring the accuracy of the output results of the recommendation model, that is, by inputting the association information of the target change component in the component change request into the recommendation model with high accuracy, an accurate target task strategy for the target change component is obtained.

[0190] The above is a schematic scheme of a task strategy recommendation device for component changes in this embodiment. It should be noted that the technical scheme of the task strategy recommendation device for component changes and the technical scheme of the task strategy recommendation method for component changes mentioned above belong to the same concept. For details not described in detail in the technical scheme of the task strategy recommendation device for component changes, please refer to the description of the technical scheme of the task strategy recommendation method for component changes mentioned above.

[0191] Corresponding to the above method embodiment, the present application also provides an embodiment of a task strategy recommendation device for vehicle parts change, Figure 8A schematic structural diagram of a task strategy recommendation device for vehicle part change provided by an embodiment of the present application is shown.

[0192] like Figure 8 As shown, the device comprises:

[0193] The second request acquisition module 802 is configured to acquire a vehicle part change request, wherein the vehicle part change request includes change association information of a target vehicle part;

[0194] The second obtaining module 804 is configured to input the change association information into a pre-trained recommendation model to obtain a target task strategy for a target vehicle part, wherein the recommendation model is trained based on sample association information and a label task strategy of a plurality of sample vehicle parts, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task;

[0195] The recommendation module 806 is configured to recommend executing the target vehicle part change task based on the target task strategy.

[0196] The above is a schematic scheme of a task strategy recommendation device for vehicle part change of this embodiment. It should be noted that the technical scheme of the task strategy recommendation device for vehicle part change and the technical scheme of the task strategy recommendation method for vehicle part change belong to the same concept, and the details not described in detail in the technical scheme of the task strategy recommendation device for vehicle part change can be referred to the description of the technical scheme of the task strategy recommendation method for vehicle part change.

[0197] Corresponding to the above method embodiment, the present application also provides a training device embodiment of a recommendation model, Fig. 9 FIG. 1 is a schematic diagram showing a structure of a training device for a recommendation model provided by an embodiment of the present application. Fig. 9 As shown, the device comprises:

[0198] The sample set acquisition module 902 is configured to acquire a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on the matching relationship between the task strategy keywords in the sample text and each sample association data item, and the sample text is text related to the sample change task;

[0199] An extraction module 904 is configured to extract sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of the multiple sample change components, and the first label task strategy is the label task strategy of the first sample change component;

[0200] An input module 906 is configured to input the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy;

[0201] The adjustment module 908 is configured to adjust the model parameters of the recommendation model based on the prediction task strategy and the first label task strategy, and return to execute the step of extracting the sample association information of the first sample change component and the first label task strategy from the sample set until the training stop condition is reached to obtain a recommendation model that has completed training.

[0202] The above is a schematic scheme of a training device for a recommendation model of this embodiment. It should be noted that the technical scheme of the training device for the recommendation model and the technical scheme of the training method for the recommendation model described above belong to the same concept, and the details not described in detail in the technical scheme of the training device for the recommendation model can be found in the description of the technical scheme of the training method for the recommendation model described above.

[0203] Fig.10 The block diagram of a computing device provided by an embodiment of the present application is shown. The components of the computing device 1000 include but are not limited to a memory 1010 and a processor 1020. The processor 1020 is connected to the memory 1010 via a bus 1030, and the database 1050 is used to store data.

[0204] The computing device 1000 also includes an access device 1040 that enables the computing device 1000 to communicate via one or more networks 1060. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 1040 may include one or more of any type of network interface (e.g., a network interface card (NIC)) of wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a world-wide interoperability for microwave access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, and a near field communication (NFC).

[0205] In one embodiment of the present application, the above components of the computing device 1000 and Fig.10 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Fig.10 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of the present application. Those skilled in the art may add or replace other components as needed.

[0206] The computing device 1000 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 1000 may also be a mobile or stationary server.

[0207] Among them, the processor 1020 is used to execute the following computer executable instructions, which, when executed by the processor, implement the steps of the above-mentioned task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model.

[0208] The above is a schematic scheme of a computing device of this embodiment. It should be noted that the technical scheme of the computing device and the technical schemes of the task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method for the recommendation model are of the same concept. For details not described in detail in the technical scheme of the computing device, please refer to the description of the technical schemes of the task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method for the recommendation model.

[0209] An embodiment of the present application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the above-mentioned task strategy recommendation method for component changes, the task strategy recommendation method for vehicle part changes, and the training method of the recommendation model.

[0210] The above is a schematic scheme of a computer-readable storage medium of this embodiment. It should be noted that the technical scheme of the storage medium and the technical scheme of the task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model are of the same concept. For details not described in detail in the technical scheme of the storage medium, please refer to the description of the technical scheme of the task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model.

[0211] An embodiment of the present application also provides a computer program, wherein when the computer program is executed in a computer, the computer is caused to execute the steps of the above-mentioned task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model.

[0212] The above is a schematic scheme of a computer program of this embodiment. It should be noted that the technical scheme of the computer program and the technical schemes of the task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model are of the same concept. For details not described in detail in the technical scheme of the computer program, please refer to the description of the technical schemes of the task strategy recommendation method for component changes, the task strategy recommendation method for vehicle parts changes, and the training method of the recommendation model.

[0213] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0214] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0215] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present application are not limited by the described action sequence, because according to the embodiments of the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of the present application.

[0216] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0217] The preferred embodiments of the present application disclosed above are only used to help explain the present application. The optional embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of the present application. The present application selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of the present application, so that those skilled in the art can understand and use the present application well.

Claims

1. A task strategy recommendation method for component changes, characterized in that: include: Obtaining a component change request, wherein the component change request includes associated information of a target change component; The association information is input into a pre-trained recommendation model to obtain a target task strategy for the target change component, wherein the recommendation model is trained based on sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between task strategy keywords in a sample text and each sample association data item, and the sample text is text related to the sample change task.

2. The method according to claim 1, characterized in that Before inputting the associated information into a pre-trained recommendation model to obtain a target task strategy for the target change component, the method further includes: Acquire a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components; Extracting sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of the multiple sample change components, and the first label task strategy is the label task strategy of the first sample change component; Inputting the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy; Based on the prediction task strategy and the first label task strategy, the model parameters of the recommendation model are adjusted, and the step of extracting the sample association information of the first sample change component and the first label task strategy from the sample set is returned to execute until the training stop condition is reached to obtain a recommendation model that has completed training.

3. The method according to claim 2, characterized in that The obtaining of the sample set comprises: Acquire sample association information of a plurality of sample change components and sample texts related to a plurality of sample change tasks, wherein the sample association information includes at least one sample association data item; Perform keyword extraction on a sample text related to a first sample change task to obtain at least one task strategy keyword corresponding to the first sample change task, wherein the first sample change task is any one of the multiple sample change tasks; Matching the sample-related data items of the first sample change component with at least one task strategy keyword corresponding to each sample change task, respectively, to determine the label task strategy of the first sample change component, wherein the first sample change component is any one of the multiple sample change components; According to the sample association information and label task strategy of each sample change component, a sample set is obtained by combining.

4. The method according to claim 3, characterized in that The step of extracting keywords from the sample text related to the first sample change task to obtain at least one task strategy keyword corresponding to the first sample change task includes: Extracting word features of each word in the sample text related to the first sample change task; The word features of the words are analyzed to determine at least one task strategy keyword corresponding to the first sample change task.

5. The method according to claim 4, characterized in that The analyzing the word features of each of the words to determine at least one task strategy keyword corresponding to the first sample change task includes: Clustering the words based on the word features of the words to obtain at least one cluster corresponding to the first sample change task; At least one task strategy keyword corresponding to the first sample change task is determined based on the at least one cluster.

6. The method according to claim 3, characterized in that The step of matching the sample-related data items of the first sample change component with at least one task strategy keyword corresponding to each sample change task to determine the label task strategy of the first sample change component includes: Determine data item features of sample-related data items of the first sample change component and word features of at least one task strategy keyword corresponding to each sample change task; Calculating a first similarity between the first sample change component and each sample change task according to a data item feature of the sample-related data item of the first sample change component and a word feature of at least one task strategy keyword corresponding to each sample change task; Based on at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than a first preset threshold, a label task strategy of the first sample change component is determined.

7. The method according to claim 6, characterized in that The determining of the label task strategy of the first sample change component based on at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than a first preset threshold comprises: Determine at least one task strategy keyword corresponding to the target sample change task whose first similarity is greater than a first preset threshold; Performing alignment processing on the at least one task strategy keyword and the sample associated data item of the first sample change component; Based on the alignment processing result, a label task strategy of the first sample change component is generated.

8. The method according to claim 7, characterized in that The aligning process of the at least one task strategy keyword and the sample associated data item of the first sample change component includes: Calculating the second similarity between each task strategy keyword and each sample-related data item according to the word feature of each task strategy keyword and the data item feature of each sample-related data item; Align the task strategy keywords and the sample-related data items whose second similarity is greater than a second preset threshold.

9. A task strategy recommendation method for vehicle parts change, characterized in that: include: Acquire a vehicle part change request, wherein the vehicle part change request includes change-related information of a target vehicle part; Inputting the change association information into a pre-trained recommendation model to obtain a target task strategy for the target vehicle part, wherein the recommendation model is trained based on sample association information and label task strategy of multiple sample vehicle parts, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task; Based on the target task strategy, it is recommended to execute the target vehicle part change task.

10. A training method for a recommendation model, characterized in that: include: Acquire a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task; Extracting sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of the multiple sample change components, and the first label task strategy is the label task strategy of the first sample change component; Inputting the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy; Based on the prediction task strategy and the first label task strategy, the model parameters of the recommendation model are adjusted, and the step of extracting the sample association information of the first sample change component and the first label task strategy from the sample set is returned to execute until the training stop condition is reached to obtain a recommendation model that has completed training.

11. A task strategy recommendation device for component changes, characterized in that: include: A first request acquisition module is configured to acquire a component change request, wherein the component change request includes associated information of a target change component; The first acquisition module is configured to input the association information into a pre-trained recommendation model to obtain a target task strategy for the target change component, wherein the recommendation model is trained based on sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between task strategy keywords in a sample text and each sample association data item, and the sample text is text related to the sample change task.

12. A task strategy recommendation device for vehicle parts change, characterized in that: include: A second request acquisition module is configured to acquire a vehicle part change request, wherein the vehicle part change request includes change association information of a target vehicle part; a second obtaining module, configured to input the change association information into a pre-trained recommendation model to obtain a target task strategy for the target vehicle part, wherein the recommendation model is trained based on sample association information and label task strategy of multiple sample vehicle parts, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task; The recommendation module is configured to recommend executing a target vehicle part change task based on the target task strategy.

13. A training device for a recommendation model, characterized in that: include: A sample set acquisition module is configured to acquire a sample set, wherein the sample set includes sample association information and label task strategies of multiple sample change components, the sample association information includes at least one sample association data item, the label task strategy is determined based on a matching relationship between a task strategy keyword in a sample text and each sample association data item, and the sample text is text related to the sample change task; an extraction module, configured to extract sample association information and a first label task strategy of a first sample change component from the sample set, wherein the first sample change component is any one of the multiple sample change components, and the first label task strategy is the label task strategy of the first sample change component; An input module, configured to input the sample association information of the first sample change component into the recommendation model to obtain a prediction task strategy; The adjustment module is configured to adjust the model parameters of the recommendation model based on the prediction task strategy and the first label task strategy, and return to execute the step of extracting the sample association information of the first sample change component and the first label task strategy from the sample set until the training stop condition is reached to obtain a recommendation model that has completed training.

14. A computing device, characterized in that include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method according to any one of claims 1 to 10 are implemented.

15. A computer-readable storage medium, characterized in that: It stores computer executable instructions, which, when executed by a processor, implement the steps of the method described in any one of claims 1 to 10.

Images