Model acquisition method, device, electronic device, storage medium and program product

Abstract

The present application provides a model acquisition method, device, electronic device, storage medium and program product. The method includes: acquiring sample text data sent by a terminal device; processing the sample text data through an initial classification model to obtain an actual processing result; the actual processing result includes at least one actual text entity and the type of each actual text entity; sending the actual processing result to the terminal device, and receiving an adjusted processing result sent by the terminal device after adjusting the actual processing result; training the initial classification model according to the sample text data and the adjusted processing result until a target classification model is obtained through training. The method for training a model in the present application can improve the processing capability of the target classification model, thereby improving the accuracy of the text entities and the types of text entities obtained by processing using the target classification model.

Description

Technical Field

[0001] The present application relates to the field of model training, and in particular to a model acquisition method, device, electronic device, storage medium and program product. Background Art

[0002] With the rapid development of information technology, the amount of text data has exploded. In order to efficiently perform text annotation work, the target classification model can be used to perform text entity recognition and text entity classification on text data.

[0003] In the related art, a neural network model is usually trained directly based on pre-acquired historical text data and processing results corresponding to the historical text data to obtain a target classification model.

[0004] However, the target classification model in the related technology is obtained by training the neural network model based on the processing results obtained from historical text data and manual annotation. There may be a problem of low processing ability of the target classification model, which may lead to inaccurate text entities and text entity types obtained by using the target classification model. Summary of the invention

[0005] The present application provides a model acquisition method, device, electronic device, storage medium and program product to solve the problem of low target classification model processing capability in the model acquisition method in the related art.

[0006] In a first aspect, the present application provides a model acquisition method, comprising:

[0007] Obtain sample text data sent by the terminal device;

[0008] Processing the sample text data through the initial classification model to obtain an actual processing result; the actual processing result includes at least one actual text entity and the type of each actual text entity;

[0009] Sending the actual processing result to the terminal device, and receiving the adjusted processing result sent by the terminal device after adjusting the actual processing result;

[0010] The initial classification model is trained based on the sample text data and the adjustment processing results until the target classification model is obtained through training.

[0011] In this solution, the electronic device obtains sample text data sent by the terminal device. The electronic device can process the sample text data through the initial classification model to obtain an actual processing result (including at least one actual text entity and the type of each actual text entity). The electronic device can send the actual processing result to the terminal device, and receive the adjustment processing result sent by the terminal device after adjusting the actual processing result. The electronic device can train the initial classification model according to the sample text data and the adjustment processing result until the target classification model is obtained through training. By the above-mentioned terminal device continuously adjusting the actual processing result, the electronic device continuously iteratively trains the initial classification model based on the adjustment processing result, so as to achieve effective training of the initial classification model. The processing capability of the target classification model is improved, thereby improving the accuracy of the text entities and the types of text entities obtained by processing using the target classification model.

[0012] In one implementation, the initial classification model is trained according to the sample text data and the adjustment processing result until a target classification model is obtained through training, including:

[0013] According to the i-th sample text data and the i-th adjustment processing result, the i-th initial classification model is trained to obtain a trained i-th initial classification model;

[0014] Determine whether the processing accuracy of the trained i-th initial classification model reaches a first preset accuracy;

[0015] If yes, the trained i-th initial classification model is determined as the target classification model;

[0016] If not, the trained i-th initial classification model is used as the i+1-th initial classification model, and the i+1-th sample text data sent by the terminal device is processed according to the i+1-th initial classification model to obtain the i+1-th actual processing result, and the i+1-th actual processing result is sent to the terminal device, and the i+1-th adjusted processing result sent by the terminal device after adjusting the i+1-th actual processing result is received;

[0017] Among them, i is 1, 2, ... in sequence until the processing accuracy of the trained i-th initial classification model reaches the first preset accuracy; the first sample text data is the sample text data, the first adjustment processing result is the adjustment processing result; the first initial classification model is the initial classification model.

[0018] In this solution, in this embodiment, during each iterative training process, the ith actual processing result is continuously adjusted, and then the ith adjustment processing result and the ith sample text data are used to train the ith initial classification model to obtain the trained ith initial classification model. In this way, the model training effect is improved, thereby improving the processing capacity of the target classification model, and further improving the accuracy of obtaining text entities and text entity types by processing using the target classification model.

[0019] In one implementation, the method further includes:

[0020] Acquire the mth training data sent by the terminal device; wherein the mth training data includes the mth training text data and the mth training processing result;

[0021] According to the mth training data, the mth neural network model is trained to obtain a trained mth neural network model;

[0022] Determine whether the trained mth neural network model meets the preset conditions;

[0023] If yes, the trained mth neural network model is determined as the initial classification model;

[0024] If not, the trained mth neural network model is used as the m+1th neural network model;

[0025] Among them, m is 1, 2, ... in sequence until the trained mth neural network model meets the preset conditions.

[0026] In this solution, in this embodiment, before iteratively training the initial classification model, an iterative method can be used to train the mth neural network model using the mth training data (the mth training text data and the mth training processing result) to obtain the initial classification model. In this way, the initial classification model can be quickly and accurately obtained, thereby improving the speed and accuracy of obtaining the target classification model.

[0027] In one implementation, determining whether the trained mth neural network model meets a preset condition includes:

[0028] Obtain the processing accuracy of the trained mth neural network model;

[0029] Determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy;

[0030] If yes, it is determined that the trained mth neural network model meets the preset conditions;

[0031] If not, it is determined that the trained mth neural network model does not meet the preset conditions.

[0032] In this solution, the electronic device can determine whether the trained mth neural network model can be used as the initial classification model by comparing the processing accuracy of the trained mth neural network model with the second preset accuracy. In this way, the initial classification model can be quickly and accurately acquired, thereby improving the speed and accuracy of acquiring the target classification model.

[0033] In one implementation, determining whether the trained mth neural network model meets a preset condition includes:

[0034] Obtain the processing accuracy of the trained mth neural network model;

[0035] Determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy and whether new training data sent by the terminal device is not obtained;

[0036] If yes, it is determined that the trained mth neural network model meets the preset conditions;

[0037] If not, it is determined that the trained mth neural network model does not meet the preset conditions.

[0038] In this solution, the electronic device can determine whether the trained mth neural network model can be used as the initial classification model by comparing the processing accuracy of the trained mth neural network model with the second preset accuracy, and determining whether the electronic device has obtained new training data. On the one hand, the initial classification model can be quickly and accurately acquired, thereby improving the rate and accuracy of acquiring the target classification model. On the other hand, the neural network model can be trained in time when new training data arrives, thereby achieving real-time training, that is, achieving continuous optimization of the initial classification model, thereby achieving continuous optimization of the target classification model.

[0039] In one implementation, the method further includes:

[0040] Obtaining text data to be processed sent by the terminal device;

[0041] According to the target classification model, the text data to be processed is processed to obtain a target processing result; the target processing result includes at least one target text entity and the type of each target text entity;

[0042] Performing screening processing on the target processing result to obtain the screening processing result;

[0043] Sending the screening result to the terminal device, and receiving the acceptance result sent by the terminal device, which is obtained by performing acceptance processing on the screening result;

[0044] When the acceptance result indicates that the acceptance is passed, the target processing result is displayed.

[0045] In this scheme, in this embodiment, after the model training is completed, that is, after the electronic device obtains the target classification model, the electronic device can obtain the text data to be processed sent by the terminal device. The electronic device can process the text data to be processed according to the target classification model to obtain the target processing result (including at least one target text entity and the type of each target text entity). The electronic device can screen the target processing result to obtain the screening processing result, and send the screening processing result to the terminal device, and receive the acceptance result sent by the terminal device for the acceptance processing of the screening processing result. The electronic device can display the target processing result when the acceptance result indicates that the acceptance is passed. In the above manner, the target processing result output by the target classification model can be sent to the terminal device for acceptance by the terminal device, thereby improving the accuracy of the target processing result.

[0046] In a second aspect, the present application provides a model acquisition device, comprising:

[0047] The transceiver module is used to obtain the sample text data sent by the terminal device;

[0048] A processing module, used to process the sample text data through the initial classification model to obtain an actual processing result; the actual processing result includes at least one actual text entity and the type of each actual text entity;

[0049] The transceiver module is further used to send the actual processing result to the terminal device, and receive the adjusted processing result sent by the terminal device after adjusting the actual processing result;

[0050] The processing module is also used to train the initial classification model according to the sample text data and the adjustment processing results until the target classification model is obtained through training.

[0051] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0052] In one implementation, the processing module is specifically configured to:

[0053] According to the i-th sample text data and the i-th adjustment processing result, the i-th initial classification model is trained to obtain a trained i-th initial classification model;

[0054] Determine whether the processing accuracy of the trained i-th initial classification model reaches a first preset accuracy;

[0055] If yes, the trained i-th initial classification model is determined as the target classification model;

[0056] If not, the trained i-th initial classification model is used as the i+1-th initial classification model, and the i+1-th sample text data sent by the terminal device is processed according to the i+1-th initial classification model to obtain the i+1-th actual processing result, and the i+1-th actual processing result is sent to the terminal device, and the i+1-th adjusted processing result sent by the terminal device after adjusting the i+1-th actual processing result is received;

[0057] Among them, i is 1, 2,... in sequence, until the processing accuracy of the trained i-th initial classification model reaches the first preset accuracy; the first sample text data is the sample text data; the first adjustment processing result is the adjustment processing result; the first initial classification model is the initial classification model.

[0058] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0059] In one implementation,

[0060] The transceiver module is further used to obtain the mth training data sent by the terminal device; wherein the mth training data includes the mth training text data and the mth training processing result;

[0061] The processing module is further used to perform training processing on the mth neural network model according to the mth training data to obtain a trained mth neural network model;

[0062] The processing module is also used to determine whether the trained mth neural network model meets the preset conditions;

[0063] The processing module is further used for determining the trained mth neural network model as the initial classification model if yes;

[0064] The processing module is further configured to, if not, use the trained mth neural network model as the m+1th neural network model;

[0065] Among them, m is 1, 2, ... in sequence until the trained mth neural network model meets the preset conditions.

[0066] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0067] In one implementation, the processing module is specifically configured to:

[0068] Obtain the processing accuracy of the trained mth neural network model;

[0069] Determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy;

[0070] If yes, it is determined that the trained mth neural network model meets the preset conditions;

[0071] If not, it is determined that the trained mth neural network model does not meet the preset conditions.

[0072] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0073] In one implementation, the processing module is specifically configured to:

[0074] Obtain the processing accuracy of the trained mth neural network model;

[0075] Determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy and whether new training data sent by the terminal device is not obtained;

[0076] If yes, it is determined that the trained mth neural network model meets the preset conditions;

[0077] If not, it is determined that the trained mth neural network model does not meet the preset conditions.

[0078] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0079] In one implementation,

[0080] The transceiver module is also used to obtain the text data to be processed sent by the terminal device;

[0081] The processing module is further used to process the text data to be processed according to the target classification model to obtain a target processing result; the target processing result includes at least one target text entity and the type of each target text entity;

[0082] The processing module is also used to screen the target processing results to obtain the screening processing results;

[0083] The transceiver module is also used to send the screening process result to the terminal device, and receive the acceptance result sent by the terminal device, which is obtained by performing acceptance processing on the screening process result;

[0084] The processing module is also used to display the target processing result when the acceptance result indicates that the acceptance is passed.

[0085] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0086] In a third aspect, the present application provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0087] Memory stores computer-executable instructions;

[0088] The processor executes the computer-executable instructions stored in the memory to implement the method according to the first aspect.

[0089] The electronic device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0090] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method of the first aspect.

[0091] When the computer-executable instructions in the computer-readable storage medium provided in the embodiment of the present application are executed by the processor, the technical solution shown in the above method embodiment can be implemented. The implementation principle and beneficial effects are similar and will not be repeated here.

[0092] In a fifth aspect, the present application provides a computer program product, comprising a computer program, which implements the method of the first aspect when executed by a processor.

[0093] When the computer program in the computer program product provided in the embodiment of the present application is executed by a processor, the technical solution shown in the above method embodiment can be implemented. The implementation principle and beneficial effects are similar and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0095] Figure 1 A schematic diagram of a scenario of a model acquisition method provided in an embodiment of the present application;

[0096] Figure 2 A schematic diagram of a flow chart of a first embodiment of a model acquisition method provided in an embodiment of the present application;

[0097] Figure 3a A schematic diagram of a flow chart of a second embodiment of a model acquisition method provided in an embodiment of the present application;

[0098] Figure 3b A schematic diagram of obtaining training text entities and types of training text entities provided in an embodiment of the present application;

[0099] Figure 3c A schematic diagram of displaying sub-training text data provided in an embodiment of the present application;

[0100] Figure 4 A schematic diagram of a flow chart of a third embodiment of a model acquisition method provided in an embodiment of the present application;

[0101] Figure 5 A schematic diagram of a flow chart of a fourth embodiment of a model acquisition method provided in an embodiment of the present application;

[0102] Figure 6 A schematic diagram of the structure of a model acquisition device provided in an embodiment of the present application;

[0103] Figure 7 A structural diagram of an electronic device provided in this application.

[0104] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0105] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0106] It should be noted that the model acquisition method, device, equipment, medium and program product of the present application can be used in the field of model training technology, and can also be used in any field other than model training. The present application does not limit the application field of the model acquisition method, device, equipment, medium and program product.

[0107] Glossary:

[0108] Natural language processing: refers to the field of computer science and artificial intelligence that enables electronic devices to understand, interpret and generate human language.

[0109] Real-time training: refers to the process by which the model can learn and adjust based on new data instantly.

[0110] Text entity recognition: A task in natural language processing that aims to identify entities with specific meanings from text data, such as names of people, places, organizations, etc.

[0111] Annotation: In text data processing, annotation refers to the process of identifying specific elements in text data and classifying them.

[0112] With the rapid development of information technology, the amount of text data has exploded. In order to efficiently perform text annotation work, the target classification model can be used to perform text entity recognition and entity classification on text data.

[0113] In the related art, an initial classification model is usually trained directly based on pre-acquired historical text data and processing results corresponding to the historical text data to obtain a target classification model.

[0114] However, the target classification model in the related technology is obtained by training the initial classification model based on the processing results obtained from historical text data and manual annotation. There may be a problem of low processing ability of the target classification model, which may lead to inaccurate entities and entity types obtained by using the target classification model.

[0115] Based on the above technical problems, the technical concept of the embodiments of the present application is as follows: the electronic device obtains sample text data sent by the terminal device. The electronic device can process the sample text data through the initial classification model to obtain an actual processing result (including at least one actual text entity and the type of each actual text entity). The electronic device can send the actual processing result to the terminal device, and receive the adjustment processing result sent by the terminal device after adjusting the actual processing result. The electronic device can train the initial classification model according to the sample text data and the adjustment processing result until the target classification model is obtained through training.

[0116] Through the above-mentioned model training method, the processing capability of the acquired target classification model can be improved, thereby improving the accuracy of text entities and types of text entities obtained by processing using the target classification model.

[0117] Below, a model acquisition method provided in an embodiment of the present application is described in detail.

[0118] To facilitate understanding, first combine Figure 1 The application scenarios involved in the embodiments of the present application are described.

[0119] Figure 1 A schematic diagram of a scenario of a model acquisition method provided in an embodiment of the present application.

[0120] like Figure 1As shown, the scenario includes a terminal device 10 and an electronic device 20. The electronic device may be a server, a server cluster, or a terminal device (such as a computer, etc.), which is not limited in the present embodiment.

[0121] In this application scenario, the electronic device 20 can obtain the sample text data sent by the terminal device 10.

[0122] The electronic device 20 can process the sample text data through the initial classification model to obtain an actual processing result, wherein the actual processing result includes at least one actual text entity and the type of each actual text entity.

[0123] The electronic device 20 may send the actual processing result to the terminal device 10 , and receive the adjusted processing result sent by the terminal device 10 after adjusting the actual processing result.

[0124] The electronic device 20 can perform training processing on the initial classification model according to the sample text data and the adjustment processing results until a target classification model is obtained through training.

[0125] It should be noted that Figure 1 This is only a schematic diagram of an application scenario provided by the embodiment of the present application. Figure 1 The actual form of the various components included in the Figure 1 The interaction mode between the components is limited, and in the application of the solution, it can be set according to actual needs.

[0126] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0127] Figure 2 A flow chart of a first embodiment of a model acquisition method provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the method comprises the following steps:

[0128] S201: Obtain sample text data sent by a terminal device.

[0129] In this embodiment, the electronic device can obtain the sample text data sent by the terminal device.

[0130] S202: Processing the sample text data through the initial classification model to obtain actual processing results.

[0131] In this embodiment, the electronic device may acquire an initial classification model in advance.

[0132] The electronic device can process the sample text data through the initial classification model to obtain actual processing results.

[0133] The actual processing result includes at least one actual text entity and the type of each actual text entity.

[0134] S203: Sending the actual processing result to the terminal device, and receiving the adjusted processing result sent by the terminal device after adjusting the actual processing result.

[0135] In this embodiment, the electronic device may send the actual processing result to the terminal device.

[0136] The terminal device can adjust the actual processing result to obtain the adjustment processing result. In one implementation, the terminal device can display at least one actual text entity and the type of each actual text entity. In one implementation, the terminal device can also display the sample text data. The terminal device can obtain an adjustment request input by the user. The adjustment request is determined by the user based on at least one actual text entity and the type of each actual text entity. The terminal device can obtain the adjustment processing result according to the adjustment request.

[0137] The electronic device can obtain the adjustment processing result sent by the terminal device.

[0138] The terminal device can send the adjustment processing result to the electronic device.

[0139] S204: Training the initial classification model according to the sample text data and the adjustment processing result until a target classification model is obtained through training.

[0140] In this embodiment, the electronic device can perform training processing on the initial classification model according to the sample text data and the adjustment processing result, until the target classification model is obtained through training.

[0141] Next, a process is described in which an electronic device trains an initial classification model according to sample text data and adjustment processing results until a target classification model is obtained through training.

[0142] In one implementation,

[0143] The electronic device can train the i-th initial classification model according to the i-th sample text data and the i-th adjustment processing result to obtain a trained i-th initial classification model. It should be noted that, in one implementation, the i-th sample text data may include multiple sub-sample text data, and accordingly, the i-th adjustment processing result includes the sub-adjustment processing results corresponding to each sub-sample text data.

[0144] The electronic device can determine whether the processing accuracy of the trained i-th initial classification model reaches a first preset accuracy (also referred to as a calling standard value, or a pushing standard value).

[0145] If so, the electronic device may determine the trained i-th initial classification model as the target classification model.

[0146] If not, the electronic device can use the trained i-th initial classification model as the i+1-th initial classification model, and process the i+1-th sample text data sent by the terminal device according to the i+1-th initial classification model to obtain the i+1-th actual processing result, and send the i+1-th actual processing result to the terminal device, and receive the i+1-th adjusted processing result sent by the terminal device after adjusting the i+1-th actual processing result.

[0147] Among them, i is 1, 2, ... in sequence until the processing accuracy of the trained i-th initial classification model reaches the first preset accuracy; the first sample text data is the sample text data, the first adjustment processing result is the adjustment processing result; the first initial classification model is the initial classification model.

[0148] Beneficial effects of this embodiment: In this embodiment, the electronic device obtains sample text data sent by the terminal device. The electronic device can process the sample text data through the initial classification model to obtain an actual processing result (including at least one actual text entity and the type of each actual text entity). The electronic device can send the actual processing result to the terminal device, and receive the adjustment processing result sent by the terminal device after adjusting the actual processing result. The electronic device can train the initial classification model according to the sample text data and the adjustment processing result until the target classification model is obtained through training. By the above-mentioned method of continuously adjusting the actual processing result by the terminal device to obtain the adjustment processing result, and then the electronic device trains the initial classification model according to the adjustment processing result, the processing capacity of the target classification model is improved, thereby improving the accuracy and efficiency of obtaining the text entity and the type of the text entity by using the target classification model to process.

[0149] The following describes the process of obtaining the initial classification model through method embodiment 2.

[0150] Figure 3a A flow chart of a second embodiment of a model acquisition method provided in an embodiment of the present application is shown in FIG. Figure 3a As shown, the method comprises the following steps:

[0151] S301: Obtain the mth training data sent by the terminal device.

[0152] In this embodiment, the electronic device can obtain the mth training data (also referred to as a training set) sent by the terminal device. The mth training data includes the mth training text data and the mth training processing result. It should be noted that the mth training text data includes at least one sub-training text data, and the mth training processing result includes the sub-training processing results corresponding to each sub-training text data.

[0153] Next, a process in which a terminal device obtains a sub-training text data and a sub-training processing result corresponding to the sub-training text data is described.

[0154] In one implementation,

[0155] Figure 3b A schematic diagram of obtaining training text entities and types of training text entities provided in an embodiment of the present application. Figure 3b As shown, for any sub-training text data, the terminal device can obtain the sub-training text data and display the sub-training text data. Figure 3c A schematic diagram of displaying sub-training text data provided in an embodiment of the present application. In one implementation, the sub-training text data may be data in JSON format.

[0156] The terminal device may obtain the training text entity selected by the user in response to the user's first mouse click operation. For example, the training text entity may be the aircraft model R1B1.

[0157] The terminal device may display at least one type to be selected in response to the user's second mouse click operation.

[0158] The terminal device may obtain the candidate type selected by the user in response to the user's third mouse click operation, and determine the candidate type as the type of the training text entity. For example, the type of the training text entity may be a cargo plane type.

[0159] In addition, it should be noted that in one implementation, in the process of the terminal device acquiring the mth training data, the terminal device can determine that the mth training data has been acquired when the number of acquired sub-training text data and the sub-training processing results corresponding to the sub-training text data reaches a first preset number.

[0160] In addition, in one implementation, the electronic device can also obtain the mth test data (test set) sent by the terminal device. The mth test data includes the mth test text data and the mth test processing result. It should be noted that the mth test text data includes at least one sub-test text data, and the mth test processing result includes the sub-test processing results corresponding to each sub-test text data. In addition, it should be noted that, in one implementation, in the process of the terminal device obtaining the mth test data, the terminal device can determine that the mth test data has been obtained when the number of the obtained sub-test text data and the sub-test processing results corresponding to the sub-test text data reaches a second preset number.

[0161] In one implementation, the sum of the first preset number and the second preset number may be a multiple of N. N may be a positive integer.

[0162] S302: Perform training processing on the mth neural network model according to the mth training data to obtain a trained mth neural network model.

[0163] In this embodiment, the electronic device can perform training processing on the mth neural network model according to the mth training data to obtain a trained mth neural network model.

[0164] S303: Determine whether the trained mth neural network model meets preset conditions.

[0165] In this embodiment, the electronic device can determine whether the trained mth neural network model meets the preset conditions.

[0166] If yes, execute S304;

[0167] If not, execute S305.

[0168] Next, a process of an electronic device determining whether a trained mth neural network model meets preset conditions is described.

[0169] In one implementation,

[0170] The electronic device can obtain the processing accuracy of the trained mth neural network model. In one implementation, during the analysis of any sub-test text data in the mth test data, the electronic device can obtain the actual processing result of the sub-test text data according to the trained mth neural network model. The electronic device can add 1 to the correct number when it recognizes that the actual processing result of the sub-test text data is consistent with the sub-test processing result corresponding to the sub-test text data. After the electronic device has completed the analysis of all the sub-test text data, it can determine the processing accuracy based on the correct number and the number of all sub-test text data.

[0171] The electronic device can determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy (also referred to as an evaluation standard value).

[0172] If so, the electronic device can determine that the trained mth neural network model meets the preset conditions.

[0173] If not, the electronic device may determine that the trained mth neural network model does not meet the preset conditions.

[0174] In one implementation,

[0175] The electronic device can obtain the processing accuracy of the trained mth neural network model.

[0176] The electronic device can determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy and whether new training data sent by the terminal device is not obtained.

[0177] If so, the electronic device can determine that the trained mth neural network model meets the preset conditions.

[0178] If not, the electronic device may determine that the trained mth neural network model does not meet the preset conditions.

[0179] S304: Determine the trained mth neural network model as the initial classification model.

[0180] In this embodiment, the electronic device may determine the trained mth neural network model as the initial classification model when it is determined that the trained mth neural network model meets a preset condition.

[0181] S305: Use the trained mth neural network model as the m+1th neural network model.

[0182] In this embodiment, when the electronic device determines that the trained mth neural network model does not meet the preset conditions, the trained mth neural network model can be used as the (m+1)th neural network model.

[0183] It should be noted that the electronic device may return to execute S301 after executing S305. It should also be noted that when returning to execute S301, the value of m changes. In other words, m is 1, 2, ... in sequence until the trained mth neural network model meets the preset conditions.

[0184] Beneficial effects of this embodiment: In this embodiment, the electronic device can obtain the mth training data sent by the terminal device; wherein the mth training data includes the mth training text data and the mth training processing result. The electronic device can perform training processing on the mth neural network model according to the mth training data to obtain a trained mth neural network model. The electronic device can determine the trained mth neural network model as the initial classification model when it is determined that the trained mth neural network model meets the preset conditions; the electronic device can use the trained mth neural network model as the m+1th neural network model when it is determined that the trained mth neural network model does not meet the preset conditions. Through the above-mentioned iterative training method, the initial classification model can be quickly obtained, thereby improving the rate and accuracy of obtaining the target classification model.

[0185] Figure 4 A flow chart of a model acquisition method embodiment 3 provided in the present application is shown in FIG. Figure 4 As shown, the method comprises the following steps:

[0186] S401: Obtain text data to be processed sent by a terminal device.

[0187] In this embodiment, after obtaining the target classification model, the electronic device can obtain the text data to be processed sent by the terminal device.

[0188] S402: Process the text data to be processed according to the target classification model to obtain a target processing result.

[0189] In this embodiment, after obtaining the text data to be processed, the electronic device can process the text data to be processed according to the target classification model to obtain a target processing result.

[0190] The target processing result includes at least one target text entity and the type of each target text entity.

[0191] For example, the target text entity may be an aircraft model A1B1, and the type of the target text entity may be a fighter aircraft type. For another example, the target text entity may be an aircraft model C1D1, and the type of the target text entity may be a commercial passenger aircraft type.

[0192] S403: Screening the target processing result to obtain a screening processing result.

[0193] In this embodiment, after obtaining the target processing result, the electronic device may perform screening processing on the target processing result to obtain the screening processing result, wherein the target processing result includes at least one target text entity and the type of each target text entity.

[0194] In one implementation,

[0195] The electronic device can filter and obtain at least one filtered text entity and the type of each filtered text entity from the target processing result according to a preset ratio.

[0196] In one implementation,

[0197] The electronic device can filter out at least one filtered text entity and the type of each filtered text entity from the target processing result according to a preset number.

[0198] S404: Sending the screening process result to the terminal device, and receiving the acceptance result sent by the terminal device, which is obtained by performing acceptance processing on the screening process result.

[0199] In this embodiment, the electronic device may send a screening process result to the terminal device, wherein the screening process result includes at least one screening text entity and the type of each screening text entity.

[0200] The terminal device can perform acceptance processing on the screening processing results to obtain acceptance results. In one implementation, the terminal device can display at least one screening text entity and the type of each screening text entity. In one implementation, the terminal device can also display the text data to be processed. The terminal device can obtain an acceptance result input request input by the user. Among them, the acceptance result input request is obtained by the user accepting at least one screening text entity and the type of each screening text entity based on the text data to be processed. The terminal device can generate an acceptance result according to the acceptance result input request.

[0201] The terminal device may send the acceptance result to the electronic device. Accordingly, the electronic device may obtain the acceptance result sent by the terminal device. The acceptance result includes acceptance passed or acceptance failed.

[0202] It should be noted that, in one implementation,

[0203] When the acceptance result indicates that the acceptance fails, the electronic device may adjust the first preset accuracy until the acceptance result indicates that the acceptance passes.

[0204] The following describes a process in which, when an acceptance result of the electronic device indicates that the acceptance has failed, the electronic device adjusts the first preset accuracy until the acceptance result indicates that the acceptance has passed.

[0205] In one implementation,

[0206] The electronic device may determine whether the nth acceptance result indicates acceptance passed.

[0207] If not, the electronic device adjusts the nth first preset accuracy to obtain the (n+1)th first preset accuracy.

[0208] The electronic device obtains the (n+1)th text data to be processed sent by the terminal device.

[0209] The electronic device processes the n+1th text data to be processed according to the n+1th target classification model to obtain the n+1th target processing result, wherein the n+1th target classification model is a model obtained according to the n+1th first preset accuracy rate.

[0210] The electronic device may perform screening processing on the n+1th target processing result to obtain the n+1th screening processing result.

[0211] The electronic device can send the n+1th screening process result to the terminal device, and receive the n+1th acceptance result sent by the terminal device, which is obtained by performing acceptance processing on the n+1th screening process result.

[0212] Here, n is 1, 2, ... in sequence, until the nth acceptance result indicates that the acceptance is passed.

[0213] It should be noted that the nth acceptance result is the aforementioned acceptance result; the nth first preset accuracy rate is the aforementioned first preset accuracy rate; and the nth target classification model is the target classification model obtained by the aforementioned method embodiment.

[0214] Next, a process of adjusting the nth first preset accuracy rate by the electronic device is described.

[0215] In one implementation, the electronic device may adjust the nth first preset accuracy rate according to a preset adjustment value when the nth acceptance result indicates that the acceptance fails.

[0216] In one implementation, the electronic device may adjust the nth first preset accuracy rate according to a preset adjustment ratio when the nth acceptance result indicates that the acceptance fails.

[0217] S405: When the acceptance result indicates that the acceptance is passed, the target processing result is displayed.

[0218] In this embodiment, the electronic device may display the target processing result when the acceptance result indicates that the acceptance is passed.

[0219] In one implementation, the electronic device may control a display interface of the electronic device to display the target processing result when the acceptance result indicates that the acceptance is passed.

[0220] In one implementation, the electronic device may control the display interface of the terminal device to display the target processing result when the acceptance result indicates that the acceptance is passed.

[0221] Beneficial effects of this embodiment: In this embodiment, after the model training is completed, that is, after the electronic device obtains the target classification model, the electronic device can obtain the text data to be processed sent by the terminal device. The electronic device can process the text data to be processed according to the target classification model to obtain the target processing result (including at least one target text entity and the type of each target text entity). The electronic device can filter the target processing result to obtain the filtering result, send the filtering result to the terminal device, and receive the acceptance result sent by the terminal device for the acceptance of the filtering result. The electronic device can display the target processing result when the acceptance result indicates that the acceptance is passed. In the above manner, the target processing result output by the target classification model can be sent to the terminal device for acceptance by the terminal device, thereby improving the accuracy of the target processing result.

[0222] Figure 5 A flow chart of a fourth embodiment of a model acquisition method provided in an embodiment of the present application is shown in FIG. Figure 5 As shown, the method comprises the following steps:

[0223] S501: Obtain sample text data sent by a terminal device.

[0224] In this embodiment, the electronic device can obtain the sample text data sent by the terminal device.

[0225] S502: Processing the sample text data through the initial classification model to obtain actual processing results.

[0226] In this embodiment, the electronic device can process the sample text data through the initial classification model to obtain an actual processing result, wherein the actual processing result includes at least one actual text entity and the type of each actual text entity.

[0227] Next, the process of an electronic device acquiring an initial classification model is described.

[0228] In one implementation,

[0229] The electronic device can obtain the mth training data sent by the terminal device. The electronic device can train the mth neural network model according to the mth training data to obtain the trained mth neural network model. The electronic device can determine whether the trained mth neural network model meets the preset conditions. If so, the electronic device can determine the trained mth neural network model as the initial classification model. If not, the electronic device can use the trained mth neural network model as the m+1th neural network model. Among them, m is 1, 2, ..., until the trained mth neural network model meets the preset conditions.

[0230] Next, a process of an electronic device determining whether a trained mth neural network model meets preset conditions is described.

[0231] In one implementation,

[0232] The electronic device may obtain the processing accuracy of the trained mth neural network model. The electronic device may determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy. If so, the electronic device may determine that the trained mth neural network model meets the preset conditions. If not, the electronic device may determine that the trained mth neural network model does not meet the preset conditions.

[0233] In one implementation,

[0234] The electronic device may obtain the processing accuracy of the trained mth neural network model. The electronic device may determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy and whether new training data sent by the terminal device is not obtained. If so, the electronic device may determine that the trained mth neural network model meets the preset conditions. If not, the electronic device may determine that the trained mth neural network model does not meet the preset conditions.

[0235] S503: Sending the actual processing result to the terminal device, and receiving the adjusted processing result sent by the terminal device after adjusting the actual processing result.

[0236] In this embodiment, after obtaining the actual processing result, the electronic device may send the actual processing result to the terminal device.

[0237] The terminal device can adjust the actual processing result to obtain an adjusted processing result.

[0238] The electronic device can obtain the adjustment processing result sent by the terminal device.

[0239] S504: Training the initial classification model according to the sample text data and the adjustment processing result until a target classification model is obtained through training.

[0240] In this embodiment, the electronic device can perform training processing on the initial classification model according to the sample text data and the adjustment processing result, until the target classification model is obtained through training.

[0241] In one implementation,

[0242] The electronic device can perform training processing on the i-th initial classification model according to the i-th sample text data and the i-th adjustment processing result to obtain a trained i-th initial classification model.

[0243] The electronic device can determine whether the processing accuracy of the trained i-th initial classification model reaches a first preset accuracy (also referred to as a calling standard value, or a pushing standard value).

[0244] If so, the electronic device may determine the trained i-th initial classification model as the target classification model.

[0245] If not, the electronic device can use the trained i-th initial classification model as the i+1-th initial classification model, and process the i+1-th sample text data sent by the terminal device according to the i+1-th initial classification model to obtain the i+1-th actual processing result, and send the i+1-th actual processing result to the terminal device, and receive the i+1-th adjusted processing result sent by the terminal device after adjusting the i+1-th actual processing result.

[0246] Among them, i is 1, 2,... in sequence, until the processing accuracy of the trained i-th initial classification model reaches the first preset accuracy; the first sample text data is the sample text data; the first adjustment processing result is the adjustment processing result; the first initial classification model is the initial classification model.

[0247] S505: Obtain the text data to be processed sent by the terminal device.

[0248] In this embodiment, the electronic device can obtain the text data to be processed sent by the terminal device.

[0249] S506: Process the text data to be processed according to the target classification model to obtain a target processing result.

[0250] In this embodiment, the electronic device can process the text data to be processed according to the target classification model to obtain a target processing result.

[0251] The target processing result includes at least one target text entity and the type of each target text entity.

[0252] S507: Screening the target processing result to obtain a screening processing result.

[0253] In this embodiment, the electronic device can perform screening processing on the target processing result to obtain the screening processing result.

[0254] S508: Send the screening result to the terminal device, and receive the acceptance result sent by the terminal device, which is obtained by performing acceptance processing on the screening result.

[0255] In this embodiment, the electronic device can send the screening process result to the terminal device, and receive the acceptance result sent by the terminal device, which is obtained by performing acceptance processing on the screening process result.

[0256] S509: When the acceptance result indicates that the acceptance is passed, the target processing result is displayed.

[0257] In this embodiment, the electronic device may display the target processing result when the acceptance result indicates that the acceptance is passed.

[0258] Beneficial effects of this embodiment: In this embodiment, the terminal device can continuously adjust the actual processing result to obtain the adjusted processing result, and the electronic device can then train the initial classification model based on the adjusted processing result to obtain the target classification model. After the electronic device obtains the target classification model, the electronic device can obtain the text data to be processed sent by the terminal device. The electronic device can process the text data to be processed according to the target classification model to obtain the target processing result (including at least one target text entity and the type of each target text entity). The electronic device can filter the target processing result to obtain the filtering processing result, send the filtering processing result to the terminal device, and receive the acceptance result sent by the terminal device for the acceptance of the filtering processing result. The electronic device can display the target processing result when the acceptance result indicates that the acceptance is passed. In the above manner, the target processing result output by the target classification model can be sent to the terminal device for acceptance by the terminal device, thereby improving the accuracy of the target processing result.

[0259] Figure 6 This is a schematic diagram of the structure of a model acquisition device provided in an embodiment of the present application. Figure 6 As shown, the model acquisition device 60 may include: a transceiver module 61 and a processing module 62 .

[0260] The transceiver module 61 is used to obtain the sample text data sent by the terminal device;

[0261] The processing module 62 is used to process the sample text data through the initial classification model to obtain an actual processing result; the actual processing result includes at least one actual text entity and the type of each actual text entity;

[0262] The transceiver module 61 is further used to send the actual processing result to the terminal device, and receive the adjusted processing result sent by the terminal device after adjusting the actual processing result;

[0263] The processing module 62 is further used to train the initial classification model according to the sample text data and the adjustment processing results until a target classification model is obtained through training.

[0264] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0265] In one implementation, the processing module 62 is specifically configured to:

[0266] According to the i-th sample text data and the i-th adjustment processing result, the i-th initial classification model is trained to obtain a trained i-th initial classification model;

[0267] Determine whether the processing accuracy of the trained i-th initial classification model reaches a first preset accuracy;

[0268] If yes, the trained i-th initial classification model is determined as the target classification model;

[0269] If not, the trained i-th initial classification model is used as the i+1-th initial classification model, and the i+1-th sample text data sent by the terminal device is processed according to the i+1-th initial classification model to obtain the i+1-th actual processing result, and the i+1-th actual processing result is sent to the terminal device, and the i+1-th adjusted processing result sent by the terminal device after adjusting the i+1-th actual processing result is received;

[0270] Among them, i is 1, 2,... in sequence, until the processing accuracy of the trained i-th initial classification model reaches the first preset accuracy; the first sample text data is the sample text data; the first adjustment processing result is the adjustment processing result; the first initial classification model is the initial classification model.

[0271] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0272] In one implementation,

[0273] The transceiver module 61 is further used to obtain the mth training data sent by the terminal device; wherein the mth training data includes the mth training text data and the mth training processing result;

[0274] The processing module 62 is further used to perform training processing on the mth neural network model according to the mth training data to obtain a trained mth neural network model;

[0275] The processing module 62 is further used to determine whether the trained mth neural network model meets the preset conditions;

[0276] The processing module 62 is further configured to determine the trained mth neural network model as the initial classification model if yes;

[0277] The processing module 62 is further configured to, if not, use the trained mth neural network model as the m+1th neural network model;

[0278] Among them, m is 1, 2, ... in sequence until the trained mth neural network model meets the preset conditions.

[0279] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0280] In one implementation, the processing module 62 is specifically configured to:

[0281] Obtain the processing accuracy of the trained mth neural network model;

[0282] Determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy;

[0283] If yes, it is determined that the trained mth neural network model meets the preset conditions;

[0284] If not, it is determined that the trained mth neural network model does not meet the preset conditions.

[0285] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0286] In one implementation, the processing module 62 is specifically configured to:

[0287] Obtain the processing accuracy of the trained mth neural network model;

[0288] Determine whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy and whether new training data sent by the terminal device is not obtained;

[0289] If yes, it is determined that the trained mth neural network model meets the preset conditions;

[0290] If not, it is determined that the trained mth neural network model does not meet the preset conditions.

[0291] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0292] In one implementation,

[0293] The transceiver module 61 is also used to obtain the text data to be processed sent by the terminal device;

[0294] The processing module 62 is further used to process the text data to be processed according to the target classification model to obtain a target processing result; the target processing result includes at least one target text entity and the type of each target text entity;

[0295] The processing module 62 is also used to perform screening processing on the target processing result to obtain the screening processing result;

[0296] The transceiver module 61 is further used to send the screening result to the terminal device, and receive the acceptance result sent by the terminal device, which is obtained by performing acceptance processing on the screening result;

[0297] The processing module 62 is also used to display the target processing result when the acceptance result indicates that the acceptance is passed.

[0298] The model acquisition device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0299] Figure 7 This is a structural diagram of an electronic device provided in this application. Figure 7 As shown, the electronic device 70 includes a processor 71 and a memory 72. The processor 71 is in communication connection with the memory 72, and the memory 72 is used to store computer-executable instructions; the processor 71 is configured to execute the technical solution in any of the above method embodiments by executing the computer-executable instructions stored in the memory 72.

[0300] Optionally, the memory 72 may be independent or integrated with the processor 71. Optionally, when the memory 72 is a device independent of the processor 71, the electronic device 70 may further include: a bus 73 for connecting the above devices.

[0301] The electronic device is used to execute the technical solution in any of the aforementioned method embodiments, and its implementation principle and technical effect are similar and will not be repeated here.

[0302] An embodiment of the present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the technical solution provided by any of the aforementioned method embodiments.

[0303] An embodiment of the present application also provides a computer program product, including a computer program, which is used to implement the technical solution provided by the aforementioned method embodiment when executed by a processor.

[0304] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0305] It should be further noted that, although the various steps in the flowchart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowchart may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0306] It should be understood that the above-mentioned device embodiments are only illustrative, and the device of the present application can also be implemented in other ways. For example, the division of units / modules in the above-mentioned embodiments is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units, modules or components can be combined, or can be integrated into another system, or some features can be ignored or not executed.

[0307] In addition, unless otherwise specified, each functional unit / module in each embodiment of the present application may be integrated into one unit / module, each unit / module may exist physically separately, or two or more units / modules may be integrated together. The above-mentioned integrated unit / module may be implemented in the form of hardware or in the form of a software program module.

[0308] If the integrated unit / module is implemented in the form of hardware, the hardware may be a digital circuit, an analog circuit, etc. The physical implementation of the hardware structure includes but is not limited to transistors, memristors, etc. If not specifically stated, the processor may be any appropriate hardware processor, such as a CPU, a GPU, an FPGA, a DSP, an ASIC, etc. If not specifically stated, the storage unit may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc.

[0309] If the integrated unit / module is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0310] In the above embodiments, the description of each embodiment has its own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant description of other embodiments. The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0311] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0312] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A model acquisition method, characterized in that: include: Obtain sample text data sent by the terminal device; Processing the sample text data through an initial classification model to obtain an actual processing result; The actual processing result includes at least one actual text entity and the type of each actual text entity; Sending the actual processing result to the terminal device, and receiving an adjusted processing result sent by the terminal device after adjusting the actual processing result; The initial classification model is trained according to the sample text data and the adjustment processing result until a target classification model is obtained through training.

2. The method according to claim 1, characterized in that: The training process of the initial classification model according to the sample text data and the adjustment process result until a target classification model is obtained through training includes: According to the i-th sample text data and the i-th adjustment processing result, the i-th initial classification model is trained to obtain a trained i-th initial classification model; Determining whether the processing accuracy of the trained i-th initial classification model reaches a first preset accuracy; If yes, the trained i-th initial classification model is determined as the target classification model; If not, the trained i-th initial classification model is used as the i+1-th initial classification model, and the i+1-th sample text data sent by the terminal device is processed according to the i+1-th initial classification model to obtain the i+1-th actual processing result, and the i+1-th actual processing result is sent to the terminal device, and the i+1-th adjusted processing result sent by the terminal device after adjusting the i+1-th actual processing result is received; Among them, i is 1, 2,... in sequence, until the processing accuracy of the trained i-th initial classification model reaches the first preset accuracy; the first sample text data is the sample text data, the first adjustment processing result is the adjustment processing result; the first initial classification model is the initial classification model.

3. The method according to claim 2, characterized in that The method further comprises: Acquire the mth training data sent by the terminal device; wherein the mth training data includes the mth training text data and the mth training processing result; According to the mth training data, the mth neural network model is trained to obtain a trained mth neural network model; Determine whether the trained mth neural network model meets a preset condition; If yes, the trained mth neural network model is determined as the initial classification model; If not, the trained mth neural network model is used as the m+1th neural network model; Among them, m is 1, 2, ... in sequence, until the trained mth neural network model meets the preset conditions.

4. The method according to claim 3, characterized in that The determining whether the trained mth neural network model meets a preset condition comprises: Obtaining the processing accuracy of the trained mth neural network model; Determining whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy; If so, determining that the trained mth neural network model meets the preset condition; If not, it is determined that the trained mth neural network model does not meet the preset condition.

5. The method according to claim 3, characterized in that: The determining whether the trained mth neural network model meets a preset condition comprises: Obtaining the processing accuracy of the trained mth neural network model; Determining whether the processing accuracy of the trained mth neural network model reaches a second preset accuracy and whether new training data sent by the terminal device is not obtained; If so, determining that the trained mth neural network model meets the preset condition; If not, it is determined that the trained mth neural network model does not meet the preset condition.

6. The method according to any one of claims 1 to 5, characterized in that: The method further comprises: Acquire the text data to be processed sent by the terminal device; According to the target classification model, the text data to be processed is processed to obtain a target processing result; the target processing result includes at least one target text entity and the type of each target text entity; Performing screening processing on the target processing result to obtain a screening processing result; Sending the screening process result to the terminal device, and receiving an acceptance result sent by the terminal device, which is obtained by performing acceptance processing on the screening process result; When the acceptance result indicates that the acceptance is passed, the target processing result is displayed.

7. A model acquisition device, characterized in that: include: The transceiver module is used to obtain the sample text data sent by the terminal device; A processing module, used for processing the sample text data through an initial classification model to obtain an actual processing result; the actual processing result includes at least one actual text entity and the type of each actual text entity; The transceiver module is further used to send the actual processing result to the terminal device, and receive the adjusted processing result sent by the terminal device after adjusting the actual processing result; The processing module is further used to train the initial classification model according to the sample text data and the adjustment processing result until a target classification model is obtained through training.

8. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 6 when executed by a processor.

10. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 6 when being executed by a processor.

Images