Object classification method and device, equipment, medium and product

Abstract

The invention discloses an object classification method, device and equipment, a medium and a product. The method comprises the steps of obtaining at least one piece of financial associated information corresponding to a to-be-classified object; for at least one piece of financial association information, determining financial association features corresponding to the financial association information according to the data type corresponding to the financial association information; and processing the at least one financial association feature according to an object classification model obtained by pre-training, and obtaining an object type corresponding to the to-be-classified object. According to the technical scheme, the effect of accurately classifying the category of the object by comprehensively fitting the financial association features based on the neural network model is achieved, and the prediction accuracy of the financial service period stage where the object is located is improved.

Description

Technical Field

[0001] The present invention relates to the field of financial technology, and in particular to an object classification method, device, equipment, medium and product. Background Art

[0002] In the transaction of banks, users are the main body of transactions. In today's networked information age, transaction flow is the foundation, and transaction flow is brought by users. However, with the entry of other financial service industries outside banks, the user groups targeted by banking business have begun to lose, resulting in a decrease in transaction flow. Therefore, while banks are increasing the promotion of their own businesses, they also need to maintain existing users, so as to reduce the decrease in transaction flow caused by the loss of existing users.

[0003] In the related art, user classification mainly relies on the professional experience of business personnel to analyze the user information stored in the database, and determine the user's category based on the analysis results. This user classification method is prone to misjudgment, resulting in inaccurate classification results. At the same time, manual operation also consumes unnecessary time and manpower costs, which does not meet actual application needs. Summary of the invention

[0004] The present invention provides an object classification method, device, equipment, medium and product to achieve the effect of accurately classifying the category to which the object belongs based on the comprehensive fitting of financial association characteristics by a neural network model, and improve the prediction accuracy of the financial service cycle stage of the object.

[0005] According to one aspect of the present invention, there is provided an object classification method, the method comprising:

[0006] Acquire at least one item of financial association information corresponding to the object to be classified; wherein the financial association information is at least used to characterize the financial operations performed by the object to be classified; the financial association information includes at least one of object association information, controllable resource information, resource transaction information and financial preference information;

[0007] For at least one item of the financial-related information, determining a financial-related feature corresponding to the financial-related information according to a data type corresponding to the financial-related information;

[0008] At least one of the financial association features is processed according to a pre-trained object classification model to obtain an object type corresponding to the object to be classified; wherein the object classification model is trained based on historical financial association features and object labels corresponding to multiple sample objects; the object type is used to indicate the stage of the financial service cycle of the object to be classified; the object type includes a potential transaction object or a transaction object that is easy to lose.

[0009] According to another aspect of the present invention, there is provided an object classification device, the device comprising:

[0010] An information acquisition module, used to acquire at least one item of financial related information corresponding to the object to be classified; wherein the financial related information is at least used to characterize the financial operations performed by the object to be classified; the financial related information includes at least one of object related information, controllable resource information, resource transaction information and financial preference information;

[0011] a feature determination module, configured to determine, for at least one item of the financial association information, a financial association feature corresponding to the financial association information according to a data type corresponding to the financial association information;

[0012] An object classification module is used to process at least one of the financial association features according to a pre-trained object classification model and obtain an object type corresponding to the object to be classified; wherein the object classification model is trained based on historical financial association features and object labels corresponding to multiple sample objects; the object type is used to indicate the stage of the financial service cycle of the object to be classified; the object type includes a potential transaction object or a transaction object that is easy to lose.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the object classification method described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the object classification method described in any embodiment of the present invention when executed.

[0018] According to another aspect of the present invention, a computer program product is provided. The computer program product includes a computer program. When the computer program is executed by a processor, the object classification method according to any embodiment of the present invention is implemented.

[0019] The technical solution of the embodiment of the present invention obtains at least one item of financial-related information corresponding to the object to be classified; further, for the at least one item of financial-related information, determines the financial-related features corresponding to the financial-related information according to the data type corresponding to the financial-related information; processes the at least one financial-related feature according to the pre-trained object classification model, and obtains the object type corresponding to the object to be classified, thereby solving the problem that the object classification method in the related art is laborious and has low classification accuracy, and achieves the effect of accurately classifying the category to which the object belongs based on the comprehensive fitting of the financial-related features by the neural network model, and improves the prediction accuracy of the financial service cycle stage of the object, thereby providing a data basis for effectively extending the financial service cycle of the object.

[0020] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 is a flow chart of an object classification method provided according to the first embodiment of the present invention;

[0023] Figure 2 is a flow chart of an object classification method provided according to Embodiment 2 of the present invention;

[0024] Figure 3 is a schematic diagram of the structure of an object classification device provided according to Embodiment 3 of the present invention;

[0025] Figure 4 It is a schematic diagram of the structure of an electronic device for implementing the object classification method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] Embodiment 1

[0029] Figure 1 This is a flow chart of an object classification method provided by the first embodiment of the present invention. This embodiment is applicable to the case where objects with transaction records are classified. The method can be executed by an object classification device. The object classification device can be implemented in the form of hardware and / or software. The object classification device can be configured in a terminal and / or a server. Figure 1 As shown, the method includes:

[0030] S110: Obtain at least one item of financial related information corresponding to the object to be classified.

[0031] It should be noted that before using the technical solution of the embodiment of the present invention, the user has been informed of the information to be obtained and its purpose, etc., in accordance with the law, and the user's authorization has been obtained.

[0032] Among them, the object to be classified may be an object with transaction records and to be classified. Optionally, the object to be classified may be a user who has paid a service price for any service (such as financial goods or financial services) in the corresponding area. The financial association information is at least used to characterize the financial operations performed by the object to be classified. The financial association information can be understood as information associated with the financial operations performed by the object. The financial association information may include any information associated with the financial operation, and optionally, at least includes at least one of object association information, controllable resource information, resource transaction information and financial preference information. The object association information may be basic information associated with the object itself. Exemplarily, the object association information may include the level of the object in the corresponding area and information characterizing the basic status of the object. The controllable resource information may be used to characterize the resource status that the object to be classified can control. The resource transaction information may be understood as information characterizing the resource transaction operation performed by the object. Optionally, the resource transaction information may include the transaction operation type, the transaction occurrence time, and the name of the resource recipient or resource provider that has the resource transaction operation with the object to be classified. The financial preference information may be understood as information characterizing the object's preference orientation in financial goods or financial services. Optionally, the financial preference information may include any financial items or financial services that can be provided to the user.

[0033] In this embodiment, at least one object to be classified can be determined from multiple objects with transaction records, and the determined at least one object can be used as the object to be classified. Furthermore, object association information, controllable resource information, resource transaction information, financial preference information and other information that can characterize the financial operations performed by the object to be classified can be obtained from the pre-stored object information, and the obtained information can be used as at least one financial association information corresponding to the object to be classified. Furthermore, the at least one financial association information that has been obtained can be used as the classification basis for the object to be classified to classify the object to be classified.

[0034] S120: For at least one item of financial-related information, determine a financial-related feature corresponding to the financial-related information according to a data type corresponding to the financial-related information.

[0035] Among them, the data type refers to an abstract representation used to define the properties and structure of data in storage units such as variables and parameters. Different data types correspond to different storage requirements and processing methods. Optionally, the data types include continuous type, categorical type, and text type. The financial correlation feature can be used to characterize the key features included in the financial correlation information. In this embodiment, the financial correlation feature may include a vector representation obtained by mapping at least part of the information in the financial correlation information to a continuous vector space. It can be understood that the vector representation can be a vector of a preset dimension, which can characterize the features contained in the information.

[0036] In this embodiment, in order to effectively reduce the data dimension while retaining the key features included in the information to improve the information processing efficiency, when obtaining at least one financial-related information corresponding to the object to be classified, for the at least one financial-related information, the data type corresponding to the financial-related information can be determined according to the data representation form of the financial-related information, and further, the financial-related information can be preprocessed according to the data type corresponding to the financial-related information. Furthermore, the information obtained after the preprocessing can be used as the financial-related feature corresponding to the financial-related information.

[0037] Optionally, determining financial correlation features corresponding to the financial correlation information based on the data type corresponding to the financial correlation information includes: determining a preprocessing method corresponding to the financial correlation information based on the data type corresponding to the financial correlation information; and preprocessing the financial correlation information according to the preprocessing method corresponding to the financial correlation information to obtain the financial correlation features corresponding to the financial correlation information.

[0038] Among them, the data type includes at least one of continuous type, categorical type and text type. It can be understood that continuous data represents values ​​within a range, these values ​​can be any real numbers, and can be infinitely subdivided within the defined range. Categorical data refers to a data type with discrete and limited possible values. Categorical data is usually used to represent classified information. Text data refers to data composed of characters or strings. The preprocessing method includes at least one of interpolation processing, standardization processing and vectorization processing. In this embodiment, different interpolation processing methods can be used for different data types. Optionally, the interpolation processing at least includes a mean interpolation method or a mode interpolation method.

[0039] As an optional implementation of this embodiment, a data type corresponding to the financial-related information may be determined. Further, at least one preprocessing method corresponding to the data type may be determined based on a mapping relationship between a preset data type and a preprocessing method. Thereafter, the financial-related information may be preprocessed based on the determined preprocessing method, and the information obtained after the preprocessing may be used as a financial-related feature corresponding to the financial-related information.

[0040] Exemplarily, when the data type corresponding to the financial-related information is a categorical type, the preprocessing method corresponding to the categorical type may include a mode interpolation method and a vectorization method according to a mapping relationship between a preset data type and a preprocessing method. Further, the financial-related information may be interpolated using the mode interpolation method to obtain the interpolated financial-related information. Further, the interpolated financial-related information may be vectorized according to a preset vectorization method, and the information obtained after the vectorization method is used as the financial-related feature corresponding to the financial-related information.

[0041] Exemplarily, when the data type corresponding to the financial-related information is a text type, the preprocessing method corresponding to the text type can be determined to be vectorized processing according to the mapping relationship between the preset data type and the preprocessing method. Further, the financial-related information can be vectorized using the preset vectorized processing method, and the information obtained after the vectorized processing is used as the financial-related feature corresponding to the financial-related information.

[0042] Exemplarily, when the data type corresponding to the financial-related information is continuous, the preprocessing method corresponding to the continuous type may include a mean interpolation method and a standardization method according to the mapping relationship between the preset data type and the preprocessing method. Further, the financial-related information may be interpolated using the mean interpolation method to obtain the interpolated financial-related information. Further, the interpolated financial-related information may be standardized, and the information obtained after the standardization process may be used as the financial-related feature corresponding to the financial-related information.

[0043] S130: Process at least one financial-related feature according to a pre-trained object classification model, and obtain an object type corresponding to the object to be classified.

[0044] Among them, the object classification model can be a neural network model that takes the financial association feature as an input object to classify the corresponding to-be-classified object based on the financial association feature. The object classification model can be a deep learning model. More specifically, the object classification model can be a binary classification model. In this embodiment, the object classification model can be set according to actual needs, for example, it can include a fully connected neural network, etc. The object classification model is trained based on the historical financial association features and object labels corresponding to multiple sample objects. Optionally, the object classification model can be trained based on positive samples and negative samples constructed based on the historical financial association features corresponding to multiple sample objects. The object type can be used to characterize the category to which the object belongs. In this embodiment, the object type is used to indicate the stage of the financial service cycle in which the to-be-classified object is located. Optionally, the object type includes a potential transaction object or an easy-to-lose transaction object. It can be understood that a potential transaction object can be an object that has a demand for a certain type of financial item or financial service and can pay a service acquisition price for the financial item or financial service. An easy-to-lose transaction object can be an object that is easily lost or lost during a transaction.

[0045] In this embodiment, when at least one financial association feature corresponding to the object to be classified is obtained, the at least one financial association feature can be input into a pre-trained object classification model. Then, the at least one financial association feature can be processed based on the object classification model to classify the object to be classified. Then, the object type corresponding to the object to be classified can be output.

[0046] In this embodiment, in order to enable the object to be classified to obtain a better financial service experience and to extend the financial service cycle of the object to be classified, after obtaining the object type corresponding to the object to be classified, it also includes: determining the processing method corresponding to the object to be classified according to the object type corresponding to the object to be classified, and processing the object to be classified based on the processing method.

[0047] The processing methods may include introducing the details of the relevant financial services and activating the loss warning mechanism. The financial service cycle may represent the entire process from the subject's understanding of any financial service or the subject's inclusion in the development of a platform for financial services until the subject's business relationship with the platform is completely terminated.

[0048] As an optional implementation of this embodiment, when the object type corresponding to the object to be classified is a potential transaction object, the processing method corresponding to the object to be classified can be determined to be to introduce the detailed information of the relevant financial services. Further, the detailed information of the relevant financial services can be displayed to the object to be classified or sent to the object terminal associated with the object to be classified, so that the object to be classified can understand the detailed information of the relevant financial services and want to pay the corresponding service acquisition price for the financial services. In turn, the effect of extending the financial service cycle of the object to be classified can be achieved.

[0049] As another optional implementation of this embodiment, when the object type corresponding to the object to be classified is a transaction object that is easy to lose, it can be determined that the processing method corresponding to the object to be classified is to start the loss warning mechanism. Further, the loss warning mechanism can be started for the object to be classified, so as to play a role in saving the object to be classified when the loss warning mechanism is started. In addition, the effect of extending the financial service cycle of the object to be classified can be achieved.

[0050] The technical solution of the embodiment of the present invention obtains at least one item of financial-related information corresponding to the object to be classified; further, for the at least one item of financial-related information, determines the financial-related features corresponding to the financial-related information according to the data type corresponding to the financial-related information; processes the at least one financial-related feature according to the pre-trained object classification model, and obtains the object type corresponding to the object to be classified, thereby solving the problem that the object classification method in the related art is laborious and has low classification accuracy, and achieves the effect of accurately classifying the category to which the object belongs based on the comprehensive fitting of the financial-related features by the neural network model, and improves the prediction accuracy of the financial service cycle stage of the object, thereby providing a data basis for effectively extending the financial service cycle of the object.

[0051] Embodiment 2

[0052] Figure 2 This is a flow chart of an object classification method provided by the second embodiment of the present invention. Based on the above embodiment, before applying the object classification model, the training samples involved in training the object classification model can be determined first, and the object classification model can be obtained based on the training samples. Then, the object classification is performed based on the object classification model. The specific implementation method can refer to the technical solution of this embodiment. Among them, the technical terms that are the same or similar to the above embodiment are not repeated here.

[0053] like Figure 2 As shown, the method includes:

[0054] S210: Determine training samples for training the object classification model.

[0055] It should be noted that before applying the object classification model provided in this embodiment, the pre-constructed neural network model can be trained in a supervised or unsupervised manner. Before training the neural network model, multiple training samples can be constructed to train the model based on multiple training samples. In order to improve the classification accuracy of the object classification model, as many and rich training samples as possible can be constructed.

[0056] The training samples may include positive samples and negative samples. The positive samples may be constructed based on sample-related features and object labels corresponding to sample objects that are prone to loss. The negative samples may be constructed based on sample-related features and object labels corresponding to potential sample objects.

[0057] In this embodiment, before training to obtain the object classification model, positive samples and negative samples for training the object classification model may be constructed first. Further, the training samples for training the object classification model may be determined based on the positive samples and negative samples.

[0058] Optionally, determining training samples for the training object classification model includes: obtaining historical financial association features and object labels associated with multiple sample objects within a preset time period; and constructing positive samples and negative samples in the training samples based on the multiple historical financial association features and object labels.

[0059] The preset time period may be any time period, and may be 1 month, 2 months, or 3 months, etc. The sample object may be an object having transaction records stored in a relevant platform. The historical financial association feature may be used to characterize the features included in the historical financial association information corresponding to the sample object. The historical financial association information is at least used to characterize the financial operations performed by the sample object. The historical financial association information includes at least one of object association information, resource transaction information, controllable resource information, and financial preference information. The object tag may be used to characterize the actual category to which the sample object belongs. The object tag may include a potential transaction object or an object that is easy to churn.

[0060] As an optional implementation of this embodiment, historical financial association features and object labels corresponding to multiple sample objects within a preset time period may be obtained, and the sample objects may be classified based on the object labels corresponding to the multiple sample objects. Furthermore, positive samples and negative samples in the training samples may be constructed based on the multiple historical financial association features and object labels.

[0061] In actual applications, in the scenarios of potential transaction object mining and loss warning analysis, the number of positive sample objects accounts for a low proportion of the total stored objects. When training an object classification model based on the stored positive and negative samples as training samples, the model may be overfitted and have poor generalization, which may lead to a low classification accuracy of the trained object classification model.

[0062] In view of the above situation, in this embodiment, positive samples and negative samples in the training samples are constructed based on multiple historical financial association features and object labels, including: when the object label is a potential transaction object, based on the object label and the corresponding historical financial association features, the negative samples in the training samples are constructed; when the object label is a transaction object that is easy to churn, based on the object label and the corresponding historical financial association features, the positive samples in the training samples are constructed; based on the positive samples, simulated positive samples are generated, and the simulated positive samples are used as positive samples in the training samples.

[0063] As an optional implementation of this embodiment, when the object label is a potential transaction object, a negative sample can be constructed based on the object label and the historical financial association characteristics of the sample object corresponding to the object label to obtain at least one negative sample. When the object label is a transaction object that is easy to churn, a positive sample can be constructed based on the object label and the historical financial association characteristics of the sample object corresponding to the object label to obtain at least one positive sample. Further, at least one simulated positive sample can be generated based on at least one positive sample, and the simulated positive sample can be used as a positive sample in the training sample. Furthermore, the constructed training sample can include positive samples and negative samples, and the positive sample can be a sample constructed based on the historical financial association characteristics, or a generated simulated positive sample.

[0064] In this embodiment, when a positive sample is obtained, the positive sample can be processed according to a preset sample generation method to obtain a simulated positive sample. The sample generation method can be any generation method, and optionally, the simulated positive sample is generated based on a pre-trained sample generation model.

[0065] Optionally, generating simulated positive samples based on positive samples includes: inputting historical financial correlation features in the positive samples into a pre-trained generation model to obtain simulated correlation features corresponding to the positive samples; and constructing simulated positive samples based on the simulated correlation features and object labels in the positive samples.

[0066] The generative model may be a neural network model that takes real features as input objects to generate simulated features based on real features. In this embodiment, the generative model may be set according to actual needs, for example, it may include a generative adversarial network, etc. The generative model may be obtained by training a pre-constructed generative adversarial network based on positive samples corresponding to the sample objects. The simulated association features may be financial association features generated based on the generative model.

[0067] As an optional implementation of the first embodiment of this invention, for at least one positive sample, the historical financial association features in the positive sample can be input into a pre-trained generation model, and then the historical financial association features are processed based on the generation model to generate simulated association features corresponding to the positive sample. Furthermore, a sample can be constructed based on the simulated association features and the object label in the positive sample, and the constructed sample is used as a simulated positive sample.

[0068] In this embodiment, when a training sample is obtained, the model to be trained can be trained based on the training sample to obtain an object classification model.

[0069] Optionally, an object classification model is obtained based on training samples, including: training samples based on the model to be trained to obtain a loss value; correcting model parameters in the model to be trained based on the loss value, taking the convergence of the loss function in the model to be trained as a training goal, and obtaining an object classification model.

[0070] The model to be trained may be a neural network model whose model parameters are default values ​​or initial values. The loss value may be a numerical value representing the degree of difference between the predicted output and the actual output. The loss function may be a function representing the degree of difference between the predicted output and the actual output. The loss function may be any loss function, and optionally, may be a binary cross entropy loss function.

[0071] As an optional implementation of this embodiment, for multiple training samples, the positive samples and negative samples in the training samples can be input into the model to be trained to obtain the object prediction type corresponding to the positive sample and the object prediction type corresponding to the negative sample. Afterwards, the loss value can be determined according to the object prediction type and object label corresponding to the positive sample and the object prediction type and object label corresponding to the negative sample. Further, the model parameters in the model to be trained can be corrected based on the loss value, and then the training error of the loss function in the model to be trained, that is, the loss parameter, can be used as a condition for detecting whether the current loss function has reached convergence, for example, the training error is less than the preset error or the error change trend tends to be stable, or whether the current model iteration number is equal to the preset number of times. If the convergence condition is detected, for example, the training error of the loss function is less than the preset error or the error change tends to be stable, it indicates that the training of the model to be trained is completed, and at this time, the iterative training can be stopped. If it is detected that the convergence condition is not reached at present, other training samples can be further obtained to train the model to be trained until the training error of the loss function is within the preset range. When the training error of the loss function reaches convergence, the trained model to be trained can be used as the object classification model.

[0072] S220: Obtain at least one item of financial related information corresponding to the object to be classified.

[0073] S230: For at least one item of financial-related information, determine a financial-related feature corresponding to the financial-related information according to a data type corresponding to the financial-related information.

[0074] S240: Process at least one financial-related feature according to a pre-trained object classification model, and obtain an object type corresponding to the object to be classified.

[0075] The technical solution of the embodiment of the present invention determines the training samples participating in the training object classification model; makes the number of constructed positive samples and negative samples more balanced, thereby improving the generalization ability and prediction accuracy of the model, and further enables the trained object classification model to accurately classify objects based on financial association features.

[0076] Embodiment 3

[0077] Figure 3 Schematic diagram of the structure of an object classification device provided by Embodiment 3 of the present invention. Figure 3 As shown, the device includes: an information acquisition module 310, a feature determination module 320 and an object classification module 330.

[0078] Among them, the information acquisition module 310 is used to obtain at least one financial association information corresponding to the object to be classified; wherein the financial association information is at least used to characterize the financial operations performed by the object to be classified; the financial association information at least includes at least one of object association information, controllable resource information, resource transaction information and financial preference information; the feature determination module 320 is used to determine the financial association feature corresponding to the financial association information according to the data type corresponding to the financial association information for at least one item of the financial association information; the object classification module 330 is used to process at least one of the financial association features according to the pre-trained object classification model and obtain the object type corresponding to the object to be classified; wherein the object classification model is trained based on the historical financial association features and object labels corresponding to multiple sample objects; the object type is used to indicate the stage of the financial service cycle of the object to be classified; the object type includes potential transaction objects or easy-to-lose transaction objects.

[0079] The technical solution of the embodiment of the present invention obtains at least one item of financial-related information corresponding to the object to be classified; further, for the at least one item of financial-related information, determines the financial-related features corresponding to the financial-related information according to the data type corresponding to the financial-related information; processes the at least one financial-related feature according to the pre-trained object classification model, and obtains the object type corresponding to the object to be classified, thereby solving the problem that the object classification method in the related art is laborious and has low classification accuracy, and achieves the effect of accurately classifying the category to which the object belongs based on the comprehensive fitting of the financial-related features by the neural network model, and improves the prediction accuracy of the financial service cycle stage of the object, thereby providing a data basis for effectively extending the financial service cycle of the object.

[0080] Optionally, the device further includes: an object processing module.

[0081] The object processing module is used to determine a processing method corresponding to the object to be classified according to the object type corresponding to the object to be classified, and process the object to be classified based on the processing method.

[0082] Optionally, the feature determination module 320 includes: a mode determination unit and a feature determination unit.

[0083] a method determination unit, configured to determine a preprocessing method corresponding to the financial-related information according to a data type corresponding to the financial-related information; wherein the data type includes at least one of a continuous type, a categorical type, and a text type; and the preprocessing method includes at least one of an interpolation process, a normalization process, and a vectorization process;

[0084] The feature determination unit is used to pre-process the financial association information according to a pre-processing method corresponding to the financial association information to obtain a financial association feature corresponding to the financial association information.

[0085] Optionally, the device further includes: a training sample determination module.

[0086] A training sample determination module, used to determine the training samples involved in training the object classification model;

[0087] The training sample determination module includes: a feature acquisition unit and a sample construction unit.

[0088] A feature acquisition unit, used to acquire historical financial related features and object labels associated with multiple sample objects within a preset time period;

[0089] The sample construction unit is used to construct positive samples and negative samples in the training samples according to the plurality of historical financial association features and the object labels.

[0090] Optionally, the sample construction unit includes: a negative sample construction subunit, a positive sample construction subunit and a sample generation subunit.

[0091] A negative sample construction subunit, used for constructing a negative sample in the training sample based on the object label and the corresponding historical financial association features when the object label is a potential transaction object;

[0092] A positive sample construction subunit, configured to construct a positive sample in the training sample based on the object label and corresponding historical financial association features when the object label is a transaction object that is prone to churn;

[0093] The sample generation subunit is used to generate a simulated positive sample based on the positive sample, and use the simulated positive sample as a positive sample in the training sample.

[0094] Optionally, a sample generation subunit is specifically used to input the historical financial correlation features in the positive sample into a pre-trained generation model to obtain simulated correlation features corresponding to the positive sample; and construct a simulated positive sample based on the simulated correlation features and the object labels in the positive sample.

[0095] Optionally, the device further includes: a sample processing module and a model determination module.

[0096] A sample processing module, used to process the training sample based on the model to be trained to obtain a loss value;

[0097] The model determination module is used to modify the model parameters in the model to be trained based on the loss value, take the convergence of the loss function in the model to be trained as the training goal, and obtain the object classification model.

[0098] The object classification device provided in the embodiment of the present invention can execute the object classification method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0099] Embodiment 4

[0100] Figure 4A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0101] like Figure 4 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0102] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0103] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as an object classification method.

[0104] In some embodiments, the object classification method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the object classification method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the object classification method in any other appropriate manner (e.g., by means of firmware).

[0105] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0106] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0107] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0108] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0109] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0110] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0111] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0112] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for object classification, characterized in that: include: Acquire at least one item of financial association information corresponding to the object to be classified; wherein the financial association information is at least used to characterize the financial operations performed by the object to be classified; the financial association information includes at least one of object association information, controllable resource information, resource transaction information and financial preference information; For at least one item of the financial-related information, determining a financial-related feature corresponding to the financial-related information according to a data type corresponding to the financial-related information; Processing at least one of the financial-related features according to a pre-trained object classification model, and obtaining an object type corresponding to the object to be classified; Among them, the object classification model is trained based on historical financial association features and object labels corresponding to multiple sample objects; the object type is used to indicate the stage of the financial service cycle of the object to be classified; the object type includes potential transaction objects or easy-to-lose transaction objects.

2. The object classification method according to claim 1, characterized in that: Also includes: A processing method corresponding to the object to be classified is determined according to the object type corresponding to the object to be classified, and the object to be classified is processed based on the processing method.

3. The object classification method according to claim 1, characterized in that: The determining the financial association feature corresponding to the financial association information according to the data type corresponding to the financial association information includes: Determining a preprocessing method corresponding to the financial-related information according to a data type corresponding to the financial-related information; wherein the data type includes at least one of a continuous type, a categorical type, and a text type; and the preprocessing method includes at least one of an interpolation process, a normalization process, and a vectorization process; The financial association information is preprocessed according to a preprocessing method corresponding to the financial association information to obtain a financial association feature corresponding to the financial association information.

4. The object classification method according to claim 1, characterized in that: Also includes: Determining training samples involved in training the object classification model; The determining of the training samples involved in training the object classification model includes: Obtaining historical financial correlation features and object labels associated with multiple sample objects within a preset time period; Positive samples and negative samples in training samples are constructed according to the plurality of historical financial association features and the object labels.

5. The object classification method according to claim 4, characterized in that: The constructing positive samples and negative samples in the training samples according to the plurality of historical financial association features and the object labels comprises: In the case where the object label is a potential transaction object, constructing a negative sample in the training sample based on the object label and the corresponding historical financial association feature; In the case where the object label is a transaction object that is easy to churn, constructing a positive sample in the training sample based on the object label and the corresponding historical financial association feature; A simulated positive sample is generated based on the positive sample, and the simulated positive sample is used as a positive sample in the training sample.

6. The object classification method according to claim 5, characterized in that: The generating a simulated positive sample based on the positive sample comprises: Inputting the historical financial correlation features in the positive sample into the pre-trained generative model to obtain simulated correlation features corresponding to the positive sample; Based on the simulated association features and the object labels in the positive samples, simulated positive samples are constructed.

7. The object classification method according to claim 4, characterized in that: Also includes: Processing the training samples based on the model to be trained to obtain a loss value; The model parameters in the model to be trained are modified based on the loss value, and the convergence of the loss function in the model to be trained is taken as the training goal to obtain the object classification model.

8. An object classification device, characterized in that: include: An information acquisition module, used to acquire at least one item of financial related information corresponding to the object to be classified; wherein the financial related information is at least used to characterize the financial operations performed by the object to be classified; the financial related information includes at least one of object related information, controllable resource information, resource transaction information and financial preference information; a feature determination module, configured to determine, for at least one item of the financial association information, a financial association feature corresponding to the financial association information according to a data type corresponding to the financial association information; An object classification module is used to process at least one of the financial association features according to a pre-trained object classification model and obtain an object type corresponding to the object to be classified; wherein the object classification model is trained based on historical financial association features and object labels corresponding to multiple sample objects; the object type is used to indicate the stage of the financial service cycle of the object to be classified; the object type includes a potential transaction object or a transaction object that is easy to lose.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the object classification method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the object classification method according to any one of claims 1 to 7 when executed.

Images