Information Sending Method, Device, Electronic Device and Computer Readable Medium

Abstract

Embodiments of the present disclosure disclose an information sending method, apparatus, electronic device, and computer-readable medium. A specific implementation of the method includes: obtaining a value transfer data sequence; splitting each value transfer data to generate a value transfer category data set, thereby obtaining a value transfer category data set sequence; classifying each value transfer category data to generate a value transfer category data sequence set; for each value transfer category data sequence, performing a first generation step: generating a future value transfer category data sequence; generating data authenticity information; generating actual predicted value voucher information; in response to determining that the actual predicted value voucher information is greater than a preset value, determining corresponding value processing behavior information; and sending the value processing behavior information and the actual predicted value voucher information to a use terminal. This implementation can accurately and efficiently generate value voucher information for a target user.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technologies, and more particularly, to an information sending method, apparatus, electronic device, and computer-readable medium. Background Art

[0002] Currently, in various scenarios, determining information about value vouchers for target users has become a major technology under current research. For determining user value voucher information, the commonly adopted method is to directly generate the user value voucher information corresponding to the target user based on the historical value transfer data sequence through a relevant user value voucher information generation model.

[0003] However, when using the above method to generate user value voucher information, the following technical problems often exist:

[0004] First, the learned value transfer feature information for the target user is less, resulting in the generated user value voucher information being inaccurate.

[0005] Second, when the generation of user value voucher information is not accurate enough, how to handle the value voucher appeal for the target user is crucial, which greatly affects the user's value voucher information and subsequent executable value operations.

[0006] The above information disclosed in this background art section is only used to enhance the understanding of the background of the inventive concept, and thus, it may include information that does not form the prior art known to those of ordinary skill in the art in this country. Summary of the Invention

[0007] This content part of the present disclosure is used to introduce the inventive concept in a brief form, which will be described in detail in the subsequent detailed implementation part. This content part of the present disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0008] Some embodiments of the present disclosure propose an information sending method, apparatus, electronic device, and computer-readable medium to solve one or more of the technical problems mentioned in the above background art section.

[0009] In a first aspect, some embodiments of the present disclosure provide a method for sending information, including: obtaining a value transfer data sequence of a target user within a historical time period; splitting each value transfer data in the value transfer data sequence according to a preset value transfer category set to generate a value transfer category data set, thereby obtaining a value transfer category data set sequence; classifying the value transfer category data in the value transfer category data set sequence to generate a value transfer category data sequence set; for each value transfer category data sequence in the value transfer category data sequence set, performing a first generation step: generating a future value transfer category data sequence within a future time period according to the value transfer category data sequence by using a future value transfer category data prediction model; generating data authenticity information for the future value transfer category data sequence by using a discriminant model; generating actual predicted value voucher information corresponding to the target user at the current time according to the value transfer category data sequence set; determining corresponding value processing behavior information in response to determining that the actual predicted value voucher information is greater than a preset value; and sending the value processing behavior information and the actual predicted value voucher information to a usage terminal corresponding to the target user.

[0010] In a second aspect, some embodiments of the present disclosure provide an information sending device, including: an obtaining unit configured to obtain a value transfer data sequence of a target user within a historical time period; a data splitting unit configured to split each value transfer data in the value transfer data sequence according to a preset value transfer category set to generate a value transfer category data set, thereby obtaining a value transfer category data set sequence; a classification unit configured to classify the value transfer category data in the value transfer category data set sequence to generate a value transfer category data sequence set; an execution unit configured to, for each value transfer category data sequence in the value transfer category data sequence set, perform a first generation step: generating a future value transfer category data sequence within a future time period according to the value transfer category data sequence by using a future value transfer category data prediction model; generating data authenticity information for the future value transfer category data sequence by using a discriminant model; a generation unit configured to generate actual predicted value voucher information corresponding to the target user at the current time according to the value transfer category data sequence set; a determination unit configured to determine corresponding value processing behavior information in response to determining that the actual predicted value voucher information is greater than a preset value; and a sending unit configured to send the value processing behavior information and the actual predicted value voucher information to a usage terminal corresponding to the target user.

[0011] In a third aspect, some embodiments of the present disclosure provide an electronic device, including: one or more processors; a storage device storing one or more programs thereon, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner of the first aspect.

[0012] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium storing a computer program thereon, wherein when the program is executed by a processor, the method described in any implementation manner of the first aspect is implemented.

[0013] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: The information sending method according to some embodiments of the present disclosure accurately and efficiently generates value voucher information for a target user. Specifically, the reason for the lack of accuracy of the relevant value voucher information is that there is less learned value transfer characteristic information for the target user, resulting in inaccurate generated user value voucher information. Based on this, the information sending method according to some embodiments of the present disclosure first obtains a value transfer data sequence within a historical time period for the target user as a basic data set for subsequent generation of value voucher information. Then, according to a pre-set value transfer category set, each value transfer data in the above value transfer data sequence is split to generate a value transfer category data set, obtaining a value transfer category data set sequence, so as to facilitate subsequent generation of corresponding value transfer category data in the future time for each value transfer category, greatly expanding the value transfer characteristic information for the target user in the future time period. Next, the data of each value transfer category in the above value transfer category data set sequence is classified to generate a value transfer category data sequence set, so as to facilitate subsequent generation of future value transfer category data. Then, for each value transfer category data sequence in the above value transfer category data sequence set, a first generation step is performed: First step, according to the above value transfer category data sequence, using a future value transfer category data prediction model, a future value transfer category data sequence within a future time period can be accurately generated to obtain prediction data for each value transfer category of the target user in the future time period. Second step, using a discriminant model, data authenticity information for the above future value transfer category data sequence can be accurately generated to ensure the accuracy of the generated future value transfer category data sequence and avoid learning the value transfer characteristic information of the target user in each value transfer category in the future time period. Immediately afterwards, according to the above value transfer category data sequence set, the actual predicted value voucher information corresponding to the above target user at the current time can be accurately generated. Furthermore, in response to determining that the above actual predicted value voucher information is greater than a preset value, corresponding value processing behavior information is determined, so as to accurately perform corresponding value processing behavior for the actual predicted value voucher information of the target user. Finally, the above value processing behavior information and the above actual predicted value voucher information are sent to the usage terminal corresponding to the above target user, so as to facilitate reminder of voucher information for value transfer. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In combination with the accompanying drawings and with reference to the following specific embodiments, the above and other features, advantages and aspects of the various embodiments of the present disclosure will become more obvious. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the elements and elements are not necessarily drawn to scale.

[0015] Figure 1 is a flowchart of some embodiments of the information sending method according to the present disclosure;

[0016] Figure 2 is a schematic structural diagram of some embodiments of the information sending device according to the present disclosure;

[0017] Figure 3 is a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed implementation manners

[0018] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not used to limit the protection scope of the present disclosure.

[0019] In addition, it should be noted that for the sake of convenience of description, only parts related to the relevant invention are shown in the drawings. Without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

[0020] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence relationship of the functions performed by these devices, modules or units.

[0021] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly stated in the context, it should be understood as "one or more".

[0022] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only for illustrative purposes and are not used to limit the scope of these messages or information.

[0023] The present disclosure will be described in detail below with reference to the drawings and in combination with the embodiments.

[0024] Referring to Figure 1 , a flow 100 of some embodiments of the information sending method according to the present disclosure is shown. The information sending method includes the following steps:

[0025] Step 101, obtaining a value transfer data sequence of a historical time period for a target user.

[0026] In some embodiments, the execution subject of the above information sending method may obtain a value transfer data sequence of a target user within a historical time period through a wired connection or a wireless connection. The target user may be the user for whom value voucher information is to be generated. In practice, for the credit investigation scenario, the value voucher information may be the user's credit value. The user's credit value can indirectly represent the degree of favorable evaluation of value transfer. The historical time period may be a pre-determined time period before the current time. For example, the historical time period may be within 3 years from the current time. Each value transfer data in the value transfer data sequence has a corresponding historical time within the historical time period. The value transfer data may be the operation data of the target user's value transfer operation. In practice, for the credit investigation scenario, the value transfer data may include the target user's transaction behavior data, may also include the target user's loan data, and may also be bank statement data. It is not specifically limited, and the value transfer data may include various data related to value transfer in the credit investigation scenario.

[0027] Step 102: According to the pre-set value transfer category set, perform data splitting on each value transfer data in the above value transfer data sequence to generate a value transfer category data set, and obtain a value transfer category data set sequence.

[0028] In some embodiments, the above execution subject may perform data splitting on each value transfer data in the above value transfer data sequence according to the pre-set value transfer category set to generate a value transfer category data set, and obtain a value transfer category data set sequence. Each value transfer category in the value transfer category set may represent the source category corresponding to the value transfer data. For example, the value transfer category set may include, but is not limited to, at least one of the following: bank statement category, app application consumption record category, social behavior category, public record category.

[0029] Step 103: Perform data category classification on each value transfer category data in the above value transfer category data set sequence to generate a value transfer category data sequence set.

[0030] In some embodiments, the above execution subject may perform data category classification on each value transfer category data in the above value transfer category data set sequence to generate a value transfer category data sequence set. Among them, each value transfer data corresponding to each value transfer category data sequence in the value transfer category data sequence set has the same value transfer category. The value transfer category data sequence may include empty value transfer category data.

[0031] Step 104: For each value transfer category data sequence in the above value transfer category data sequence set, execute the first generation step:

[0032] Step 1041: Based on the above value transfer category data sequence, use the future value transfer category data prediction model to generate a future value transfer category data sequence within a future time period.

[0033] In some embodiments, the above execution entity may, based on the above value transfer category data sequence, use the future value transfer category data prediction model to generate a future value transfer category data sequence within a future time period. Among them, the future value transfer category data prediction model may be a time series neural network model for generating future value transfer category data. In practice, the future value transfer category data may be value transfer category data at a future time. The value transfer category data may be data information under the value transfer category. For example, for a value transfer category data sequence corresponding to a bank transfer category, the corresponding future value transfer category data sequence may be a future bank statement data sequence in a future time period. The future time period may be a time period in a pre-determined future scenario. In practice, the duration corresponding to the future time period may be the same as the duration corresponding to the historical time period. In practice, the future value transfer category data prediction model may be a seq2seq model.

[0034] Step 1042: Use a discriminant model to generate data authenticity information for the above future value transfer category data sequence.

[0035] In some embodiments, the above execution entity may use a discriminant model to generate data authenticity information for the above future value transfer category data sequence. Among them, the discriminant model may be a neural network model for discriminating the data authenticity of an input data set. The output of the discriminant model is data authenticity information. The data authenticity information may be a numerical value between 0 and 1, or it may be label data. For example, the data authenticity information may be one of the following: very real, relatively real, average authenticity, poor authenticity, very untrue. In practice, the discriminant model may be an LSTM model.

[0036] Step 105: Generate the actual predicted value voucher information corresponding to the above target user at the current time according to the above value transfer category data sequence set.

[0037] In some embodiments, the above execution entity may generate the actual predicted value voucher information corresponding to the above target user at the current time according to the above value transfer category data sequence set. Among them, the actual predicted value voucher information may be the most accurate predicted value voucher information corresponding to the target user.

[0038] In some optional implementation manners of some embodiments, the above generating the actual predicted value voucher information corresponding to the above target user at the current time according to the above value transfer category data sequence set may include the following steps:

[0039] In the first step, according to the above value transfer category data sequence set, generate the first user value voucher information corresponding to the above target user. Among them, the first user value voucher information can be the user credit information of the target user.

[0040] In the second step, according to the above value transfer category data set sequence, the obtained future value transfer category data sequence set and the data authenticity information set, generate the second user value voucher information corresponding to the above target user. Among them, the second user value voucher information can be the user credit information of the target user.

[0041] In the third step, according to the above first user value voucher information and the above second user value voucher information, generate the actual predicted value voucher information corresponding to the above target user at the current time.

[0042] In some optional implementation manners of some embodiments, the generating the first user value voucher information corresponding to the above target user according to the above value transfer category data sequence set may include the following steps:

[0043] In the first step, input each value transfer category data sequence in the above value transfer category data sequence set into a category data change information generation model to generate category data change information, and obtain a category data change information set. Among them, the category data change information generation model can be a neural network model that is pre-trained to output category data change information. The category data change information can characterize the data change degree of each value transfer category data in the value transfer category data sequence. For example, the category data change information can characterize that each value transfer category data in the value transfer category data sequence is continuously decreasing. The category data change information generation model can be a recurrent neural network model. The category data change information can be information in numerical form. The category data change information can be a negative number or a positive number. A negative number indicates that each value transfer category data is continuously decreasing. A positive number indicates that each value transfer category data is continuously increasing.

[0044] In the second step, determine the importance degree information of each value transfer category in the above value transfer category set relative to the user value voucher information, and obtain an importance degree information set. Among them, the importance degree information can be a numerical value between 0 and 1. The larger the numerical value, the higher the category feature importance degree of the value transfer category relative to the user value voucher information. Among them, the importance degree information set can be determined in advance by relevant experts.

[0045] In the third step, according to the above category data change information set and the above importance degree information set, generate a substantial data change information set for the above value transfer category data sequence set.

[0046] As an example, the above-mentioned execution entity may perform weighted summation on the category data change information in the category data change information set and the importance information in the importance information set to generate substantial data change information, and obtain a substantial data change information set.

[0047] Fourth step, generate the overall data feature information corresponding to each value transfer category data set in the above-mentioned value transfer category data set sequence, and obtain an overall data feature information sequence. Among them, the overall data feature information can represent the overall feature information of the value transfer data corresponding to each value transfer category at the corresponding time.

[0048] As an example, the above-mentioned execution entity may generate graph structure data corresponding to the value transfer category data set. Then, input the graph structure data into a graph feature extraction model to generate overall data feature information, and obtain an overall data feature information sequence.

[0049] Fifth step, input the above-mentioned substantial data change information set and the above-mentioned overall data feature information sequence into a self-attention mechanism model to generate attention information.

[0050] Sixth step, input the above-mentioned attention information into a pre-trained user value voucher information output layer to output first user value voucher information. Among them, the user value voucher information output layer may be a fully connected layer.

[0051] In some optional implementation manners of some embodiments, generating the second user value voucher information corresponding to the target user according to the above-mentioned value transfer category data set sequence, the obtained future value transfer category data sequence set, and the data authenticity information set may include the following steps:

[0052] First step, set the data authenticity information corresponding to each value transfer category data in the above-mentioned value transfer category data set sequence to a predetermined value. In practice, the predetermined value may be the value 1. The predetermined value may represent that the data authenticity of each value transfer category data in the value transfer category data set sequence is 100%.

[0053] Second step, for each value transfer category data set in the above-mentioned value transfer category data set sequence, perform the following second generation steps:

[0054] Sub-step 1, vectorize the above-mentioned value transfer category data set to generate a value transfer category vector.

[0055] As an example, first, the above-mentioned execution entity can input each value transfer data in the value transfer data set into a vector generation model to generate value transfer data vectors. Then, the various value transfer vectors in the obtained value transfer data vector set are combined in a predetermined manner to generate a combined vector, and a value transfer category vector is obtained.

[0056] Sub-step 2: Numerically vectorize the above-mentioned predetermined value to generate a numerical vector.

[0057] As an example, the above-mentioned execution entity can input the predetermined value into a vector generation model for numerical vectorization to generate a numerical vector.

[0058] Sub-step 3: Concatenate the above-mentioned numerical vector to the target position of the above-mentioned value transfer category vector to obtain a first concatenated vector. The target position can be a pre-determined position. For example, the target position can be the rightmost position.

[0059] Third step: Perform data time normalization on each future value transfer category data in the above-mentioned future value transfer category data sequence set to generate a future value transfer category data set sequence. Among them, the future value transfer category data in the future value transfer category data set in the future value transfer category data set sequence corresponds to the same future time.

[0060] Fourth step: For each future value transfer category data set in the above-mentioned future value transfer category data set sequence, perform the following third generation step:

[0061] Sub-step 1: Vectorize the above-mentioned future value transfer category data set to generate a future value transfer category vector. Details are not elaborated here. Refer to the generation of the value transfer category vector.

[0062] Sub-step 2: Vectorize the above-mentioned data authenticity information set to generate a data authenticity vector. Details are not elaborated here. Refer to the generation of the numerical vector.

[0063] Sub-step 3: Concatenate the above-mentioned data authenticity vector to the target position of the above-mentioned future value transfer category vector to obtain a second concatenated vector;

[0064] Fifth step: Input the obtained first concatenated vector sequence and second concatenated vector sequence into a user value voucher information generation model to generate second user value voucher information. The user value voucher information generation model can be a neural network model for generating user value voucher information. For example, the user value voucher information generation model can be an LSTM model.

[0065] In some alternative implementations of some embodiments, the above data vectorization of the value transfer category data set to generate a value transfer category vector may include the following steps:

[0066] First, perform data vector conversion on each value transfer category data in the above value transfer category data set to generate a category vector, obtaining a category vector set.

[0067] As an example, the above execution entity may use a vector conversion model to perform data vector conversion on each value transfer category data in the above value transfer category data set to generate a category vector, obtaining a category vector set.

[0068] Second, determine at least one category data group having a value transfer association relationship in the above value transfer category data set. Among them, the value transfer association relationship may be a relationship having a value transfer association. For example, the value transfer association relationship may be a superior-subordinate relationship of data sources. There is a corresponding value transfer association relationship between each category data in each category data group of the at least one category data group.

[0069] Third, for each category data group in the above at least one category data group, perform the following fifth generation step:

[0070] Sub-step 1, determine the category vector group corresponding to the above category data group. Among them, there is a one-to-one correspondence between the category data in the category data group and the category vectors in the category vector group.

[0071] Sub-step 2, input the above category vector group into a category data association representation information generation model to generate associated data association representation information. Among them, the category data association representation information generation model may be a neural network model for generating category data association representation information. The associated data association representation information may represent the representation relationship reflected by each associated data in the category data group corresponding to the category vector group. For example, the associated data association representation information may be that the target user's turnover is continuously decreasing. In practice, the category data association representation information generation model may be a multi-layer cascaded convolutional neural network.

[0072] Sub-step 3, generate an association representation vector representing the above category data group and the above associated data association representation information. The specific implementation manner may refer to the generation of the category vector.

[0073] Fourth, combine the above category vector set and at least one association representation vector to generate a value transfer category vector.

[0074] In some alternative implementations of some embodiments, generating the second user value voucher information corresponding to the target user based on the above value transfer category data set sequence, the obtained future value transfer category data sequence set, and the data authenticity information set may include the following steps:

[0075] First step, determine the historical time sequence corresponding to the above value transfer category data set sequence. Among them, there is a one-to-one correspondence between the value transfer category data sets in the value transfer category data set sequence and the historical times in the historical time sequence.

[0076] Second step, perform data co-temporal placement on each future value transfer category data in the above future value transfer category data sequence set to generate a future value transfer category data set sequence.

[0077] Third step, determine the future time sequence corresponding to the above future value transfer category data set sequence. Among them, there is a one-to-one correspondence between the future value transfer category data sets in the future value transfer category data set sequence and the future times in the future time sequence.

[0078] Fourth step, combine the above historical time sequence and future time sequence to generate a time sequence.

[0079] Fifth step, determine at least one time range centered on the current time. The time range may include: a historical time subsequence, the current time, and a future time subsequence. Each of the at least one time range is different. The time center corresponding to each time range is the current time.

[0080] Sixth step, for each of the at least one time range, perform the following fourth generation step:

[0081] Sub-step 1, determine the time subsequence in the above time sequence corresponding to the above time range. Among them, the central time corresponding to the above time subsequence is the above current time.

[0082] Sub-step 2, determine the value transfer category data set subsequence and the future value transfer category data set subsequence corresponding to the above time subsequence.

[0083] Sub-step 3, generate candidate user value voucher information for the above target user according to the above value transfer category data set subsequence and future value transfer category data set subsequence.

[0084] As an example, the above execution entity may use a user value voucher information generation model to generate candidate user value voucher information for the above target user according to the above value transfer category data set subsequence and future value transfer category data set subsequence.

[0085] Step 7: Determine the information weight corresponding to each of the at least one time range above to obtain at least one information weight. The information weight can characterize the importance degree of the data characteristics of the category data set corresponding to the time range.

[0086] Step 8: Perform a weighted summation process on the at least one information weight and the at least one candidate user value voucher information obtained above to obtain a weighted summation value, which is used as the second user value voucher information.

[0087] In some optional implementation manners of some embodiments, generating the actual predicted value voucher information of the target user at the current time according to the first user value voucher information and the second user value voucher information above may include the following steps:

[0088] Step 1: Generate third user value voucher information according to the future value transfer category data sequence set and the data authenticity information set above.

[0089] As an example, the execution subject may input the future value transfer category data sequence set and the data authenticity information set above into a pre-trained future user value voucher information generation model for future data to generate third user value voucher information. In practice, the future user value voucher information generation model may be a Transformer model.

[0090] Step 2: Obtain the feature focus direction information corresponding to the predicted value voucher information. Among them, the feature focus direction information can characterize the direction information of which features the value voucher information is more inclined to emphasize. For example, the feature focus direction information may be the direction information more inclined to bank statement related data features.

[0091] Step 3: Perform a weighted summation process on the first user value voucher information, the second user value voucher information, and the third user value voucher information according to the feature focus direction information above to generate the actual predicted value voucher information.

[0092] As an example, first, according to the feature focus direction information, determine the weight information set corresponding to the first user value voucher information, the second user value voucher information, and the third user value voucher information above. Then, perform a weighted summation on the first user value voucher information, the second user value voucher information, the third user value voucher information, and each weight information in the weight information set to generate the actual predicted value voucher information.

[0093] Step 106: In response to determining that the actual predicted value voucher information is greater than a preset value, determine the corresponding value processing behavior information.

[0094] In some embodiments, in response to determining that the above actual predicted value voucher information is greater than a preset value, the above execution entity may determine corresponding value processing behavior information. Among them, the value processing behavior information may be the behavior information of the value processing behavior. For example, the value processing behavior information may be the behavior information of increasing the lending operation for the target user.

[0095] As an example, the above execution entity may determine the value processing behavior information corresponding to the actual predicted value voucher information by querying the corresponding value processing behavior association table.

[0096] Step 107, send the above value processing behavior information and the above actual predicted value voucher information to the usage terminal corresponding to the above target user.

[0097] In some embodiments, the above execution entity may send the above value processing behavior information and the above actual predicted value voucher information to the usage terminal corresponding to the above target user.

[0098] In some optional implementation manners of some embodiments, after step 107, the steps may include:

[0099] First step, in response to receiving value voucher appeal information for the target user, perform material analysis on the appeal materials corresponding to the value voucher appeal information to generate appeal text information in a target format. Among them, the value voucher appeal information may be information for appealing against the value voucher information. In practice, the value voucher appeal information may be credit appeal information. In practice, the value voucher appeal information may be information in a predetermined format. The predetermined format may be a predetermined text format. In practice, there are also predetermined requirements for the format of the value voucher appeal information. That is, it is required that the target user supplement the corresponding appeal materials in each sub-file of the predetermined document.

[0100] As an example, the above execution entity may perform material analysis on the appeal materials corresponding to the value voucher appeal information through a predetermined analysis format to generate appeal text information in a target format.

[0101] Second step, determine the appeal feature information set of the above appeal text information under a predetermined appeal feature set. Among them, there is a one-to-one correspondence between the predetermined appeal features in the predetermined appeal feature set and the appeal feature information in the appeal feature information set, and the appeal feature information may be the feature content under the predetermined appeal feature. In practice, each of the predetermined appeal features in the predetermined appeal feature set may be preset. For example, the predetermined appeal feature set may include: value transaction records, credit guarantors, credit certificates, value processing behavior correction information.

[0102] Step 3: Generate an appeal score based on the above appeal feature information set. Among them, the appeal score can represent the effectiveness of the value voucher appeal information. That is, the higher the appeal score, the higher the success probability of the appeal for the target user. That is, the more effective the restoration of the value voucher for the target user. That is, the more effective the credit restoration for the target user.

[0103] As an example, the above-mentioned execution entity can generate an appeal score based on the appeal feature information set using a regression model. For example, the regression model can be an SVM model or a linear regression model.

[0104] Step 4: Adjust the above actual predicted value voucher information according to the above appeal score to generate adjusted value voucher information.

[0105] As an example, the above-mentioned execution entity determines the appeal interval corresponding to the appeal score as the target appeal interval. Then, determine the value voucher growth score corresponding to the above target appeal interval. Finally, add the above actual predicted value voucher information and the above value voucher growth score to generate adjusted value voucher information.

[0106] Step 5: Replace the above actual predicted value voucher information in the above usage terminal with the above adjusted value voucher information.

[0107] Optionally, generating an appeal score based on the above appeal feature information set includes the following steps:

[0108] Step 1: Determine the feature level corresponding to each appeal feature in the above appeal feature set. Among them, the feature level represents the importance of the appeal feature relative to the effectiveness of the appeal. The higher the feature level, the higher the degree of influence on the effectiveness of the appeal.

[0109] Step 2: Perform feature information and feature level classification on each appeal feature information in the appeal feature information set according to the feature level corresponding to each appeal feature to generate an appeal feature information group and obtain an appeal feature information group set. Among them, the feature levels corresponding to each appeal feature information in the appeal feature information group are the same.

[0110] Step 3: Input each appeal feature information group in the above appeal feature information group set into the regression model to generate an initial appeal score and obtain an initial appeal score group.

[0111] Step 4: Determine the appeal score influence degree information preset for each feature level in the feature level set. Among them, the appeal score influence degree information represents the importance of the feature corresponding to the feature level. The appeal score influence degree information is a value between 0 and 1.

[0112] In the fifth step, perform a multiplication and summation process on the degree of impact on the appeal score in the information set of the degree of impact on the appeal score and the initial appeal score in the initial appeal score group to generate the above-mentioned appeal score.

[0113] The content in the above "in some alternative implementation manners of some embodiments", as an inventive point of the present disclosure, solves the technical problem mentioned in the background art, "When the generation of user value voucher information is not accurate enough, how to handle the value voucher appeal for the target user is crucial, which greatly affects the user's value voucher information and subsequent executable value operations." Based on this, the present disclosure, first, parses the materials of the value voucher appeal information to accurately obtain the appeal text information in the target format. Then, according to the appeal feature information set corresponding to the appeal text information, the appeal score can be accurately generated to quantitatively adjust the adjusted value voucher information to accurately quantify the value voucher of the target user.

[0114] In some alternative implementation manners of some embodiments, the above-mentioned sending the above-mentioned value processing behavior information and the above-mentioned actual predicted value voucher information to the usage terminal corresponding to the above-mentioned target user may include the following steps:

[0115] In the first step, generate a value voucher description image according to the above-mentioned value processing behavior information and the above-mentioned actual predicted value voucher information. Among them, the value voucher description image may be an image for explaining value voucher-related information in the form of an image.

[0116] As an example, the above-mentioned execution entity may fill the above-mentioned value processing behavior information, the above-mentioned actual predicted value voucher information, and the user information corresponding to the target user into the value voucher description image template to generate the value voucher description image.

[0117] In the second step, send the link corresponding to the above-mentioned value voucher description image to the usage terminal corresponding to the above-mentioned target user.

[0118] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: The information sending method according to some embodiments of the present disclosure accurately and efficiently generates value voucher information for a target user. Specifically, the reason for the inaccuracy of the relevant value voucher information is that: the learned value transfer characteristic information for the target user is less, resulting in the inaccuracy of the generated user value voucher information. Based on this, the information sending method according to some embodiments of the present disclosure, first, obtains a value transfer data sequence within a historical time period for the target user as a basic data set for subsequent generation of value voucher information. Then, according to a preset value transfer category set, each value transfer data in the above value transfer data sequence is split to generate a value transfer category data set, obtaining a value transfer category data set sequence, so as to facilitate subsequent generation of corresponding value transfer category data in the future time for each value transfer category, expanding the value transfer data for the target user in the future time period, and greatly expanding the value transfer characteristic information for the target user. Next, the various value transfer category data in the above value transfer category data set sequence are classified by data category to generate a value transfer category data sequence set, so as to facilitate subsequent generation of future value transfer category data. Then, for each value transfer category data sequence in the above value transfer category data sequence set, a first generation step is executed: First step, according to the above value transfer category data sequence, using a future value transfer category data prediction model, a future value transfer category data sequence within a future time period can be accurately generated to obtain prediction data for each value transfer category of the target user in the future time period. Second step, using a discriminant model, data authenticity information for the above future value transfer category data sequence can be accurately generated to ensure the accuracy of the generated future value transfer category data sequence and avoid learning the value transfer characteristic information of the target user in each value transfer category in the future time period. Immediately afterwards, according to the above value transfer category data sequence set, the actual predicted value voucher information corresponding to the above target user at the current time can be accurately generated. Furthermore, in response to determining that the above actual predicted value voucher information is greater than a preset value, corresponding value processing behavior information is determined, so as to accurately execute the corresponding value processing behavior for the actual predicted value voucher information of the target user. Finally, the above value processing behavior information and the above actual predicted value voucher information are sent to the corresponding usage terminal of the above target user for voucher information reminder of value transfer.

[0119] Further referring to Figure 2 , as an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of an information sending device. These device embodiments correspond to Figure 1 the method embodiments shown, and the information sending device can be specifically applied to various electronic devices.

[0120] As Figure 2 shown, an information sending device 200 includes: an obtaining unit 201, a data splitting unit 202, an arranging unit 203, an executing unit 204, a generating unit 205, a determining unit 206, and a sending unit 207. Among them, the obtaining unit 201 is configured to obtain a value transfer data sequence of a target user within a historical time period; the data splitting unit 202 is configured to perform data splitting on each value transfer data in the value transfer data sequence according to a preset value transfer category set to generate a value transfer category data set, and obtain a value transfer category data set sequence; the arranging unit 203 is configured to perform data category arrangement on each value transfer category data in the value transfer category data set sequence to generate a value transfer category data sequence set; the executing unit 204 is configured to perform a first generating step on each value transfer category data sequence in the value transfer category data sequence set: according to the value transfer category data sequence, use a future value transfer category data prediction model to generate a future value transfer category data sequence within a future time period; use a discriminant model to generate data authenticity information for the future value transfer category data sequence; the generating unit 205 is configured to generate actual predicted value voucher information of the target user at the current time according to the value transfer category data sequence set; the determining unit 206 is configured to determine corresponding value processing behavior information in response to determining that the actual predicted value voucher information is greater than a preset value; the sending unit 207 is configured to send the value processing behavior information and the actual predicted value voucher information to a usage terminal corresponding to the target user.

[0121] It can be understood that the units described in the information sending device 200 correspond to the respective steps in the method described in the reference Figure 1 description. Therefore, the operations, features, and beneficial effects described above for the method also apply to the information sending device 200 and the units included therein, and will not be elaborated here.

[0122] Next, refer to Figure 3 , which shows a schematic structural diagram of an electronic device (e.g., an electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is only an example and should not bring any limitation to the functions and usage scope of the embodiments of the present disclosure.

[0123] As Figure 3As shown, the electronic device 300 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 301, which may perform various appropriate actions and processes according to a program stored in the read-only memory (ROM) 302 or a program loaded from the storage device 308 into the random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. The input / output (I / O) interface 305 is also connected to the bus 304.

[0124] Generally, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 may allow the electronic device 300 to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 3 the electronic device 300 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices may be implemented or had. Figure 3 Each block shown in the figure may represent one device or, as needed, multiple devices.

[0125] In particular, according to some embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for performing the methods shown in the flowcharts. In such some embodiments, the computer program may be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above functions defined in the methods of some embodiments of the present disclosure are executed.

[0126] It should be noted that in some embodiments of the present disclosure, the above-mentioned computer-readable medium may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, 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 above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium may also be any computer-readable medium other than the computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0127] In some embodiments, the client and the server can communicate using any currently known or future-developed network protocol such as HTTP (HyperText Transfer Protocol), and can be interconnected with digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include local area networks ("LANs"), wide area networks ("WANs"), the Internet (e.g., the Internet), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed network.

[0128] The above computer-readable medium may be included in the above electronic device; or it may exist independently without being assembled into the electronic device. The above computer-readable medium carries one or more programs. When the above one or more programs are executed by the electronic device, the electronic device is caused to: obtain a value transfer data sequence of a target user within a historical time period; according to a preset set of value transfer categories, perform data splitting on each value transfer data in the above value transfer data sequence to generate a value transfer category data set, and obtain a value transfer category data set sequence; perform data category classification on each value transfer category data in the above value transfer category data set sequence to generate a value transfer category data sequence set. For each value transfer category data sequence in the above value transfer category data sequence set, perform a first generation step: according to the above value transfer category data sequence, use a future value transfer category data prediction model to generate a future value transfer category data sequence within a future time period; use a discriminant model to generate data authenticity information for the above future value transfer category data sequence; according to the above value transfer category data sequence set, generate actual prediction value voucher information corresponding to the above target user at the current time; in response to determining that the above actual prediction value voucher information is greater than a preset value, determine corresponding value processing behavior information; send the above value processing behavior information and the above actual prediction value voucher information to the usage terminal corresponding to the above target user.

[0129] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages or combinations thereof. The above programming languages include object-oriented programming languages - such as Java, Smalltalk, C++; and also include conventional procedural programming languages - such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network - including a local area network (LAN) or a wide area network (WAN) - or may be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0130] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a portion of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks may occur in a different order than noted in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or by a combination of dedicated hardware and computer instructions.

[0131] The units described in some embodiments of the present disclosure can be implemented in software or in hardware. The described units can also be provided in a processor. For example, it can be described as: a processor includes an acquisition unit, a data splitting unit, a homing unit, an execution unit, a generation unit, a determination unit, and a sending unit. Among them, the names of these units do not constitute a limitation on the unit itself in some cases. For example, the sending unit can also be described as "the unit that sends the above-mentioned value processing behavior information and the above-mentioned actual predicted value voucher information to the usage terminal corresponding to the above-mentioned target user".

[0132] The functions described above can be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that can be used include: Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), Systems on Chip (SOCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0133] The above description is only some preferred embodiments of the present disclosure and an explanation of the technical principles applied. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, technical solutions formed by mutually replacing the above features with technical features (but not limited to) having similar functions disclosed in the embodiments of the present disclosure.

Claims

1. A method for sending information, comprising: Acquire a value flow data sequence within a historical time period for a target user, wherein each value flow data in the value flow data sequence has a corresponding historical time within the historical time period, and the value flow data is operation data of a value flow operation performed by the target user; According to a preset value flow category set, each value flow data in the value flow data sequence is split to generate a value flow category data set, thereby obtaining a value flow category data set sequence; Arranging the value circulation category data in the value circulation category data set sequence by data category to generate a value circulation category data sequence set; For each value flow category data sequence in the value flow category data sequence set, a first generation step is performed: According to the value circulation category data sequence, using the future value circulation category data prediction model, a future value circulation category data sequence in a future time period is generated; Using a discriminant model, generating data authenticity information for the future value flow category data sequence; According to the value flow category data sequence set, the actual predicted value voucher information corresponding to the target user at the current time is generated, wherein: The generating the actual predicted value voucher information corresponding to the target user at the current time according to the value flow category data sequence set includes: generating the first user value voucher information corresponding to the target user according to the value flow category data sequence set; generating the second user value voucher information corresponding to the target user according to the value flow category data set sequence, the obtained future value flow category data sequence set and the data authenticity information set; generating the actual predicted value voucher information corresponding to the target user at the current time according to the first user value voucher information and the second user value voucher information, wherein, The generating of the first user value credential information corresponding to the target user according to the value flow category data sequence set includes: inputting each value flow category data sequence in the value flow category data sequence set into a category data change information generating model to generate category data change information and obtain a category data change information set; determining the importance information of each value flow category in the value flow category set relative to the user value credential information and obtain an importance information set; generating a substantial data change information set for the value flow category data sequence set according to the category data change information set and the importance information set; generating overall data feature information corresponding to each value flow category data set in the value flow category data set sequence and obtain an overall data feature information sequence; inputting the substantial data change information set and the overall data feature information sequence into a self-attention mechanism model to generate attention information; inputting the attention information into a pre-trained user value credential information output layer to output the first user value credential information; In response to determining that the actual predicted value voucher information is greater than a preset value, determining corresponding value processing behavior information; The value processing behavior information and the actual predicted value voucher information are sent to a user terminal corresponding to the target user.

2. The method according to claim 1, wherein: The generating the second user value credential information corresponding to the target user according to the value circulation category data set sequence, the obtained future value circulation category data sequence set and the data authenticity information set includes: Setting the data authenticity information corresponding to each value circulation category data in the value circulation category data set sequence to a predetermined value; For each value flow category data set in the value flow category data set sequence, the following second generation step is performed: Performing data vectorization on the value flow category data set to generate a value flow category vector; Performing numerical vectorization on the predetermined numerical value to generate a numerical vector; splicing the numerical vector to the target position of the value flow category vector to obtain a first splicing vector; Performing data synchronization for each future value circulation category data in the future value circulation category data sequence set to generate a future value circulation category data set sequence; For each future value transfer category data set in the future value transfer category data set sequence, the following third generation step is performed: Performing data vectorization on the future value flow category data set to generate a future value flow category vector; Performing data vectorization on the data authenticity information set to generate a data authenticity vector; Splicing the data authenticity vector to the target position of the future value flow category vector to obtain a second splicing vector; The obtained first concatenated vector sequence and the second concatenated vector sequence are input into a user value credential information generation model to generate second user value credential information.

3. The method according to claim 2, wherein: The generating the second user value credential information corresponding to the target user according to the value circulation category data set sequence, the obtained future value circulation category data sequence set and the data authenticity information set includes: Determine the historical time series corresponding to the value flow category data set sequence; Performing data synchronization for each future value circulation category data in the future value circulation category data sequence set to generate a future value circulation category data set sequence; Determine a future time series corresponding to the future value flow category data set sequence; Combining the historical time series and the future time series to generate a time series; determining at least one time range centered on the current time; For each time range of the at least one time range, the following fourth generating step is performed: Determine a time subsequence in the time series that corresponds to the time range, wherein a central time corresponding to the time subsequence is the current time; Determine a value circulation category data set subsequence and a future value circulation category data set subsequence corresponding to the time subsequence; Generating candidate user value credential information for the target user according to the value circulation category data set subsequence and the future value circulation category data set subsequence; Determine an information weight corresponding to each time range in the at least one time range to obtain at least one information weight; The at least one information weight and the obtained at least one candidate user value credential information are weighted and summed to obtain a weighted sum value as the second user value credential information.

4. The method according to claim 3, wherein: Generating actual predicted value voucher information corresponding to the target user at the current time according to the first user value voucher information and the second user value voucher information includes: Generate third user value voucher information according to the future value circulation category data sequence set and the data authenticity information set; Obtain feature focus direction information corresponding to the predicted value voucher information; According to the feature emphasis direction information, a weighted summation process is performed on the first user value credential information, the second user value credential information and the third user value credential information to generate actual predicted value credential information.

5. The method according to claim 4, wherein: The step of vectorizing the value flow category data set to generate a value flow category vector includes: Performing data vector conversion on each value flow category data in the value flow category data set to generate a category vector, thereby obtaining a category vector set; Determine at least one category data group having a value flow association relationship in the value flow category data set; For each of the at least one category data group, the following fifth generating step is performed: Determine a category vector group corresponding to the category data group; Inputting the category vector group into a category data association representation information generation model to generate association data association representation information; Generate an association representation vector corresponding to the association representation information of the association data; The category vector set and at least one associated representation vector are combined to generate a value flow category vector.

6. An information sending device, comprising: An acquisition unit is configured to acquire a value flow data sequence within a historical time period for a target user, wherein each value flow data in the value flow data sequence has a corresponding historical time within the historical time period, and the value flow data is operation data of a value flow operation performed by the target user; A data splitting unit is configured to split each value flow data in the value flow data sequence according to a preset value flow category set to generate a value flow category data set, thereby obtaining a value flow category data set sequence; a placement unit configured to perform data category placement on each value flow category data in the value flow category data set sequence to generate a value flow category data sequence set; The execution unit is configured to execute a first generation step for each value flow category data sequence in the value flow category data sequence set: based on the value flow category data sequence, using a future value flow category data prediction model to generate a future value flow category data sequence in a future time period; using a discriminant model to generate data authenticity information for the future value flow category data sequence; A generating unit is configured to generate actual predicted value voucher information corresponding to the target user at the current time according to the value flow category data sequence set, wherein the generating of the actual predicted value voucher information corresponding to the target user at the current time according to the value flow category data sequence set includes: generating first user value voucher information corresponding to the target user according to the value flow category data sequence set; generating second user value voucher information corresponding to the target user according to the value flow category data set sequence, the obtained future value flow category data sequence set and the data authenticity information set; generating actual predicted value voucher information corresponding to the target user at the current time according to the first user value voucher information and the second user value voucher information, wherein the generating of the first user value voucher information corresponding to the target user according to the value flow category data sequence set includes: Input each value flow category data sequence in the value flow category data sequence set into the category data change information generation model to generate category data change information and obtain a category data change information set; determine the importance information of each value flow category in the value flow category set relative to the user value credential information to obtain an importance information set; generate a substantial data change information set for the value flow category data sequence set based on the category data change information set and the importance information set; generate overall data feature information corresponding to each value flow category data set in the value flow category data set sequence to obtain an overall data feature information sequence; input the substantial data change information set and the overall data feature information sequence into the self-attention mechanism model to generate attention information; input the attention information into a pre-trained user value credential information output layer to output the first user value credential information; A determination unit configured to determine corresponding value processing behavior information in response to determining that the actual predicted value voucher information is greater than a preset value; The sending unit is configured to send the value processing behavior information and the actual predicted value voucher information to a user terminal corresponding to the target user.

7. An electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 5.

8. A computer readable medium having a computer program stored thereon, wherein: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images