Business recommendation method and device, computer device and storage medium

Abstract

The application relates to a business recommendation method and device, computer equipment and a storage medium. The method comprises the following steps: obtaining current investment business data; calculating a correlation coefficient between the current investment business data and original input data; wherein the original input data is training data of a to-be-migrated model; inputting the current investment business data into the to-be-migrated model to obtain a first recommendation result; inputting the current investment business data into a business recommendation model to obtain a second recommendation result; wherein the business recommendation model is obtained by training historical investment business data; processing the correlation coefficient, the first recommendation result and the second recommendation result to output a target recommendation result; and the target recommendation result is used for recommending investment business. The target recommendation result obtained in the method can learn the knowledge in the to-be-migrated model, mine data from multiple angles, sufficiently utilize the model resources of the to-be-migrated model, and improve the prediction effect of the recommendation result.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a business recommendation method, apparatus, computer device, and storage medium. Background Technology

[0002] In the current banking system, corporate clients often need to reserve a certain amount of funds for working capital. These funds are not always immediately available, so most clients choose to use them for short-term investments to increase their income. Current research explores many solutions using artificial intelligence modeling for decision-making and investment. These typically involve modeling from a single data source. However, different data sources contain different data dimensions and information, making different models incompatible. When new data is added, the existing model becomes unusable, leading to a waste of model resources. Summary of the Invention

[0003] Therefore, it is necessary to provide a business recommendation method, apparatus, computer equipment, and storage medium that can utilize existing models to address the aforementioned technical problems.

[0004] Firstly, this application provides a business recommendation method. The method includes: acquiring current investment business data; wherein the investment business data includes user profile data and investment record data; calculating a correlation coefficient between the current investment business data and original input data; wherein the original input data is training data for a transfer model; inputting the current investment business data into the transfer model to obtain a first recommendation result; inputting the current investment business data into a business recommendation model to obtain a second recommendation result; wherein the business recommendation model is trained using historical investment business data; processing the correlation coefficient, the first recommendation result, and the second recommendation result, and outputting a target recommendation result; the target recommendation result is used to recommend investment businesses.

[0005] In one embodiment, the step of calculating the correlation coefficient between the current investment business data and the original input data includes: performing dimensional transformation processing on the current investment business data to obtain a current feature vector; obtaining a training feature vector of the original input data; wherein the training feature vector includes multiple input feature vectors, and the current feature vector has the same data dimension as the input feature vectors; calculating the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient.

[0006] In one embodiment, the step of performing dimensional transformation processing on the current investment business data to obtain the current feature vector includes: inputting the current investment business data into a fully connected neural network for dimensional transformation processing to obtain the current feature vector.

[0007] In one embodiment, the step of calculating the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient includes: calculating the similarity between the current feature vector and each of the input feature vectors respectively to obtain multiple similarity values; calculating the average of the multiple similarity values, and using the average as the correlation coefficient.

[0008] In one embodiment, the step of calculating the similarity between the current feature vector and each of the input feature vectors includes: calculating the Euclidean similarity between the current feature vector and each of the input feature vectors.

[0009] In one embodiment, the step of processing the correlation coefficient, the first recommendation result, and the second recommendation result to output a target recommendation result includes: obtaining a relevant recommendation result; the relevant recommendation result includes the product of the correlation coefficient and the first recommendation result; and determining the sum of the relevant recommendation result and the second recommendation result as the target recommendation result.

[0010] In one embodiment, the method further includes: inputting historical investment business data into a long short-term memory network for model training to obtain the business recommendation model.

[0011] Secondly, this application also provides a business recommendation device. The device includes:

[0012] The data acquisition module is used to acquire current investment business data; wherein, the investment business data includes user profile data and investment record data;

[0013] The first calculation module is used to calculate the correlation coefficient between the current investment business data and the original input data; wherein, the original input data is the training data of the model to be transferred.

[0014] The first prediction module is used to input the current investment business data into the model to be migrated to obtain a first recommendation result;

[0015] The second prediction module is used to input the current investment business data into the business recommendation model to obtain a second recommendation result; wherein, the business recommendation model is trained from the historical investment business data;

[0016] The second calculation module is used to process the correlation coefficient, the first recommendation result, and the second recommendation result, and output the target recommendation result; the target recommendation result is used to recommend investment business.

[0017] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor. The memory stores a computer program, and the processor, when executing the computer program, performs the following steps: acquiring current investment business data; wherein the investment business data includes user profile data and investment record data; calculating the correlation coefficient between the current investment business data and original input data; wherein the original input data is training data for a model to be transferred; inputting the current investment business data into the model to be transferred to obtain a first recommendation result; inputting the current investment business data into a business recommendation model to obtain a second recommendation result; wherein the business recommendation model is trained using historical investment business data; processing the correlation coefficient, the first recommendation result, and the second recommendation result, and outputting a target recommendation result; the target recommendation result is used to recommend investment businesses.

[0018] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps: acquiring current investment business data; wherein the investment business data includes user profile data and investment record data; calculating the correlation coefficient between the current investment business data and original input data; wherein the original input data is training data for a model to be transferred; inputting the current investment business data into the model to be transferred to obtain a first recommendation result; inputting the current investment business data into a business recommendation model to obtain a second recommendation result; wherein the business recommendation model is trained using historical investment business data; processing the correlation coefficient, the first recommendation result, and the second recommendation result, and outputting a target recommendation result; the target recommendation result is used to recommend investment businesses.

[0019] The aforementioned business recommendation method, apparatus, computer equipment, and storage medium input current investment business data into the model to be transferred and the business recommendation model respectively, thereby obtaining a first recommendation result and a second recommendation result. Simultaneously, the correlation coefficient between the current investment business data and the original input data is calculated to determine the difference between the current investment data and the original input data. Since the model to be transferred is trained using the original input data, by processing the correlation coefficient and the first recommendation result, and combining it with the second recommendation result, the target recommendation result can learn the knowledge from the model to be transferred, mining the data from multiple perspectives to fully utilize the model resources of the model to be transferred and improve the predictive performance of the recommendation result. Attached Figure Description

[0020] Figure 1 This is a diagram illustrating the application environment of a business recommendation method in one embodiment.

[0021] Figure 2 This is a flowchart illustrating a business recommendation method in one embodiment;

[0022] Figure 3 This is a flowchart illustrating the process of calculating the correlation coefficient between current investment business data and original input data in one embodiment.

[0023] Figure 4 This is a flowchart illustrating the process of calculating the correlation coefficient between the current feature vector and the training feature vector in one embodiment.

[0024] Figure 5 This is a flowchart illustrating the process of processing the correlation coefficient, the first recommendation result, and the second recommendation result to output the target recommendation result in one embodiment.

[0025] Figure 6 This is a schematic diagram of the structure of a long short-term memory network in one embodiment;

[0026] Figure 7 This is a schematic diagram of a service recommendation device in one embodiment;

[0027] Figure 8 This is an internal structural diagram of a computer device in one embodiment;

[0028] Figure 9 This is a diagram of the internal structure of a computer device in another embodiment. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0030] The business recommendation method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on another network server.

[0031] Taking the application of the business recommendation method to server 104 as an example, server 104 can obtain current investment business data from terminal 102 or from the data storage system. This investment business data includes user profile data and investment record data. After obtaining the current investment business data, server 104 calculates the correlation coefficient between the current investment business data and the original input data, where the original input data is the training data for the model to be migrated. The current investment business data is then input into the model to be migrated to obtain a first recommendation result; the current investment business data is then input into the business recommendation model to obtain a second recommendation result, where the business recommendation model is trained using historical investment business data. Finally, server 104 processes the correlation coefficient, the first recommendation result, and the second recommendation result, and outputs a target recommendation result, which is used to recommend investment businesses. Upon receiving the target recommendation result, terminal 102 will then recommend investment businesses.

[0032] In some other embodiments, the business recommendation method can also be applied to terminal 102, which can obtain current investment business data collected from server 104. After obtaining the current investment business data, terminal 102 performs calculations on its local processor to calculate the correlation coefficient between the current investment business data and the original input data, where the original input data is the training data of the model to be transferred. The current investment business data is then input into the model to be transferred to obtain a first recommendation result; the current investment business data is then input into the business recommendation model to obtain a second recommendation result, where the business recommendation model is trained using historical investment business data. Finally, terminal 102 processes the correlation coefficient, the first recommendation result, and the second recommendation result, and outputs a target recommendation result, which terminal 102 uses to recommend investment businesses based on the target recommendation result.

[0033] It should be noted that terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0034] In one embodiment, such as Figure 2 As shown, a business recommendation method is provided, which can be applied to... Figure 1 Taking a server or terminal as an example, the following steps are included:

[0035] Step S100: Obtain current investment business data; wherein, investment business data includes user profile data and investment record data.

[0036] The investment business data includes collected user profile data for current users. This user profile data includes basic customer information (such as gender, age, and place of residence), basic account information (such as account opening address, bank, account type, and account opening date), and basic transfer information (such as amount, date, and transfer method). Investment record data includes the amount and duration of holding a particular financial product. It can be understood that investment record data can include data on a user's holdings of financial products at different points in time.

[0037] Specifically, current investment business data can be stored in the terminal. The server can respond to the terminal's instructions. When investment business recommendations are initiated, the server can obtain the current user's investment business data from the terminal and store the obtained investment business data in the storage unit. When data processing is required, the server retrieves volatile storage resources from the storage unit for the central processing unit to perform calculations.

[0038] Step S200: Calculate the correlation coefficient between the current investment business data and the original input data; wherein, the original input data is the training data of the model to be transferred.

[0039] The model to be transferred is an existing, pre-trained prediction model. The data type of the original input data used to train the model can be the same as or partially the same as the data type of the investment business data, and the classification of its prediction results can also be the same as or partially the same as the classification of the business recommendation model. For example, the original input data only uses the account balances, transfer records, investment amounts, and investment times of multiple customers. This data is input into the model to be transferred for training to obtain the predicted probabilities for three specific investment businesses. It is understood that the model architecture of the model to be transferred can be the same as or different from the architecture of the business recommendation model, as long as the prediction results of the two models can be combined.

[0040] Specifically, there are several methods for calculating the correlation coefficient between current investment business data and the original input data. The original input data serves as the training data for the model to be transferred, including data from multiple users. When calculating the correlation coefficient, features can be summarized and extracted from the data of multiple users, and the extracted representative data can be compared with the current investment business data to obtain the correlation coefficient. Alternatively, the current investment business data can be calculated separately with the data of each user in the original input data, and all the calculation results can be summarized to obtain the correlation coefficient. The correlation coefficient reflects the degree of correlation between the current investment business data and the original input data.

[0041] Step S300: Input the current investment business data into the model to be migrated to obtain the first recommendation result.

[0042] Specifically, the acquired current investment business data is input into the trained transfer model. After analysis and processing by the model, the first recommendation result is obtained. It can be understood that the first recommendation result can be a sequence of probability values ​​for different investment businesses, calculated based on the current investment business data. For example, it could be: Investment business A, 0.75; Investment business B, 0.2; Investment business C, 0.05. The first recommendation result varies depending on the transfer model.

[0043] Step S400: Input the current investment business data into the business recommendation model to obtain the second recommendation result; wherein, the business recommendation model is trained from historical investment business data.

[0044] Specifically, the business recommendation model is trained using historical investment data. This historical investment data comprises investment transactions from multiple users at different points in time, with each historical data point sharing the same data type as the current investment data. Since the business recommendation model is trained directly on this historical data, inputting the current investment data into the model yields a relatively accurate second recommendation that predicts the type of investment the current user will purchase.

[0045] Step S500: Process the correlation coefficient, the first recommendation result, and the second recommendation result, and output the target recommendation result; the target recommendation result is used to recommend investment business.

[0046] The calculated correlation coefficient, first recommendation result, and second recommendation result are weighted to obtain the corresponding target recommendation result. For example, the weight of the first recommendation result in the target recommendation result is calculated based on the correlation coefficient. Different weights result in varying degrees of influence of the first recommendation result on the target recommendation result. For instance, when the current investment data and the original input data have the same data type, the correlation is highest, the calculated correlation coefficient is closer to 1, and the first recommendation result obtained by the transfer model is more accurate. In this case, the weights of the first and second recommendation results in the target recommendation result can be set to 1:1, meaning that the first and second recommendation results have equal influence on the target recommendation result. Conversely, when the data types of the current investment data and the original input data differ significantly, the influence of the first recommendation result on the target recommendation result is smaller. During the calculation of the target recommendation result, the existing trained transfer model can be directly utilized, allowing the knowledge learned by the transfer model to be incorporated into the target recommendation result, enabling prediction from multiple perspectives and improving the predictive performance of the target recommendation result.

[0047] The target recommendation results are used to recommend investment opportunities. Specifically, the output target recommendation results have the same data type as the first and second recommendation results, consisting of a sequence of predicted probability values ​​for different investment opportunities. When recommending investment opportunities, the system can select the investment opportunity with the highest probability value, or it can sort the investment opportunities according to their probability values ​​and recommend them to the user for selection. The specific method of recommending investment opportunities can be changed according to actual needs and is not limited here.

[0048] In the aforementioned business recommendation method, current investment business data is input into both the model to be transferred and the business recommendation model to obtain a first recommendation result and a second recommendation result. Simultaneously, the correlation coefficient between the current investment business data and the original input data is calculated to determine the difference between the current investment data and the original input data. Since the model to be transferred is trained using the original input data, by processing the correlation coefficient and the first recommendation result, and combining it with the second recommendation result, the target recommendation result can learn the knowledge from the model to be transferred. This allows for multi-faceted data mining, fully utilizing the model resources of the model to be transferred and improving the predictive performance of the recommendation results.

[0049] In one embodiment, such as Figure 3 As shown, the steps for calculating the correlation coefficient between the current investment data and the original input data include:

[0050] Step S210: Perform dimensional transformation on the current investment business data to obtain the current feature vector.

[0051] Specifically, in this embodiment, the data types of the current investment business data and the original input data are different, making it inconvenient to calculate the correlation coefficient. Therefore, it is necessary to perform dimensionality transformation on the current investment business data, that is, to upgrade or reduce the dimensionality of the current investment business data so that the resulting current feature vector has the same dimension as the original input data. For example, if the current investment business data is a three-dimensional feature (u, d, n), while the original input data of the model to be transferred is two-dimensional, then it is necessary to perform dimensionality reduction on the current investment business data to convert the current feature vector into a two-dimensional feature (u, d*n), which facilitates the subsequent calculation of the correlation coefficient.

[0052] Step S220: Obtain the training feature vector of the original input data; wherein, the training feature vector includes multiple input feature vectors, and the current feature vector has the same data dimension as the input feature vector.

[0053] Specifically, during model training, the original input data is converted into training feature vectors in matrix form. These training feature vectors include multiple input feature vectors, each with the same dimensionality as the current feature vector. It's understandable that since the model to be transferred has already been trained on the original input data, its training vector features can be stored in a data storage system and directly accessed when needed.

[0054] Step S230: Calculate the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient.

[0055] Specifically, this embodiment obtains the correlation coefficient by calculating the similarity between the current feature vector and the training feature vector. Similarity can be calculated by determining the distance between two vectors; the closer the distance, the greater the similarity. Similarity can be calculated using several methods, including Euclidean distance, cosine similarity, Spearman correlation coefficient, log-likelihood similarity, and Manhattan distance. It is understood that the calculated similarity can be used directly as the correlation coefficient, or it can be normalized before being used as the correlation coefficient.

[0056] In one embodiment, the step of performing dimensional transformation processing on the current investment business data to obtain the current feature vector includes: inputting the current investment business data into a fully connected neural network for dimensional transformation processing to obtain the current feature vector.

[0057] Specifically, this embodiment uses a fully connected neural network for dimensionality transformation. In this network, any node in the previous layer is connected to all nodes in the current layer. By adjusting the number of nodes in the output layer, the dimensionality of the input current investment business data can be transformed, resulting in a current feature vector with the same dimension as the input feature vector. In some other embodiments, backpropagation neural networks or convolutional neural networks can also be used for dimensionality transformation.

[0058] In one embodiment, such as Figure 4 As shown, the steps for calculating the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient include:

[0059] Step S231: Calculate the similarity between the current feature vector and each input feature vector to obtain multiple similarity values.

[0060] Specifically, in this embodiment, the input feature vectors in the training feature vectors are obtained sequentially, and the similarity between the current feature vector and each input feature vector is calculated respectively, thereby obtaining multiple similarity values. Similarity calculation can be performed using the following methods: Euclidean distance, cosine similarity, Spearman correlation coefficient, log-likelihood similarity, and Manhattan distance, etc.

[0061] Step S232: Calculate the average of multiple similarity values ​​and use this average as the correlation coefficient. Specifically, in this embodiment, the multiple similarity values ​​are summed and divided by the number of similarity values ​​to obtain the average, i.e., the correlation coefficient. The correlation system calculated in this way can reflect the overall correlation between the current investment business data and the original input data for training the transfer model. In some other embodiments, the average of multiple input feature vectors can be calculated first, and then the similarity between this average and the current feature vector can be calculated to obtain the same correlation coefficient.

[0062] In one embodiment, the step of calculating the similarity between the current feature vector and each input feature vector includes: calculating the Euclidean similarity between the current feature vector and each input feature vector.

[0063] Specifically, this embodiment calculates the similarity between the current feature vector and the input feature vector using Euclidean similarity. First, the Euclidean distance is calculated using the following formula:

[0064]

[0065] Here, x and y are two vectors in n-dimensional space, namely the current feature vector and the input feature vector, respectively. Then, the Euclidean similarity s is calculated using the Euclidean distance d, where s = 1 / (1+d). The Euclidean similarity ranges from [0,1], with a value closer to 1 indicating a smaller Euclidean distance, meaning the two vectors are closer and more similar. The correlation coefficient calculated using Euclidean similarity also ranges from [0,1], facilitating its calculation with the predicted probability values ​​of different investment activities.

[0066] In one embodiment, such as Figure 5 As shown, the steps for processing the correlation coefficient, the first recommendation result, and the second recommendation result to output the target recommendation result include:

[0067] Step S510: Obtain relevant recommendation results; the relevant recommendation results include the product of the correlation coefficient and the first recommendation result.

[0068] Specifically, in this embodiment, the larger the correlation coefficient, the more similar the investment business data is to the original input data. For example, it can be calculated using Euclidean similarity or cosine similarity. When processing the correlation coefficient and the first recommendation result, the correlation coefficient is directly multiplied by the first recommendation result to obtain the relevant recommendation result. For a specific example, if the calculated correlation coefficient is 0.6, the first recommendation result is: Investment business A, 0.6; Investment business B, 0.3; Investment business C, 0.1. The relevant recommendation result obtained after calculation is: Investment business A, 0.36; Investment business B, 0.18; Investment business C, 0.06.

[0069] Step S520: The sum of the relevant recommendation results and the second recommendation results is determined as the target recommendation result.

[0070] Specifically, since the business recommendation model is trained based on historical investment business data, its prediction results are relatively accurate and have a significant impact on the target recommendation result. Therefore, the calculated relevant recommendation result is directly added to the second recommendation result to obtain the target recommendation result. For example, the relevant recommendation results are: Investment Business A, 0.36; Investment Business B, 0.18; Investment Business C, 0.06. The second recommendation result is: Investment Business A, 0.2; Investment Business B, 0.7; Investment Business C, 0.1. The calculated target recommendation result is: Investment Business A, 0.56; Investment Business B, 0.88; Investment Business C, 0.16. At this point, when recommending investment businesses based on the target recommendation result, Investment Business B will be recommended first.

[0071] In one embodiment, the business recommendation method further includes: inputting historical investment business data into a long short-term memory network for model training to obtain a business recommendation model.

[0072] Specifically, this embodiment uses a Long Short-Term Memory (LSTM) network to perform time-series modeling of historical business investment data. Its structure is as follows: Figure 6 As shown in the figure, σ corresponds to the sigmoid function, τ corresponds to the tanh function, and ⊕ represents matrix addition. This involves matrix dot product. When modeling using a Long Short-Term Memory (LSTM) network, each user's business investment data, after being processed by a nonlinear function, generates two variables, c and h. c represents the business parameter information passed from the previous day's investment data to the current time; h represents the hidden variable resulting from the current day's investment data processed by the nonlinear function. In the model, h from the last day is taken as the final prediction feature. After passing through a simple fully connected network, the second recommendation result can be obtained.

[0073] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0074] Based on the same inventive concept, this application also provides a business recommendation apparatus for implementing the aforementioned business recommendation method. The solution provided by this apparatus is similar to the implementation described in the above method; therefore, the specific limitations in one or more business recommendation apparatus embodiments provided below can be found in the limitations of the business recommendation method described above, and will not be repeated here.

[0075] In one embodiment, such as Figure 7 As shown, a business recommendation device is provided, including:

[0076] The data acquisition module 710 is used to acquire current investment business data; wherein, the investment business data includes user profile data and investment record data;

[0077] The first calculation module 720 is used to calculate the correlation coefficient between the current investment business data and the original input data; wherein, the original input data is the training data of the model to be transferred.

[0078] The first prediction module 730 is used to input the current investment business data into the model to be migrated to obtain the first recommendation result;

[0079] The second prediction module 740 is used to input the current investment business data into the business recommendation model to obtain the second recommendation result; wherein, the business recommendation model is trained from historical investment business data;

[0080] The second calculation module 750 is used to process the correlation coefficient, the first recommendation result, and the second recommendation result, and output the target recommendation result; the target recommendation result is used to recommend investment business.

[0081] In the aforementioned business recommendation device, current investment business data is input into both the model to be transferred and the business recommendation model to obtain a first recommendation result and a second recommendation result. Simultaneously, the correlation coefficient between the current investment business data and the original input data is calculated to determine the difference between the current investment data and the original input data. Since the model to be transferred is trained using the original input data, by processing the correlation coefficient and the first recommendation result, and combining it with the second recommendation result, the target recommendation result can learn the knowledge from the model to be transferred. This allows for multi-faceted data mining, fully utilizing the model resources of the model to be transferred and improving the predictive performance of the recommendation results.

[0082] In one embodiment, the first computing module 720 includes:

[0083] The dimension transformation unit is used to perform dimension transformation processing on the current investment business data to obtain the current feature vector.

[0084] The vector acquisition unit is used to acquire the training feature vector of the original input data; wherein the training feature vector includes multiple input feature vectors, and the current feature vector has the same data dimension as the input feature vectors;

[0085] The correlation coefficient acquisition unit is used to calculate the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient.

[0086] In one embodiment, the dimension transformation unit is used to input the current investment business data into a fully connected neural network for dimension transformation processing to obtain the current feature vector.

[0087] In one embodiment, the correlation coefficient acquisition unit is used to calculate the similarity between the current feature vector and each input feature vector to obtain multiple similarity values; calculate the average of the multiple similarity values, and use the average as the correlation coefficient.

[0088] In one embodiment, the correlation coefficient acquisition unit is used to calculate the Euclidean similarity between the current feature vector and each input feature vector.

[0089] In one embodiment, the second calculation module 750 is used to obtain relevant recommendation results; the relevant recommendation results include the product of the correlation coefficient and the first recommendation result; the sum of the relevant recommendation results and the second recommendation result is determined as the target recommendation result.

[0090] In one embodiment, the business recommendation device further includes a model acquisition module, used to input historical investment business data into a long short-term memory network for model training to obtain a business recommendation model.

[0091] Each module in the aforementioned business recommendation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0092] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores investment business data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a business recommendation method.

[0093] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 9As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a business recommendation method. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0094] Those skilled in the art will understand that Figure 8 , Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0095] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: acquiring current investment business data; wherein the investment business data includes user profile data and investment record data; calculating the correlation coefficient between the current investment business data and the original input data; wherein the original input data is training data for a model to be transferred; inputting the current investment business data into the model to be transferred to obtain a first recommendation result; inputting the current investment business data into a business recommendation model to obtain a second recommendation result; wherein the business recommendation model is trained using historical investment business data; processing the correlation coefficient, the first recommendation result, and the second recommendation result, and outputting a target recommendation result; the target recommendation result is used to recommend investment businesses.

[0096] In the aforementioned computer equipment, current investment business data is input into the model to be transferred and the business recommendation model, respectively, to obtain a first recommendation result and a second recommendation result. Simultaneously, the correlation coefficient between the current investment business data and the original input data is calculated to determine the difference between the current investment data and the original input data. Since the model to be transferred is trained using the original input data, by processing the correlation coefficient and the first recommendation result, and combining it with the second recommendation result, the target recommendation result can learn the knowledge from the model to be transferred. This allows for multi-faceted data mining, fully utilizing the model resources of the model to be transferred and improving the predictive performance of the recommendation results.

[0097] In one embodiment, when the processor executes the computer program, it further performs the following steps: calculating the correlation coefficient between the current investment business data and the original input data, including: performing dimensional transformation processing on the current investment business data to obtain the current feature vector; obtaining the training feature vector of the original input data; wherein the training feature vector includes multiple input feature vectors, and the current feature vector has the same data dimension as the input feature vector; calculating the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient.

[0098] In one embodiment, when the processor executes the computer program, it further implements the following steps: the step of performing dimensional transformation processing on the current investment business data to obtain the current feature vector, including: inputting the current investment business data into a fully connected neural network for dimensional transformation processing to obtain the current feature vector.

[0099] In one embodiment, when the processor executes the computer program, it further implements the following steps: calculating the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient, including: calculating the similarity between the current feature vector and each input feature vector respectively to obtain multiple similarity values; calculating the average of the multiple similarity values ​​and using the average as the correlation coefficient.

[0100] In one embodiment, when the processor executes the computer program, it further performs the following steps: calculating the similarity between the current feature vector and each input feature vector, including calculating the Euclidean similarity between the current feature vector and each input feature vector.

[0101] In one embodiment, when the processor executes the computer program, it further implements the following steps: processing the correlation coefficient, the first recommendation result, and the second recommendation result, and outputting the target recommendation result, including: obtaining the relevant recommendation result; the relevant recommendation result includes the product of the correlation coefficient and the first recommendation result; and determining the sum of the relevant recommendation result and the second recommendation result as the target recommendation result.

[0102] In one embodiment, when the processor executes the computer program, it also performs the following steps: inputting historical investment business data into a long short-term memory network for model training to obtain a business recommendation model.

[0103] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When executed by a processor, the computer program performs the following steps: acquiring current investment business data; wherein the investment business data includes user profile data and investment record data; calculating the correlation coefficient between the current investment business data and the original input data; wherein the original input data is the training data of the model to be transferred; inputting the current investment business data into the model to be transferred to obtain a first recommendation result; inputting the current investment business data into a business recommendation model to obtain a second recommendation result; wherein the business recommendation model is trained from historical investment business data; processing the correlation coefficient, the first recommendation result, and the second recommendation result, and outputting a target recommendation result; the target recommendation result is used to recommend investment businesses.

[0104] In the aforementioned computer-readable storage medium, when the computer program is executed by a processor, it inputs current investment business data into the model to be transferred and the business recommendation model respectively, thereby obtaining a first recommendation result and a second recommendation result. Simultaneously, it calculates the correlation coefficient between the current investment business data and the original input data to obtain the difference between the current investment data and the original input data. Since the model to be transferred is trained using the original input data, by processing the correlation coefficient and the first recommendation result, and combining it with the second recommendation result, the target recommendation result can learn the knowledge in the model to be transferred, mining the data from multiple perspectives to fully utilize the model resources of the model to be transferred and improve the predictive performance of the recommendation result.

[0105] In one embodiment, when the computer program is executed by the processor, it further implements the following steps: the step of calculating the correlation coefficient between the current investment business data and the original input data, including: performing dimensional transformation processing on the current investment business data to obtain the current feature vector; obtaining the training feature vector of the original input data; wherein the training feature vector includes multiple input feature vectors, and the current feature vector has the same data dimension as the input feature vector; calculating the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient.

[0106] In one embodiment, when the computer program is executed by the processor, it further implements the following steps: performing dimensional transformation processing on the current investment business data to obtain the current feature vector, including: inputting the current investment business data into a fully connected neural network for dimensional transformation processing to obtain the current feature vector.

[0107] In one embodiment, when the computer program is executed by the processor, it further implements the following steps: calculating the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient, including: calculating the similarity between the current feature vector and each input feature vector respectively to obtain multiple similarity values; calculating the average of the multiple similarity values ​​and using the average as the correlation coefficient.

[0108] In one embodiment, when the computer program is executed by the processor, it further implements the following steps: calculating the similarity between the current feature vector and each input feature vector, including: calculating the Euclidean similarity between the current feature vector and each input feature vector.

[0109] In one embodiment, when the computer program is executed by the processor, it further implements the following steps: processing the correlation coefficient, the first recommendation result, and the second recommendation result, and outputting the target recommendation result, including: obtaining the relevant recommendation result; the relevant recommendation result includes the product of the correlation coefficient and the first recommendation result; and determining the sum of the relevant recommendation result and the second recommendation result as the target recommendation result.

[0110] In one embodiment, when the computer program is executed by the processor, it also performs the following steps: inputting historical investment business data into a long short-term memory network for model training to obtain a business recommendation model.

[0111] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0112] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0113] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0114] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A business recommendation method, characterized in that, The method includes: Obtain current investment business data; wherein, the investment business data includes user profile data and investment record data; Calculate the correlation coefficient between the current investment business data and the original input data; wherein, the original input data is the training data of the model to be transferred. The current investment business data is input into the model to be migrated to obtain the first recommendation result; The current investment business data is input into the business recommendation model to obtain a second recommendation result; wherein, the business recommendation model is trained using historical investment business data; The correlation coefficient, the first recommendation result, and the second recommendation result are processed to output a target recommendation result; the target recommendation result is used to recommend investment opportunities. The step of calculating the correlation coefficient between the current investment business data and the original input data includes: The current investment business data is transformed to obtain the current feature vector; Obtain the training feature vector of the original input data; wherein the training feature vector includes multiple input feature vectors, and the current feature vector has the same data dimension as the input feature vectors; Calculate the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient; The step of processing the correlation coefficient, the first recommendation result, and the second recommendation result to output the target recommendation result includes: Obtain relevant recommendation results; the relevant recommendation results include the product of the correlation coefficient and the first recommendation result; The sum of the relevant recommendation result and the second recommendation result is determined as the target recommendation result.

2. The method according to claim 1, characterized in that, The step of performing dimensional transformation on the current investment business data to obtain the current feature vector includes: The current investment business data is input into a fully connected neural network for dimensionality transformation to obtain the current feature vector.

3. The method according to claim 1, characterized in that, The step of calculating the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient includes: Calculate the similarity between the current feature vector and each of the input feature vectors to obtain multiple similarity values; Calculate the average of the multiple similarity values, and use the average as the correlation coefficient.

4. The method according to claim 3, characterized in that, The step of calculating the similarity between the current feature vector and each of the input feature vectors includes: Calculate the Euclidean similarity between the current feature vector and each of the input feature vectors.

5. The method according to claim 1, characterized in that, The method further includes: The historical investment business data is input into a long short-term memory network for model training to obtain the business recommendation model.

6. A business recommendation device, characterized in that, The device includes: The data acquisition module is used to acquire current investment business data; wherein, the investment business data includes user profile data and investment record data; The first calculation module is used to calculate the correlation coefficient between the current investment business data and the original input data; wherein, the original input data is the training data of the model to be transferred. The first prediction module is used to input the current investment business data into the model to be migrated to obtain a first recommendation result; The second prediction module is used to input the current investment business data into the business recommendation model to obtain a second recommendation result; wherein, the business recommendation model is trained from the historical investment business data; The second calculation module is used to process the correlation coefficient, the first recommendation result, and the second recommendation result, and output the target recommendation result; the target recommendation result is used to recommend investment business. The first calculation module is specifically used to perform dimensional transformation processing on the current investment business data to obtain a current feature vector; obtain the training feature vector of the original input data; wherein, the training feature vector includes multiple input feature vectors, and the current feature vector has the same data dimension as the input feature vector; calculate the similarity between the current feature vector and the training feature vector to obtain the correlation coefficient; The second calculation module is specifically used to obtain relevant recommendation results; the relevant recommendation results include the product of the correlation coefficient and the first recommendation result; the sum of the relevant recommendation results and the second recommendation result is determined as the target recommendation result.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images