Business event prediction method and device

By processing historical business data to obtain feature vectors of time period dimensions and training classification models, the problem of low prediction accuracy caused by not taking into account timing characteristics in the prior art is solved, and the accuracy and user experience of business event prediction are improved.

CN119089254APending Publication Date: 2024-12-06BEIJING JINGDONG QIANSHITECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310661482.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-06-06
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

In the prior art, the business event prediction model does not consider the timing characteristics of the business event, resulting in low prediction accuracy and poor user experience.

Method used

By obtaining historical business data in the target business scenario, the feature vectors in multiple time period dimensions are obtained, the training set and verification set are selected according to the order of time periods, the classification model is trained, and the business event information to be predicted is received. Perform classification calculations and output prediction results.

Benefits of technology

It improves the accuracy of business event prediction, improves the user experience, and overcomes the problem of low prediction accuracy caused by unacceptable timing characteristics.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089254A_ABST
    Figure CN119089254A_ABST
Patent Text Reader

Abstract

The invention discloses a business event prediction method and device, and relates to the technical field of computers. A specific embodiment of the method comprises the steps of obtaining historical business data in a target business scene, and processing the historical business data to obtain feature vectors in multiple time period dimensions; according to the sequence of the time periods, selecting a training set and a verification set from the feature vectors to train the classification model; and receiving to-be-predicted target business event information, inputting a target business event type and a target time period indicated by the target business event information into the optimized and trained classification model for classification calculation, and outputting a prediction result corresponding to a target business event. According to the embodiment, the accuracy of business event prediction is improved, and the user experience is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a business event prediction method and device. Background Art

[0002] With the development of Internet technology, the application scenarios of business systems are gradually expanding. Based on the full-process data tracking of business events by business systems, how to predict business events is a hot topic in current research.

[0003] In the related technology, historical business data is usually processed by removing outliers and then directly input into the prediction model for calculation to achieve the prediction of business events. However, since the input data does not take into account the time series characteristics of business event triggering, the prediction accuracy is low and the user experience is poor. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a business event prediction method and device, which can process historical business data based on time series characteristics, improve the accuracy of business event prediction, and enhance user experience.

[0005] To achieve the above object, according to one aspect of an embodiment of the present invention, a service event prediction method is provided, comprising:

[0006] Obtain historical business data under the target business scenario, and process the historical business data to obtain feature vectors under multiple time period dimensions;

[0007] According to the order of the time periods, a training set and a validation set are selected from the feature vector to train the classification model;

[0008] Receive target business event information to be predicted, input the target business event type and target time period indicated by the target business event information into the optimized and trained classification model for classification calculation, and output the prediction result corresponding to the target business event.

[0009] Furthermore, before the step of processing the historical business data to obtain feature vectors in multiple time period dimensions, the method further includes:

[0010] Determine the characteristic data of the historical business data of each business event type in different time periods; wherein the characteristic data indicates the triggering times and triggering duration of the corresponding business event type in the corresponding time period.

[0011] Furthermore, the step of processing the historical business data to obtain feature vectors in multiple time period dimensions also includes:

[0012] Process the feature data to obtain feature matrices in multiple time period dimensions;

[0013] The feature matrix is ​​normalized and filled with null values ​​to obtain the feature vector.

[0014] Furthermore, the step of processing the feature data to obtain a feature matrix of multiple time period dimensions also includes:

[0015] Determine the domain and event subject information of the corresponding business event type based on the characteristic data;

[0016] Calculate at least one of the following processing results: the number of triggers of the corresponding business event type, the coefficient of variation corresponding to the trigger duration, the difference in coefficient of variation between adjacent time periods, the mean within multiple time periods, and the variance within multiple time periods, so as to complete the processing of the feature data;

[0017] The processing result of each business event type is used as the feature column to construct a feature matrix of feature data in multiple time period dimensions.

[0018] Furthermore, the method further comprises:

[0019] Obtain the actual triggering conditions of the target business events and optimize the classification model based on the actual triggering conditions.

[0020] According to another aspect of an embodiment of the present invention, there is provided a service event prediction device, characterized in that it includes:

[0021] An acquisition module is used to acquire historical business data under a target business scenario and process the historical business data to obtain feature vectors under multiple time period dimensions;

[0022] A training module is used to select a training set and a validation set from the feature vector according to the sequence of time periods to train the classification model;

[0023] The prediction module is used to receive the target business event information to be predicted, input the target business event type and target time period indicated by the target business event information into the optimized and trained classification model for classification calculation, and output the prediction result corresponding to the target business event.

[0024] Furthermore, the device further includes a feature data determination module, which is used to:

[0025] Determine the characteristic data of the historical business data of each business event type in different time periods; wherein the characteristic data indicates the triggering times and triggering duration of the corresponding business event type in the corresponding time period.

[0026] Furthermore, the acquisition module is also used to:

[0027] Determine the domain and event subject information of the corresponding business event type based on the characteristic data;

[0028] Calculate at least one of the following processing results: the number of triggers of the corresponding business event type, the coefficient of variation corresponding to the trigger duration, the difference in coefficient of variation between adjacent time periods, the mean within multiple time periods, and the variance within multiple time periods, so as to complete the processing of the feature data;

[0029] The processing result of each business event type is used as the feature column to construct a feature matrix of feature data in multiple time period dimensions.

[0030] According to another aspect of an embodiment of the present invention, there is provided an electronic device for predicting a business event, characterized in that it includes:

[0031] one or more processors;

[0032] a storage device for storing one or more programs,

[0033] When one or more programs are executed by one or more processors, the one or more processors implement any of the business event prediction methods described above.

[0034] According to another aspect of an embodiment of the present invention, there is provided a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, any of the above-mentioned business event prediction methods is implemented.

[0035] An embodiment of the above invention has the following advantages or beneficial effects: because historical business data under the target business scenario is obtained and processed to obtain feature vectors under multiple time period dimensions; according to the order of the time periods, a training set and a validation set are selected from the feature vector to train the classification model; the target business event information to be predicted is received, and the target business event type and target time period indicated by the target business event information are input into the optimized trained classification model for classification calculation, and the prediction result corresponding to the target business event is output. Therefore, the technical problem in the related technology that the input data does not take into account the timing characteristics of the business event triggering, resulting in low prediction accuracy and poor user experience is overcome, thereby achieving the technical effect of being able to process historical business data based on timing characteristics, improving the accuracy of business event prediction, and improving user experience.

[0036] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0038] Figure 1 is a schematic diagram of the main process of a business event prediction method provided according to an embodiment of the present invention;

[0039] Figure 2 is a schematic diagram of the main process of a business event prediction method provided according to yet another embodiment of the present invention;

[0040] Figure 3 is a schematic diagram of main modules of a business event prediction device provided according to an embodiment of the present invention;

[0041] Figure 4 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;

[0042] Figure 5 It is a schematic diagram of the structure of a computer system of a terminal device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0044] Figure 1 FIG. 1 is a schematic diagram of the main process of a business event prediction method according to an embodiment of the present invention; Figure 1 As shown, the business event prediction method provided by the embodiment of the present invention mainly includes:

[0045] Step S101, historical business data in a target business scenario is acquired, and the historical business data is processed to obtain feature vectors in multiple time period dimensions.

[0046] Specifically, the selection of the target business scenario needs to correspond to the subsequent business events to be predicted. For example, whether the employee leave event of the business event to be predicted will be triggered in the next month, the corresponding business scenario is the employee's attendance record; if the business event to be predicted is the website user churn rate, the corresponding business scenario is the website's registration and deregistration record; if the business event to be predicted is whether the user's payment overdue is triggered, the corresponding business scenario is the user's historical payment record. The above description is only an example, and the specific target business scenario and business event can be adjusted according to actual conditions.

[0047] According to an embodiment of the present invention, before the step of processing the historical business data to obtain feature vectors in multiple time period dimensions, the method further includes:

[0048] Eliminate outliers in historical business data.

[0049] By eliminating outliers, more accurate statistical quantities can be obtained, which in turn helps to improve the accuracy of subsequent predictions.

[0050] Based on the above technical means, the present invention not only eliminates outliers in historical business data, but also processes the historical business data to obtain feature vectors under multiple time period dimensions, thereby realizing the processing of historical business data based on time series characteristics, thereby more accurately predicting the triggering probability of business events in a future period of time, thereby improving user experience.

[0051] Further, according to an embodiment of the present invention, before the step of processing the historical business data to obtain feature vectors in multiple time period dimensions, the method further includes:

[0052] Determine the characteristic data of the historical business data of each business event type in different time periods; wherein the characteristic data indicates the triggering times and triggering duration of the corresponding business event type in the corresponding time period.

[0053] Specifically, different time periods can be divided into different time period intervals such as days, weeks, months, and years, so as to more accurately understand the triggering frequency of different business event types in different time period intervals, and then in the subsequent prediction of the business event to be predicted, the future time period to be predicted of the business event to be predicted can be combined to achieve more accurate prediction.

[0054] Preferably, according to an embodiment of the present invention, the step of processing the historical business data to obtain feature vectors in multiple time period dimensions further includes:

[0055] Process the feature data to obtain feature matrices in multiple time period dimensions;

[0056] The feature matrix is ​​normalized and filled with null values ​​to obtain the feature vector.

[0057] By determining the feature matrix for the feature data of each business event type determined above in different time periods, and then obtaining the feature vector through normalization and null value filling, the data calculated in the input classification model can be uniformly divided into time periods, which helps to improve the efficiency of business event prediction.

[0058] Further, according to an embodiment of the present invention, the step of processing the feature data to obtain a feature matrix of multiple time period dimensions further includes:

[0059] Determine the domain and event subject information of the corresponding business event type based on the characteristic data;

[0060] Calculate at least one of the following processing results: the number of triggers of the corresponding business event type, the coefficient of variation corresponding to the trigger duration, the difference in coefficient of variation between adjacent time periods, the mean within multiple time periods, and the variance within multiple time periods, so as to complete the processing of the feature data;

[0061] The processing result of each business event type is used as the feature column to construct a feature matrix of feature data in multiple time period dimensions.

[0062] According to an embodiment of the present invention, before the step of processing the characteristic data, determining the field to which the corresponding business event type belongs and the event subject information helps to more accurately divide the object information that triggers the corresponding business event (including the triggering subject information and the belonging information of the triggering subject). It can not only realize the trigger probability prediction of a certain business event corresponding to a certain user in a period of time in the future, but also make probability predictions on the corresponding business events of a certain department (field), further expanding the applicable scenarios of business event prediction.

[0063] Step S102, selecting a training set and a validation set from the feature vector according to the sequence of the time periods to train the classification model.

[0064] From the feature vector, the feature data in the earlier time period is preferentially selected as the training set, and the feature data in the later time period is selected as the validation set, so as to realize the training of the classification model.

[0065] According to a specific implementation of the embodiment of the present invention, taking the prediction of whether a certain business event is triggered within the next three months as an example, it is assumed that the prediction label of the business event triggering is Y.

[0066] The logic of setting Y: the number of days from the business event triggering time is Y` = the time point of marking the business event triggering field - the time dt of the feature matrix. If Y` is less than 90 days in the training set within the specified time period, set Y = 1; if Y` is greater than or equal to 90 days, set Y = 0. Based on this, the classification model is trained.

[0067] Step S103, receiving target business event information to be predicted, inputting the target business event type and target time period indicated by the target business event information into the optimized and trained classification model for classification calculation, and outputting the prediction result corresponding to the target business event.

[0068] In order to further ensure the prediction accuracy, the target time period corresponding to the target business event to be predicted should be selected within the middle time period of the multiple time periods counted above, and the predicted time period should not be too long to ensure the prediction accuracy.

[0069] Furthermore, according to an embodiment of the present invention, the method further includes:

[0070] Obtain the actual triggering conditions of the target business events and optimize the classification model based on the actual triggering conditions.

[0071] Through the above settings, the classification model is optimized to further improve the accuracy of business event prediction.

[0072] According to the technical solution of the embodiment of the present invention, the historical business data in the target business scenario is obtained, and the historical business data is processed to obtain feature vectors in multiple time period dimensions; according to the sequence of the time periods, a training set and a validation set are selected from the feature vector to train the classification model; the target business event information to be predicted is received, and the target business event type and the target time period indicated by the target business event information are input into the optimized trained classification model for classification calculation, and a prediction result corresponding to the target business event is output. Therefore, the technical problem in the related technology that the input data does not take into account the timing characteristics of the business event triggering, resulting in low prediction accuracy and poor user experience is overcome, thereby achieving the technical effect of being able to process historical business data based on timing characteristics, improving the accuracy of business event prediction, and enhancing user experience.

[0073] Figure 2 FIG. 4 is a schematic diagram of the main process of a business event prediction method according to another embodiment of the present invention; Figure 2 As shown, the business event prediction method provided by the embodiment of the present invention mainly includes:

[0074] Step S201, obtaining historical business data under a target business scenario, and removing abnormal values ​​in the historical business data.

[0075] Specifically, the selection of the target business scenario needs to correspond to the subsequent business events to be predicted. For example, whether the employee leave event of the business event to be predicted will be triggered in the next month, the corresponding business scenario is the employee's attendance record; if the business event to be predicted is the website user churn rate, the corresponding business scenario is the website's registration and deregistration record; if the business event to be predicted is whether the user's payment overdue is triggered, the corresponding business scenario is the user's historical payment record. The above description is only an example, and the specific target business scenario and business event can be adjusted according to actual conditions.

[0076] By eliminating outliers, more accurate statistical quantities can be obtained, which in turn helps to improve the accuracy of subsequent predictions.

[0077] Step S202, determining characteristic data of historical business data of each business event type in different time periods; wherein the characteristic data indicates the triggering times and triggering duration of the corresponding business event type in the corresponding time period.

[0078] Different time periods can be divided into different time intervals such as days, weeks, months, and years, so as to more accurately understand the triggering frequency of different business event types in different time intervals, and then in the subsequent prediction of the business event to be predicted, the future time period to be predicted of the business event to be predicted can be combined to achieve more accurate prediction.

[0079] Step S203, processing the feature data to obtain a feature matrix of multiple time period dimensions; normalizing and filling the feature matrix with null values ​​to obtain a feature vector.

[0080] By determining the feature matrix for the feature data of each business event type determined above in different time periods, and then obtaining the feature vector through normalization and null value filling, the data calculated in the input classification model can be uniformly divided into time periods, which helps to improve the efficiency of business event prediction.

[0081] Based on the above technical means, the present invention not only eliminates outliers in historical business data, but also processes the historical business data to obtain feature vectors under multiple time period dimensions, thereby realizing the processing of historical business data based on time series characteristics, thereby more accurately predicting the triggering probability of business events in the future period of time, thereby improving user experience.

[0082] Step S204, based on the feature data, determine the field and event subject information of the corresponding business event type; calculate the number of triggering times of the corresponding business event type, the coefficient of variation corresponding to the triggering duration, the difference in coefficient of variation in adjacent time periods, the mean in multiple time periods, and the variance in multiple time periods to complete the processing of the feature data; use the processing result of each business event type as a feature column to construct a feature matrix of the feature data in multiple time period dimensions.

[0083] According to the embodiments of the present invention, determining the field to which the corresponding business event type belongs and the event subject information helps to more accurately divide the object information that triggers the corresponding business event (including the triggering subject information and the belonging information of the triggering subject). It can not only realize the trigger probability prediction of a certain business event corresponding to a certain user in a period of time in the future, but also make probability predictions for the corresponding business events of a certain department (field), further expanding the applicable scenarios of business event predictions.

[0084] According to a specific implementation of the embodiment of the present invention, the feature data divided by historical business data may be processed in depth on a monthly or daily basis in the following manner:

[0085] Business fields with daily cycles (e.g. "Whether to take leave" field) are used to calculate the number of occurrences or duration in a month;

[0086] Calculate the coefficient of variation of the monthly frequency or duration of a business field

[0087] Calculate the difference between the coefficient of variation of a business field this month and last month.

[0088] Calculate the mean and variance of a business field from the beginning of this year to the present.

[0089] Calculate the relative position of a business field within a certain range (different fields, such as different departments, different company ages, etc.) (the relative position of a specific value within a certain range of values);

[0090] Calculate the difference between a business field and the mean in d;

[0091] Map the business fields of the quarterly dimension or annual dimension to the monthly dimension or daily dimension.

[0092] Each of the above processing methods obtains a feature column, and then forms a feature matrix according to the sequence of time periods.

[0093] Step S205 , selecting a training set and a validation set from the feature vector according to the sequence of the time periods to train the classification model.

[0094] From the feature vector, the feature data in the earlier time period is preferentially selected as the training set, and the feature data in the later time period is selected as the validation set, so as to realize the training of the classification model.

[0095] According to a specific implementation of the embodiment of the present invention, taking the prediction of whether a certain business event is triggered within the next three months as an example, it is assumed that the prediction label of the business event triggering is Y.

[0096] The logic of setting Y: the number of days from the business event triggering time is Y` = the time point of marking the business event triggering field - the time dt of the feature matrix. If Y` is less than 90 days in the training set within the specified time period, set Y = 1; if Y` is greater than or equal to 90 days, set Y = 0. Based on this, the classification model is trained.

[0097] Step S206, receiving target business event information to be predicted, inputting the target business event type and target time period indicated by the target business event information into the optimized trained classification model for classification calculation, and outputting the prediction result corresponding to the target business event.

[0098] In order to further ensure the prediction accuracy, the target time period corresponding to the target business event to be predicted should be selected within the middle time period of the multiple time periods counted above, and the predicted time period should not be too long to ensure the prediction accuracy.

[0099] Furthermore, according to an embodiment of the present invention, the method further includes:

[0100] Obtain the actual triggering conditions of the target business events and optimize the classification model based on the actual triggering conditions.

[0101] Through the above settings, the classification model is optimized to further improve the accuracy of business event prediction.

[0102] According to the technical solution of the embodiment of the present invention, the historical business data in the target business scenario is obtained, and the historical business data is processed to obtain feature vectors in multiple time period dimensions; according to the sequence of the time periods, a training set and a validation set are selected from the feature vector to train the classification model; the target business event information to be predicted is obtained, and the target business event type and the target time period indicated by the target business event information are input into the optimized trained classification model for classification calculation, and the prediction result corresponding to the target business event is output. Therefore, the technical problem in the related technology that the input data does not take into account the timing characteristics of the business event triggering, resulting in low prediction accuracy and poor user experience is overcome, thereby achieving the technical effect of being able to process historical business data based on timing characteristics, improving the accuracy of business event prediction, and enhancing user experience.

[0103] Figure 3 is a schematic diagram of the main modules of the service event prediction device provided according to an embodiment of the present invention; Figure 3 As shown, the business event prediction device 300 provided by the embodiment of the present invention mainly includes:

[0104] The acquisition module 301 is used to acquire historical business data in a target business scenario and process the historical business data to obtain feature vectors in multiple time period dimensions.

[0105] Specifically, the selection of the target business scenario needs to correspond to the subsequent business events to be predicted. For example, whether the employee leave event of the business event to be predicted will be triggered in the next month, the corresponding business scenario is the employee's attendance record; if the business event to be predicted is the website user churn rate, the corresponding business scenario is the website's registration and deregistration record; if the business event to be predicted is whether the user's payment overdue is triggered, the corresponding business scenario is the user's historical payment record. The above description is only an example, and the specific target business scenario and business event can be adjusted according to actual conditions.

[0106] According to an embodiment of the present invention, the business event prediction device 300 further includes an outlier processing module. Before the step of processing the historical business data to obtain feature vectors under multiple time period dimensions, the outlier processing module is used to:

[0107] Eliminate outliers in historical business data.

[0108] By eliminating outliers, more accurate statistical quantities can be obtained, which in turn helps to improve the accuracy of subsequent predictions.

[0109] Based on the above technical means, the present invention not only eliminates outliers in historical business data, but also processes the historical business data to obtain feature vectors under multiple time period dimensions, thereby realizing the processing of historical business data based on time series characteristics, thereby more accurately predicting the triggering probability of business events in a future period of time, thereby improving user experience.

[0110] Further, according to an embodiment of the present invention, before the step of processing the historical business data to obtain feature vectors in multiple time period dimensions, the acquisition module 301 is further used to:

[0111] Determine the characteristic data of the historical business data of each business event type in different time periods; wherein the characteristic data indicates the triggering times and triggering duration of the corresponding business event type in the corresponding time period.

[0112] Specifically, different time periods can be divided into different time period intervals such as days, weeks, months, and years, so as to more accurately understand the triggering frequency of different business event types in different time period intervals, and then in the subsequent prediction of the business event to be predicted, the future time period to be predicted of the business event to be predicted can be combined to achieve more accurate prediction.

[0113] Preferably, according to an embodiment of the present invention, the acquisition module 301 is further used for:

[0114] Process the feature data to obtain feature matrices in multiple time period dimensions;

[0115] The feature matrix is ​​normalized and filled with null values ​​to obtain the feature vector.

[0116] By determining the feature matrix for the feature data of each business event type determined above in different time periods, and then obtaining the feature vector through normalization and null value filling, the data calculated in the input classification model can be uniformly divided into time periods, which helps to improve the efficiency of business event prediction.

[0117] Further, according to an embodiment of the present invention, the acquisition module 301 is further used for:

[0118] Determine the domain and event subject information of the corresponding business event type based on the characteristic data;

[0119] Calculate at least one of the following processing results: the number of triggers of the corresponding business event type, the coefficient of variation corresponding to the trigger duration, the difference in coefficient of variation between adjacent time periods, the mean within multiple time periods, and the variance within multiple time periods, so as to complete the processing of the feature data;

[0120] The processing result of each business event type is used as the feature column to construct a feature matrix of feature data in multiple time period dimensions.

[0121] According to an embodiment of the present invention, before the step of processing the characteristic data, determining the department to which the corresponding business event type belongs and the event subject information helps to more accurately divide the object information that triggers the corresponding business event (including the triggering subject information and the information to which the triggering subject belongs). It can not only realize the trigger probability prediction of a certain business event corresponding to a certain user in a period of time in the future, but also make probability predictions on the corresponding business events of a certain department (field), further expanding the applicable scenarios of business event predictions.

[0122] The training module 302 is used to select a training set and a validation set from the feature vector according to the sequence of time periods to train the classification model.

[0123] From the feature vector, the feature data in the earlier time period is preferentially selected as the training set, and the feature data in the later time period is selected as the validation set, so as to realize the training of the classification model.

[0124] The prediction module 303 is used to receive target business event information to be predicted, input the target business event type and target time period indicated by the target business event information into the optimized and trained classification model for classification calculation, and output the prediction result corresponding to the target business event.

[0125] In order to further ensure the prediction accuracy, the target time period corresponding to the target business event to be predicted should be selected within the middle time period of the multiple time periods counted above, and the predicted time period should not be too long to ensure the prediction accuracy.

[0126] Furthermore, according to an embodiment of the present invention, the business event prediction device 300 further includes an optimization module, which is used to:

[0127] Obtain the actual triggering conditions of the target business events and optimize the classification model based on the actual triggering conditions.

[0128] Through the above settings, the classification model is optimized to further improve the accuracy of business event prediction.

[0129] According to the technical solution of the embodiment of the present invention, the historical business data in the target business scenario is obtained, and the historical business data is processed to obtain feature vectors in multiple time period dimensions; according to the sequence of the time periods, a training set and a validation set are selected from the feature vector to train the classification model; the target business event information to be predicted is received, and the target business event type and the target time period indicated by the target business event information are input into the optimized trained classification model for classification calculation, and a prediction result corresponding to the target business event is output. Therefore, the technical problem in the related technology that the input data does not take into account the timing characteristics of the business event triggering, resulting in low prediction accuracy and poor user experience is overcome, thereby achieving the technical effect of being able to process historical business data based on timing characteristics, improving the accuracy of business event prediction, and enhancing user experience.

[0130] Figure 4 An exemplary system architecture 400 is shown to which the business event prediction method or business event prediction device according to the embodiment of the present invention can be applied.

[0131] like Figure 4 As shown, system architecture 400 may include terminal devices 401, 402, 403, network 404 and server 405. Network 404 is used to provide a medium for communication links between terminal devices 401, 402, 403 and server 405. Network 404 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0132] Users can use terminal devices 401, 402, 403 to interact with server 405 through network 404 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 401, 402, 403, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only as examples).

[0133] The terminal devices 401 , 402 , and 403 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0134] The server 405 may be a server that provides various services, such as a server (for example only) that performs (testing / data processing) on ​​the user using the terminal devices 401, 402, and 403. The server may analyze and process the received historical business data and other data, and feed back the processing results (such as the prediction results corresponding to the target business event) to the terminal device.

[0135] It should be noted that the business event prediction method provided in the embodiment of the present invention is generally executed by the server 405 , and accordingly, the business event prediction device is generally disposed in the server 405 .

[0136] It should be understood that Figure 4 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0137] Reference below Figure 5 , which shows a schematic diagram of the structure of a computer system 500 of a terminal device or a server suitable for implementing an embodiment of the present invention. Figure 5 The terminal device or server shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0138] like Figure 5 As shown, the computer system 500 includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage part 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the system 500 are also stored. The CPU 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0139] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed, so that a computer program read therefrom is installed into the storage section 508 as needed.

[0140] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 509, and / or installed from the removable medium 511. When the computer program is executed by the central processing unit (CPU) 501, the above-mentioned functions defined in the system of the present invention are executed.

[0141] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with 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 the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may 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 using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0142] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0143] The modules involved in the embodiments of the present invention may be implemented in software or hardware. The modules described may also be provided in a processor, for example, may be described as: a processor including an acquisition module, a training module and a prediction module. The names of these modules do not, in some cases, constitute limitations on the modules themselves, for example, the acquisition module may also be described as "a module for acquiring historical business data in a target business scenario, processing the historical business data, and obtaining feature vectors in multiple time period dimensions".

[0144] As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device includes: obtaining historical business data under the target business scenario, processing the historical business data to obtain feature vectors under multiple time period dimensions; selecting a training set and a validation set from the feature vector according to the sequence of time periods to train a classification model; receiving target business event information to be predicted, inputting the target business event type and target time period indicated by the target business event information into the optimized trained classification model for classification calculation, and outputting the prediction result corresponding to the target business event.

[0145] According to the technical solution of the embodiment of the present invention, the historical business data in the target business scenario is obtained, and the historical business data is processed to obtain feature vectors in multiple time period dimensions; according to the sequence of the time periods, a training set and a validation set are selected from the feature vector to train the classification model; the target business event information to be predicted is received, and the target business event type and the target time period indicated by the target business event information are input into the optimized trained classification model for classification calculation, and a prediction result corresponding to the target business event is output. Therefore, the technical problem in the related technology that the input data does not take into account the timing characteristics of the business event triggering, resulting in low prediction accuracy and poor user experience is overcome, thereby achieving the technical effect of being able to process historical business data based on timing characteristics, improving the accuracy of business event prediction, and enhancing user experience.

[0146] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A business event prediction method, characterized in that: include: Acquire historical business data under a target business scenario, and process the historical business data to obtain feature vectors under multiple time period dimensions; Selecting a training set and a validation set from the feature vectors according to the sequence of the time periods to train the classification model; Receive target business event information to be predicted, input the target business event type and target time period indicated by the target business event information into the optimized and trained classification model for classification calculation, and output the prediction result corresponding to the target business event.

2. The business event prediction method according to claim 1, characterized in that: Before the step of processing the historical business data to obtain feature vectors in multiple time period dimensions, the method further includes: Determine characteristic data of historical business data of each business event type in different time periods; wherein the characteristic data indicates the number of triggering times and triggering duration of the corresponding business event type in the corresponding time period.

3. The business event prediction method according to claim 2, characterized in that: The step of processing the historical business data to obtain feature vectors in multiple time period dimensions also includes: Processing the characteristic data to obtain a characteristic matrix of multiple time period dimensions; The feature matrix is ​​normalized and filled with null values ​​to obtain a feature vector.

4. The business event prediction method according to claim 3, characterized in that: The step of processing the feature data to obtain a feature matrix of multiple time period dimensions also includes: Determine the domain and event subject information of the corresponding business event type according to the characteristic data; Calculate at least one of the following processing results: the number of triggers of the corresponding business event type, the coefficient of variation corresponding to the trigger duration, the difference in coefficient of variation in adjacent time periods, the mean in multiple time periods, and the variance in multiple time periods, so as to complete the processing of the feature data; The processing result of each business event type is used as a feature column to construct a feature matrix of the feature data in multiple time period dimensions.

5. The business event prediction method according to claim 1, characterized in that: The method further comprises: The actual triggering condition of the target business event is obtained, and the classification model is optimized according to the actual triggering condition.

6. A business event prediction device, characterized in that: include: An acquisition module is used to acquire historical business data in a target business scenario and process the historical business data to obtain feature vectors in multiple time period dimensions; A training module, used for selecting a training set and a validation set from the feature vectors according to the sequence of time periods, so as to train the classification model; The prediction module is used to receive the target business event information to be predicted, input the target business event type and target time period indicated by the target business event information into the optimized and trained classification model for classification calculation, and output the prediction result corresponding to the target business event.

7. The business event prediction device according to claim 6, characterized in that: The device further includes a feature data determination module, which is used to: Determine characteristic data of historical business data of each business event type in different time periods; wherein the characteristic data indicates the number of triggering times and triggering duration of the corresponding business event type in the corresponding time period.

8. The business event prediction device according to claim 7, characterized in that: The acquisition module is further used for: Determine the domain and event subject information of the corresponding business event type according to the characteristic data; Calculate at least one of the following processing results: the number of triggers of the corresponding business event type, the coefficient of variation corresponding to the trigger duration, the difference in coefficient of variation in adjacent time periods, the mean in multiple time periods, and the variance in multiple time periods, so as to complete the processing of the feature data; The processing result of each business event type is used as a feature column to construct a feature matrix of the feature data in multiple time period dimensions.

9. An electronic device for predicting business events, characterized in that: include: one or more processors; a storage device for storing one or more programs, 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.

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