Data prediction method and device, storage medium and electronic equipment

Abstract

The invention provides a data prediction method and device, a storage medium and electronic equipment, and the method comprises the steps: responding to a data recommendation instruction in a data recommendation service, obtaining platform service data related to the data recommendation instruction in a service platform, extracting feature data in the platform service data according to the service type of the data recommendation service, and carrying out the feature data prediction according to the extracted feature data; predicting a growth change trend based on the feature data, generating a first prediction result, obtaining event data of a target event identifier selected by the data recommendation instruction, adjusting the first prediction result based on a time parameter and an event influence value in the event data, and generating a second prediction result, and generating a recommendation result corresponding to the data recommendation instruction based on the second prediction result, and outputting the recommendation result on the service platform, by adopting the specification, the flexibility of predicting the event and the accuracy of determining the influence of the event are improved.

Description

Technical Field

[0001] This specification relates to the field of computer technology, and more specifically, to a data prediction method, apparatus, storage medium, and electronic device in the field of computer technology. Background Technology

[0002] Nowadays, with the continuous development of the economy and the times, in order to facilitate the planning of future tasks, it is necessary to make predictions based on task data so as to reasonably plan future goals. Therefore, how to make data predictions has become a problem that needs to be solved. Summary of the Invention

[0003] This specification provides a data prediction method, apparatus, storage medium, and electronic device. The method can adjust the prediction results generated by the model by pre-setting event parameters to determine the degree of influence of the event on the prediction results. By combining human-computer interaction and model prediction, it improves the flexibility of event prediction and the accuracy of determining the impact of the event.

[0004] Firstly, a data prediction method is provided, which includes: In response to a data recommendation instruction in the data recommendation service, the platform service data related to the data recommendation instruction is obtained from the service platform; Based on the service type of the data recommendation service, feature data is extracted from the platform service data, and the growth trend is predicted based on the feature data to generate a first prediction result. The event data of the target event identifier selected by the data recommendation instruction is obtained. The first prediction result is adjusted based on the time parameter and event impact value in the event data to generate a second prediction result. The event impact value represents the degree of influence of the event data. The time parameter is used to adjust the target time span and target unit duration of the first prediction result. Based on the second prediction result, a recommendation result corresponding to the data recommendation instruction is generated, and the recommendation result is output on the service platform.

[0005] The above technical solution enables the acquisition of platform service data corresponding to the data recommendation instruction triggered in the data recommendation service within the service platform. Based on the service type, the platform service data is used to predict the recommendation effect and generate a first prediction result. Then, the first prediction result is adjusted based on the selected event to generate a second prediction result. Finally, a recommendation result is generated and output based on the second prediction result. By using pre-set event parameters, the prediction result generated by the model is adjusted to determine the degree of influence of the event on the prediction result. Furthermore, by combining human-computer interaction and model prediction, the flexibility of event prediction and the accuracy of determining the event's impact are improved.

[0006] In conjunction with the first aspect, in some possible implementations, the step of extracting feature data from the platform service data based on the service type of the data recommendation service, predicting the growth trend based on the feature data, and generating a first prediction result includes: Based on the service type of the data recommendation service, feature extraction is performed on the platform service data to obtain the feature data corresponding to the platform service data; The feature data and the preset prediction duration are input into a pre-trained prediction model for prediction, generating a first prediction result corresponding to the feature data. The preset prediction duration is the time span of the first prediction result.

[0007] In combination with the first aspect and the above implementation methods, in some possible implementation methods, the step of obtaining event data of the target event identifier selected by the data recommendation instruction, adjusting the first prediction result based on the time parameter and event impact value in the event data, and generating a second prediction result includes: Obtain the event data of the target event identifier selected by the data recommendation instruction, and determine the time parameter and event impact value in the event data; Adjust the preset unit duration of the first prediction result to the target unit duration; The predicted value of the first prediction result is adjusted based on the target time span and the event impact value to generate a second prediction result.

[0008] The above technical solution obtains event data corresponding to the target event identifier of the data recommendation instruction, and adjusts the first prediction result based on the event data to obtain the second prediction result. This achieves the fusion of manually set parameters and model prediction results, thereby achieving a comprehensive analysis effect and improving the flexibility and accuracy of event prediction.

[0009] In combination with the first aspect and the above implementation methods, in some possible implementation methods, obtaining the event data of the target event identifier selected by the data recommendation instruction, and determining the time parameter and event impact value in the event data, includes: In response to the selection instruction in the data recommendation instruction, the target event identifier corresponding to the selection instruction is determined, and the event data corresponding to the target event identifier is obtained. The selection instruction is an instruction generated by selecting an event identifier in the event selection component displayed by the service platform. The event selection component includes at least one event identifier. The event data is analyzed to determine the time parameters and event impact values ​​within the event data.

[0010] The above technical solution provides an event selection component for users to choose from, enabling users to intuitively select event identifiers according to actual needs, thereby improving the convenience of users in selecting events to determine the impact of the event on the first prediction result.

[0011] In combination with the first aspect and the above implementation methods, in some possible implementation methods, prior to the response to the selection instruction in the data recommendation instruction, the method further includes: The service platform displays a data interaction component and an event creation component. The data interaction component includes a time setting component and a numerical input component. In response to a creation operation for the event creation component, determine the event identifier corresponding to the creation operation; The time parameter and event impact value of the event identifier are set through the data interaction component, and the time parameter and event impact value are determined as the event data corresponding to the event identifier; Add the event identifier to the event selection component of the service platform.

[0012] By using the above technical solution, event identifiers are pre-created, and the set time parameters and event impact values ​​are associated with the event identifiers. The created event identifiers are then added to the event selection component, thereby improving the convenience for users to determine the impact of event data on the first prediction result.

[0013] In combination with the first aspect and the above implementation methods, in some possible implementation methods, the step of setting the time parameter and event impact value of the event identifier through the data interaction component, and determining the time parameter and event impact value as the event data corresponding to the event identifier, includes: In response to a first setting operation for the time setting component, the time span and unit duration corresponding to the first setting operation are determined, and the time span and the unit duration are determined as the time parameters of the event identifier; In response to a second setting operation for the numerical input component, determine the event impact value corresponding to the second setting operation; The time parameter and event impact value are determined as the event data corresponding to the event identifier.

[0014] In combination with the first aspect and the above implementation methods, in some possible implementation methods, setting the time parameter and event impact value of the event identifier, and determining the time parameter and event impact value as the event data corresponding to the event identifier, includes: Obtain the event planning parameters of the event identifier, input the event planning parameters into the pre-trained incremental model, and obtain the event impact value and time parameters corresponding to the event identifier. The event planning parameters include the event duration and event content corresponding to the event identifier.

[0015] The above technical solution outputs event impact values ​​based on the pre-trained incremental model and event planning parameters, enabling the accuracy of the event impact values ​​output by the incremental model to be close to that of human judgment, thereby improving the efficiency of event data setting while ensuring data accuracy.

[0016] In combination with the first aspect and the above implementation methods, in some possible implementation methods, adjusting the predicted value of the first prediction result based on the target time span and the event impact value to generate a second prediction result includes: The target adjustment range of the first prediction result is determined based on the target time span; Based on the event impact value, the first predicted value of each target unit duration within the target adjustment range in the first prediction result is adjusted to obtain the second predicted value; Based on the second predicted value of each unit duration in the target time span, the second prediction result corresponding to the event data is obtained.

[0017] In combination with the first aspect and the above implementation methods, in some possible implementation methods, generating the recommendation result corresponding to the data recommendation instruction based on the second prediction result and outputting the recommendation result in the service platform includes: Based on the preset data format of the data recommendation instruction, the second prediction result is converted into the target prediction result; Based on the target prediction result and the generation log of the second prediction result, a recommendation result corresponding to the data recommendation instruction is generated. The generation log includes the first prediction result and the numerical adjustment process for generating the second prediction result based on the first prediction result. The recommendation results are output on the service platform.

[0018] The above technical solution outputs a second prediction result and a generation log on the service platform. This allows users to clearly understand the process by which the event data corresponding to each target event identifier is adjusted to the first prediction result by viewing the generation log, thereby improving the accuracy and convenience for users to determine the impact of events.

[0019] Secondly, a data prediction apparatus is provided, the apparatus comprising: The data acquisition unit is used to respond to the data recommendation instruction in the data recommendation service and acquire platform service data related to the data recommendation instruction from the service platform; The first prediction unit is used to extract feature data from the platform service data based on the service type of the data recommendation service, predict the growth trend based on the feature data, and generate a first prediction result. The second prediction unit is used to acquire event data of the target event identifier selected by the data recommendation instruction, adjust the first prediction result based on the time parameter and event impact value in the event data, and generate a second prediction result. The event impact value characterizes the degree of influence of the event data, and the time parameter is used to adjust the target time span and target unit duration of the first prediction result. The result output unit is used to generate a recommendation result corresponding to the data recommendation instruction based on the second prediction result, and output the recommendation result on the service platform.

[0020] Thirdly, a computer program product is provided, comprising: a computer program that, when executed by a processor of an electronic device, causes the computer to perform the method described in the first aspect or any possible implementation thereof.

[0021] Fourthly, a computer-readable storage medium is provided that stores computer program code, which, when executed on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation thereof.

[0022] Fifthly, an electronic device is provided, comprising: a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to execute the steps of the method described above. Attached Figure Description

[0023] Figure 1 This is a system architecture diagram of a data prediction method provided in the embodiments of this specification; Figure 2 This is a flowchart illustrating a data prediction method provided in an embodiment of this specification; Figure 3This is a flowchart illustrating a data prediction method provided in an embodiment of this specification; Figure 4 This is an example schematic diagram of a first prediction result provided in the embodiments of this specification; Figure 5 This is an example schematic diagram of a second prediction result provided in the embodiments of this specification; Figure 6 This is a schematic diagram of the structure of a data prediction device provided in the embodiments of this specification; Figure 7 This is a schematic diagram of the structure of an electronic device provided in the embodiments of this specification. Detailed Implementation

[0024] The technical solutions in this specification will now be described clearly and in detail with reference to the accompanying drawings. In the description of the embodiments in this specification, unless otherwise stated, " / " indicates "or," for example, A / B can mean A or B. "And / or" in the text is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of the embodiments in this specification, "multiple" refers to two or more than two.

[0025] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0026] Figure 1 This is a system architecture diagram of a data prediction method provided in the embodiments of this specification. For example... Figure 1 As shown in the embodiments of this specification, the data prediction method provided can be applied to terminal devices to realize the process of adjusting prediction results according to platform services and outputting recommendation results. The system structure provided in the embodiments of this specification mainly includes terminal device 10, service platform 20, and recommendation results 30. Terminal device 10 can be a device with data processing capabilities, such as a personal computer, smartphone, tablet computer, server, etc. Service platform 20 can be a platform for storing platform service data 21 for various objects, specifically a database, a cloud server, or a service set up on terminal device 10. Platform service data 21 can be all data related to the service type included in the service platform, such as the historical asset balance of the transaction party, the user group composition of the transaction party, and the historical events of the transaction party. Recommendation results 30 can be output to the user, facilitating the user's viewing of the prediction result adjustment process and the content of the second prediction result.

[0027] In related technologies, when making data predictions, the methods used are either predictions based on expert experience or predictions based on algorithms. Prediction based on expert experience is inefficient and highly subjective, while predictions based on algorithms cannot cope with unexpected events. Both methods result in inaccurate predictions, which means that the prediction effect is insufficient to meet the prediction requirements.

[0028] In the embodiments described in this specification, the terminal device 10 responds to a data recommendation instruction in a data recommendation service, obtains platform service data 21 related to the data recommendation instruction from the service platform 20, extracts feature data from the platform service data 21 according to the service type of the data recommendation service, predicts the growth trend based on the feature data, generates a first prediction result, obtains event data of the target event identifier selected by the data recommendation instruction, adjusts the first prediction result based on the time parameters and event impact values ​​in the event data, generates a second prediction result, generates a recommendation result corresponding to the data recommendation instruction based on the second prediction result, and outputs the recommendation result 30 on the service platform. Thus, by using pre-set event parameters, the prediction result generated by the model is adjusted to determine the degree of influence of the event on the prediction result. In this way, by combining human-computer interaction and model prediction, the flexibility of predicting events and the accuracy of determining the impact of events are improved.

[0029] based on Figure 1 The system architecture shown below will be combined with Figures 2-5 This document provides a detailed description of the data prediction methods provided in the embodiments of this specification.

[0030] Please see Figure 2 This is a flowchart illustrating a data prediction method provided in an embodiment of this specification. Figure 2 As shown, the method in the embodiments of this specification may include the following steps S102-S108.

[0031] S102, in response to the data recommendation instruction in the data recommendation service, obtain platform service data related to the data recommendation instruction from the service platform; In one embodiment, the terminal device responds to a data recommendation service triggered by a user on a data platform by obtaining a data recommendation instruction from the data recommendation service. The service platform can be a platform located on the terminal device used to provide users with data prediction functions. The data recommendation service can be a service used to provide users with data prediction and data analysis, such as predicting the financial status of a transaction or the revenue from advertising. The data recommendation instruction can be an instruction generated when the data recommendation service is triggered on the service platform, causing the service platform to respond to the data recommendation instruction by acquiring and predicting data. The platform service data related to the data recommendation instruction is obtained from the service platform. The platform service data can be data related to the recommendation purpose indicated by the data recommendation instruction. For example, if the data recommendation instruction is used to trigger a prediction of a transaction's financial status, then the platform service data could be the transaction's historical asset balance, the transaction's user base composition, and the transaction's historical events.

[0032] S104. Based on the service type of the data recommendation service, extract feature data from the platform service data, predict the growth trend based on the feature data, and generate the first prediction result. In one embodiment, feature extraction is performed on platform service data based on the service type of the data recommendation service to obtain feature data from the platform service data. The service type can represent the type of prediction the data recommendation service makes, such as predicting the financial status of a transaction party or the revenue from advertising. Feature data can be data related to the service type; the data extracted from the service recommendation data for the same transaction party differs depending on the service type. For example, if the service type is predicting the financial status of a transaction party, the feature data extracted from the platform service data could be the user group composition of the transaction party, the event data corresponding to the transaction party and historical events (specifically, the duration and revenue of historical events), and a pre-set prediction duration to accurately predict as needed. The pre-set prediction duration can represent the prediction period, such as 12 months, indicating that predictions will be made for the next 12 months. Based on the feature data, a growth trend is predicted to generate a first prediction result. The growth trend can represent the overall trend within a specific period (e.g., monthly, annually). The first prediction result can be data representing the growth trend of transactions corresponding to the transaction type. The data format of the first prediction result can be a line graph, a table, etc., which can be set according to the actual situation.

[0033] Specifically, the first prediction result may include a preset unit duration and the predicted value corresponding to each preset unit duration. The preset unit duration can be the minimum time span used in the first prediction result, such as one day, one week, or one month. The predicted value corresponding to each preset unit duration can be a value representing the prediction obtained for each preset unit duration. For example, if the service type is advertising revenue prediction and the preset unit duration is one day, then the predicted value for each preset unit duration can be the revenue change for each day.

[0034] S106, Obtain event data of the target event identifier selected by the data recommendation instruction, adjust the first prediction result based on the time parameter and event impact value in the event data, and generate the second prediction result; In one embodiment, a target event identifier selected in the data recommendation instruction is determined. The target event identifier can be an identifier used to identify the event. The characters of the target event identifier can be set according to the actual situation, such as event titles like "March Exhibition" or "April TV Advertising," or codes like "1," "2," or "3." Event data corresponding to the target event identifier is obtained. Based on the time parameters and event impact values ​​in the event data, the first prediction result is adjusted to generate a second prediction result. The event impact value can represent the degree of influence of the event data, such as adjusting the predicted value in the first prediction result upwards or downwards. The time parameters can be the target time span and target unit duration used to adjust the first prediction result. The target time span can represent the total duration of the prediction, and the target unit duration can be the duration indicating the preset unit duration for adjusting the first prediction result. For example, if the preset unit duration is one day and the target unit duration is one month, then the minimum time span of the generated second prediction result is one month. The second prediction result can be data that reflects the impact of each event after comprehensive analysis of the event data corresponding to the target event identifier.

[0035] Understandably, since the unit duration has changed from the preset unit duration to the previous target unit duration, the method for generating the predicted value corresponding to each target unit duration can be as follows: sum the predicted values ​​of the target unit durations to obtain the unit predicted sum, and then adjust the unit predicted sum according to the event impact value to obtain the target unit predicted value corresponding to each target unit duration in the second preset result. It should be noted that, besides summation, the unit predicted sum of the target unit duration can also be determined using statistical methods such as the median or total, depending on the actual situation.

[0036] S108: Generate recommendation results corresponding to the data recommendation instructions based on the second prediction results, and output the recommendation results on the service platform; In one embodiment, after obtaining the second prediction result corresponding to the data recommendation instruction, a recommendation result corresponding to the data recommendation instruction is generated based on the second prediction result and the log data corresponding to the process of generating the second prediction result, and the recommendation result is output on the service platform. The log data may include the first prediction result and data on the process of adjusting the first prediction result according to the event data corresponding to the target event identifier. The recommendation result can be output to the user, making it convenient for the user to view the prediction result adjustment process and the content of the second prediction result. By outputting the recommendation result on the service platform, the user can clearly and accurately view the impact of the event corresponding to the target event identifier on the prediction result.

[0037] The information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.) and signals involved in the embodiments of this specification are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions. For example, the platform service data and event data involved in this specification were obtained with full authorization.

[0038] In the embodiments of this specification, platform service data corresponding to the data recommendation instruction in the triggered data recommendation service is obtained from the service platform. A first prediction result is generated by predicting the recommendation effect of the platform service data according to the service type. Then, the first prediction result is adjusted according to the selected event to generate a second prediction result. A recommendation result is generated and output according to the second prediction result. In this way, the prediction result generated by the model is adjusted by pre-set event parameters to determine the degree of influence of the event on the prediction result. By combining human-computer interaction and model prediction, the flexibility of predicting events and the accuracy of determining the impact of events are improved.

[0039] The following will combine Figure 3 The present invention provides a detailed description of a data prediction method provided in the embodiments of this specification, using specific examples.

[0040] S202, in response to the data recommendation instruction in the data recommendation service, obtain platform service data related to the data recommendation instruction from the service platform; In one embodiment, the terminal device responds to a data recommendation service triggered by a user on a data platform by obtaining a data recommendation instruction from the data recommendation service. The service platform can be a platform located on the terminal device used to provide users with data prediction functions. The data recommendation service can be a service used to provide users with data prediction and data analysis, such as predicting the financial status of a transaction or the revenue from advertising. The data recommendation instruction can be an instruction generated when the data recommendation service is triggered on the service platform, causing the service platform to respond to the data recommendation instruction by acquiring and predicting data. The platform service data related to the data recommendation instruction is obtained from the service platform. The platform service data can be data related to the recommendation purpose indicated by the data recommendation instruction. For example, if the data recommendation instruction is used to trigger a prediction of a transaction's financial status, then the platform service data could be the transaction's historical asset balance, the transaction's user base composition, and the transaction's historical events.

[0041] S204. Based on the service type of the data recommendation service, feature extraction is performed on the platform service data to obtain the feature data corresponding to the platform service data. In one embodiment, feature extraction is performed on platform service data based on the service type of the data recommendation service to obtain feature data from the platform service data. The service type can represent the type of prediction the data recommendation service makes, such as predicting the financial status of a transaction party or the revenue from advertising. Feature data can be data related to the service type; the data extracted from the service recommendation data for the same transaction party differs depending on the service type. For example, if the service type is predicting the financial status of a transaction party, the feature data extracted from the platform service data could be the user group composition of the transaction party, the event data corresponding to the transaction party and historical events (specifically, the duration and revenue of historical events), and a pre-set prediction duration to accurately predict as needed. The pre-set prediction duration can represent the duration of the prediction, such as 12 months, indicating that predictions will be made for data over the next 12 months.

[0042] S206, Input the feature data and the preset prediction time into the pre-trained prediction model to make a prediction and generate the first prediction result corresponding to the feature data; In one embodiment, feature data and a preset prediction duration are input into a pre-trained prediction model. The prediction model then predicts the growth trend based on the feature data, generating a first prediction result corresponding to the feature data. The prediction model can be a digital model with data prediction capabilities. The growth trend can represent the overall trend within a specific period (e.g., monthly, annually). The first prediction result can be data representing the growth trend of a transaction type. The data format of the first prediction result can be a line graph, a table, etc., which can be set according to the actual situation.

[0043] Specifically, the first prediction result may include a preset unit duration and the predicted value corresponding to each preset unit duration. The preset unit duration can be the minimum time span used in the first prediction result, such as one day, one week, or one month. The predicted value corresponding to each preset unit duration can be a value representing the prediction obtained for each preset unit duration. For example, if the service type is advertising revenue prediction and the preset unit duration is one day, then the predicted value for each preset unit duration can be the revenue change for each day.

[0044] For example, such as Figure 4 As shown, Figure 4 The data format of the first prediction result is a line graph. The preset unit duration of the first prediction result is one day, and the preset prediction duration is one year. The prediction value corresponding to each preset unit duration is displayed on the line graph of the first prediction result.

[0045] S208, Obtain the event data of the target event identifier selected by the data recommendation instruction, and determine the time parameters and event impact values ​​in the event data; In one embodiment, in response to a selection instruction in a data recommendation instruction, a target event identifier corresponding to the selection instruction is determined. The target event identifier can be an identifier used to identify the event. The characters of the target event identifier can be set according to the actual situation, such as event titles like "March Exhibition" or "April TV Advertising," or codes like "1," "2," or "3." Event data corresponding to the target event identifier is obtained. The selection instruction is generated by selecting an event identifier from an event selection component displayed on the service platform. The event selection component includes at least one event identifier. The event data is parsed to determine the time parameters and event impact values ​​in the event data. The event impact value can characterize the degree of influence of the event data, such as adjusting the predicted value in the first prediction result upwards or downwards. The time parameters can be the target time span and target unit duration used to adjust the first prediction result. The target time span can characterize the total prediction duration, and the target unit duration can be the duration indicating the preset unit duration for adjusting the first prediction result. For example, if the preset unit duration is one day and the target unit duration is one month, then the minimum time span of the generated second prediction result is one month.

[0046] To improve the response efficiency of data prediction, event identifiers can be pre-created in the event selection component. These identifiers can be created by displaying a data interaction component and an event creation component on the service platform. The event creation component is used to create new events. The data interaction component can be a numerical input component for setting values ​​for the created events, and a time setting component for setting time parameters.

[0047] Furthermore, in response to a creation operation performed on the event creation component, an event identifier corresponding to the creation operation is determined. The event identifier can be determined by selecting from identifiers provided by the event creation component or by manually entering an identifier, depending on the specific circumstances. The creation operation can be any operation related to event creation performed within the event creation component, specifically including clicks, data input, etc.

[0048] Furthermore, the time parameters and event impact value of the event identifier are set through the data interaction component, and these parameters and impact value are then determined as the event data corresponding to the event identifier. Specifically, the time parameters and event impact value can be set as follows: In response to a first setting operation on the time setting component, the time span and unit duration corresponding to the first setting operation are determined, and these time span and unit duration are then determined as the time parameters of the event identifier. In response to a second setting operation on the numerical input component, the event impact value corresponding to the second setting operation is determined. The time parameters and event impact value are then determined as the event data corresponding to the event identifier. The event impact value can be a value determined in advance by the user based on expert experience, used to adjust the first prediction result.

[0049] Optionally, besides having users determine the event impact value based on expert experience, another method for determining the event impact value is to obtain the event planning parameters of the event identifier. These parameters can include the event duration and event content corresponding to the event identifier. The event planning parameters are then input into a pre-trained incremental model to obtain the event impact value and time parameters corresponding to the event identifier. The incremental model can be one that simulates expert experience in analysis, determining the event impact value based on the event duration and content. It is understandable that because the incremental model is trained based on expert experience, the accuracy of the event impact value output by the incremental model can approach the requirements of human judgment, thereby improving the efficiency of event data setup while ensuring data accuracy.

[0050] For example, if a user plans to create an event in March, April, and May, holding exhibitions in Beijing, Shanghai, and Guangzhou respectively, with each city hosting the exhibition for one month in sequence, and the main purpose of the exhibitions is to promote the company's products, then the event identifier can be "3-City Exhibition". The time planning parameters are as follows: the event duration is three months (March, April, and May), and the event content includes holding exhibitions in Beijing, Shanghai, and Guangzhou to promote the company's products.

[0051] Furthermore, the event identifier is added to the event selection component of the service platform. This means that after establishing the relationship between the event identifier and the event parameters, the event identifier is added to the event selection component of the service platform, and the corresponding time parameters and event impact values ​​are stored in the service platform so that users can retrieve the corresponding time parameters and event impact values ​​by selecting the event identifier.

[0052] S210, adjust the preset unit duration of the first prediction result to the target unit duration; S212, Adjust the predicted value of the first prediction result based on the target time span and the event impact value to generate the second prediction result; In one embodiment, the target unit duration in the event data corresponding to the target event identifier is adjusted to the preset unit duration of the first prediction result, so that the subsequently generated prediction results meet the user's desired unit duration viewing method. It is understood that since adjusting the unit duration involves adjusting the first prediction result, the user can set the unit duration according to actual needs when creating an event, ensuring that the adjusted unit duration is the duration the user requires. Then, based on the event impact value, the predicted values ​​within the target time span in the first prediction result are adjusted to generate a second prediction result. The second prediction result can be data reflecting the impact of each event, obtained through comprehensive analysis of the event data corresponding to the target event identifier.

[0053] It should be noted that if the target event identifier selected by the user includes multiple event identifiers, the unit duration of the second prediction result can be determined in the following ways: according to the order in which the target event identifiers are selected, the preset unit duration is adjusted sequentially, and the target unit duration corresponding to the last target event identifier is determined as the unit duration of the second prediction result; or, the mode of each target unit duration in each target event identifier is determined as the unit duration of the second prediction result; or, the duration preset by the user is determined as the unit duration of the second prediction result. The specific settings can be made according to the actual situation.

[0054] For example, such as Figure 5 As shown, Figure 5 The first prediction result is adjusted using the event data identified by the target event to obtain... Figure 5 The second prediction result is a graph showing the change in the predicted value compared to the first prediction result, with a target unit of time of one month and a time span of one year.

[0055] S214, Generate recommendation results corresponding to the data recommendation instructions based on the second prediction results, and output the recommendation results on the service platform; In one embodiment, the second prediction result is converted into the target prediction result based on a preset data format of the data recommendation instruction. The preset data format can be the data format set by the user when triggering the data recommendation service, such as a line graph or table. It is understood that if the user does not set a data format, a feasible approach is to determine the default format pre-set by the service platform as the preset data format. Based on the generation logs of the target prediction result and the second prediction result, a recommendation result corresponding to the data recommendation instruction is generated. The generation log includes the first prediction result and the numerical adjustment process for generating the second prediction result based on the first prediction result. The recommendation result is output to the service platform. The recommendation result can be used to output to the user, facilitating the user's viewing of the prediction result adjustment process and the content of the second prediction result. By outputting the recommendation result to the service platform, the user can clearly and accurately view the impact of the event corresponding to the target event identifier on the prediction result.

[0056] Understandably, if only the second prediction result is output on the service platform, it will be like... Figure 5 The second prediction result shown cannot intuitively provide users with the impact of each target event identifier on the first prediction result. However, by outputting the second prediction result and the generation log on the service platform, users can clearly understand the process by which the event data corresponding to each target event identifier adjusts the first prediction result by viewing the generation log. This includes adjusting the first prediction result by unit duration and adjusting the predicted value of each target event identifier on the first prediction result, thereby determining whether the event corresponding to that target event identifier needs to be executed.

[0057] For example, the event corresponding to the target event identifier is advertising using "Plan A". The second prediction result indicates that the revenue brought by the advertising is an increase of "0.3" in profit, while the revenue brought by advertising using "Plan B" is an increase of "0.2" in profit. However, advertising using both "Plan A" and "Plan B" at the same time only increases the profit by "0.1". Therefore, the user can determine which plan to choose for advertising based on the second prediction results obtained by selecting the target event identifier of "Plan A", selecting the target event identifier of "Plan B", and selecting the target event identifier of both "Plan A" and "Plan B" at the same time.

[0058] In the embodiments of this specification, platform service data corresponding to the data recommendation instruction triggered in the data recommendation service is obtained from the service platform. A first prediction result is generated by predicting the recommendation effect of the platform service data based on the service type. Then, the first prediction result is adjusted based on the selected event to generate a second prediction result. Finally, a recommendation result is generated and output based on the second prediction result. This allows for the adjustment of the prediction result generated by the model using pre-set event parameters, thereby determining the degree of influence of the event on the prediction result. By combining human-computer interaction and model prediction, the flexibility of event prediction and the accuracy of determining the event's impact are improved. Furthermore, by providing an event selection component for users to choose from, users can intuitively select event identifiers according to their actual needs, thus improving the convenience for users to select events and determine their impact on the first prediction result. In addition, by pre-creating event identifiers, setting time parameters and event impact values ​​through manual or model calculations, and establishing a relationship between the event identifiers and the created event identifiers, the event identifiers are added to the event selection component, thereby improving the convenience for users to determine the impact of event data on the first prediction result. Furthermore, by outputting the second prediction result and the generation log on the service platform, users can clearly understand the process by which the event data corresponding to each target event identifier adjusts the first prediction result by viewing the generation log, thereby improving the accuracy and convenience for users to determine the impact of events.

[0059] based on Figure 1 The system architecture will be discussed below. Figure 6 This specification provides a detailed description of the data prediction device provided in the embodiments. It should be noted that... Figure 6 The data prediction device 1 in this specification is used to perform the data prediction function described herein. Figures 2-5 The methods shown in the embodiments are illustrated for ease of explanation, showing only the parts related to the embodiments of this specification. For specific technical details not disclosed, please refer to this specification. Figures 2-5 The example shown.

[0060] The data acquisition unit 11 is used to acquire platform service data related to the data recommendation instruction in the service platform in response to the data recommendation instruction in the data recommendation service. The first prediction unit 12 is used to extract feature data from the platform service data based on the service type of the data recommendation service, predict the growth trend based on the feature data, and generate a first prediction result. The second prediction unit 13 is used to acquire event data of the target event identifier selected by the data recommendation instruction, adjust the first prediction result based on the time parameter and event impact value in the event data, and generate a second prediction result. The event impact value characterizes the degree of influence of the event data, and the time parameter is used to adjust the target time span and target unit duration of the first prediction result. Result output unit 14 is used to generate a recommendation result corresponding to the data recommendation instruction based on the second prediction result, and output the recommendation result on the service platform.

[0061] Optionally, the first prediction unit 12 is also used for: Based on the service type of the data recommendation service, feature extraction is performed on the platform service data to obtain the feature data corresponding to the platform service data; The feature data and the preset prediction duration are input into a pre-trained prediction model for prediction, generating a first prediction result corresponding to the feature data. The preset prediction duration is the time span of the first prediction result.

[0062] Optionally, the second prediction unit 13 is also used for: Obtain the event data of the target event identifier selected by the data recommendation instruction, and determine the time parameter and event impact value in the event data; Adjust the preset unit duration of the first prediction result to the target unit duration; The predicted value of the first prediction result is adjusted based on the target time span and the event impact value to generate a second prediction result.

[0063] Optionally, the second prediction unit 13 is also used for: In response to the selection instruction in the data recommendation instruction, the target event identifier corresponding to the selection instruction is determined, and the event data corresponding to the target event identifier is obtained. The selection instruction is an instruction generated by selecting an event identifier in the event selection component displayed by the service platform. The event selection component includes at least one event identifier. The event data is analyzed to determine the time parameters and event impact values ​​within the event data.

[0064] Optionally, the data prediction device 1 further includes an event creation unit 15, used for: The service platform displays a data interaction component and an event creation component. The data interaction component includes a time setting component and a numerical input component. In response to a creation operation for the event creation component, determine the event identifier corresponding to the creation operation; The time parameter and event impact value of the event identifier are set through the data interaction component, and the time parameter and event impact value are determined as the event data corresponding to the event identifier; Add the event identifier to the event selection component of the service platform.

[0065] Optionally, event creation unit 15 is also used for: In response to a first setting operation for the time setting component, the time span and unit duration corresponding to the first setting operation are determined, and the time span and the unit duration are determined as the time parameters of the event identifier; In response to a second setting operation for the numerical input component, determine the event impact value corresponding to the second setting operation; The time parameter and event impact value are determined as the event data corresponding to the event identifier.

[0066] Optionally, event creation unit 15 is also used for: Obtain the event planning parameters of the event identifier, input the event planning parameters into the pre-trained incremental model, and obtain the event impact value and time parameters corresponding to the event identifier. The event planning parameters include the event duration and event content corresponding to the event identifier.

[0067] Optionally, the second prediction unit 13 is also used for: The target adjustment range of the first prediction result is determined based on the target time span; Based on the event impact value, the first predicted value of each target unit duration within the target adjustment range in the first prediction result is adjusted to obtain the second predicted value; Based on the second predicted value of each unit duration in the target time span, the second prediction result corresponding to the event data is obtained.

[0068] Optionally, the result output unit 14 is also used for: Based on the preset data format of the data recommendation instruction, the second prediction result is converted into the target prediction result; Based on the target prediction result and the generation log of the second prediction result, a recommendation result corresponding to the data recommendation instruction is generated. The generation log includes the first prediction result and the numerical adjustment process for generating the second prediction result based on the first prediction result. The recommendation results are output on the service platform.

[0069] In the embodiments of this specification, platform service data corresponding to the data recommendation instruction triggered in the data recommendation service is obtained from the service platform. A first prediction result is generated by predicting the recommendation effect of the platform service data based on the service type. Then, the first prediction result is adjusted based on the selected event to generate a second prediction result. Finally, a recommendation result is generated and output based on the second prediction result. This allows for the adjustment of the prediction result generated by the model using pre-set event parameters, thereby determining the degree of influence of the event on the prediction result. By combining human-computer interaction and model prediction, the flexibility of event prediction and the accuracy of determining the event's impact are improved. Furthermore, by providing an event selection component for users to choose from, users can intuitively select event identifiers according to their actual needs, thus improving the convenience for users to select events and determine their impact on the first prediction result. In addition, by pre-creating event identifiers, setting time parameters and event impact values ​​through manual or model calculations, and establishing a relationship between the event identifiers and the created event identifiers, the event identifiers are added to the event selection component, thereby improving the convenience for users to determine the impact of event data on the first prediction result. Furthermore, by outputting the second prediction result and the generation log on the service platform, users can clearly understand the process by which the event data corresponding to each target event identifier adjusts the first prediction result by viewing the generation log, thereby improving the accuracy and convenience for users to determine the impact of events.

[0070] This specification also provides a computer storage medium that can store multiple program instructions, which are adapted to be loaded and executed by a processor as described above. Figures 1-5 The method steps of the illustrated embodiment can be found in the following documentation for detailed execution. Figures 1-5 The specific details of the illustrated embodiments will not be elaborated here.

[0071] This specification also provides an embodiment of a computer program product, which stores at least one instruction, the at least one instruction being loaded and executed by the processor as described above. Figures 1-5 The item recommendation model training method described in the illustrated embodiment can be found in the following documentation for a detailed execution process. Figures 1-5 The specific details of the illustrated embodiments will not be elaborated here.

[0072] Please see Figure 7 This document provides a schematic diagram of the structure of an electronic device as an embodiment of the present specification. Figure 7As shown, the electronic device 1000 may include: at least one processor 1001, such as a CPU; at least one network interface 1004; an input / output interface 1003; a memory 1005; and at least one communication bus 1002. The communication bus 1002 is used to enable communication between these components. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be high-speed RAM or non-volatile memory, such as at least one disk drive. Optionally, the memory 1005 may also be at least one storage device located remotely from the aforementioned processor 1001. Figure 7 As shown, the memory 1005, which serves as a computer storage medium, may include an operating system, a network communication module, an input / output interface module, and a data prediction application.

[0073] exist Figure 7 In the electronic device 1000 shown, the input / output interface 1003 is mainly used to provide an input interface for the user and to obtain the user input data.

[0074] In one embodiment, processor 1001 can be used to invoke a data prediction application stored in memory 1005 and specifically perform the following operations: In response to a data recommendation instruction in the data recommendation service, the platform service data related to the data recommendation instruction is obtained from the service platform; Based on the service type of the data recommendation service, feature data is extracted from the platform service data, and the growth trend is predicted based on the feature data to generate a first prediction result. The event data of the target event identifier selected by the data recommendation instruction is obtained. The first prediction result is adjusted based on the time parameter and event impact value in the event data to generate a second prediction result. The event impact value represents the degree of influence of the event data. The time parameter is used to adjust the target time span and target unit duration of the first prediction result. Based on the second prediction result, a recommendation result corresponding to the data recommendation instruction is generated, and the recommendation result is output on the service platform.

[0075] Optionally, when the processor 1001 executes the service type recommendation service based on the data, extracts feature data from the platform service data, predicts the growth trend based on the feature data, and generates a first prediction result, it specifically performs the following operations: Based on the service type of the data recommendation service, feature extraction is performed on the platform service data to obtain the feature data corresponding to the platform service data; The feature data and the preset prediction duration are input into a pre-trained prediction model for prediction, generating a first prediction result corresponding to the feature data. The preset prediction duration is the time span of the first prediction result.

[0076] Optionally, when the processor 1001 executes the operation of obtaining event data of the target event identifier selected by the data recommendation instruction, adjusting the first prediction result based on the time parameters and event impact values ​​in the event data, and generating the second prediction result, it specifically performs the following operations: Obtain the event data of the target event identifier selected by the data recommendation instruction, and determine the time parameter and event impact value in the event data; Adjust the preset unit duration of the first prediction result to the target unit duration; The predicted value of the first prediction result is adjusted based on the target time span and the event impact value to generate a second prediction result.

[0077] Optionally, when the processor 1001 executes the event data of the target event identifier selected by the data recommendation instruction and determines the time parameter and event impact value in the event data, it specifically performs the following operations: In response to the selection instruction in the data recommendation instruction, the target event identifier corresponding to the selection instruction is determined, and the event data corresponding to the target event identifier is obtained. The selection instruction is an instruction generated by selecting an event identifier in the event selection component displayed by the service platform. The event selection component includes at least one event identifier. The event data is analyzed to determine the time parameters and event impact values ​​within the event data.

[0078] Optionally, before executing the selection instruction in response to the data recommendation instruction, the processor 1001 also performs the following operations: The service platform displays a data interaction component and an event creation component. The data interaction component includes a time setting component and a numerical input component. In response to a creation operation for the event creation component, determine the event identifier corresponding to the creation operation; The time parameter and event impact value of the event identifier are set through the data interaction component, and the time parameter and event impact value are determined as the event data corresponding to the event identifier; Add the event identifier to the event selection component of the service platform.

[0079] Optionally, when the processor 1001 executes the setting of the time parameter and event impact value of the event identifier through the data interaction component, and determines the time parameter and event impact value as the event data corresponding to the event identifier, it specifically performs the following operations: In response to a first setting operation for the time setting component, the time span and unit duration corresponding to the first setting operation are determined, and the time span and the unit duration are determined as the time parameters of the event identifier; In response to a second setting operation for the numerical input component, determine the event impact value corresponding to the second setting operation; The time parameter and event impact value are determined as the event data corresponding to the event identifier.

[0080] Optionally, when the processor 1001 executes the setting of the time parameter and event impact value of the event identifier, and determines the time parameter and event impact value as the event data corresponding to the event identifier, it specifically performs the following operations: Obtain the event planning parameters of the event identifier, input the event planning parameters into the pre-trained incremental model, and obtain the event impact value and time parameters corresponding to the event identifier. The event planning parameters include the event duration and event content corresponding to the event identifier.

[0081] Optionally, when the processor 1001 adjusts the predicted value of the first prediction result based on the target time span and the event impact value to generate the second prediction result, it specifically performs the following operations: The target adjustment range of the first prediction result is determined based on the target time span; Based on the event impact value, the first predicted value of each target unit duration within the target adjustment range in the first prediction result is adjusted to obtain the second predicted value; Based on the second predicted value of each unit duration in the target time span, the second prediction result corresponding to the event data is obtained.

[0082] Optionally, when the processor 1001 executes the recommendation result corresponding to the data recommendation instruction generated based on the second prediction result and outputs the recommendation result in the service platform, it specifically performs the following operations: Based on the preset data format of the data recommendation instruction, the second prediction result is converted into the target prediction result; Based on the target prediction result and the generation log of the second prediction result, a recommendation result corresponding to the data recommendation instruction is generated. The generation log includes the first prediction result and the numerical adjustment process for generating the second prediction result based on the first prediction result. The recommendation results are output on the service platform.

[0083] In the embodiments of this specification, platform service data corresponding to the data recommendation instruction triggered in the data recommendation service is obtained from the service platform. A first prediction result is generated by predicting the recommendation effect of the platform service data based on the service type. Then, the first prediction result is adjusted based on the selected event to generate a second prediction result. Finally, a recommendation result is generated and output based on the second prediction result. This allows for the adjustment of the prediction result generated by the model using pre-set event parameters, thereby determining the degree of influence of the event on the prediction result. By combining human-computer interaction and model prediction, the flexibility of event prediction and the accuracy of determining the event's impact are improved. Furthermore, by providing an event selection component for users to choose from, users can intuitively select event identifiers according to their actual needs, thus improving the convenience for users to select events and determine their impact on the first prediction result. In addition, by pre-creating event identifiers, setting time parameters and event impact values ​​through manual or model calculations, and establishing a relationship between the event identifiers and the created event identifiers, the event identifiers are added to the event selection component, thereby improving the convenience for users to determine the impact of event data on the first prediction result. Furthermore, by outputting the second prediction result and the generation log on the service platform, users can clearly understand the process by which the event data corresponding to each target event identifier adjusts the first prediction result by viewing the generation log, thereby improving the accuracy and convenience for users to determine the impact of events.

[0084] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0085] The above-disclosed embodiments are merely preferred embodiments of this specification and should not be construed as limiting the scope of this specification. Therefore, any equivalent variations made in accordance with the claims of this specification shall still fall within the scope of this specification.

Claims

1. A data prediction method, characterized in that, The method includes: In response to a data recommendation instruction in the data recommendation service, the platform service data related to the data recommendation instruction is obtained from the service platform; Based on the service type of the data recommendation service, feature data is extracted from the platform service data, and the growth trend is predicted based on the feature data to generate a first prediction result. The event data of the target event identifier selected by the data recommendation instruction is obtained. The first prediction result is adjusted based on the time parameter and event impact value in the event data to generate a second prediction result. The event impact value represents the degree of influence of the event data. The time parameter is used to adjust the target time span and target unit duration of the first prediction result. Based on the second prediction result, a recommendation result corresponding to the data recommendation instruction is generated, and the recommendation result is output on the service platform.

2. The method according to claim 1, characterized in that, The step of recommending service types based on the data, extracting feature data from the platform service data, predicting growth trends based on the feature data, and generating a first prediction result includes: Based on the service type of the data recommendation service, feature extraction is performed on the platform service data to obtain the feature data corresponding to the platform service data; The feature data and the preset prediction duration are input into a pre-trained prediction model for prediction, generating a first prediction result corresponding to the feature data. The preset prediction duration is the time span of the first prediction result.

3. The method according to claim 1, characterized in that, The process of obtaining event data of the target event identifier selected by the data recommendation instruction, adjusting the first prediction result based on the time parameters and event impact values ​​in the event data, and generating a second prediction result includes: Obtain the event data of the target event identifier selected by the data recommendation instruction, and determine the time parameter and event impact value in the event data; Adjust the preset unit duration of the first prediction result to the target unit duration; The predicted value of the first prediction result is adjusted based on the target time span and the event impact value to generate a second prediction result.

4. The method according to claim 3, characterized in that, The step of obtaining event data of the target event identifier selected by the data recommendation instruction, and determining the time parameters and event impact values ​​in the event data, includes: In response to the selection instruction in the data recommendation instruction, the target event identifier corresponding to the selection instruction is determined, and the event data corresponding to the target event identifier is obtained. The selection instruction is an instruction generated by selecting an event identifier in the event selection component displayed by the service platform. The event selection component includes at least one event identifier. The event data is analyzed to determine the time parameters and event impact values ​​within the event data.

5. The method according to claim 4, characterized in that, Prior to the selection instruction in the data recommendation instruction, the method further includes: The service platform displays a data interaction component and an event creation component. The data interaction component includes a time setting component and a numerical input component. In response to a creation operation for the event creation component, determine the event identifier corresponding to the creation operation; The time parameter and event impact value of the event identifier are set through the data interaction component, and the time parameter and event impact value are determined as the event data corresponding to the event identifier; Add the event identifier to the event selection component of the service platform.

6. The method according to claim 5, characterized in that, The step of setting the time parameter and event impact value of the event identifier through the data interaction component, and determining the time parameter and event impact value as the event data corresponding to the event identifier, includes: In response to a first setting operation for the time setting component, the time span and unit duration corresponding to the first setting operation are determined, and the time span and the unit duration are determined as the time parameters of the event identifier; In response to a second setting operation for the numerical input component, determine the event impact value corresponding to the second setting operation; The time parameter and event impact value are determined as the event data corresponding to the event identifier.

7. The method according to claim 5, characterized in that, The step of setting the time parameter and event impact value of the event identifier, and determining the time parameter and event impact value as the event data corresponding to the event identifier, includes: Obtain the event planning parameters of the event identifier, input the event planning parameters into the pre-trained incremental model, and obtain the event impact value and time parameters corresponding to the event identifier. The event planning parameters include the event duration and event content corresponding to the event identifier.

8. The method according to claim 3, characterized in that, The step of adjusting the predicted value of the first prediction result based on the target time span and the event impact value to generate a second prediction result includes: The target adjustment range of the first prediction result is determined based on the target time span; Based on the event impact value, the first predicted value of each target unit duration within the target adjustment range in the first prediction result is adjusted to obtain the second predicted value; Based on the second predicted value of each unit duration in the target time span, the second prediction result corresponding to the event data is obtained.

9. The method according to claim 1, characterized in that, The step of generating a recommendation result corresponding to the data recommendation instruction based on the second prediction result, and outputting the recommendation result in the service platform, includes: Based on the preset data format of the data recommendation instruction, the second prediction result is converted into the target prediction result; Based on the target prediction result and the generation log of the second prediction result, a recommendation result corresponding to the data recommendation instruction is generated. The generation log includes the first prediction result and the numerical adjustment process for generating the second prediction result based on the first prediction result. The recommendation results are output on the service platform.

10. A data prediction device, characterized in that, The device includes: The data acquisition unit is used to respond to the data recommendation instruction in the data recommendation service and acquire platform service data related to the data recommendation instruction from the service platform; The first prediction unit is used to extract feature data from the platform service data based on the service type of the data recommendation service, predict the growth trend based on the feature data, and generate a first prediction result. The second prediction unit is used to acquire event data of the target event identifier selected by the data recommendation instruction, adjust the first prediction result based on the time parameter and event impact value in the event data, and generate a second prediction result. The event impact value characterizes the degree of influence of the event data, and the time parameter is used to adjust the target time span and target unit duration of the first prediction result. The result output unit is used to generate a recommendation result corresponding to the data recommendation instruction based on the second prediction result, and output the recommendation result on the service platform.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program code that, when executed, implements the method as described in any one of claims 1 to 9.

12. An electronic device, characterized in that, include: A processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to execute the steps of the method as claimed in any one of claims 1 to 9.

13. A computer program product, characterized in that, include: A computer program, when executed by a processor of an electronic device, causes the processor to perform the steps of the method as described in any one of claims 1 to 9.

Images