Information adjusting method and device, electronic equipment, medium and product
By receiving information adjustment requests and processing data through a filtering and hierarchical classification model, the problem of low accuracy in information adjustment is solved, achieving dynamic adaptation and accurate information adjustment, which is applicable to the field of data processing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA CONSTRUCTION BANK
- Filing Date
- 2025-12-25
- Publication Date
- 2026-04-14
AI Technical Summary
Existing technologies have low accuracy in information adjustment and cannot adapt to the influence of various factors, resulting in discrepancies between the adjustment results and the true information.
By receiving information adjustment requests, the system obtains the current information of the target object and multiple candidate data. Based on the target precision, it filters out multi-dimensional target data and uses a hierarchical classification model to adjust the information, determine the target level and adjustment parameters, and achieve dynamic adjustment.
It improves the accuracy of information adjustment, ensures that the adjustment results are consistent with the current state of the target object, adapt to its multi-dimensional changes, and improve the accuracy of decision-making.
Smart Images
Figure CN121858892A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of data processing technology, and in particular to a method, apparatus, electronic device, medium and product for adjusting information. Background Technology
[0002] In current data processing, with the rapid development of data analysis technology, data can be analyzed comprehensively and accurately, and accurate decisions can be made based on the analysis results.
[0003] In related technologies, data related to the target object is obtained through static rules, and information is adjusted based on the data.
[0004] However, this method suffers from low adjustment accuracy. Summary of the Invention
[0005] This application provides methods, apparatus, electronic devices, media, and products for adjusting information, in order to improve the accuracy of information adjustment.
[0006] In a first aspect, embodiments of this application provide an information adjustment method, comprising: receiving an information adjustment request, the information adjustment request including a target object identifier and a target precision; according to the information adjustment request, obtaining a target object corresponding to the target object identifier, determining current information corresponding to the target object, multiple candidate data, and time attributes corresponding to the multiple candidate data respectively, the multiple candidate data corresponding to multiple dimensions; determining a target time period corresponding to the target precision, filtering the multiple candidate data according to the target time period and the time attributes to obtain multiple target data, the times corresponding to the multiple target data all being within the target time period; inputting the multiple target data into a hierarchical classification model to obtain a target level corresponding to the target object, the hierarchical classification model being obtained by model training based on sample data; determining target adjustment parameters corresponding to the target level, adjusting the current information according to the target adjustment parameters to obtain target information.
[0007] In one possible implementation, determining the current information corresponding to the target object, multiple candidate data, and the time attributes corresponding to the multiple candidate data respectively includes: determining a query template; generating a query instruction based on the target object identifier and the query template; performing query processing from the database according to the query instruction to obtain the current information corresponding to the target object and the multiple candidate data; and determining the time attributes corresponding to the multiple candidate data respectively from the metadata of the multiple candidate data.
[0008] In one possible implementation, determining the target time period corresponding to the target precision includes: determining the current time and the mapping relationship between precision and duration; determining the target duration based on the target precision and the mapping relationship; and determining the target time period based on the target duration, with the current time as the endpoint.
[0009] In one possible implementation, inputting the plurality of target data into a hierarchical classification model to obtain the target level corresponding to the target object includes: performing feature engineering processing on the plurality of target data to obtain a plurality of data features; concatenating the plurality of data features to obtain a feature vector; inputting the feature vector into the hierarchical classification model to obtain a probability distribution of the probability of belonging to each level; and determining the target level based on the probability distribution.
[0010] In one possible implementation, determining the target adjustment parameters corresponding to the target level includes: determining the current lifecycle of the target object and the initial adjustment parameters corresponding to the target level; and determining the target adjustment parameters based on the current lifecycle and the initial adjustment parameters.
[0011] In one possible implementation, determining the target adjustment parameter based on the current lifecycle and the initial adjustment parameter includes: determining a stability parameter of the target object based on the current lifecycle; determining a correction value corresponding to the stability parameter in response to the stability parameter being greater than or equal to a preset threshold; and correcting the initial adjustment parameter based on the correction value to obtain the target adjustment parameter.
[0012] In one possible implementation, the method further includes: acquiring multiple historical data and historical levels corresponding to the multiple historical data respectively; using the multiple historical data and the historical levels as samples, training a lightweight gradient boosting algorithm model to obtain the hierarchical classification model.
[0013] Secondly, embodiments of this application provide an information adjustment device, comprising: a receiving module, configured to receive an information adjustment request, the information adjustment request including a target object identifier and a target precision; an acquisition module, configured to acquire, according to the information adjustment request, a target object corresponding to the target object identifier, determine current information corresponding to the target object, multiple candidate data, and time attributes corresponding to the multiple candidate data respectively, the multiple candidate data corresponding to multiple dimensions; a filtering module, configured to determine a target time period corresponding to the target precision, and perform filtering processing on the multiple candidate data according to the target time period and the time attributes to obtain multiple target data, the times corresponding to the multiple target data being all within the target time period; a classification module, configured to input the multiple target data into a hierarchical classification model to obtain a target level corresponding to the target object, the hierarchical classification model being obtained by model training based on sample data; and an adjustment module, configured to determine target adjustment parameters corresponding to the target level, and perform adjustment processing on the current information according to the target adjustment parameters to obtain target information.
[0014] In one possible implementation, the acquisition module is specifically used to determine a query template; the acquisition module is also specifically used to generate a query instruction based on the target object identifier and the query template; the acquisition module is also specifically used to perform query processing from the database according to the query instruction to obtain the current information corresponding to the target object and the plurality of candidate data; the acquisition module is also specifically used to determine the time attributes corresponding to the plurality of candidate data from the metadata of the plurality of candidate data.
[0015] In one possible implementation, the filtering module is specifically used to determine the current time and the mapping relationship between precision and duration; the filtering module is further specifically used to determine the target duration based on the target precision and the mapping relationship; the filtering module is further specifically used to determine the target time period based on the target duration, with the current time as the endpoint.
[0016] In one possible implementation, the classification module is specifically used to perform feature engineering processing on the multiple target data respectively to obtain multiple data features; the classification module is also specifically used to concatenate the multiple data features to obtain a feature vector; the classification module is also specifically used to input the feature vector into the hierarchical classification model to obtain the probability distribution of the probability of belonging to each level; the classification module is also specifically used to determine the target level according to the probability distribution.
[0017] In one possible implementation, the adjustment module is specifically used to determine the current lifecycle of the target object and the initial adjustment parameters corresponding to the target level; the adjustment module is also specifically used to determine the target adjustment parameters based on the current lifecycle and the initial adjustment parameters.
[0018] In one possible implementation, the adjustment module is further configured to determine the stability parameter of the target object based on the current lifecycle; the adjustment module is further configured to determine a correction value corresponding to the stability parameter in response to the stability parameter being greater than or equal to a preset threshold; the adjustment module is further configured to perform correction processing on the initial adjustment parameter based on the correction value to obtain the target adjustment parameter.
[0019] In one possible implementation, the apparatus further includes: a training module, configured to acquire multiple historical data and historical levels corresponding to the multiple historical data respectively; the training module is further configured to use the multiple historical data and the historical levels as samples to train a lightweight gradient boosting algorithm model to obtain the hierarchical classification model.
[0020] Thirdly, embodiments of this application provide an information adjustment device, including: a memory and a processor;
[0021] The memory stores computer-executed instructions;
[0022] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.
[0023] Fourthly, embodiments of this application provide a non-volatile computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.
[0024] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.
[0025] The information adjustment method, apparatus, electronic device, medium, and product provided in this application include: receiving an information adjustment request, the information adjustment request including a target object identifier and a target precision; according to the information adjustment request, obtaining the target object corresponding to the target object identifier, determining the current information corresponding to the target object, multiple candidate data, and time attributes corresponding to the multiple candidate data, the multiple candidate data corresponding to multiple dimensions; determining the target time period corresponding to the target precision, filtering the multiple candidate data according to the target time period and the time attributes to obtain multiple target data, the times corresponding to the multiple target data all being within the target time period; inputting the multiple target data into a hierarchical classification model to obtain the target level corresponding to the target object, the hierarchical classification model being obtained by model training based on sample data; determining the target adjustment parameters corresponding to the target level, adjusting the current information according to the target adjustment parameters to obtain target information. The above scheme improves the accuracy of information adjustment by obtaining multi-dimensional and reliable target data through target precision filtering and by adjusting information through a hierarchical classification model and target data. Attached Figure Description
[0026] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.
[0027] Figure 1 This is a schematic diagram illustrating an application scenario of an information adjustment method provided in an embodiment of this application.
[0028] Figure 2 A flowchart illustrating an information adjustment method provided in an embodiment of this application;
[0029] Figure 3 A flowchart illustrating another method for adjusting information provided in an embodiment of this application;
[0030] Figure 4 A schematic diagram illustrating the determination of the target hierarchy provided in an embodiment of this application;
[0031] Figure 5 A schematic diagram illustrating parameter corrections provided in the embodiments of this application;
[0032] Figure 6 This is a schematic diagram of the structure of an information adjustment device provided in an embodiment of this application;
[0033] Figure 7 A schematic diagram of another information adjustment device provided in an embodiment of this application;
[0034] Figure 8This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.
[0035] The accompanying drawings have illustrated specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to specific embodiments. Detailed Implementation
[0036] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.
[0037] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.
[0038] It should be noted that the phrase "at...time" in the embodiments of this application can refer to the instant at which a certain situation occurs, or to a period of time after the occurrence of a certain situation; the embodiments of this application do not specifically limit this. Furthermore, the display interface provided in the embodiments of this application is merely an example, and the display interface may include more or less content.
[0039] The collection, storage, use, processing, transmission, provision, and disclosure of financial data or user data involved in the technical solution of this application all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.
[0040] Furthermore, the technical solution involved in this application, which involves big data analysis of user information (including but not limited to personal biometrics, identity data, consumption data, asset data, electronic terminal operation data, etc.) and the use of artificial intelligence technology for automated decision-making, and makes decisions that have a significant impact on personal rights based on the results of automated decision-making, provides users with corresponding operation entry points for users to choose to agree to or reject the results of automated decision-making; if the user chooses to reject, the process will proceed to the expert decision-making process.
[0041] It should be noted that in the embodiments of this application, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, it does not mean that the applicant has used or necessarily used the solution.
[0042] It should be noted that the methods, apparatus, electronic devices, media, and products for adjusting the information in this application can be used in the field of data processing, or in any field other than data processing. The application fields of the methods, apparatus, electronic devices, media, and products for adjusting the information in this application are not limited.
[0043] Figure 1 This is a schematic diagram illustrating an application scenario of an information adjustment method provided in an embodiment of this application. An example is given based on the illustrated scenario: adjusting current information to obtain target information, and then using the target information to participate in object decision-making.
[0044] In related technologies, information is adjusted according to fixed rules and static patterns. However, in actual use, information is affected by various factors, and the adjustment results obtained by adjusting information according to static rules differ from the actual information, leading to the technical problem of low adjustment accuracy.
[0045] The information adjustment method provided in this application obtains multi-dimensional and reliable target data through target precision filtering, thereby improving the accuracy of information adjustment.
[0046] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.
[0047] Figure 2 A flowchart illustrating an information adjustment method provided in this application embodiment, the method comprising the following steps:
[0048] S201. Receive information adjustment request. The information adjustment request includes the target object identifier and the target precision.
[0049] The information can be financial information, such as bank account information or housing provident fund account information. The object can be a user, account, or device, and the information is adjusted to adapt to the user's current state.
[0050] For example, information adjustment is triggered by an information adjustment request. The target object identifier is used to accurately indicate the entity whose information needs to be adjusted.
[0051] For example, target precision is used to quantify the stringency of adjustments. Compared to fixed, static rules, information can be flexibly adjusted based on target precision to ensure the adjustment results meet expectations.
[0052] S202. Based on the information adjustment request, obtain the target object corresponding to the target object identifier, determine the current information corresponding to the target object, multiple candidate data, and the time attributes corresponding to the multiple candidate data respectively, with the multiple candidate data corresponding to multiple dimensions.
[0053] For example, the current information is the information before the adjustment, and the current information can be financial information.
[0054] With the help of scenario examples, it can be shown that as time goes by and user status changes, the current information may no longer be relevant. Information can be adjusted to make it more relevant, thereby enabling accurate decision-making based on the information.
[0055] For example, multiple candidate data are information-related data, and changes in the candidate data will be reflected in the information.
[0056] For example, by using candidate data from multiple dimensions, the state of the target object can be comprehensively analyzed, thereby enabling accurate information adjustments.
[0057] Optional, the data types of multiple dimensions include, but are not limited to, at least one of the following: risk attribute data, liquidity demand data, value contribution data, or resource scale data.
[0058] S203. Determine the target time period corresponding to the target accuracy. Based on the target time period and time attribute, filter and process multiple candidate data to obtain multiple target data. The time corresponding to each of the multiple target data is within the target time period.
[0059] For example, the time period range is determined based on the target precision. The higher the precision, the more data closer to the current moment is needed (i.e., higher timeliness), because data closer to the current moment better reflects the current state of the target object. For example, a precision level of "high" maps to "the last 7 days," and a precision level of "medium" maps to "the last 30 days."
[0060] For example, the time attribute of each target data includes the time corresponding to the target data, which can be the time when the target data was generated.
[0061] For example, filtering can be performed by target time periods to ensure that the timeliness of the target data meets the accuracy requirements and to avoid the time window being too short and lacking statistical significance, thereby improving the accuracy of information adjustment.
[0062] S204. Input multiple target data into the hierarchical classification model to obtain the target hierarchy corresponding to the target object. The hierarchical classification model is obtained by training the model based on the sample data.
[0063] For example, a hierarchical classification model is a trained machine learning model (such as a classifier). It learns a complex mapping relationship from data features to a specific hierarchy by studying a large amount of known historical sample data.
[0064] For example, by inputting multiple target data points (i.e., purified, multi-dimensional feature vectors) into the model, the model will output a predicted target level. The target level is a comprehensive evaluation of the target object in its current state, which is more comprehensive, objective, and accurate than a single indicator.
[0065] S205. Determine the target adjustment parameters corresponding to the target level, and adjust the current information according to the target adjustment parameters to obtain the target information.
[0066] For example, the target level is converted into specific executable values to obtain target adjustment parameters. These parameters are then used to adjust the current information, resulting in target information adapted to the current state of the target object.
[0067] With the help of scenario examples, the target adjustment parameters accurately quantify the information of the current state adaptation of the target object. Based on the target adjustment parameters, it can be determined how to adjust and by how much, so as to accurately adjust the information.
[0068] The information adjustment method provided in this application embodiment receives an information adjustment request, which includes a target object identifier and a target precision. Based on the information adjustment request, it obtains the target object corresponding to the target object identifier, determines the current information corresponding to the target object, multiple candidate data, and the time attributes corresponding to each candidate data, with the candidate data corresponding to multiple dimensions. It determines the target time period corresponding to the target precision, and filters the candidate data based on the target time period and time attributes to obtain multiple target data, where the times corresponding to each target data are all within the target time period. It inputs the multiple target data into a hierarchical classification model to obtain the target level corresponding to the target object; the hierarchical classification model is obtained through model training based on sample data. It determines the target adjustment parameters corresponding to the target level, and adjusts the current information based on the target adjustment parameters to obtain the target information. This solution improves the accuracy of information adjustment by obtaining multi-dimensional and reliable target data through target precision filtering and by adjusting information using a hierarchical classification model and target data.
[0069] Based on any of the above embodiments, the following, in conjunction with Figure 3 The detailed process of adjusting the information is explained.
[0070] Figure 3 This is a flowchart illustrating another method for adjusting information provided in an embodiment of this application. Figure 3 As shown, the method includes:
[0071] S301. Receive information adjustment request, which includes target object identifier and target precision.
[0072] It should be noted that the execution process of S301 is the same as that of S201, and will not be repeated here.
[0073] S302. Based on the information adjustment request, obtain the target object corresponding to the target object identifier, determine the current information corresponding to the target object, multiple candidate data, and the time attributes corresponding to the multiple candidate data respectively, with the multiple candidate data corresponding to multiple dimensions.
[0074] One feasible implementation method is to determine the current information, multiple candidate data, and time attributes through the following steps: determining a query template; generating a query instruction based on the target object identifier and the query template; performing query processing from the database according to the query instruction to obtain the current information and multiple candidate data corresponding to the target object; and determining the time attributes corresponding to the multiple candidate data from the metadata of the multiple candidate data.
[0075] For example, a query template is a pre-configured framework or script that defines the data fields, data types, data sources, or filtering conditions for the query. Query templates standardize queries and ensure consistency.
[0076] For example, the query template includes multiple placeholders. The target object identifier is filled into the placeholders to obtain the query instruction. The query instruction is a complete instruction that can be recognized and executed by the database, thereby enabling dynamic adjustment of the query logic.
[0077] For example, the database engine executes a query and returns a structured result set. The result set contains the current information corresponding to the target object identifier and multiple candidate data (all historical data records that conform to the template definition). The current information is the information of the target object before adjustment.
[0078] For example, metadata describing the data of candidate data attributes can be generated and maintained by a database management system. Metadata is unique and immutable, and retrieving time attributes from metadata can improve the reliability of time attributes.
[0079] In this feasible implementation, the accuracy of data acquisition is ensured through a templated and automated data query process, thereby improving the accuracy of information adjustment.
[0080] S303. Determine the target time period corresponding to the target accuracy. Based on the target time period and time attribute, filter and process multiple candidate data to obtain multiple target data. The time corresponding to each of the multiple target data is within the target time period.
[0081] One feasible implementation method is to determine the target time period by: determining the current time and the mapping relationship between precision and duration; determining the target duration based on the target precision and the mapping relationship; and determining the target time period based on the target duration, with the current time as the endpoint.
[0082] For example, the current moment is the moment when the information is adjusted, and the current moment is used as a benchmark to determine the target duration.
[0083] For example, the mapping relationship between precision and duration binds different levels of precision with corresponding durations to accurately determine the duration corresponding to a given precision. Through a lookup table operation, the target duration corresponding to the target precision can be determined from the mapping relationship.
[0084] For example, the current time is used as the end point, the start time is determined according to the target duration, and the time period between the start time and the end time is determined as the target time period.
[0085] To illustrate with a scenario example, if the current time is November 24th and the target duration is 10 days, then the target period is determined to be from November 14th to November 24th.
[0086] In this feasible implementation, by establishing a configurable mapping relationship, the target time period can be accurately determined, thereby improving the accuracy of data acquisition and, consequently, the accuracy of information adjustment.
[0087] S304. Perform feature engineering on multiple target data separately to obtain multiple data features.
[0088] Optionally, feature engineering includes, but is not limited to, at least one of the following: normalization, standardization, missing value handling, or feature construction.
[0089] For example, feature engineering can transform raw target data into high-quality numerical features suitable for machine learning models, thereby improving model performance and stability.
[0090] S305. Consolidate multiple data features to obtain a feature vector.
[0091] For example, multiple data features from different dimensions are concatenated into a one-dimensional feature vector in a predetermined order.
[0092] Based on the above implementation methods, structured, multi-feature data is converted into a standard mathematical input form that can be accepted by machine learning models, so that hierarchical classification models can make accurate predictions.
[0093] S306. Input the feature vector into the hierarchical classification model to obtain the probability distribution of the probability belonging to each level.
[0094] For example, feature vectors are input into a pre-trained hierarchical classification model. The model performs internal calculations and outputs a probability distribution. This probability distribution quantifies the likelihood of a target object belonging to each preset level. Compared to directly outputting the predicted level, this approach accurately reflects the confidence level of the hierarchical classification model's judgment.
[0095] S307. Determine the target level based on the probability distribution.
[0096] For example, the final target level is derived from the probability distribution based on a preset decision rule. For instance, the level with the highest probability value in the probability distribution can be used as the target level.
[0097] Below, in conjunction with Figure 4 Explain the determination of the target level.
[0098] Figure 4 This is a schematic diagram illustrating the determination of the target hierarchy as provided in an embodiment of this application. For example... Figure 4 As shown, the data features corresponding to multiple target data are determined, and the features are concatenated to obtain a feature vector. The feature vector is input into a hierarchical classification model, which classifies the feature vector and outputs multiple levels corresponding to the feature vector: level A, level B, and level C, as well as the corresponding probabilities A, B, and C, respectively.
[0099] Based on the above implementation method, the probability result output by the model, which contains uncertainty, is transformed into a clear decision label that can be used in subsequent processes.
[0100] S308. Determine the target adjustment parameters corresponding to the target level, and adjust the current information according to the target adjustment parameters to obtain the target information.
[0101] One feasible implementation method is to determine the target adjustment parameters by: determining the current lifecycle of the target object and the initial adjustment parameters corresponding to the target level; and determining the target adjustment parameters based on the current lifecycle and the initial adjustment parameters.
[0102] For example, the current lifecycle is essentially a label used to identify the stage the target object is in, such as the introduction stage, growth stage, maturity stage, or decline stage. Different stages can represent different levels of stability and development trends of the target object.
[0103] Optionally, the current lifecycle can be determined based on information such as the duration of the target object's existence or its recent activity level.
[0104] In this feasible implementation, by combining the current lifecycle, static attributes (i.e., initial adjustment parameters) and dynamic trends (lifecycle stages) can be combined to improve the accuracy of information adjustment.
[0105] One feasible implementation method is to determine the target adjustment parameters by: determining the stability parameters of the target object based on the current lifecycle; determining the correction value corresponding to the stability parameters in response to the stability parameters being greater than or equal to a preset threshold; and correcting the initial adjustment parameters based on the correction value to obtain the target adjustment parameters.
[0106] For example, the current lifecycle is the context of the stability parameters.
[0107] For example, the stability parameter is used to identify the stability of the target object's continued use of the service. The larger the stability parameter value, the more likely the target object is to continue using the service.
[0108] For example, if the stability parameter is greater than or equal to a preset threshold, it indicates that the target object tends to continue using the business logic. Based on this, a corresponding correction value is determined according to the specific value of the stability parameter.
[0109] For example, if the stability parameter is less than a preset threshold, it indicates that the target object tends to end the task, so no further correction processing is performed to reduce overhead.
[0110] Below, in conjunction with Figure 5 Explanation of parameter correction.
[0111] Figure 5 This is a schematic diagram illustrating parameter corrections provided in an embodiment of this application. For example... Figure 5 As shown, the stability parameter corresponding to the current lifecycle is determined, and it is checked whether the stability parameter is greater than or equal to a preset threshold. If not, no parameter correction is performed, and the initial adjustment parameter is determined as the target adjustment parameter. If yes, the initial adjustment parameter is corrected according to the stability parameter to obtain the target adjustment parameter.
[0112] In this feasible implementation, the stability of the target object is quantified by a correction value, thereby adjusting the initial adjustment parameters to match the state of the target object and improve the accuracy of information adjustment.
[0113] One feasible implementation method for adjusting information may include: acquiring multiple historical data and the historical levels corresponding to the multiple historical data; using the multiple historical data and historical levels as samples, training the lightweight gradient boosting algorithm model to obtain a hierarchical classification model.
[0114] For example, labeled samples are constructed, with the historical level as the label, to supervise the training of the lightweight gradient boosting algorithm model.
[0115] Among them, the Light Gradient Boosting Machine (LightGBM) is an efficient gradient boosting framework based on the decision tree algorithm, which features fast training speed, low memory consumption, and the ability to effectively handle large-scale data.
[0116] Optionally, sample data can be input into an initialized lightweight gradient boosting algorithm model. The model continuously adjusts its internal parameters (i.e., the structure of the decision tree) through iterative learning (e.g., gradient descent) to minimize the difference (i.e., the loss function) between the model's predicted output (predicted level) and the true label (historical level).
[0117] In this feasible implementation, supervised model training using a lightweight gradient boosting algorithm can automatically capture the nonlinear relationships and interaction effects between multi-dimensional features and complex hierarchies, thereby improving the classification accuracy of the hierarchical classification model.
[0118] Figure 6 This is a schematic diagram of an information adjustment device provided in an embodiment of this application. Figure 6 As shown, the information adjustment device 60 may include: a receiving module 61, an acquisition module 62, a filtering module 63, a classification module 64, and an adjustment module 65.
[0119] The receiving module 61 is used to receive information adjustment requests, which include the target object identifier and the target precision.
[0120] The acquisition module 62 is used to obtain the target object corresponding to the target object identifier according to the information adjustment request, determine the current information corresponding to the target object, multiple candidate data, and the time attributes corresponding to the multiple candidate data respectively, and the multiple candidate data correspond to multiple dimensions.
[0121] The filtering module 63 is used to determine the target time period corresponding to the target precision. Based on the target time period and time attribute, multiple candidate data are filtered to obtain multiple target data. The time corresponding to each of the multiple target data is within the target time period.
[0122] The classification module 64 is used to input multiple target data into the hierarchical classification model to obtain the target hierarchy corresponding to the target object. The hierarchical classification model is obtained by training the model based on the sample data.
[0123] The adjustment module 65 is used to determine the target adjustment parameters corresponding to the target level, and to adjust the current information according to the target adjustment parameters to obtain the target information.
[0124] Optionally, the receiving module 61 can perform... Figure 2 S201 in the embodiment.
[0125] Optionally, module 62 can be executed. Figure 2 S202 in the embodiment.
[0126] Optionally, the filtering module 63 can perform... Figure 2 S203 in the embodiment.
[0127] Optionally, the classification module 64 can be executed. Figure 2 S204 in the embodiment.
[0128] Optionally, adjustment module 65 can be executed. Figure 2 S205 in the embodiment.
[0129] It should be noted that the information adjustment device shown in the embodiments of this application can execute the technical solution shown in the above method embodiments, and its implementation principle and beneficial effects are similar, so they will not be described again here.
[0130] In one possible implementation, the acquisition module 62 is specifically used for:
[0131] Select the query template;
[0132] Generate a query instruction based on the target object identifier and the query template;
[0133] Based on the query command, the database is queried to obtain the current information of the target object and multiple candidate data.
[0134] From the metadata of multiple candidate data, determine the time attributes corresponding to each candidate data.
[0135] In one possible implementation, the filtering module 63 is specifically used for:
[0136] Determine the current time and the mapping relationship between precision and duration;
[0137] Determine the target duration based on the target accuracy and mapping relationship;
[0138] Based on the target duration, the target time period is determined with the current moment as the endpoint.
[0139] In one possible implementation, the classification module 64 is specifically used for:
[0140] Multiple target data are subjected to feature engineering processing to obtain multiple data features;
[0141] Multiple data features are concatenated to obtain a feature vector;
[0142] Input the feature vector into the hierarchical classification model to obtain the probability distribution of the probability belonging to each level;
[0143] Determine the target level based on the probability distribution.
[0144] In one possible implementation, the adjustment module 65 is specifically used for:
[0145] Determine the current lifecycle of the target object and the initial adjustment parameters corresponding to the target level;
[0146] Determine the target adjustment parameters based on the current lifecycle and initial adjustment parameters.
[0147] In one possible implementation, the adjustment module 65 is specifically used for:
[0148] Determine the stability parameters of the target object based on its current lifecycle;
[0149] If the stability parameter is greater than or equal to a preset threshold, then the correction value corresponding to the stability parameter is determined.
[0150] The initial adjustment parameters are corrected based on the correction values to obtain the target adjustment parameters.
[0151] Figure 7 This is a schematic diagram of another information adjustment device provided in an embodiment of this application. Figure 6 Based on the illustrated embodiments, as Figure 7 As shown, the information adjustment device 60 also includes a training module 66.
[0152] Training module 66 is used for:
[0153] Retrieve multiple historical data sets and the corresponding historical levels for each set of historical data;
[0154] Using multiple historical data points and historical hierarchies as samples, a lightweight gradient boosting algorithm model is trained to obtain a hierarchical classification model.
[0155] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, such as... Figure 8 As shown, the electronic device includes:
[0156] The electronic device includes a processor 291 and a memory 292; it may also include a communication interface 293 and a bus 294. The processor 291, memory 292, and communication interface 293 can communicate with each other via the bus 294. The communication interface 293 can be used for information transmission. The processor 291 can invoke logical instructions stored in the memory 292 to execute the methods of the above embodiments.
[0157] Furthermore, the logic instructions in the aforementioned memory 292 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.
[0158] The memory 292, as a non-volatile computer-readable storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this application. The processor 291 executes functional applications and data processing by running the software programs, instructions, and modules stored in the memory 292, that is, it implements the methods in the above-described method embodiments.
[0159] The memory 292 may include a program storage area and a data storage area. The program storage area may store the operating system and application programs required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 292 may include high-speed random access memory and may also include non-volatile memory.
[0160] This application provides a non-volatile computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the method as described in the foregoing embodiments.
[0161] This application provides a computer program product, including a computer program that, when executed by a processor, implements the method as described in the foregoing embodiments.
[0162] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.
[0163] It should be further noted that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.
[0164] It should be understood that the above-described device embodiments are merely illustrative, and the device of this application can also be implemented in other ways. For example, the division of units / modules in the above embodiments is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units, modules, or components may be combined, or integrated into another system, or some features may be ignored or not executed.
[0165] Furthermore, unless otherwise specified, the functional units / modules in the various embodiments of this application can be integrated into one unit / module, or each unit / module can exist physically separately, or two or more units / modules can be integrated together. The integrated units / modules described above can be implemented in hardware or as software program modules.
[0166] When the integrated unit / module is implemented in hardware, the hardware can be digital circuits, analog circuits, etc. The physical implementation of the hardware structure includes, but is not limited to, transistors, memristors, etc. The processor can be any suitable hardware processor, such as CPU, GPU, FPGA, DSP, and ASIC. The storage unit can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc.
[0167] If the integrated unit / module is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.
[0168] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.
[0169] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the claims.
[0170] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.
Claims
1. A method for adjusting information, characterized in that, include: Receive an information adjustment request, the information adjustment request including the target object identifier and the target precision; Based on the information adjustment request, obtain the target object corresponding to the target object identifier, determine the current information corresponding to the target object, multiple candidate data, and the time attributes corresponding to the multiple candidate data respectively, wherein the multiple candidate data correspond to multiple dimensions; Determine the target time period corresponding to the target precision, and filter the multiple candidate data according to the target time period and the time attribute to obtain multiple target data, wherein the time corresponding to each of the multiple target data is within the target time period; The multiple target data are input into a hierarchical classification model to obtain the target hierarchy corresponding to the target object. The hierarchical classification model is obtained by training the model based on the sample data. Determine the target adjustment parameters corresponding to the target level, and adjust the current information according to the target adjustment parameters to obtain the target information.
2. The method according to claim 1, characterized in that, Determining the current information corresponding to the target object, multiple candidate data, and the time attributes corresponding to the multiple candidate data respectively includes: Select the query template; Generate a query instruction based on the target object identifier and the query template; According to the query instruction, a query is performed from the database to obtain the current information corresponding to the target object and the multiple candidate data; From the metadata of multiple candidate data, determine the time attribute corresponding to each of the multiple candidate data.
3. The method according to claim 1, characterized in that, Determining the target time period corresponding to the target accuracy includes: Determine the current time and the mapping relationship between precision and duration; The target duration is determined based on the target accuracy and the mapping relationship; Based on the target duration, the target time period is determined with the current time as the endpoint.
4. The method according to claim 1, characterized in that, Inputting the multiple target data into a hierarchical classification model yields the target hierarchy corresponding to the target object, including: The multiple target data are subjected to feature engineering processing to obtain multiple data features; The multiple data features are concatenated to obtain a feature vector; The feature vector is input into the hierarchical classification model to obtain the probability distribution of the probability belonging to each level; The target level is determined based on the probability distribution.
5. The method according to claim 1, characterized in that, Determining the target adjustment parameters corresponding to the target level includes: Determine the current lifecycle of the target object and the initial adjustment parameters corresponding to the target level; The target adjustment parameters are determined based on the current lifecycle and the initial adjustment parameters.
6. The method according to claim 5, characterized in that, Based on the current lifecycle and the initial adjustment parameters, the target adjustment parameters are determined, including: Based on the current lifecycle, determine the stability parameters of the target object; In response to the stability parameter being greater than or equal to a preset threshold, a correction value corresponding to the stability parameter is determined; The initial adjustment parameters are corrected based on the correction value to obtain the target adjustment parameters.
7. The method according to any one of claims 1-6, characterized in that, The method further includes: Obtain multiple historical data sets and the corresponding historical levels for each of the multiple historical data sets; Using the aforementioned historical data and historical levels as samples, the lightweight gradient boosting algorithm model is trained to obtain the hierarchical classification model.
8. An information adjustment device, characterized in that, include: The receiving module is used to receive information adjustment requests, wherein the information adjustment requests include target object identifier and target precision; The acquisition module is used to adjust the request according to the information, acquire the target object corresponding to the target object identifier, determine the current information corresponding to the target object, multiple candidate data, and the time attributes corresponding to the multiple candidate data respectively, wherein the multiple candidate data correspond to multiple dimensions; The filtering module is used to determine the target time period corresponding to the target precision, and to filter the multiple candidate data according to the target time period and the time attribute to obtain multiple target data, wherein the time corresponding to each of the multiple target data is within the target time period; The classification module is used to input the multiple target data into the hierarchical classification model to obtain the target level corresponding to the target object. The hierarchical classification model is obtained by training the model based on the sample data. The adjustment module is used to determine the target adjustment parameters corresponding to the target level, and to adjust the current information according to the target adjustment parameters to obtain the target information.
9. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1-7.
10. A non-volatile computer-readable storage medium, characterized in that, The non-volatile computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.
11. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1-7.