Data processing method and electronic device

Abstract

To determine at least one influence factor influencing a target attribute parameter.SOLUTION: A computer performs data processing on a time-series dataset including a plurality of time-series data items whose elements are time and a plurality of attribute parameters corresponding to the time. The data processing includes: acquiring a time-series dataset; acquiring a target attribute parameter, which is an attribute parameter determined by being influenced by some of the other attribute parameters included in the plurality of time-series data items and is a target of analysis of the influence, from the attribute parameters in the plurality of time-series data items; determining an influence factor by inputting the acquired target attribute parameter and time-series dataset to a time-series model for determining the influence factor indicating the attribute parameter influencing the target attribute parameter in time series among the attribute parameters included in the plurality of time-series data items, and the time corresponding to the attribute parameter; and outputting the determined influence factor.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] TECHNICAL FIELD Embodiments of the present disclosure relate generally to the field of computers, and more particularly to data processing methods and electronic devices. [Background technology]

[0002] With the development of technology, causal analysis has been widely applied in various fields, such as retail, energy control, and intervention in abnormal events. Causal analysis can lead to relevant decision-making. However, existing solutions usually require manual determination of factors such as the cause of an event occurrence, which, on the one hand, excessively increases human costs, and, on the other hand, creates uncertainty due to manual determination and is prone to errors. As a result, it is difficult to achieve better decisions. Summary of the Invention [Problem to be solved by the invention]

[0003] According to an exemplary embodiment of the present disclosure, a data processing solution is provided that can determine at least one influencing factor that influences a target attribute parameter. [Means for solving the problem]

[0004] A data processing method according to a first aspect of the present disclosure is a data processing method for executing, by a computer, processing on a time series dataset including a plurality of time series data items, each of which has as elements a time and a plurality of attribute parameters corresponding to the time, the method including: acquiring the time series dataset; acquiring a target attribute parameter from the attribute parameters included in the plurality of time series data items, the target attribute parameter being determined by being influenced by some of the other attribute parameters included in the plurality of time series data items, and being the subject of the influence analysis; determining at least one influence factor by inputting the acquired target attribute parameter and the time series dataset into a pre-given time series model to determine at least one attribute parameter among the attribute parameters included in the plurality of time series data items that influences the target attribute parameter in a time series and a time corresponding to the at least one attribute parameter; and outputting the determined at least one influence factor.

[0005] A data processing method according to a second aspect of the present disclosure is a data processing method for executing, by a computer, processing on a time series dataset related to sales records, the time series dataset including a plurality of time series data items each having a plurality of attribute parameters as elements, the element being time and the plurality of attribute parameters including at least one of purchase price, selling price, sales volume, inventory volume, or customer view count corresponding to the time, the data processing method including: acquiring the time series dataset; acquiring the inventory volume from the attribute parameters included in the plurality of time series data items, the inventory volume being an attribute parameter determined by being influenced by some of the other attribute parameters included in the plurality of time series data items; determining at least one influence factor by inputting the acquired inventory volume and the time series dataset into a pre-given time series model to determine at least one influence factor indicating at least one attribute parameter included in the plurality of time series data items that influences the inventory volume in a time series and the time corresponding to the at least one attribute parameter; and determining the inventory volume for the next hour based on the determined at least one influence factor.

[0006] A data processing method according to a third aspect of the present disclosure is a data processing method for executing, by a computer, processing on a time series dataset related to power of the IoT device, the time series dataset including a plurality of time series data items each having as elements a time and a plurality of attribute parameters including at least one of an electricity price, a power demand, a voltage, a current, a temperature, a humidity, a barometric pressure, or a power consumption of the IoT device corresponding to the time, the data processing method including: acquiring the time series dataset; acquiring the power consumption, which is an attribute parameter determined by being influenced by some of the other attribute parameters included in the plurality of time series data items, from the attribute parameters included in the plurality of time series data items; determining at least one influence factor by inputting the acquired power consumption and the time series dataset into a pre-given time series model to determine at least one influence factor indicating at least one attribute parameter included in the plurality of time series data items that influences the power consumption in a time series and a time corresponding to the at least one attribute parameter; and determining the power consumption for a next hour based on the determined at least one influence factor.

[0007] A data processing method according to a fourth aspect of the present disclosure is a data processing method for executing, by a computer, processing on a time series dataset collected from a social network, the time series dataset including a plurality of time series data items each having a time and a plurality of attribute parameters including at least one of user identification information corresponding to the time, a user's post, a number of views of the user's post, a number of comments on the user's post, or an abnormality indication, the method including: acquiring the time series dataset; acquiring the abnormality indication from the attribute parameters included in the plurality of time series data items, the abnormality indication being an attribute parameter determined by being influenced by some of other attribute parameters included in the plurality of time series data items; determining at least one influence factor by inputting the acquired abnormality indication and the time series dataset into a pre-given time series model to determine at least one influence factor indicating at least one attribute parameter included in the plurality of time series data items that influences the abnormality indication in a time series and a time corresponding to the at least one attribute parameter; and outputting presentation information indicating a user at an abnormality risk based on the determined at least one influence factor.

[0008] In a fifth aspect of the present disclosure, there is provided an electronic device, the electronic device including at least one processor unit and at least one memory coupled to the at least one processor unit and storing instructions executed by the at least one processor unit, the instructions, when executed by the at least one processor unit, causing the electronic device to perform a method according to any one of the first to fourth aspects of the present disclosure.

[0009] In a sixth aspect of the present disclosure, there is provided an electronic device, the electronic device including a memory and a processor, the memory for storing one or more computer instructions, the one or more computer instructions being executed by the processor to implement a method according to any one of the first to fourth aspects of the present disclosure.

[0010] In a seventh aspect of the present disclosure, there is provided a computer-readable storage medium having machine-readable instructions stored thereon that, when executed by a device, cause the device to perform a method according to any one of the first to fourth aspects of the present disclosure.

[0011] In an eighth aspect of the present disclosure, there is provided a computer program product, the computer program product including computer-readable instructions that, when executed by a processor, implement a method according to any one of the first to fourth aspects of the present disclosure.

[0012] A ninth aspect of the present disclosure provides an electronic device, the electronic device including a processing circuit device configured to perform a method according to any one of the first to fourth aspects of the present disclosure.

[0013] The Summary of the Invention is intended to introduce a series of concepts in a simplified manner, which will be further described in the following embodiments. The descriptions in the Summary of the Invention are not intended to identify key or necessary features of the disclosure, nor are they intended to limit the scope of the disclosure. Other features of the disclosure should be readily apparent from the following description. [Brief explanation of the drawings]

[0014] These and other features, advantages, and aspects of the embodiments of the present disclosure will become more apparent from the following detailed description taken in conjunction with the drawings, in which like or similar reference numerals indicate like or similar elements. [Figure 1] 1 illustrates a block diagram of an exemplary environment according to an embodiment of the present disclosure. [Figure 2] 1 shows a schematic flow chart of a process used for data processing according to some embodiments of the present disclosure. [Figure 3] FIG. 1 shows a schematic block diagram of an exemplary system according to some embodiments of the present disclosure. [Figure 4] FIG. 1 shows a schematic block diagram of the interaction between a user interface module and a time series analysis engine, according to some embodiments of the present disclosure. [Figure 5A] 1 shows a schematic diagram of a first time relationship diagram according to some embodiments of the present disclosure. [Figure 5B] 10 shows a schematic diagram of a second time relationship diagram according to some embodiments of the present disclosure. [Figure 6] 10 shows an example flowchart of a process for obtaining a second time relationship diagram according to some embodiments of the present disclosure. [Figure 7] 1 shows a schematic flowchart of a process for determining key elements according to some embodiments of the present disclosure. [Figure 8A] FIG. 1 shows a schematic diagram of determining the relationship between any two numerical intervals a and b according to some embodiments of the present disclosure. [Figure 8B] FIG. 1 shows a schematic diagram of determining two numerical intervals based on the ordering of three intervals a, b, and c, according to some embodiments of the present disclosure. [Figure 9] 1 illustrates a display of multiple influences according to some embodiments of the present disclosure. [Figure 10A] 10 shows a schematic diagram of the relationship of sales volume influence over time to target attribute parameters, according to some embodiments of the present disclosure. [Figure 10B] 10 shows a schematic diagram of the relationship of influence of user views to target attribute parameters over time, according to some embodiments of the present disclosure; [Figure 11] 1 shows a schematic flowchart of a process for determining response relationships according to some embodiments of the present disclosure. [Figure 12A] 10 shows a schematic diagram of the influence of sales volume adjustment on target attribute parameters over time, according to some embodiments of the present disclosure. [Figure 12B] 10 shows a schematic diagram of cumulative change over time in cumulative impact of sales volume adjustments on target attribute parameters, according to some embodiments of the present disclosure; [Figure 13A]10 shows a schematic diagram of the influence of user view adjustments on target attribute parameters over time, according to some embodiments of the present disclosure; [Figure 13B] 10 shows a schematic diagram of cumulative change over time of cumulative impact of user view adjustments on target attribute parameters, according to some embodiments of the present disclosure; [Figure 14] 1 illustrates a schematic flowchart of a process for determining inventory levels, according to some embodiments of the present disclosure. [Figure 15] 1 illustrates a schematic flowchart of a process for determining device uptime, according to some embodiments of the present disclosure. [Figure 16] 10 shows a schematic diagram of the time course of collected attribute parameters, in accordance with some embodiments of the present disclosure; [Figure 17] 1 illustrates a schematic flowchart of a process for determining an abnormal event, according to some embodiments of the present disclosure. [Figure 18A] FIG. 1 shows a schematic diagram for determining abnormality risk, according to some embodiments of the present disclosure. [Figure 18B] FIG. 1 shows a schematic diagram for determining abnormality risk, according to some embodiments of the present disclosure. [Figure 18C] FIG. 1 shows a schematic diagram for determining abnormality risk, according to some embodiments of the present disclosure. [Figure 19] FIG. 1 shows a block diagram of an exemplary device in which embodiments of the present disclosure can be implemented. DETAILED DESCRIPTION OF THE INVENTION

[0015] Hereinafter, embodiments of the present disclosure will be described in more detail with reference to the drawings. Although the drawings show several embodiments of the present disclosure, it should be understood that the present disclosure can be realized in various forms and should not be construed as being limited to the embodiments described herein. Rather, these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should also be understood that the drawings and embodiments of the present disclosure are merely illustrative and are not intended to limit the scope of protection of the present disclosure.

[0016] In describing embodiments of the present disclosure, the terms "comprising" and similar terms should be understood as open-ended, i.e., "including, but not limited to." The term "based on" should be understood as "based at least in part on." The terms "one embodiment" or "the embodiment" should be understood as "at least one embodiment." The terms "first," "second," etc. may refer to different or the same object. Other explicit and implicit definitions may also be included in the following text.

[0017] The individual methods and processes described in the embodiments of the present disclosure may be applied to various electronic devices such as terminal devices, network devices, etc. Furthermore, the embodiments of the present disclosure may be executed in test devices such as signal generators, signal analyzers, spectrum analyzers, network analyzers, test terminal devices, test network devices, and channel emulators.

[0018] In describing embodiments of the present disclosure, the term "circuit" may refer to a hardware circuit and / or a combination of a hardware circuit and software. For example, a circuit may be a combination of analog and / or digital hardware circuitry and software / firmware. As another example, a circuit may be any portion of a hardware processor that includes software. A hardware processor includes digital signal processor(s), software, and memory(s), which cooperate to enable an apparatus, such as a computing device, to operate and perform various functions. As yet another example, a circuit may be a hardware circuit and / or processor, such as a microprocessor or portion of a microprocessor, that requires software / firmware for operation but may not require software if software is not required for operation. As used herein, the term "circuit" includes implementations of only a hardware circuit or processor(s) or a portion of a hardware circuit or processor(s) along with its (or their) associated software and / or firmware.

[0019] Agents need to make a series of decisions in a variety of problems across many domains. Although agents have access to large datasets, it is still difficult to make more accurate decisions. For example, the datasets to be collected, the determination of sufficiently predictable patterns, and the relative importance of individual variables are usually manually set by humans, such as data scientists or domain managers. This results in a heavy reliance on humans for decision-making, which is highly inefficient on the one hand and results in insufficient decision-making accuracy on the other.

[0020] In view of this, embodiments of the present disclosure provide a data processing solution to solve one or more of the above-mentioned problems and / or other potential problems. The solution can utilize a time series model to determine at least one influencing factor affecting a target attribute parameter. This eliminates manual specification and improves processing efficiency. Furthermore, the influencing factor determined by the time series model can be used in subsequent decision-making, resulting in more accurate decision-making.

[0021] 1 illustrates a block diagram of an exemplary environment 100 according to an embodiment of the present disclosure. It should be understood that the environment 100 illustrated in FIG. 1 is merely one example in which embodiments of the present disclosure may be implemented and is not intended to limit the scope of the present disclosure. Embodiments of the present disclosure apply to other systems or architectures as well.

[0022] 1 , the environment 100 may include a computing device 110. The computing device 110 may be any device having computing capabilities. The computing device 110 may include, but is not limited to, a personal computer, a server computer, a handheld or laptop device, a mobile device (such as a cell phone, a personal digital assistant (PDA), a media player, etc.), a wearable device, a consumer electronics appliance, a minicomputer, a mainframe, a distributed computing system, a cloud computing resource, etc. It should be understood that the computing device 110 may or may not have sufficient computing resources for model training, taking into account factors such as cost.

[0023] The computing device 110 may be configured to obtain the time series data set 120 and output at least one influence factor 140. The determination regarding the at least one influence factor 140 may be performed by the time series model 130.

[0024] By way of example, the time series dataset 120 may be input by a user or retrieved from a storage device, although the present disclosure is not limited in this respect. In some embodiments, the time series dataset 120 may include multiple time series data items, each of which may include a time and corresponding attribute parameters. Optionally, different data items may have different times. Optionally, the attribute parameters may be referred to as parameters, variables, or other names in some examples, although the present disclosure is not limited in this respect.

[0025] In some examples, embodiments of the present disclosure may be used in the retail field, such as for retail demand forecasting. For example, each data item in the time-series dataset 120 may include a time and a corresponding number of attribute parameters. The attribute parameters may be one or more of a purchase price, a sale price, a sales volume, an inventory volume, a customer view count, or the like.

[0026] In some examples, embodiments of the present disclosure may be used in the energy field, for example, for power control of Internet of Things (IoT) devices. For example, each data item in the time-series dataset 120 may include a time and a corresponding number of attribute parameters. The attribute parameters may be one or more of an electricity price, a power demand, a voltage, a current, a temperature, a humidity, an air pressure, or a power consumption of an IoT device.

[0027] In some examples, embodiments of the present disclosure may be used for intervention in abnormal events, such as predicting abnormal events, such as suicides, based on information on social networks. For example, each data item in the time series dataset 120 may include a time and a corresponding number of attribute parameters. The attribute parameters may be one or more of a user identification, a user post, a number of views of the user post, a number of comments on the user post, or an abnormality indication.

[0028] It should be understood that the above-mentioned scenarios are merely illustrative and do not limit the scope of the present disclosure. The embodiments of the present disclosure may be applied to various fields where similar problems exist, and will not be listed here. Furthermore, the "action" in the embodiments of the present disclosure may be referred to as, for example, "decision-making," and the present disclosure is not limited thereto.

[0029] It can be understood that time series can be used to represent data of spatiotemporal processes in the real world. When making decisions, predictions can be made using correlations of a certain time length, so the accuracy of decision-making can be improved by determining a "certain time length" as the minimum sufficient statistic. The minimum sufficient statistic can be understood as follows: correlations smaller than a certain time length are not sufficient for prediction, and correlations larger than a certain time length may be contaminated with noise and may contain a certain amount of redundant information.

[0030] In some examples, the correlation between different times in a time series may be represented by a graph (e.g., a causal graph), where nodes of the graph represent certain attribute parameters at certain times. Pointing from one node to another indicates a causal relationship between the two nodes, and a linear relationship can be depicted by reinforcing the quantitative relationship of the edges. This is described in detail below in connection with Figures 2 to 17.

[0031] 2 illustrates a schematic flowchart of a process 200 used for data processing according to some embodiments of the present disclosure. For example, the process 200 may be performed by the computing device 110 illustrated in FIG. 1. It should be understood that the process 200 may further include additional blocks not illustrated and / or omit some illustrated blocks. The scope of the present disclosure is not limited in this respect.

[0032] In block 210, a time series dataset is acquired. The time series dataset includes a plurality of time series data items, each of which includes a time and a corresponding plurality of attribute parameters. In block 220, a target attribute parameter is acquired. The target attribute parameter is at least one of the plurality of attribute parameters. In block 230, at least one influence factor of the target attribute parameter is determined based on the time series model. The at least one influence factor indicates at least one attribute parameter that influences the target attribute parameter and at least one time corresponding to the at least one attribute parameter. In block 240, the at least one influence factor is output.

[0033] In some embodiments, the data items in the time series dataset include:

number

number

[0034] [Table 1]

[0035] As can be understood, the time series dataset shown in Table 1 above is merely an example, and in actual applications, the time series dataset may be represented in other formats, and the present disclosure is not limited in this regard.

[0036] In some embodiments, a user-inputted time series dataset and a user-inputted target attribute parameter may be obtained, for example, the target attribute parameter may be a parameter that the user is interested in or a parameter that the user intends to make a decision on.

[0037] In some other embodiments, an initial time series data set input by a user may be obtained, and after data processing, the time series data set of block 210 may be obtained. Optionally, the data processing may include removing noise and / or generating augmented data items, etc.

[0038] In some embodiments of the present disclosure, the time series model may be obtained by pre-training, for example, pre-constructed by a data scientist, or obtained by other means, although the present disclosure is not limited in this respect. Optionally, the time series model may be determined from multiple candidate models.

[0039] For example, different influence factors among the at least one influence factor determined in block 230 may have different influences on the target attribute parameter. Optionally, the influence may be expressed in the form of an interval, for example, including a maximum influence value and a minimum influence value.

[0040] In some embodiments, when there are multiple influence factors determined in block 230, the multiple influence factors may be ordered based on their influence, and some or all of the multiple influence factors may be output based on the ordering. For example, one or more influence factors with the greatest influence may be determined.

[0041] As can be seen, factor analysis is a data analysis technique that can be used to determine the impact of influencing factors on target attribute parameters. For example, it can determine the importance of product quality or product price on customer attitudes toward a brand. Therefore, factor analysis enables advance prediction and more accurate decision-making.

[0042] In some embodiments of the present disclosure, for a particular influence element among the at least one influence element determined in block 230, e.g., for a first influence element indicating a first attribute parameter, a change in influence of the first attribute parameter on the target attribute parameter over time may be determined. Optionally, the first influence element may be the influence element with the greatest influence determined through ordering.

[0043] In some embodiments, the influence of adjusting a particular attribute parameter on a target attribute parameter may be determined. For example, an adjustment instruction for a second attribute parameter may be obtained, and based on the adjustment instruction, the influence of the second attribute parameter on the target attribute parameter versus time may be determined.

[0044] FIG. 3 illustrates a schematic block diagram of an exemplary system 300 according to some embodiments of the present disclosure. As illustrated in FIG. 3, the system 300 includes a user interface module 310 and a time series analysis engine 320. The user interface module 310 may transmit an input 301 to the time series analysis engine 320, which may perform various operations to obtain an output 302. Additionally or optionally, the time series analysis engine 320 may further transmit the output 302 to the user interface module 310 for display to a user. Illustratively, the time series analysis engine 320 may include a relationship discovery module 322 and a model-based analysis module 323. Additionally or optionally, the time series analysis engine 320 may further include a data processing module 321.

[0045] 4 shows a schematic block diagram of the interaction between a user interface module 310 and a time series analysis engine 320 according to some embodiments of the present disclosure. As shown in FIG. 4, the user interface module 310 includes an input interface 311, a data preparation interface 312, a model selection interface 313, a relationship discovery interface 314, and a model-based analysis interface 315. As shown in FIG. 4, the time series analysis engine 320 includes a data processing module 321, a relationship discovery module 322, and a model-based analysis module 323. The data processing module 321 includes a basic data preprocessing submodule 3211 and a data augmentation submodule 3212. The relationship discovery module 322 includes a model selection submodule 3221, a time series modeling submodule 3222, and a model validation submodule 3223. The model-based analysis module 323 includes a key element identification submodule 3231 and an element analysis submodule 3232.

[0046] For example, the relationship discovery module 322 may use causal discovery techniques to automatically search for potential patterns in a given dataset. Causal discovery techniques may include, but are not limited to, the Peter-Clark (PC) algorithm, Greedy Equivalent Search (GES), Linear Non-Gaussian Model (LinGAM), or Causal Additive Model (CAM). In an embodiment of the present disclosure, a model may be constructed based on a Bayesian network. A Bayesian network can identify individual direct and indirect causes of a target, imply causal relationships, and draw causal conclusions.

[0047] In some examples, embodiments of the present disclosure may perform analysis based on a time series model, which may optionally be a causal model, and may be represented, for example, in the form of a graph.

[0048] In some embodiments, the input interface 311 may be used to receive a time series dataset. For example, the time series dataset may be provided to the data processing module 321. In some embodiments, the input interface 311 may be used to receive configuration information for model selection. For example, the configuration information for model selection may include user suggestions for model selection, structural information of the time series model, etc. For example, the configuration information for model selection may be provided to the model selection sub-module 3221. In some embodiments, the input interface 311 may be used to receive configuration information for analysis. For example, the configuration information for analysis may include user requests and / or target attribute parameter instructions, etc. Optionally, for example, the configuration information for analysis may be provided to the key element identification sub-module 3231.

[0049] In some embodiments, the data preparation interface 312 may be used to receive configuration information related to data processing. For example, the configuration information related to data processing may indicate one or more parameters of data transformation, missing data values, outlier handling, data augmentation, etc. For example, the configuration information related to data processing may be provided to the data processing module 321, such as to the basic data preprocessing sub-module 3211 and the data augmentation sub-module 3212.

[0050] In some embodiments, the model selection interface 313 may be used to receive configuration information related to model selection. For example, the configuration information related to model selection may indicate criteria for model selection, etc. For example, the configuration information related to model selection may be provided to the model selection sub-module 3221.

[0051] In some embodiments, the relationship discovery interface 314 may be used to receive configuration information related to time series modeling and / or configuration information related to validation. For example, the configuration information related to time series modeling may be provided to the time series modeling sub-module 3222. For example, the configuration information related to validation may indicate criteria for validation and may be provided to the model validation sub-module 3223.

[0052] In some embodiments, the model-based analysis interface 315 may be used to receive configuration information for element determination. For example, the configuration information for element determination may indicate target attribute parameters, the number of influencing factors that the user desires to output, one or more thresholds, adjustment instructions for the attribute parameters, etc. For example, the configuration information for element determination may be provided to a model-based analysis module 323, such as a key element identification sub-module 3231 or an element analysis sub-module 3232.

[0053] In some embodiments, the data processing module 321 may be used to process the time series dataset to obtain a processed time series dataset. Illustratively, the basic data preprocessing submodule 3211 may perform preprocessing on the time series dataset, including, for example, data cleansing, data range analysis, and missing value imputation.

[0054] In some examples, data cleansing may include removing anomalous data items. For example, if a data item indicates an age greater than 200 years, remove the item. For example, if a data item indicates a price discount greater than 100%, remove the item.

[0055] In some examples, data range analysis may include determining a range of values ​​for an attribute parameter of the plurality of data items. In some examples, data range analysis may include determining statistics based on quantiles, such as 25%, 50%, or 75%. In some examples, data range analysis may include determining the percentage of missing values ​​for an attribute parameter of the plurality of data items. In some examples, data range analysis may include determining the unique value for each attribute parameter, etc.

[0056] In some examples, if the time series dataset collection process encounters a fault, such as a sensor failure, causing the gap between the two closest times in the time series dataset to be too large, data items may be imputed to the time series dataset by missing value imputation. For example, missing value imputation may be performed based on configuration information regarding missing values ​​from the data preparation interface 312.

[0057] In this manner, the basic data preprocessing sub-module 3211 may preprocess the time series dataset and provide the preprocessed time series dataset to the data augmentation sub-module 3212.

[0058] For illustrative purposes, the basic idea of ​​data augmentation is to generate several synthetic data items that can cover a data space that has not actually been collected and maintain accurate labeling. In general situations, multiple data items in a time-series dataset can represent a uniform time series, that is, the time series is essentially collected at regular time intervals. However, in some fields, such as clinical diagnosis or advertising campaigns, the sample times are usually not consecutive or even regular. In embodiments of the present disclosure, augmented data items (also referred to as synthetic data items) may be obtained by data augmentation to enrich the time-series dataset.

[0059] In some embodiments, a data generation model may be utilized to obtain an augmented data item based on a time series dataset (e.g., a preprocessed time series dataset). The augmented time series dataset may be obtained by adding the augmented data item to the time series dataset (e.g., a preprocessed time series dataset). For example, a first data item and a second data item of the time series dataset (e.g., a preprocessed time series dataset) may be input to the data generation model to obtain an augmented data item, where the first data item has a first time, the second data item has a second time, and the augmented data item has a third time. Optionally, the data generation model may be implemented as a directed acyclic graph (DAG). The data generation model may include two sub-models that are adversarial to each other and may further include a third sub-model for determining a difference between the outputs of the two sub-models.

[0060] In this manner, the data augmentation sub-module 3212 may enrich the time series dataset and provide the augmented time series dataset to the model selection sub-module 3221 .

[0061] In some embodiments, the time series model may be user-input. For example, a data scientist may provide a time series model based on their own experience. In some embodiments, the time series model may be classified based on the type of data. For example, different data items in different time series datasets may represent different types, and corresponding time series models may be different. For example, an appropriate time series model may be determined based on the binary representation of the data items or the continuous characteristics of the data items. In some embodiments, the time series model may be performance-driven. For example, a time series model may be selected from multiple candidate models based on the relationship between the time series dataset and a predetermined dataset. Optionally, the relationship here may include the similarity between features. For example, the extracted feature may be "{%binary=10%, |feature|=100, |sample|=10,000, %missingData=10%}", which indicates that the proportion of binary data is 10%, the feature dimension is 100, the number of samples is 10,000, and the proportion of missing data is 10%. For example, a predetermined data set with the highest similarity may be determined, and then a candidate model corresponding to the predetermined data set with the highest similarity may be determined. Optionally, the determined candidate model may be further fine-tuned to obtain a time series model. For example, the fine-tuning may be performed through actual interaction with a user, but this disclosure is not limited in this respect. In some embodiments, the time series model may be determined based on an application scenario.

[0062] In this manner, the model selection sub-module 3221 may select a time series model for subsequent processing, for example, providing the best modeling settings to the time series modeling sub-module 3222.

[0063] In some examples, the time series modeling submodule 3222 may be used to establish dependencies between multiple attribute parameters across time. Optionally, one measure of the dependency may be based on covariance, such as a Vector Autoregressive (VAR) model, Granger causality, etc. Optionally, one measure of the dependency may be based on a conditional independence test, such as the Peter and Clark - Momentary Conditional Independence (PCMCI) algorithm, which explores dependencies based on lag times.

[0064] In this manner, the time series modeling sub-module 3222 may establish dependencies between multiple variables and may provide, for example, a p-lag coefficient matrix to the model validation sub-module 3223 and the key element identification sub-module 3231 .

[0065] In some embodiments, the model validation submodule 3223 may be used to evaluate the quality of the time series model, for example, based on a set of evaluation metrics. Optionally, the set of evaluation metrics may be included in the configuration information provided by the relationship discovery interface 314 for validation. Illustratively, the method for model validation is R 2 , Mean Square Error (MSE), F statistic, or other methods, and the present disclosure is not limited in this respect.

[0066] In some embodiments, the key factor identification sub-module 3231 may be used to determine a set of influencing factors that affect the target attribute parameter, where the set of influencing factors includes at least one influencing factor. In some examples, a given time series model may be used to determine all possible causal factors that affect the target attribute parameter, and a key factor set may be determined from all possible causal factors. For example, the causal factors may include time-lagged causal factors, i.e., factors at other times that are lagged relative to the current time have an effect on the target attribute parameter at the current time. For example, the causal factors may include instantaneous causal factors, i.e., other attribute parameters at the current time have an effect on the target attribute parameter at the current time.

[0067] In this manner, the key element identification sub-module 3231 may be used to identify elements that have a significant impact on the target attribute parameter, and may provide a set of key elements to the element analysis sub-module 3232, for example.

[0068] In some embodiments, the factor analysis submodule 3232 may perform forward analysis or backward analysis. By way of example, forward analysis may be used to identify disturbance responses that have the greatest future impact. For example, it may be determined which attribute parameter disturbances cause the cumulative impact on a target attribute parameter to exceed a predetermined threshold at a given time. By way of example, backward analysis may be used to understand which factor or factors have the greatest impact and are most important for a given time. For example, factors determined for the same target attribute parameter (e.g., restaurant selection) will typically be different at different given times (e.g., weekdays and weekends). Optionally, if the causal relationships are linear, the ordering of related factors across different times will be consistent.

[0069] It should be noted that, although exemplary systems according to the embodiments of the present disclosure have been described in connection with Figures 3 and 4, what is shown here is merely an example and should not be construed as a limitation on the embodiments of the present disclosure. For example, it should be understood that in an actual scenario, the data processing module 321 may be omitted, or the system may further include other modules. More detailed embodiments will be described later in connection with Figures 5 to 17.

[0070] In some embodiments, an initial time series model may be obtained, and a time series model (which may be referred to as a first time series model) may be obtained based on the initial time series model. In some examples, the time series model may be obtained based on the initial time series model and the time series dataset. Optionally, the time series model may be represented in the form of a time relationship diagram, which includes nodes and edges. For ease of explanation, the initial time series model may be referred to as a first time relationship diagram, and the time series model may be referred to as a second time relationship diagram.

[0071] For example, the first temporal relationship diagram may be input by a user or may be a diagram provided by an expert based on experience. In some examples, a test (e.g., a t-test) may be performed on each edge of the first temporal relationship diagram to obtain a value representing a significance level, referred to as a p-value. Then, based on the p-value, it may be determined whether each edge represents a causal relationship, for example, whether the edge is significantly equal to 0. If equal, the corresponding edge is retained; if not equal, the corresponding edge is deleted. The deleted edge may be considered a redundant edge in the first temporal relationship diagram. In this manner, the second temporal relationship diagram can be obtained by removing the redundant edges.

[0072] In this example, the input may be understood to include a first time relationship diagram, a target attribute parameter, and a time series dataset, and the output may be understood to include a second time relationship diagram. Optionally, the first time relationship diagram may be represented as G, as shown in FIG. 5A for example. The first time relationship diagram as shown in FIG. 5A may be input by a user (e.g., provided by an expert) or determined by an algorithm (e.g., constructed by the time series modeling submodule 3222), and the present disclosure is not limited in this respect.

[0073] The first temporal relationship diagram may represent the causal relationships between multiple attribute parameters at different times, for example, the arrows in Figure 5A represent that a cause points to an effect, for example, arrow 501 represents that there is a causal relationship between the online advertising volume at time t-1 and the online advertising volume at time t. Figure 5A also further circled the target attribute parameter "sales volume."

[0074] As can be seen, there is a lag effect in the first time relationship diagram, i.e., past information (e.g., t-1, t-2) affects the current value at time t. There is also an instantaneous effect, i.e., the remaining attribute parameters at time t have an instantaneous effect on the target attribute parameters at time t.

[0075] FIG. 6 illustrates an exemplary flowchart of a process 600 for obtaining a second temporal relationship diagram according to some embodiments of the present disclosure. In block 610, multiple time lag elements that may have a causal relationship with the target attribute parameter are determined in the first temporal relationship diagram. In block 620, some key elements of the multiple possible time lag elements are determined through statistical testing. In block 630, redundant edges are removed. For example, the input 601 of FIG. 6 includes a time series dataset (e.g., represented as X), the first temporal relationship diagram (e.g., represented as G), and the target attribute parameter (e.g., represented as T), and the output 602 of FIG. 6 includes a second temporal relationship diagram (e.g., represented as G′). In one example, the second temporal relationship diagram is as shown in FIG. 5B. Alternatively, the second temporal relationship diagram may be referred to as a temporal causal graph or other names, and the present disclosure is not limited in this respect.

[0076] Optionally, in block 610, a Structure Vector Autoregressive (SVAR) algorithm may be used, for example, assuming that all other attribute parameters (i.e., nodes or variables) have a direct edge to the target attribute parameter.

[0077] Optionally, in block 620, a t-test algorithm may be used, for example, quantizing the arithmetic mean value to obtain a p-value for each edge. Further, in block 630, if the p-value is less than a threshold, it means that it is significantly not equal to 0. If an edge is significantly not equal to 0, it means that the start node of this edge has a relatively large influence on the end node.

[0078] In some examples, the obtained second temporal relationship diagram (temporal causal graph) may include (tp+1)*K parameters and their influences on the target attribute parameters, where tp represents the time lag value (e.g., in FIG. 5B, tp=2), and K represents the number of key elements (e.g., in FIG. 5B, K=3). In this manner, the causal effect may be determined by the second temporal relationship diagram (temporal causal graph).

[0079] Thus, at least one influence factor of the target attribute parameter (e.g., sales volume in FIG. 5B) may be determined based on the second time relationship diagram. As shown in FIG. 5B, the direct influence factors on the sales volume at time t include (t-1, sales volume), (t, online advertising volume), and (t, user view volume). Each influence factor includes an attribute parameter and a corresponding time. As can be understood, the direct influence factor may indicate an edge directly connected to (t, sales volume). For example, indirect influence factors may also be included, but are not listed here.

[0080] In some embodiments, each influence element may have an influence on the target attribute parameter. In one example, the influence may be expressed in the form of a number, such as the numbers written near a particular edge in FIG. 5B . In another example, the influence may be expressed in the form of a numerical interval. In this way, the influence expressed in the form of an interval can reflect more statistical information and can be used to more accurately determine causal relationships.

[0081] Furthermore, for the plurality of influence factors, some key factors may be further determined. Figure 7 shows a schematic flowchart of a process 700 for determining key factors according to some embodiments of the present disclosure. At block 710, the influence of each influence factor of the plurality of influence factors on the target attribute parameter is determined. At block 720, the plurality of influence factors are ordered based on the influence of each influence factor. At block 730, some or all of the plurality of influence factors are output based on the ordering.

[0082] In some examples, the quantities of some or all of the influence factors to be output may be specified by a user, for example, a user may input values ​​for the quantities, e.g., represented as m, via the model-based analysis interface 315.

[0083] In this example, it may be understood that the input includes the second time relationship diagram and the quantity m, and the output includes the key m influencing elements.

[0084] Optionally, when ordering in block 720, a comparison matrix may be constructed, the influence element with the smallest (or largest) value may be determined from the comparison matrix, and then the order may be determined from the influence element with the smallest (or largest) value to complete the ordering.

[0085] In some embodiments, if the influences are expressed as numerical intervals, the ordering needs to determine relationships based on the intervals. For example, the relationships include:

number

[0086] Specifically, it is assumed that the first influence of a first influence element is represented as a first interval, and the second influence of a second influence element is represented as a second interval. The relationship between the first influence and the second influence may include the following: (1) If the upper limit value of the first interval is smaller than the lower limit value of the second interval, the first influence is smaller than the second influence (<). (2) If the first interval and the second interval have an overlapping area and the first statistic of the first interval is smaller than the second statistic of the second interval, the first influence is equal to or smaller than the second influence (<).

number

Number

[0087] As can be understood, the upper limit may also be referred to as the upper bound value, and the lower limit may also be referred to as the lower bound value. Based on the upper limit and the lower limit, an interval that can be trusted can be determined, and the true value exists within that interval that can be trusted. For example, the interval that can be trusted may represent a probability of 1-α, and α represents the significance level. For example, when α = 0.05, as α increases, the corresponding interval that can be trusted shrinks, and as α decreases, the corresponding interval that can be trusted expands.

[0088] As an illustration, in block 730, m influencing factors may be determined based on the ordering. As an example, when the influence is represented in the form of a numerical interval, FIG. 8B shows a schematic diagram for determining two numerical intervals 820 based on the ordering of three intervals a, b, c. In the illustration 822 of FIG. 8B,

Number

[0089] In some other illustrations, the ordering may be represented in a two-dimensional coordinate graph. For example, FIG. 9 shows a plurality of influences represented in coordinates, and m influencing factors may be determined therefrom.

[0090] In some embodiments of the present disclosure, the influence of a particular attribute parameter on a target attribute parameter over time may be determined. Optionally, the particular attribute parameter may be referred to as a first attribute parameter, and the first attribute parameter may be specified by a user. For example, a user may input an indication of the first attribute parameter (e.g., denoted as q) via the input interface 311 or the model-based analysis interface 315.

[0091] In this example, the input may be understood to include the second time relationship diagram, the first attribute parameter, and the target attribute parameter, and the output may be understood to include a time-dependent change in the influence of the first attribute parameter on the target attribute parameter, where the time-dependent change is expressed in the form of a vector, for example.

[0092] Specifically, the influence of the first attribute parameter on the target attribute parameter at each time may be obtained from the second time relationship diagram, and thus its change over time may be obtained. In one example, the change over time in the influence of the first attribute parameter on the target attribute parameter may be represented in the form of a curve (or a broken line). Assuming that the target attribute parameter is (t, sales volume), FIG. 10A shows a relationship 1010 of the change over time in the influence of sales volume on the target attribute parameter, and FIG. 10B shows a relationship 1020 of the change over time in the influence of user view volume on the target attribute parameter. In FIGS. 10A and 10B, the horizontal axis represents the time lag, and the time lag value represents the time unit value that lags the time of the target attribute parameter. For example, time lag 2 represents t-2. If the time unit is days, t-2 represents two days ago. As can be seen from FIG. 10A, the influence of past sales volume on current sales volume tends to decrease. As can be seen from FIG. 10B, the amount of past user views has an effect on the current sales volume both in the short term (eg, three days ago) and in the long term (eg, two weeks ago).

[0093] In some embodiments of the present disclosure, a response of the model may also be determined. Illustratively, one or more of the attribute parameters may be adjusted to determine a response of the target attribute parameter. For example, an attribute parameter that maximizes the response of the target attribute parameter may be determined. FIG. 11 shows a schematic flowchart of a process 1100 for determining a response relationship according to some embodiments of the present disclosure. At block 1110, a first adjustment instruction for a second attribute parameter of the plurality of attribute parameters is obtained. At block 1120, a relationship between the influence of the second attribute parameter on the target attribute parameter and time is determined based on the first adjustment instruction.

[0094] In some examples, the effect of adjusting the second attribute parameter on the target attribute parameter over a predetermined time period may be determined. For example, the predetermined time period may be input by a user, such as by a user entering a value for the predetermined time period, e.g., represented as D, via the model-based analysis interface 315.

[0095] In this example, the input may be understood to include the second time relationship diagram, the target attribute parameter, the predetermined period D, and the first adjustment instruction for the second attribute parameter. The output may include the change in response over the predetermined period D, represented, for example, as a vector including D elements.

[0096] In some embodiments, the relationship determined in block 1120 may be expressed as a change in influence over time (the time range is a predetermined period D) or as a cumulative change in influence over time (the time range is a predetermined period D). For example, assuming a target attribute parameter is (t, sales volume) and a second attribute parameter is sales volume, FIG. 12A shows the change in the influence of adjusting sales volume on the target attribute parameter over time, and FIG. 12B shows the cumulative change in the cumulative influence of adjusting sales volume on the target attribute parameter over time.

[0097] Optionally, the cumulative influence may further be determined to be greater than a predetermined threshold for a predetermined period of time. In conjunction with Figure 12B, assuming the predetermined threshold is 15, it can be seen that the cumulative influence for a predetermined period of time is less than the predetermined threshold.

[0098] In some other examples, two or more attribute parameters may be adjusted separately or simultaneously to determine a response to a target attribute parameter within a certain period D. For example, a second adjustment instruction for a third attribute parameter may be further obtained, and a relationship between the influence of the third attribute parameter on the target attribute parameter and the predetermined period may be determined based on the second adjustment instruction. For example, an attribute parameter having a relatively larger influence on the target attribute parameter from among the second attribute parameter and the third attribute parameter may be further determined.

[0099] For example, if the target attribute parameter is (t, sales volume) and the third attribute parameter is the number of user views, Figure 13A shows the change over time in the influence of adjusting the number of user views on the target attribute parameter, and Figure 13B shows the cumulative change over time in the cumulative influence of adjusting the number of user views on the target attribute parameter. Comparing Figure 12A and Figure 13A, it can be seen that the influence of sales volume is greater.

[0100] Additionally or optionally, a most effective time point may be further determined. The most effective time point represents the time when the optimal effect is obtained when a decision is made. In this way, when a decision is made by instructing an adjustment to an attribute parameter, the time point at which the decision is most effective may be determined based on this. For example, a threshold value for the desired effect may be determined, or the effect of the parameter adjustment may be simulated, and the time point may be determined by determining the relationship (e.g., the intersection point) between the effect of the adjustment and the threshold value for the desired effect. In this way, a more appropriate decision can be made based on this.

[0101] Through the above-mentioned embodiments combined with Figures 2 to 13B, the present disclosure provides a data processing solution, which can determine, for a target attribute parameter, at least one influencing factor on the target attribute parameter based on a time series dataset and a time series model, and further facilitates providing accurate reference information for decision-making on the target attribute parameter, thereby making the decision-making more accurate.

[0102] As can be understood, embodiments of the present disclosure can be applied to different fields. Taking the sales field as an example, FIG. 14 shows a schematic flowchart of a process 1400 for determining inventory quantity according to some embodiments of the present disclosure. In block 1410, a time series dataset related to sales records is obtained. The time series dataset includes multiple time series data items. Each time series data item includes a time and multiple corresponding attribute parameters. The multiple attribute parameters include at least one of a purchase price, a sales price, a sales volume, an inventory quantity, or a customer view count. In block 1420, at least one influencing factor on the inventory quantity is determined based on a time series model. The at least one influencing factor indicates at least one attribute parameter that affects the inventory quantity and at least one time corresponding to the at least one attribute parameter. In block 1430, the inventory quantity for the next time is determined based on the at least one influencing factor. For example, in the sales field, if the target attribute parameter is inventory quantity, an embodiment of the present disclosure may determine a key factor having a causal relationship with the inventory quantity. As can be appreciated, the schematic flowchart shown in FIG. 14 is for illustrative purposes only, and in a practical scenario, the target attribute parameter may be other variables, such as sales profit, and the present disclosure is not limited in this regard.

[0103] In another example, taking the field of green energy as an example, FIG. 15 shows a schematic flowchart of a process 1500 for determining the operating time of a device according to some embodiments of the present disclosure. In block 1510, a time series dataset related to the power of an IoT device is acquired. The time series dataset includes multiple time series data items, each of which includes a time and multiple corresponding attribute parameters. The multiple attribute parameters include at least one of an electricity price, an electricity demand, a voltage, a current, a temperature, a humidity, an air pressure, or a power consumption of the IoT device. In block 1520, at least one influencing factor of the power consumption is determined based on a time series model. The at least one influencing factor indicates at least one attribute parameter that affects the power consumption and at least one time corresponding to the at least one attribute parameter. In block 1530, the operating time of one or more IoT devices is determined based on the at least one influencing factor. For example, in the field of green energy, if the target attribute parameter is power consumption, the embodiments of the present disclosure may determine key factors having a causal relationship with the power consumption. As can be appreciated, the schematic flowchart shown in FIG. 15 is for illustrative purposes only, and in a real scenario, the target attribute parameter may be other variables, such as household electricity bill payments, and the present disclosure is not limited in this regard.

[0104] Optionally, the IoT device may be an Artificial Intelligence - Internet of Things (AIoT) device, and the device or system for energy control of the IoT device may be a Green AIoT controller. As can be seen, modern IoT devices (such as electrical appliances and cookers) are becoming increasingly sophisticated, but their energy consumption comes at a cost in the form of air pollution, water pollution, deforestation, and so on. In this context, Green Transformation refers to actions that utilize digital technology to solve these environmental problems while achieving economic growth. For example, in the home environment, Green Transformation requires a controller to maintain the home's internal environment at an appropriate level so that all household appliances can function properly, while ensuring economic growth. For example, a refrigerator consumes more electricity in a hot and humid environment. Operating conditions are improved when air conditioning and ventilation systems work together. However, if all electrical appliances are operating during peak hours, a low voltage means the refrigerator takes longer to reach the desired temperature and consumes more electricity. By using the green AIoT controller in the embodiment of the present disclosure to collect electricity bill data in real time, sense the internal temperature and humidity, collect user behavior patterns, and analyze the collected data, it is possible to determine the operation of energy consumption, thereby achieving the goal of green transformation.

[0105] For example, a time series data set may be obtained by collecting changes in multiple attribute parameters over time. Referring to Figure 16, changes over time in electricity prices, changes in demand in a city, household voltage, household current, kitchen air conditioner temperature, kitchen humidifier humidity, bedroom air conditioner temperature, and bedroom humidifier humidity are shown. Optionally, the horizontal axis represents time, which may be in minutes, hours, or other time units, although the present disclosure is not limited in this respect.

[0106] By utilizing an embodiment of the present disclosure, the total electricity consumption of a household may be assumed to be a target attribute parameter, and key factors affecting the target attribute parameter may be determined to include the electricity price one hour ago, the current electricity consumption of the kitchen air conditioner, etc. Furthermore, based on a time series dataset, it may be determined that peak demand in a city may lead to peak electricity prices. Based on this, the controller may determine the operating hours of each appliance in the household. For example, most appliances may be operated during off-peak demand times in the city.

[0107] In some other examples, taking the field of abnormal event response based on a social network as an example, FIG. 17 shows a schematic flowchart of a process 1700 for determining an abnormal event according to some embodiments of the present disclosure. In block 1710, a time series dataset collected from a social network is acquired. The time series dataset includes a plurality of time series data items, each of which includes a time and a corresponding plurality of attribute parameters. The plurality of attribute parameters include at least one of user identification information, user posts, the number of views of the user posts, the number of comments of the user posts, or an abnormality indication. In block 1720, at least one influence factor of the abnormality indication is determined based on the time series model. The at least one influence factor indicates at least one attribute parameter that affects the target attribute parameter and at least one time corresponding to the at least one attribute parameter. In block 1730, presentation information is output based on the at least one influence factor. The presentation information indicates a user at an abnormal risk.

[0108] By way of example, the abnormal indication may include indicating whether the user committed suicide, indicating whether the user committed illegal acts, etc. For ease of explanation, a user's suicide will be described below as an abnormal incident.

[0109] According to data from the World Health Organization, nearly 800,000 people commit suicide each year, a tragedy occurring roughly every 40 seconds. Due to the rare nature of suicide, available data is scarce. However, the booming internet has led to a rapid increase in the amount of data available on social networking media. Therefore, identifying the causes and consequences of suicide from social networking media and implementing effective interventions is highly desirable and feasible. A device or system used to determine abnormal incidents through social networks may be called an emotional robot. By collecting a user's timely activities, their social network updates, and influential news, fluctuations in the user's state can be captured. For example, by collecting each attribute parameter at multiple different times, a time series dataset can be obtained.

[0110] In some embodiments, the time series dataset may be divided into a first subset and a second subset. The first subset may be a suicide dataset, representing users therein who have performed a suicidal act. The second subset may be a non-suicidal dataset, representing users therein who have not performed a suicidal act. In some examples, similarity matching may further determine one or more non-suicidal users from the non-suicidal dataset that have a similarity to the suicide dataset above a threshold.

[0111] As an example, for example, the time series of 20 suicidal users and the time series of 80 non-suicidal users (normal users) may be acquired. Each user's own time series is used to construct their own time series model.

[0112] The target attribute parameter may be "mood," and may be expressed as a numerical value based on, for example, text information of a user, which may include posts, comments, etc.

[0113] As an example, we may assume that the current "mood" of one non-suicidal user (e.g., user 1) among 80 determined non-suicidal users (normal users) is expressed as follows: Mood (t) = 5 × mood (t-1) + 2 × number of views (t-1) + 5 × number of views (t-2).

[0114] For example, we may assume that the current "mood" of one suicidal user (say, user 2) among 20 suicidal users is expressed as follows: Mood (t) = -3 x number of late-night posts (t-1) - 15 x mood (t-1) - 5 x mood (t-3) + 2 x number of views (t-2).

[0115] In the above formula, mood(t) represents the current mood, t-1 represents a lag of one time unit, t-2 represents a lag of two time units, and t-3 represents a lag of three time units. For example, if the time unit is "days," t-1, t-2, and t-3 represent one day ago, two days ago, and three days ago, respectively.

[0116] 18A-18C show schematic diagrams of determining an abnormality risk according to some embodiments of the present disclosure. Specifically, for a specific user (e.g., user xyz), the specific user's time series is collected, and the specific user's mood is determined (e.g., predicted) from the specific user's time series.

[0117] Referring to FIG. 18A, assume that actual collected data for a particular user is shown as "x" and the predicted mood is shown as dashed line 1801. FIG. 18A also shows, for example, mood curves for four non-suicidal users (solid lines) and two suicidal users (dotted lines). Through analysis, it may be determined that the similarity between a particular user and suicidal users is 52% and the similarity between a particular user and non-suicidal users is 48%. As an example, assuming the threshold is 80%, 52%<80%, so no presentation information may need to be output based on FIG. 18A.

[0118] Referring to FIG. 18B, assume that actual collected data for a particular user is shown as "x" and the predicted mood is shown as dashed line 1802. FIG. 18B also shows, for example, mood curves for four non-suicidal users (solid lines) and two suicidal users (dotted lines). Through analysis, it may be determined that the similarity between a particular user and suicidal users is 73% and the similarity between a particular user and non-suicidal users is 27%. As an example, assuming the threshold is 80%, 73%<80%, so no presentation information may need to be output based on FIG. 18B.

[0119] Referring to FIG. 18C , assume that actual collected data for a particular user is shown as an "x" and the predicted mood is shown as a dashed line 1803. FIG. 18C also shows, for example, mood curves for three non-suicidal users (solid lines) and two suicidal users (dotted lines). Through analysis, it may be determined that the similarity between the particular user and suicidal users is 89% and the similarity between the particular user and non-suicidal users is 11%. As an example, assuming the threshold is 80%, 89% > 80%, and therefore, presentation information may be output based on FIG. 18C . The presentation information may indicate that the particular user is at risk of suicide.

[0120] As can be appreciated, the illustrations shown in Figures 18A-18C are merely examples, and for example, the figures may show more or fewer mood curves for non-suicidal users and suicidal users. This disclosure is not limited in this respect. Also, while the horizontal axes in Figures 18A-18C represent time (days), in a real scenario, the time units may be other time units, and this disclosure is not limited in this respect.

[0121] In this way, the embodiment of the present disclosure can determine users who may be at risk of anomalies based on the timeline of users on a social network and output suggested information, thereby facilitating timely intervention and assistance, reducing or even preventing the occurrence of tragedies.

[0122] As can be appreciated, although applicability in certain areas has been illustrated above in conjunction with Figures 14 through 18C, these areas are provided for illustrative purposes only and are not intended to limit the scope of the present invention in any way. The embodiments of the present disclosure may be applied to a variety of areas where similar problems exist, and will not be listed here one by one.

[0123] In some embodiments, the computing device includes circuitry configured to perform the following operations: acquiring a time series dataset including a plurality of time series data items, each time series data item including a time and a corresponding plurality of attribute parameters, acquiring a target attribute parameter, the target attribute parameter being at least one of the plurality of attribute parameters, determining at least one influence factor of the target attribute parameter based on a time series model, the at least one influence factor indicating at least one attribute parameter that influences the target attribute parameter and at least one time corresponding to the at least one attribute parameter, and outputting the at least one influence factor.

[0124] In some embodiments, each influence element of the at least one influence element has an influence on the target attribute parameter, and the influence is expressed in the form of an interval.

[0125] In some embodiments, the at least one influence element includes a plurality of influence elements, and the computing device includes circuitry configured to perform the following operations: determine an influence of each influence element in the plurality of influence elements on the target attribute parameter, order the plurality of influence elements based on the influence of each influence element, and output some or all of the plurality of influence elements based on the ordering.

[0126] In some embodiments, the plurality of influence elements include a first influence element and a second influence element, the first influence element having a first influence on the target attribute parameter, the second influence element having a second influence on the target attribute parameter, the first influence element being expressed as a first interval, and the second influence element being expressed as a second interval, and the computing device includes a circuit configured to perform the following operations: the first influence element is smaller than the second influence element if an upper limit value of the first interval is smaller than a lower limit value of the second interval; the first influence element is smaller than or equal to the second influence element if the first interval and the second interval have an overlapping area and the first statistic of the first interval is smaller than the second statistic of the second interval; or the first influence element is equal to the second influence element if a difference between the lower limit value of the first interval and the lower limit value of the second interval is smaller than a first threshold, the upper limit value of the first interval and the upper limit value of the second interval are smaller than a second threshold, and the difference between the first statistic and the second statistic is smaller than a third threshold.

[0127] In some embodiments, the computing device includes circuitry configured to perform the following operation: determine a change in influence of a first attribute parameter of the plurality of attribute parameters on a target attribute parameter over time.

[0128] In some embodiments, the computing device includes circuitry configured to perform operations to obtain an indication of the first attribute parameter input by a user.

[0129] In some embodiments, the computing device includes circuitry configured to perform the following operations: obtain a first adjustment instruction for a second attribute parameter of the plurality of attribute parameters; and determine, based on the first adjustment instruction, an influence of the second attribute parameter on a target attribute parameter versus time.

[0130] In some embodiments, the indication of the influence of the second attribute parameter on the target attribute parameter versus time indicates at least one of a change in the influence of the second attribute parameter on the target attribute parameter over time, or a cumulative change in the cumulative influence of the second attribute parameter on the target attribute parameter over time.

[0131] In some embodiments, the computing device includes circuitry configured to perform the following operation: determine, based on a cumulative change over time in the cumulative influence of the second attribute parameter on the target attribute parameter, whether the cumulative influence reaches a predetermined threshold for a predetermined period of time.

[0132] In some embodiments, the computing device includes circuitry configured to perform the following operations: obtain a second adjustment instruction for a third attribute parameter of the plurality of attribute parameters; and determine, based on the first adjustment instruction and the second adjustment instruction, an attribute parameter from among the second attribute parameter and the third attribute parameter that has a relatively greater influence on the target attribute parameter.

[0133] In some embodiments, the computing device includes circuitry configured to perform operations to obtain a first adjustment instruction input by a user.

[0134] In some embodiments, the computing device includes circuitry configured to perform the following operations: obtaining an initial time series model; and performing testing on the initial time series model to remove redundant edges to obtain a time series model.

[0135] In some embodiments, the computing device includes circuitry configured to perform the following operation: select a time series model from among a plurality of candidate models based on a relationship between the time series data set and a predetermined data set.

[0136] In some embodiments, the computing device includes circuitry configured to perform the following operations: obtain an initial time series data set; and process the initial time series data set by at least one of removing noise data items or generating augmented data items to obtain the time series data set.

[0137] In some embodiments, the computing device includes circuitry configured to perform the following operations: input a first data item and a second data item of the initial time series dataset into a data generation model to obtain an augmented data item, where the first data item has a first time, the second data item has a second time, and the augmented data item has a third time.

[0138] In some embodiments, the computing device includes circuitry configured to perform the following operations: acquiring a time series dataset related to sales records, the time series dataset including a plurality of time series data items, each time series data item including a time and a corresponding plurality of attribute parameters, the plurality of attribute parameters including at least one of a purchase price, a sale price, a sales amount, an inventory amount, or a number of customer views; determining at least one influencing factor on an inventory amount based on a time series model, the at least one influencing factor indicating at least one attribute parameter affecting the inventory amount and at least one time corresponding to the at least one attribute parameter; and determining an inventory amount for a next time based on the at least one influencing factor.

[0139] In some embodiments, the computing device includes circuitry configured to perform the following operations: acquiring a time series dataset related to power of an IoT device, the time series dataset including a plurality of time series data items, each time series data item including a time and a corresponding plurality of attribute parameters, the plurality of attribute parameters including at least one of an electricity price, an electricity demand, a voltage, a current, a temperature, a humidity, a barometric pressure, or a power consumption of the IoT device; determining at least one influencing factor of the power consumption based on a time series model, the at least one influencing factor indicating at least one attribute parameter affecting the power consumption and at least one time corresponding to the at least one attribute parameter; and determining an operating time of the IoT device based on the at least one influencing factor.

[0140] In some embodiments, the computing device includes circuitry configured to perform the following operations: acquiring a time series dataset collected from a social network, the time series dataset including a plurality of time series data items, each time series data item including a time and a corresponding plurality of attribute parameters, the plurality of attribute parameters including at least one of a user identification information, a user post, a number of views of the user post, a number of comments of the user post, or an anomaly indication; determining at least one influencing factor of the anomaly indication based on the time series model, the at least one influencing factor indicating at least one attribute parameter affecting a target attribute parameter and at least one time corresponding to the at least one attribute parameter; and outputting presentation information indicating a user at an anomaly risk based on the at least one influencing factor.

[0141] FIG. 19 illustrates a block diagram of an exemplary device 1900 capable of implementing embodiments of the present disclosure. For example, the computing device 110 illustrated in FIG. 1 may be implemented by the device 1900. As illustrated in FIG. 19, the device 1900 includes a central processing unit (CPU) 1901. The CPU 1901 may perform various appropriate operations and processes based on instructions of a computer program stored in a read-only memory (ROM) 1902 or loaded from a storage unit 1908 into a random access memory (RAM) 1903. The RAM 1903 may further store various programs and data necessary for the operation of the device 1900. The CPU 1901, the ROM 1902, and the RAM 1903 are connected to one another via a bus 1904. An input / output (I / O) interface 1905 is also connected to the bus 1904.

[0142] Several components of the device 1900 are connected to an I / O interface 1905. The several components include an input unit 1906 such as a keyboard, a mouse, etc.; an output unit 1907 such as various types of displays, speakers, etc.; a storage unit 1908 such as a magnetic disk, an optical disk, etc.; and a communication unit 1909 such as a network interface card, a modem, a wireless communication transceiver, etc. The communication unit 1909 enables the device 1900 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks. It should be understood that in the present disclosure, the output unit 1907 may be used to display information regarding real-time dynamic changes in user satisfaction, key factor identification information for user groups or individual users regarding satisfaction, information regarding optimization policies, and information regarding evaluation of policy implementation effects, etc.

[0143] The processor unit 1901 may be implemented by one or more processing circuits. The processor unit 1901 may be configured to perform each of the processes and operations described above. For example, in some embodiments, the aforementioned processes may be implemented as a computer software program and tangibly stored in a machine-readable medium, such as the storage unit 1908. In some embodiments, some or all of the computer program may be loaded and / or installed into the device 1900 via the ROM 1902 and / or the communication unit 1909. When the computer program is loaded into the RAM 1903 and executed by the CPU 1901, it may perform one or more steps of the processes described above.

[0144] The present disclosure may be a system, a method, and / or a computer program product, which may include a computer-readable storage medium having computer-readable program instructions stored thereon for carrying out aspects of the present disclosure.

[0145] A computer-readable storage medium may be a tangible device capable of holding and storing instructions for use by an instruction execution device. A computer-readable storage medium may be, for example, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples of computer-readable storage media include (but are not limited to) portable computer diskettes, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital versatile disk (DVD), memory sticks, floppy disks, mechanical encoder disks, punch cards or protrusion structures in grooves on which instructions are stored, and any suitable combination of the above. As used herein, a computer-readable storage medium is not to be understood as a momentary signal itself, such as, for example, radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through guided waves or other transmission media (e.g., light pulses through optical cable), or electrical signals transmitted over electrical wires.

[0146] The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to each computing / processing device, or may be downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include copper transmission cables, optical transmission cables, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. A network interface card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions to be stored in the computer-readable storage medium of each computing / processing device.

[0147] Computer program instructions for carrying out the operations of the present disclosure may be assembler directives, instruction set architecture (ISA), machine language instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages. Programming languages ​​include object-oriented programming languages ​​such as Smalltalk, C++, and general process-based programming languages ​​such as "C" or similar programming languages. The computer-readable program instructions may execute entirely on the user computer, partially on the user computer, as a separate software package, partially on the user computer and partially on a remote computer, or entirely on a remote computer or server. In the context of a remote computer, the remote computer may be connected to the user computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet Service Provider). In some embodiments, the status information of the computer-readable program instructions is used to personalize electronic circuitry, such as a programmable logic circuit, a field-programmable gate array (FPGA), or a programmable logic array (PLA), which may execute the computer-readable program instructions to implement aspects of the present disclosure.

[0148] Aspects of the present disclosure are described herein with reference to flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present disclosure. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, may be implemented by computer-readable program instructions.

[0149] These computer-readable program instructions may be provided to a processor unit of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus to produce a machine, which, when executed by the processor unit of the computer or other programmable data processing apparatus, produces an apparatus that implements the functions / acts specified in one or more blocks of the flowcharts and / or block diagrams. These computer-readable program instructions may be stored on a computer-readable storage medium. These instructions cause a computer, programmable data processing apparatus, and / or other device to operate in a particular manner. Thus, a computer-readable medium having instructions stored thereon includes an article of manufacture containing instructions for implementing the functions / acts specified in one or more blocks of the flowcharts and / or block diagrams.

[0150] The computer-readable program instructions may be loaded into a computer, other programmable data processing apparatus, or other device to cause the computer, other programmable data processing apparatus, or other device to perform a series of operational steps to generate a computer-implemented process, such that the instructions executed on the computer, other programmable data processing apparatus, or other device implement the functions / acts specified in one or more blocks of the flowcharts and / or block diagrams.

[0151] The flowcharts and block diagrams in the figures represent possible architectures, functions, and operations of systems, methods, and computer program products according to embodiments of the present disclosure. In this regard, each block in the flowcharts or block diagrams may represent a module, program segment, or part of an instruction, which includes one or more executable instructions for implementing a specified logical function. In alternative implementations, the functions depicted in the blocks may occur in a different order than depicted in the figures. For example, two consecutive blocks may actually be executed essentially in parallel, or may even be executed in the reverse order, as determined by the functionality involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented in a dedicated hardware-based system that performs the specified functions or operations, or in a combination of dedicated hardware and computer instructions.

[0152] Although each embodiment of the present disclosure has been described above, the above description is illustrative and not exhaustive, and is not limited to each of the disclosed embodiments. It is clear that those skilled in the art can make multiple modifications and changes without departing from the scope and spirit of each of the described embodiments. The terms used herein have been selected with the intention of optimally explaining the principles, actual applications, or technical improvements in the market of each embodiment, or allowing those skilled in the art to understand each embodiment disclosed in this specification.

Claims

1. 1. A data processing method for executing, by a computer, processing on a time series data set including a plurality of time series data items, each of which has a time and a plurality of attribute parameters corresponding to the time as elements, the method comprising: obtaining the time series data set; acquiring, from the attribute parameters included in the plurality of time-series data items, a target attribute parameter that is determined by being influenced by some of the other attribute parameters included in the plurality of time-series data items and that is to be a target of the influence analysis; determining at least one influencing factor by inputting the acquired target attribute parameter and the time series data set into a time series model given in advance to determine at least one influencing factor indicating at least one attribute parameter that influences the target attribute parameter in a time series manner among the attribute parameters included in the plurality of time series data items and a time corresponding to the at least one attribute parameter; and outputting the determined at least one influence factor. Data processing methods.

2. obtaining at least one target attribute parameter; determining the at least one influencing factor for each of the at least one target attribute parameter obtained; The data processing method according to claim 1 .

3. Each of the at least one influence element has an influence on the target attribute parameter, and the influence is expressed in the form of a numerical range.

3. The data processing method according to claim 1 or 2.

4. the at least one influence element includes a plurality of influence elements; Each of the at least one influencing element has an influence on the target attribute parameter; After determining the at least one influencing factor, determining an influence of each of the plurality of influencing factors on the target attribute parameter; ordering the plurality of influence factors based on a numerical range of influence of each of the plurality of influence factors; outputting one or all of the influence elements having the greatest influence among the plurality of influence elements based on the result of ordering the plurality of influence elements; further comprising: The data processing method according to claim 3 .

5. the plurality of influence factors include a first influence factor and a second influence factor; the first influencing element has a first influence on the target attribute parameter; the second influencing element has a second influence on the target attribute parameter; the first influence is expressed as a first range of values, and the second influence is expressed as a second range of values; The ordering of the plurality of influence factors includes: When the upper limit value of the first numerical range is smaller than the lower limit value of the second numerical range, it is determined that the first influence is smaller than the second influence; If the first numerical range and the second numerical range have an overlapping area and the median value of the first numerical range is smaller than the median value of the second numerical range, the first influence is determined to be equal to or less than the second influence; or When the difference between the lower limit value of the first numerical range and the lower limit value of the second numerical range is smaller than a first threshold value, the upper limit value of the first numerical range and the upper limit value of the second numerical range are smaller than a second threshold value, and the difference between the median value of the first numerical range and the median value of the second numerical range is smaller than a third threshold value, it is determined that the first influence is equal to the second influence. Including, 5. The data processing method according to claim 4.

6. and after determining the at least one influencing factor, determining a time-varying influence of a first attribute parameter among the plurality of attribute parameters on the target attribute parameter. The data processing method according to claim 3 .

7. and, before determining the change in the influence over time, acquiring a designation instruction of the first attribute parameter input by a user.

7. The data processing method according to claim 6.

8. After determining at least one influencing factor of the target attribute parameter, a first adjustment instruction is obtained for a second attribute parameter of the plurality of attribute parameters that is different from the target attribute parameter, the first adjustment instruction being to change a value of the second attribute parameter, and the second attribute parameter being included in the at least one influencing factor; determining a relationship between time and an influence of the second attribute parameter on the target attribute parameter based on the first adjustment instruction; further comprising The relationship between the influence of the second attribute parameter on the target attribute parameter and time is as follows: A change over time in the influence of the second attribute parameter on the target attribute parameter; or a cumulative change over time in the cumulative influence of the second attribute parameter on the target attribute parameter; Indicate at least one of The data processing method according to claim 3 .

9. and after determining the relationship between the influence of the second attribute parameter on the target attribute parameter and time, determining whether the cumulative influence of the second attribute parameter on the target attribute parameter is greater than a predetermined threshold value for a predetermined period of time based on a cumulative change over time in the cumulative influence of the second attribute parameter on the target attribute parameter.

9. The data processing method according to claim 8.

10. After determining a relationship between time and the influence of the second attribute parameter on the target attribute parameter, obtaining a second adjustment instruction for a third attribute parameter different from the target attribute parameter and the second attribute parameter among the plurality of attribute parameters, wherein the second adjustment instruction changes a value of the third attribute parameter, and the third attribute parameter is included in the at least one influencing element; determining which of the second attribute parameter adjusted by the first adjustment instruction and the third attribute parameter adjusted by the second adjustment instruction has a relatively greater influence on the target attribute parameter; further comprising:

9. The data processing method according to claim 8.

11. and further comprising: obtaining the first adjustment instruction input by a user after determining the at least one influencing factor of the target attribute parameter.

9. The data processing method according to claim 8.

12. Instead of determining the time series model, obtaining an initial time series model represented in the form of a time relationship diagram including nodes and edges; performing a t-test on each edge of the initial time series model to obtain a p-value, and removing redundant edges based on the p-value to obtain the time series model; further comprising:

3. The data processing method according to claim 1 or 2.

13. Instead of determining the time series model, obtaining a plurality of predetermined data sets similar to the time series data set; and selecting, as the time series model, a candidate model corresponding to a predetermined data set having the highest similarity from among a plurality of candidate models based on a similarity between the time series data set and each of the plurality of predetermined data sets.

3. The data processing method according to claim 1 or 2.

14. Obtaining the time series dataset includes: Obtaining an initial time series dataset; if the data items of the initial time series data set are not a uniform time series, data augmenting the initial time series data set to obtain a time series data set consisting of a uniform time series, thereby obtaining augmented data items to add to the initial time series data set; Including, 3. The data processing method according to claim 1 or 2.

15. The generation of the extended data item comprises: inputting a first data item having a first time and a second data item having a second time from the initial time series data set into a data generation model to obtain the extended data item having a third time that is later than the first and second times; the first time, the second time, and the third time indicate a time series in that order; 15. The data processing method according to claim 14.

16. A data processing method for executing, by a computer, processing on a time-series data set relating to sales records, the time-series data set including a plurality of time-series data items each having a time and a plurality of attribute parameters including at least one of a purchase price, a sales price, a sales amount, an inventory amount, or a customer view number corresponding to the time, the method comprising: obtaining the time series data set; acquiring, from the attribute parameters included in the plurality of time-series data items, the inventory amount, which is an attribute parameter determined by being influenced by some of the other attribute parameters included in the plurality of time-series data items; determining at least one influencing factor by inputting the acquired inventory amount and the time series data set into a time series model given in advance to determine at least one influencing factor indicating at least one attribute parameter that influences the inventory amount in a time series manner among the attribute parameters included in the plurality of time series data items and a time corresponding to the at least one attribute parameter; determining an inventory amount for a next time period based on the determined at least one influencing factor; Including, Data processing methods.

17. A data processing method for executing, by a computer, processing on a time series dataset related to power of an IoT device, the time series dataset including a plurality of time series data items each having a time and a plurality of attribute parameters including at least one of an electricity price, a power demand, a voltage, a current, a temperature, a humidity, an atmospheric pressure, or a power consumption amount of an IoT device corresponding to the time, the method comprising: obtaining the time series data set; obtaining, from the attribute parameters included in the plurality of time-series data items, the power consumption, which is an attribute parameter determined by being influenced by some of the other attribute parameters included in the plurality of time-series data items; determining at least one influencing factor by inputting the acquired power consumption and the time series data set into a time series model given in advance to determine at least one influencing factor indicating at least one attribute parameter that influences the power consumption in a time series manner among the attribute parameters included in the plurality of time series data items and a time corresponding to the at least one attribute parameter; determining a power consumption for a next time period based on the determined at least one influencing factor; Including, Data processing methods.

18. 1. A data processing method for performing, by a computer, processing on a time series data set collected from a social network, the time series data set including a plurality of time series data items each having a time and a plurality of attribute parameters including at least one of user identification information corresponding to the time, a user post, a number of views of the user post, a number of comments on the user post, or an anomaly indication, the method comprising: obtaining the time series data set; acquiring, from attribute parameters included in the plurality of time-series data items, the abnormality indication, which is an attribute parameter determined by being influenced by some of other attribute parameters included in the plurality of time-series data items; determining at least one influencing factor by inputting the acquired abnormality indication and the time series data set into a time series model given in advance to determine at least one influencing factor indicating at least one attribute parameter that influences the abnormality indication in a time series manner among the attribute parameters included in the plurality of time series data items and a time corresponding to the at least one attribute parameter; outputting presentation information indicating a user who is at risk of an abnormality based on the at least one influencing factor determined; Including, Data processing methods.

19. comprising processing circuitry configured to carry out the method of any one of claims 1 to 18. electronic equipment.

Images