Data processing method, device and equipment and readable medium

Abstract

The embodiment of the invention provides a data processing method and device, equipment and a medium. The method comprises the steps of obtaining historical supply data and a first relationship of a target object, wherein the first relationship indicates a relationship among sales volume data, a reference value attribute value and factors influencing the sales volume data and a relationship between sales volume data corresponding to different value attribute values and sales volume data corresponding to the reference value attribute value; determining a reference value attribute value based on the historical supply data; based on the historical supply data and the reference value attribute value, determining a second relationship between sales volume data corresponding to different value attribute values and sales volume data corresponding to the reference value attribute value; and adjusting the first relationship based on the reference value attribute value and the second relationship to obtain a target relationship of the target object. Therefore, the accuracy of sales volume prediction can be improved.

Description

Technical Field

[0001] Example embodiments of the present disclosure generally relate to the field of computer technology, and more particularly, to a method, apparatus, device, and computer-readable storage medium for data processing. Background Art

[0002] Product pricing has always been one of the most important business decisions in the e-commerce field. Product pricing is usually based on the price-demand relationship of the product (hereinafter referred to as the "quantity-price relationship"). For example, based on this quantity-price relationship, it can guide business and demand scenarios such as daily sales pricing, promotional pricing, clearing out near-expiring products, and increasing sales rates, thereby improving various business indicators of the company. Therefore, how to improve the accuracy of this quantity-price relationship is worthy of attention and urgent need. Summary of the invention

[0003] In a first aspect of the present disclosure, a method for data processing is provided, including: obtaining historical supply data and a first relationship of a target object, the first relationship indicating the relationship between sales data, a benchmark value attribute value and factors affecting the sales data, and the relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value; determining the benchmark value attribute value based on the historical supply data; determining the second relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value based on the historical supply data and the benchmark value attribute value; and adjusting the first relationship based on the benchmark value attribute value and the second relationship to obtain a target relationship of the target object.

[0004] In a second aspect of the present disclosure, a data processing device is provided, including: an acquisition module, configured to acquire historical supply data and a first relationship of a target object, the first relationship indicating the relationship between sales data, a benchmark value attribute value and factors affecting the sales data, and the relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value; a first determination module, configured to determine the benchmark value attribute value based on the historical supply data; a second determination module, configured to determine the second relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value based on the historical supply data and the benchmark value attribute value; and an adjustment module, configured to adjust the first relationship based on the benchmark value attribute value and the second relationship to obtain a target relationship of the target object.

[0005] In a third aspect of the present disclosure, an electronic device is provided. The electronic device includes at least one processing unit; and at least one memory, the at least one memory is coupled to the at least one processing unit and stores instructions for execution by the at least one processing unit, and when the instructions are executed by the at least one processing unit, the electronic device executes the method of the first aspect of the present disclosure.

[0006] In a fourth aspect of the present disclosure, a computer-readable storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and the computer program can be executed by a processor to perform the method according to the first aspect of the present disclosure.

[0007] In a fifth aspect of the present disclosure, a computer program product is provided, comprising computer executable instructions, wherein the computer executable instructions implement the method according to the first aspect of the present disclosure when executed by a processor.

[0008] It should be understood that the contents described in the summary of the invention are not intended to limit the key features or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In the following, in conjunction with the accompanying drawings and with reference to the following detailed description, the above and other features, advantages and aspects of various implementations of the present disclosure will become more apparent. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0010] Figure 1 A schematic diagram showing an example environment in which embodiments of the present disclosure can be implemented;

[0011] Figure 2 A schematic diagram showing the trend of price changes over time according to some embodiments of the present disclosure;

[0012] Figure 3 A schematic diagram showing a process of quantity-price relationship prediction according to some embodiments of the present disclosure;

[0013] Figure 4 A schematic diagram showing a flow sequence according to some embodiments of the present disclosure is shown;

[0014] Figure 5 A schematic diagram showing a process of real-time quantity and price prediction according to some embodiments of the present disclosure;

[0015] Figure 6 shows a schematic diagram of a gain curve according to some embodiments of the present disclosure;

[0016] Figure 7 A flowchart showing a process of data processing according to some embodiments of the present disclosure;

[0017] Figure 8 A flowchart showing an apparatus for data processing according to some embodiments of the present disclosure; and

[0018] Fig. 9 A block diagram of an electronic device is shown in which one or more embodiments of the present disclosure may be implemented. DETAILED DESCRIPTION

[0019] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0020] In the description of the embodiments of the present disclosure, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The term "some embodiments" should be understood as "at least some embodiments". Other explicit and implicit definitions may also be included below.

[0021] It should be noted that the acquisition, storage and application of user personal information involved in the technical solution of the present disclosure are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0022] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0023] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested to be performed will require obtaining and using the user's personal information, so that the user can independently choose whether to provide personal information to software or hardware such as electronic devices, applications, servers or storage media that execute operations of the technical solution of the present disclosure based on the prompt message.

[0024] As an optional but non-limiting implementation, in response to receiving an active request from the user, the prompt information is sent to the user in a manner such as a pop-up window, in which the prompt information can be presented in text form. In addition, the pop-up window can also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0025] Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.

[0026] Figure 1 A schematic diagram of an example environment 100 is shown in which embodiments of the present disclosure can be implemented.

[0027] like Figure 1 As shown, the environment 100 may include a server 100 and terminal devices 120-1, 120-2, ..., 120-N. For ease of discussion, the terminal devices 120-1, 120-2, ..., 120-N may be collectively or individually referred to as terminal devices 120.

[0028] The user can use the terminal device 120 to interact with the server 110 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 120, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0029] The server 110 may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content distribution networks, and big data and artificial intelligence platforms. The server 110 may include, for example, a computing system / server, such as a mainframe, an edge computing node, a computing device in a cloud environment, and the like.

[0030] The terminal device 120 can be any type of mobile terminal, fixed terminal or portable terminal, including mobile phones, desktop computers, laptop computers, notebook computers, netbook computers, tablet computers, media computers, multimedia tablets, handheld computers, portable game terminals, VR / AR devices, personal communication systems (PCS) devices, personal navigation devices, personal digital assistants (PDA), audio / video players, digital cameras / camcorders, positioning devices, television receivers, radio broadcast receivers, e-book devices, game devices, or any combination of the foregoing, including accessories and peripherals of these devices or any combination thereof. In some embodiments, the client device can also support any type of interface for the user (such as "wearable" circuits, etc.).

[0031] The communication connection between the server 110 and the terminal device 120 can be established in a wired manner or a wireless manner. The communication connection may include but is not limited to a Bluetooth connection, a mobile network connection, a Universal Serial Bus (USB) connection, a Wireless Fidelity (WiFi) connection, etc., and the embodiments of the present disclosure are not limited in this respect.

[0032] It should be understood that Figure 1The components and arrangements in the environment 100 shown are merely examples, and a computing system suitable for implementing the exemplary implementation described in the present disclosure may include one or more different components, other components, and / or different arrangements. The implementation of the present disclosure is not limited in this respect.

[0033] As briefly mentioned above, product pricing has always been one of the most important business decisions in the e-commerce field. Product pricing is usually based on the relationship between quantity and price. For example, based on this relationship between quantity and price, it can guide business and demand scenarios such as daily sales pricing, promotional pricing, clearing out near-expiring products, and increasing sales rates, thereby improving various business indicators of the company. Therefore, how to improve the accuracy of this relationship between quantity and price is worthy of attention and urgent need.

[0034] In some scenarios, sales can be predicted based on a separate response model, that is, price is used as a feature to learn the correlation between price and sales, that is, to learn the relationship between quantity and price, so as to predict sales.

[0035] In such a scenario, due to the interference of confounding variables, the obtained quantity-price relationship may be biased and cannot truly reflect the causal relationship between commodity demand and price, and the obtained prediction results will fluctuate greatly.

[0036] In other scenarios, sales can be predicted based on response models and counterfactual inferences. That is, through machine learning, sales at "regular prices" can be predicted, and on this basis, the sales increase brought about by promotions (i.e. counterfactual inference) can be learned to obtain the relationship between quantity and price.

[0037] In such a scenario, since the price of goods is variable, it is not possible to clearly define a fixed "regular price". For example, a certain product can have multiple promotions, and different promotion plans correspond to different transaction prices. For example, see Figure 2 , Figure 2 A schematic diagram of the trend 200 of price changes over time is shown in Figure 200. In trend 200, the "dotted line" represents the transaction price of the product, and the "solid line" represents the price of the product. It can be seen from the figure that the relative changes in the transaction price of the product relative to the price of the product are different on different dates. Moreover, the study of the relationship between quantity and price usually requires learning the demand elasticity of the transaction price (that is, the degree to which the transaction demand of the product responds to changes in its transaction price within a certain period of time) so that the operator can set the price of the product. Therefore, using the "regular price" for prediction of the same product will result in poor learning of the model.

[0038] In addition, the evaluation of the relationship between quantity and price usually includes offline methods and online indicators. For example, online indicators can be compared through A / B testing, such as the gross profit and other indicators of products with system pricing and manual pricing; the offline method mainly substitutes the actual price into the model to determine the error between the predicted sales volume and the actual sales volume. The judgment indicators may include weighted mean absolute percentage error (Weighted Mean Absolute Percentage Error, WMAPE), mean absolute error (Mean Absolute Error, MAE), etc.

[0039] It can be seen that in these methods, the price to be predicted is input into the model for prediction and evaluation, which is essentially a point prediction and evaluation, and cannot reflect the quality of price demand elasticity. At the same time, the evaluation of weighted average absolute percentage error, average absolute error, etc., can easily lead to the model focusing too much on accuracy and ignoring the learning of price demand elasticity, resulting in errors in the prediction results.

[0040] The embodiments of the present disclosure provide a data processing scheme. In various embodiments of the present disclosure, the historical supply data and the first relationship of the target object can be obtained. The first relationship can indicate the relationship between the sales data, the benchmark value attribute value and the factors affecting the sales data, and the relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value. Thus, the relationship between the sales data, the benchmark value attribute value and the factors affecting the sales data can be obtained, that is, the relationship between the value attribute value, the sales data and the factors affecting the sales data under the condition of considering the demand elasticity. Further, the benchmark value attribute value can be determined based on the historical supply data. Thus, the predicted reference point of the target relationship can be determined. Further, the second relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value can be determined based on the historical supply data and the benchmark value attribute value. Thus, the relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value can be obtained. Further, in some embodiments, the first relationship can be adjusted based on the benchmark value attribute value and the second relationship to obtain the target relationship of the target object. For example, the benchmark value attribute value in the first relationship is replaced with the benchmark value attribute value determined based on historical supply data, and the relationship between the sales data corresponding to different value attribute values ​​in the first relationship and the sales data corresponding to the benchmark value attribute value is replaced with the second relationship to obtain the target relationship of the target object. In this way, the relationship of value attribute value-sales data-factors affecting sales data (i.e., quantity-price relationship) considering the advantages and disadvantages of price demand elasticity can be obtained, thereby improving the accuracy of sales forecasting.

[0041] Some example embodiments of the present disclosure will be described below with continued reference to the accompanying drawings.

[0042] Figure 3 FIG. 3 is a schematic diagram showing a process 300 of predicting the quantity-price relationship according to some embodiments of the present disclosure. Figure 1 For example, the process of quantity-price relationship prediction may be implemented by the server 110 .

[0043] In block 310, the server 110 obtains the historical supply data of the target object from the offline data warehouse (ODW). Taking the target object as a commodity as an example, the offline data warehouse stores the historical supply data of various commodities, and the server 110 can directly obtain the required historical supply data from the offline data warehouse. For example, the historical supply data of the commodity may include, but is not limited to: attribute information, competitive product information, commodity flow, date information, commodity price, holidays, promotion information, value attribute value, sales data, etc.

[0044] In some embodiments, the server 110 may process the acquired historical supply data (for example, the historical supply data may be aggregated according to required dimensions). For example, the server 110 may aggregate the acquired historical supply data according to date (Date Time, dt), unique identifier of the commodity inventory unit (Stock Keeping Unit, sku), etc.

[0045] At block 320 , the server 110 filters the historical supply data based on quantity-price correlation.

[0046] In some embodiments, the server 110 may filter the historical supply data based on the historical value attribute value and the historical sales data in the historical supply data of the target object. For example, the historical value attribute value of the target object may be the historical price of the commodity.

[0047] For example, the server 110 may determine the correlation coefficient and significance level of the target object based on the historical value attribute value and the historical sales data, and filter the historical supply data based on the correlation coefficient and significance level.

[0048] The correlation coefficient can indicate the degree of linear correlation between the historical value attribute value and the historical sales data. The significance level can indicate the probability that there is no significant relationship between the historical value attribute value and the historical sales data.

[0049] The server 110 may calculate the correlation coefficient between the historical value attribute value and the historical sales data of the target object based on the following formula:

[0050] Where n represents the number of time points (also called sample size), X i Represents the value attribute value of the object at time i, Y i Represents the sales data of the object at time i, Represents the average value of the value attribute of the object in n samples, Represents the average sales data of the objects in n samples, and i represents the time point.

[0051] The server 110 may convert the above formula into a t-statistic. The t-statistic follows a t-distribution with n-2 degrees of freedom, where n is the number of samples.

[0052] The calculation formula of t statistic is as follows:

[0053] Among them, s represents the correlation coefficient and n represents the number of samples.

[0054] Based on the calculated t-statistic, the server 110 may look up the t-distribution to calculate the significance level p, where the value of p refers to the probability of obtaining the current t-value or a more extreme t-value under the assumption that p-0 is true.

[0055] Server 110 can filter the historical supply data based on the correlation between the historical value attribute values ​​and the historical sales data. Since the sales data and the value attribute values ​​are negatively correlated, the historical supply data corresponding to all positively correlated historical value attribute values ​​and historical sales data can be removed.

[0056] Furthermore, after removing the historical supply data corresponding to the historical value attribute values ​​and historical sales data that are positively correlated, the server 110 may further filter the remaining historical supply data based on a predetermined filtering condition. For example, the remaining historical supply data may be further filtered based on the following conditions: s i ≥0.2&p i ≤0.05,i∈1,2,3,…,n

[0057] Among them, n represents the number of samples, i represents the time point, and s i represents the correlation coefficient between the value attribute value and sales data of the object at time i, p i Represents the significance level of the value attribute value and sales data of the object at time i.

[0058] In block 330 , the server 110 performs feature engineering on the filtered historical supply data. For example, feature engineering may include, but is not limited to, missing value processing, extreme value processing, high-order feature generation, and the like.

[0059] In some embodiments, considering the impact of traffic on sales, traffic features may be added to feature data obtained by feature engineering historical supply data on the server 110 .

[0060] In some embodiments, the server 110 may obtain historical traffic information of the target object, determine current traffic information of the target object based on the historical traffic information, and add the current traffic information of the target object to the feature data.

[0061] For example, the server 110 may predict the current traffic information of the target object through a trained traffic prediction model. For example, the traffic prediction model may include, but is not limited to, time series and statistical models such as Autoregressive Integrated Moving Average Model (Arima), Holt-Winters Exponential Smoothing (Holtwinter), Long Short-Term Memory (LSTM), Transformer, and eXtreme Gradient Boosting (XGB).

[0062] Figure 4 FIG. 4 is a schematic diagram showing a flow sequence 400 according to some embodiments of the present disclosure. Taking the method of time series prediction as an example, see Figure 4 The server 110 may sort the historical traffic information of the target object in chronological order to form a traffic sequence, and organize the traffic values ​​in the traffic sequence into sample 1, sample 2, ..., sample N according to chronological order.

[0063] Take the traffic value of a product as an example. Figure 4 Each point in the data may be the traffic value of a certain product on that day (for example, page views, page clicks, number of visitors, etc.). Multiple points constitute a sequence data (i.e., a sample) as the input of the time series prediction model. The last value of the sequence is the target value. The server 110 can train the traffic prediction model through such multiple sequences and multiple target values, and after training, predict the traffic characteristics of day T+1 based on the trained traffic prediction model.

[0064] After obtaining the current flow information, the server 110 may determine the reference price attribute value based on the current flow information and the historical supply data excluding the historical commodity flow.

[0065] Continue to see Figure 3In block 350, the server 110 may determine the first prediction results of the one or more parameter models based on the current traffic information and the historical supply data excluding the historical commodity traffic through the one or more trained parameter models. In some embodiments, the first prediction results may include the corresponding one or more first sales data and one or more first value attribute values.

[0066] For example, the parameter model may include but is not limited to: linear regression model, perceptron, etc. It should be understood that one or more parameter models are based on the relationship between sales data, benchmark value attribute values ​​and factors affecting sales data, that is, E[f(X i ,d i0 )] is trained as the goal.

[0067] Taking multiple parameter models as an example, the server 110 can obtain the following prediction results through multiple parameter models: [E ψ1 ,E ψ2 ,E ψ3 ,…,E ψn ],ψ n ∈ψ

[0068] Among them, ψ represents the set of parameter models, E ψn Represents the prediction result of the nth parameter model.

[0069] Each prediction result contains the prediction of the historical transaction value attribute value of the target object, namely:

[0070] Wherein, j represents the historical transaction value attribute value. The server 110 can calculate [E ψ1 ,E ψ2 ,E ψ3 ,…,E ψn ] Establish a linear model and learn the weight β of each model’s prediction results, so as to obtain the final prediction results of multiple non-parametric models, that is, the first prediction result: E Ψ =E[β1E ψ1 +β1E ψ2 +β1E ψ3 +,…,+β n E ψn ].

[0071] In block 360, the server 110 may also determine the second prediction results of the one or more non-parametric models based on the current traffic information and the historical supply data excluding the historical commodity traffic through the one or more trained non-parametric models. In some embodiments, the second prediction results may include the corresponding one or more second sales data and one or more second value attribute values.

[0072] The non-parametric model may include but is not limited to: tree model, depth model, etc. It should be understood that one or more non-parametric models are based on the relationship between sales data, benchmark value attribute values ​​and factors affecting sales data, that is, E[f(X i ,d i0 )] is trained as the goal.

[0073] Taking multiple non-parametric models as an example, the server 110 can obtain the following prediction results through multiple non-parametric models:

[0074] Among them, Φ represents the set of non-parametric models, Represents the prediction results of the nth nonparametric model.

[0075] Each prediction result contains the prediction of the historical transaction value attribute value of the target object, namely:

[0076] Wherein, j represents the historical transaction value attribute value. The server 110 can calculate the historical transaction value attribute value based on the historical prediction accuracy of each model. A linear model is established to learn the weight α of the prediction results of each model, so as to obtain the final prediction results of multiple non-parametric models, that is, the second prediction results:

[0077] In some embodiments, the server 110 may determine a baseline value attribute value based on the first prediction result and the second prediction result. Specifically, the server 110 may determine the first sales data and the second sales data with the lowest difference based on one or more first sales and one or more second sales, and determine the baseline value attribute value based on the first value attribute value and the second value attribute value corresponding to the first sales data and the second sales data with the lowest difference, respectively.

[0078] For example, the server 110 may determine the baseline value attribute value based on the following formula:

[0079] Among them, d i0 Indicates the base value attribute value.

[0080] In block 370 , the server 110 may determine the relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value based on the historical supply data and the benchmark value attribute value by means of counterfactual inference.

[0081] In some embodiments, the server 110 may use a predetermined model (e.g., a causal inference model) to determine the relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value, i.e., the elasticity between adjacent value attribute values. For example, the causal inference model may include, but is not limited to: a model based on deep instrumental variables (DeepInstrumental Variables, DeepIV), a deep causal inference model (Dragonnet), a causal forest model, etc.

[0082] Specifically, the server 110 can use the factors affecting sales data and value attribute values ​​in historical supply data as the input of the model, use the sales data as the output to train the causal inference model, and use the trained causal inference model to predict the sales data under a certain value attribute value and a certain factor affecting sales data, thereby obtaining the relationship between the three.

[0083] In some embodiments, the server 110 may also use parameter fitting to determine the relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value.

[0084] Specifically, the server 110 may fit the historical sales data and the historical value attribute values ​​in the historical supply data to obtain a relationship curve, and select the relationship curve with the smallest error between the relationship curve and the actual sales data as the quantity-price form of the target object.

[0085] The quantity-price form can be represented by a quantity-price relationship formula. The following Table 1 shows several formulas for representing the quantity-price form: Table 1 function form parameter Linear <![CDATA[D(p)=D(1-bp) + ]]> D>0,b>0 index <![CDATA[D(p)=De -bp ]]> D>0,b>0 Constant elasticity <![CDATA[D(p)=Dp -b ]]> A>0,b>0

[0086] In Table 1, p represents the value attribute value, and D(p) represents the sales data.

[0087] After obtaining the quantity-price format, the server 110 can obtain the relationship between the historical value attribute value, the historical sales data, and the factors affecting the sales data based on the quantity-price format and the historical supply data. Then, based on multiple predetermined value attribute values ​​(e.g., multiple value attribute values ​​that are higher or lower than the benchmark value attribute value) and the relationship between the historical value attribute value, the historical sales data, and the factors affecting the sales data, determine the sales data corresponding to each of the multiple predetermined value attribute values. Further, the server 110 determines the sales data corresponding to the benchmark value attribute value based on the benchmark value attribute value and the relationship between the historical value attribute value, the historical sales data, and the factors affecting the sales data. Further, the server 110 determines the relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value based on the sales data corresponding to each of the multiple predetermined value attribute values ​​and the sales data corresponding to the benchmark value attribute value. Here, "the relationship between the historical value attribute value, the historical sales data, and the factors affecting the sales data" is sometimes also referred to as the "third relationship" in this article.

[0088] In some embodiments, the server 110 may obtain or define a relationship (i.e., the quantity-price relationship mentioned above, which may be referred to as a "candidate relationship" for the convenience of discussion). The candidate relationship may indicate the relationship between sales data, the benchmark value attribute value, and the factors affecting sales data, and the relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value. Here, "the relationship between sales data, the benchmark value attribute value, and the factors affecting sales data, and the relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value" is sometimes also referred to as the "first relationship" in this article.

[0089] In some embodiments, the candidate relations may be represented as follows: E[T i |d i ,X i ]=E[g(T i1 |d i1 ,X i )-g(T i0 |d i0 ,X i )]+E[f(X i ,d i0 )]

[0090] Among them, T i Indicates sales volume, final sales volume after promotion and the ratio of basic sales volume, etc. i Indicates the value attribute value or discount of the object (for example, discount strength, 5%, 10%, etc.), d i0 Indicates the base value attribute value, d i1Indicates the value attribute value that is higher or lower than the benchmark value attribute value, X i Indicates factors that affect sales data (for example, promotion, traffic, product attributes, etc.).

[0091] In some embodiments, the server 110 can adjust the candidate relationship based on the reference value attribute value and the relationship between the sales data corresponding to the different value attribute values ​​and the sales data corresponding to the reference value attribute value, so as to obtain the target relationship, that is, the quantity-price relationship. Here, the "relationship between the sales data corresponding to the different value attribute values ​​and the sales data corresponding to the reference value attribute value" is sometimes also referred to as the "second relationship" in this article.

[0092] For example, server 110 can replace the "benchmark value attribute value" in the candidate relationship with the benchmark value attribute value determined by the parametric model and the non-parametric model, and can replace the "relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value" in the candidate relationship with the relationship between sales data corresponding to the same value attribute value and sales data corresponding to the benchmark value attribute value obtained by a predetermined model or quantity-price relationship formula, thereby obtaining the final quantity-price relationship.

[0093] In block 381 , the server 110 may perform quantity-price prediction based on the final quantity-price relationship.

[0094] In some scenarios, the quantity-price forecast based on the final quantity-price relationship may not meet the forecasting needs, such as the temporary occurrence / cancellation of promotional activities, and the adjustment of inventory by merchants.

[0095] In some embodiments, in order to adapt to more scenarios, the server 110 can obtain real-time supply data of the target object, determine the candidate value attribute value-sales data relationship of the target object based on the historical supply data and the final quantity-price relationship, and determine the target value attribute value-sales data relationship of the target object based on the real-time supply data and the candidate value attribute value-sales data relationship.

[0096] Figure 5 A schematic diagram of a process 500 of real-time quantity and price prediction according to some embodiments of the present disclosure is shown. For example, the process of real-time quantity and price prediction may be implemented by the server 110 .

[0097] In block 510, the server 110 obtains real-time features and offline features, and trains a predetermined model based on the real-time features and offline features to learn the target value attribute value-sales data relationship. In some embodiments, the real-time features may be real-time supply data. In some embodiments, the offline features may be candidate value attribute value-sales data relationships obtained based on historical supply data and the final quantity-price relationship.

[0098] In block 520, the server 110 uses the learned target value attribute value-sales data relationship as an online reasoning framework and deploys it in a real-time service (different real-time services can be set according to different scenario requirements, for example, real-time service 1, real-time service 2, ..., real-time service N). In some embodiments, the real-time service can be deployed in the terminal device 120, or the terminal device 120 can call the real-time service, which is not limited in the embodiments of the present disclosure. In this way, the accuracy of the prediction can be improved.

[0099] Continue to see Figure 3 , in box 382, ​​the server 110 can guide price decisions based on the results of the quantity price prediction.

[0100] In block 383 , the server 110 may evaluate the results of the quantity and price prediction.

[0101] In some embodiments, the server 110 may evaluate the target relationship based on the accuracy of predicted sales volume and the accuracy of elasticity.

[0102] 1) Accuracy of sales forecast

[0103] The accuracy of predicted sales volume = 1-WMAPE, where WMAPE can be calculated based on the following formula:

[0104] Among them, Q 预测 represents the predicted sales volume of the product, Q 真实 represents the actual sales volume of the product, and D represents the dimension that the product needs to be aggregated, such as category, warehouse, city, and all products.

[0105] Since the target relationship is based on T data and predicts the sales volume of T+1, Q 真实 The actual sales volume of T+1 needs to be used. 预测 Take the quantity and price prediction results and the actual transaction price P 真实 The closest estimated price P 预测 The corresponding predicted sales volume.

[0106] For the case where the transaction volume N>1, Q 真实 and P 真实 can be calculated based on the following formula:

[0107] From the prediction results of the target relationship, find the predicted sales volume corresponding to the price closest to the actual price as Q 预测 :

[0108] Among them, m is the number of predicted points for the product given by the target relationship prediction.

[0109] 2) Elasticity accuracy

[0110] The indicators for evaluating elasticity are usually based on the uplift curve and the area under the uplift curve (Area Under Uplift Curve, AUUC). The form of the uplift curve is as follows: Figure 6 As shown, Figure 6 A gain curve 500 is shown in FIG.

[0111] Based on the observed data, for a single product, the price (treatment) is a continuous value, and a gain curve is given.

[0112] The final AUUC is the area under the cumsum gain curve. In order to evaluate whether the predicted volume-price range has good price elasticity, the elasticity error calculation method is given:

[0113] Among them, AUUC gt Based on historical data, the average treatment effect (ATE) is obtained (such as the expected difference in the effect between the treated and untreated groups in an AB experiment), and the historical real sales are sorted. The obtained Uplift curve and the calculated AUUC are used as the ground truth of the smallest available unit (Stock Keeping Unit, SKU) in inventory management; AUUC pred Based on the prediction results use Sort the historical real sales volume, obtain the Uplift curve and calculate the auuc.

[0114] For AUUC gt The higher the SKU, the lower the auuc mape, which means the target relationship has better learned price elasticity.

[0115] According to the embodiments of the present disclosure, the historical supply data and the first relationship of the target object can be obtained. The first relationship can indicate the relationship between the sales data, the benchmark value attribute value and the factors affecting the sales data, and the relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value. Thus, the relationship between the sales data, the benchmark value attribute value and the factors affecting the sales data can be obtained, that is, the relationship between the value attribute value, the sales data and the factors affecting the sales data under the condition of considering the demand elasticity. Further, the benchmark value attribute value can be determined based on the historical supply data. Thus, the predicted reference point of the target relationship can be determined. Further, the second relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value can be determined based on the historical supply data and the benchmark value attribute value. Thus, the relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value can be obtained. Further, the first relationship can be adjusted based on the benchmark value attribute value and the second relationship to obtain the target relationship of the target object. For example, the benchmark value attribute value in the first relationship is replaced with the benchmark value attribute value determined based on historical supply data, and the relationship between the sales data corresponding to different value attribute values ​​in the first relationship and the sales data corresponding to the benchmark value attribute value is replaced with the second relationship to obtain the target relationship of the target object. In this way, the relationship of value attribute value-sales data-factors affecting sales data (i.e., quantity-price relationship) considering the advantages and disadvantages of price demand elasticity can be obtained, thereby improving the accuracy of sales forecasting.

[0116] Figure 7 A flowchart of a process 700 of data processing according to some embodiments of the present disclosure is shown. The process 700 may be implemented at the server 110.

[0117] In box 710, the server 110 obtains historical supply data and a first relationship of the target object, wherein the first relationship indicates the relationship between sales data, a benchmark value attribute value and factors affecting the sales data, and the relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value.

[0118] At block 720 , the server 110 determines a baseline value attribute value based on the historical supply data.

[0119] In block 730 , the server 110 determines a second relationship between the sales data corresponding to the different value attribute values ​​and the sales data corresponding to the reference value attribute value based on the historical supply data and the reference value attribute value.

[0120] In block 740 , the server 110 adjusts the first relationship based on the baseline value attribute value and the second relationship to obtain a target relationship of the target object.

[0121] In some embodiments, determining a baseline value attribute value based on historical supply data includes: determining, through one or more trained parameter models, a first prediction result of one or more parameter models based on the historical supply data, the first prediction result including corresponding one or more first sales data and one or more first value attribute values; determining, through one or more trained non-parametric models, a second prediction result of one or more non-parametric models based on the historical supply data, the second prediction result including corresponding one or more second sales data and one or more second value attribute values; and determining a baseline value attribute value based on the first prediction result and the second prediction result.

[0122] In some embodiments, determining the baseline value attribute value based on the first prediction result and the second prediction result includes: determining the first sales data and the second sales data with the lowest difference based on one or more first sales volumes and one or more second sales volumes; and determining the baseline value attribute value based on the first value attribute value and the second value attribute value respectively corresponding to the first sales data and the second sales data with the lowest difference.

[0123] In some embodiments, determining the second relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value based on historical supply data includes: determining a third relationship based on historical supply data through a predetermined model, the third relationship indicating the relationship between historical value attribute values, historical sales data and factors affecting the sales data; determining sales data corresponding to each of the multiple predetermined value attribute values ​​based on the multiple predetermined value attribute values ​​and the third relationship; determining sales data corresponding to the benchmark value attribute value based on the benchmark value attribute value and the third relationship; and determining the second relationship based on the sales data corresponding to each of the multiple predetermined value attribute values ​​and the sales data corresponding to the benchmark value attribute value.

[0124] In some embodiments, the historical supply data includes at least historical value attribute values, historical sales data, and factors affecting the sales data, and the factors affecting the sales data include at least current traffic information of the target object.

[0125] In some embodiments, process 700 further includes: acquiring historical traffic information of the target object; and determining current traffic information of the target object based on the historical traffic information.

[0126] In some embodiments, process 700 also includes: obtaining real-time supply data of the target object; determining the candidate value attribute value-sales data relationship of the target object based on historical supply data and the target relationship; and determining the target value attribute value-sales data relationship of the target object based on the real-time supply data and the candidate value attribute value-sales data relationship.

[0127] In some embodiments, the historical supply data includes at least historical value attribute values ​​and historical sales data, and before determining the baseline value attribute value, process 700 also includes: determining the correlation coefficient and significance level of the target object based on the historical value attribute values ​​and the historical sales data, the correlation coefficient indicating the degree of linear correlation between the historical value attribute values ​​and the historical sales data, and the significance level indicating the probability that there is no significant relationship between the historical value attribute values ​​and the historical sales data; filtering the historical supply data based on the correlation coefficient and the significance level, wherein the determination of the baseline value attribute is based on the filtered historical supply data.

[0128] Figure 8 FIG. 8 is a block diagram of a data processing apparatus 800 according to some embodiments of the present disclosure. The apparatus 800 may be implemented as or include the server 110 .

[0129] The device 800 includes an acquisition module 810, which is configured to acquire historical supply data and a first relationship of a target object, wherein the first relationship indicates the relationship between sales data, a benchmark value attribute value and factors affecting the sales data, and the relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value; a first determination module 820, which is configured to determine the benchmark value attribute value based on the historical supply data; a second determination module 830, which is configured to determine the second relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the benchmark value attribute value based on the historical supply data and the benchmark value attribute value; and an adjustment module 840, which is configured to adjust the first relationship based on the benchmark value attribute value and the second relationship to obtain a target relationship of the target object.

[0130] In some embodiments, the first determination module 820 is also configured to: determine, through one or more trained parameter models, a first prediction result of one or more parameter models based on historical supply data, the first prediction result includes one or more corresponding first sales data and one or more first value attribute values; determine, through one or more trained non-parametric models, a second prediction result of one or more non-parametric models based on historical supply data, the second prediction result includes one or more corresponding second sales data and one or more second value attribute values; and determine a baseline value attribute value based on the first prediction result and the second prediction result.

[0131] In some embodiments, the first determination module 820 is further configured to: determine the first sales volume data and the second sales volume data with the lowest difference based on one or more first sales volume and one or more second sales volume; and determine the benchmark value attribute value based on the first value attribute value and the second value attribute value respectively corresponding to the first sales volume data and the second sales volume data with the lowest difference.

[0132] In some embodiments, the second determination module 830 is further configured to: determine a third relationship based on historical supply data through a predetermined model, the third relationship indicating the relationship between historical value attribute values, historical sales data and factors affecting the sales data; determine sales data corresponding to each of the multiple predetermined value attribute values ​​based on multiple predetermined value attribute values ​​and the third relationship; determine sales data corresponding to the baseline value attribute value based on the baseline value attribute value and the third relationship; and determine the second relationship based on the sales data corresponding to each of the multiple predetermined value attribute values ​​and the sales data corresponding to the baseline value attribute value.

[0133] In some embodiments, the historical supply data includes at least historical value attribute values, historical sales data, and factors affecting the sales data, and the factors affecting the sales data include at least current traffic information of the target object.

[0134] In some embodiments, the device 800 further includes a third determination module configured to obtain historical traffic information of the target object; and determine current traffic information of the target object based on the historical traffic information.

[0135] In some embodiments, the device 800 also includes a fourth determination module, which is configured to obtain real-time supply data of the target object; determine the candidate value attribute value-sales data relationship of the target object based on historical supply data and the target relationship; and determine the target value attribute value-sales data relationship of the target object based on the real-time supply data and the candidate value attribute value-sales data relationship.

[0136] In some embodiments, the historical supply data includes at least historical value attribute values ​​and historical sales data, and the device 800 also includes an evaluation module, which is configured to determine the correlation coefficient and significance level of the target object based on the historical value attribute values ​​and the historical sales data, the correlation coefficient indicates the degree of linear correlation between the historical value attribute values ​​and the historical sales data, and the significance level indicates the probability that there is no significant relationship between the historical value attribute values ​​and the historical sales data; based on the correlation coefficient and the significance level, the historical supply data is filtered, wherein the determination of the benchmark value attribute is based on the filtered historical supply data.

[0137] The modules included in the device 800 can be implemented in various ways, including software, hardware, firmware, or any combination thereof. In some embodiments, one or more units can be implemented using software and / or firmware, such as machine executable instructions stored on a storage medium. In addition to or as an alternative to machine executable instructions, some or all of the units in the device 800 can be implemented at least in part by one or more hardware logic components. As an example and not limitation, exemplary types of hardware logic components that can be used include field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0138] Fig. 9 900 is a block diagram of an electronic device in which one or more embodiments of the present disclosure may be implemented. It should be understood that Fig. 9 The electronic device 900 shown is merely exemplary and should not constitute any limitation on the functionality and scope of the embodiments described herein.

[0139] like Fig. 9 As shown, the electronic device 900 is in the form of a general electronic device. The components of the electronic device 900 may include, but are not limited to, one or more processors or processing units 910, a memory 920, a storage device 930, one or more communication units 940, one or more input devices 950, and one or more output devices 960. The processing unit 910 may be an actual or virtual processor and is capable of performing various processes according to a program stored in the memory 920. In a multi-processor system, multiple processing units execute computer executable instructions in parallel to improve the parallel processing capability of the electronic device 900.

[0140] The electronic device 900 typically includes a plurality of computer storage media. Such media can be any available media accessible to the electronic device 900, including but not limited to volatile and non-volatile media, removable and non-removable media. The memory 920 can be a volatile memory (e.g., register, cache, random access memory (RAM)), a non-volatile memory (e.g., a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), flash memory) or some combination thereof. The storage device 930 can be a removable or non-removable medium, and can include a machine-readable medium, such as a flash drive, a disk, or any other medium, which can be used to store information and / or data and can be accessed within the electronic device 900.

[0141] The electronic device 900 may further include additional removable / non-removable, volatile / non-volatile storage media. Fig. 9As shown in , a disk drive for reading or writing from a removable, non-volatile disk (e.g., a "floppy disk") and an optical drive for reading or writing from a removable, non-volatile optical disk may be provided. In these cases, each drive may be connected to the bus (not shown) by one or more data media interfaces. Memory 920 may include a computer program product 925 having one or more program modules that are configured to perform various methods or actions of various embodiments of the present disclosure.

[0142] The communication unit 940 implements communication with other electronic devices through a communication medium. Additionally, the functions of the components of the electronic device 900 can be implemented with a single computing cluster or multiple computing machines that can communicate through a communication connection. Therefore, the electronic device 900 can operate in a networked environment using a logical connection with one or more other servers, a network personal computer (PC), or another network node.

[0143] The input device 950 may be one or more input devices, such as a mouse, a keyboard, a tracking ball, etc. The output device 960 may be one or more output devices, such as a display, a speaker, a printer, etc. The electronic device 900 may also communicate with one or more external devices (not shown) through the communication unit 940 as needed, such as a storage device, a display device, etc., communicate with one or more devices that allow a user to interact with the electronic device 900, or communicate with any device that allows the electronic device 900 to communicate with one or more other electronic devices (e.g., a network card, a modem, etc.). Such communication may be performed via an input / output (I / O) interface (not shown).

[0144] According to an exemplary implementation of the present disclosure, a computer-readable storage medium is provided, on which one or more computer instructions are stored, wherein the one or more computer instructions are executed by a processor to implement the method described above. According to an exemplary implementation of the present disclosure, a computer program product is also provided, which is tangibly stored on a non-transitory computer-readable medium and includes computer-executable instructions, and the computer-executable instructions are executed by a processor to implement the method described above.

[0145] Various aspects of the present disclosure are described herein with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products implemented according to the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0146] These computer-readable program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processing unit of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0147] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0148] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple implementations of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of a module, program segment or instruction includes one or more executable instructions for realizing the logical function of the specification. In some implementations as replacements, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be realized by a special hardware-based system that performs the function or action of the specification, or can be realized by a combination of special hardware and computer instructions.

[0149] The above descriptions of various implementations of the present disclosure are exemplary, non-exhaustive, and not limited to the disclosed implementations. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described implementations. The selection of terms used herein is intended to best explain the principles of the implementations, practical applications, or improvements to the technology in the marketplace, or to enable other persons of ordinary skill in the art to understand the implementations disclosed herein.

Claims

1. A method for data processing, comprising: Acquire historical supply data of the target object and a first relationship, wherein the first relationship indicates a relationship between sales data, a reference value attribute value, and a factor affecting the sales data, and a relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the reference value attribute value; Determining the benchmark value attribute value based on the historical supply data; Determining, based on the historical supply data and the reference value attribute value, a second relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the reference value attribute value; as well as Based on the reference value attribute value and the second relationship, the first relationship is adjusted to obtain the target relationship of the target object.

2. The method of claim 1, wherein determining a baseline value attribute value based on the historical supply data comprises: Determine, by one or more trained parameter models, a first prediction result of the one or more parameter models based on the historical supply data, the first prediction result including the corresponding one or more first sales data and one or more first value attribute values; Determine, by one or more trained non-parametric models, a second prediction result of the one or more non-parametric models based on the historical supply data, the second prediction result including one or more corresponding second sales data and one or more second value attribute values; as well as The reference value attribute value is determined based on the first prediction result and the second prediction result.

3. The method according to claim 2, wherein determining the baseline value attribute value based on the first prediction result and the second prediction result comprises: Based on the one or more first sales volumes and the one or more second sales volumes, determining first sales volume data and second sales volume data having the lowest difference; as well as The reference value attribute value is determined based on the first value attribute value and the second value attribute value respectively corresponding to the first sales volume data and the second sales volume data with the lowest difference.

4. The method according to claim 1, wherein based on the historical supply data and the benchmark value attribute value, determining the second relationship between the sales data corresponding to different value attribute values ​​and the sales data corresponding to the benchmark value attribute value comprises: Determining, by a predetermined model, a third relationship based on the historical supply data, the third relationship indicating a relationship between the historical value attribute value, the historical sales data, and factors affecting the sales data; Based on the plurality of predetermined value attribute values ​​and the third relationship, determining the sales data corresponding to each of the plurality of predetermined value attribute values; Based on the reference value attribute value and the third relationship, determining the sales volume data corresponding to the reference value attribute value; as well as The second relationship is determined based on the sales data corresponding to each of the plurality of predetermined value attribute values ​​and the sales data corresponding to the reference value attribute value.

5. The method according to claim 1, wherein the historical supply data at least includes historical value attribute values, historical sales data and factors affecting the sales data, and the factors affecting the sales data at least include current traffic information of the target object.

6. The method according to claim 5, further comprising: Obtaining historical traffic information of the target object; as well as Based on the historical traffic information, the current traffic information of the target object is determined.

7. The method according to claim 1, further comprising: Get real-time supply data of target objects; Determining a candidate value attribute value-sales data relationship of the target object based on the historical supply data and the target relationship; as well as Based on the real-time supply data and the candidate value attribute value-sales data relationship, a target value attribute value-sales data relationship of the target object is determined.

8. The method according to claim 1, wherein the historical supply data comprises at least historical value attribute values ​​and historical sales data, and before determining the baseline value attribute value, the method further comprises: Based on the historical value attribute value and the historical sales data, determining a correlation coefficient and a significance level of the target object, wherein the correlation coefficient indicates the degree of linear correlation between the historical value attribute value and the historical sales data, and the significance level indicates the probability that there is no significant relationship between the historical value attribute value and the historical sales data; filtering the historical supply data based on the correlation coefficient and the significance level, The determination of the benchmark value attribute is based on the filtered historical supply data.

9. A data processing device, comprising: an acquisition module configured to acquire historical supply data of a target object and a first relationship, wherein the first relationship indicates a relationship between sales data, a reference value attribute value and a factor affecting the sales data, and a relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the reference value attribute value; A first determination module is configured to determine the baseline value attribute value based on the historical supply data; A second determination module is configured to determine a second relationship between sales data corresponding to different value attribute values ​​and sales data corresponding to the reference value attribute value based on the historical supply data and the reference value attribute value; as well as The adjustment module is configured to adjust the first relationship based on the reference value attribute value and the second relationship to obtain the target relationship of the target object.

10. An electronic device comprising: at least one processing unit; as well as At least one memory, the at least one memory is coupled to the at least one processing unit and stores instructions for execution by the at least one processing unit, the instructions, when executed by the at least one processing unit, cause the electronic device to perform the method according to any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, wherein the computer program can be executed by a processor to implement the method according to any one of claims 1 to 8.

12. A computer program product comprising computer executable instructions, wherein the computer executable instructions, when executed by a processor, implement the method according to any one of claims 1 to 8.

Images