Demand information generation method and apparatus, electronic device, and computer readable medium

Abstract

Embodiments of the present disclosure disclose a demand information generation method and device, electronic equipment and a computer readable medium. A specific implementation of the method comprises: obtaining item attribute information and an item flow information set of a target item; performing feature extraction on the item attribute information and the item flow information set respectively to generate item attribute feature information and an item flow feature information set; generating demand prior information according to the item attribute feature information; and generating demand posterior information according to the demand prior information and the item flow feature information set. The implementation is related to a smart supply chain and improves the accuracy of item demand classification.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the technical field of computer, and particularly, to a demand information generation method and device, an electronic device and a computer readable medium. BACKGROUND

[0002] In the article circulation scenario (such as a retail scenario), it is often necessary to generate demand information corresponding to the article according to the demand change characteristics of the article, so as to realize the demand classification of the article. At present, when performing demand classification, the commonly used method is: first, determining the index value capable of describing the demand of the article according to the historical circulation information of the article, and then performing demand classification by combining the index value with the threshold value set by artificial means.

[0003] However, when the above method is used, the following technical problems often exist:

[0004] The threshold value set by artificial means has strong subjectivity, thereby resulting in an inaccurate demand classification result. SUMMARY

[0005] The summary part of the present disclosure is used to introduce the concepts in a brief form, which will be described in detail in the specific embodiments part. The summary part of the present disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0006] Some embodiments of the present disclosure propose a demand information generation method and device, an electronic device and a computer readable medium to solve the technical problems mentioned in the background part.

[0007] In a first aspect, some embodiments of the present disclosure provide a demand information generation method, which comprises: obtaining article attribute information and an article circulation information set of a target article; performing feature extraction on the article attribute information and the article circulation information set respectively to generate article attribute feature information and an article circulation feature information set; generating demand prior information according to the article attribute feature information; and generating demand posterior information according to the demand prior information and the article circulation feature information set.

[0008] Optionally, the article attribute information comprises a non-numeric article attribute information set, the article attribute feature information comprises a non-numeric article attribute feature set, and the feature extraction on the article attribute information and the article circulation information set respectively to generate the article attribute feature information and the article circulation feature information set comprises: performing feature coding on each non-numeric article attribute information in the non-numeric article attribute information set to generate a non-numeric article attribute feature, thereby obtaining the non-numeric article attribute feature set.

[0009] Optionally, the item attribute feature information further comprises: a numerical item attribute feature set; and the generating demand prior information according to the item attribute feature information comprises: for each of the at least one classification category, performing the following processing steps: performing weighted sum processing on the item attribute features in the non-numerical item attribute feature set and the numerical item attribute feature set to generate weighted feature information corresponding to the classification category; performing interval compression on the weighted feature information to generate probability information corresponding to the classification category; and generating the demand prior information according to the probability information corresponding to the classification category in the at least one classification category.

[0010] Optionally, the item flow feature information set comprises: non-zero demand interval feature information; and the feature extraction on the item attribute information and the item flow information set respectively to generate item attribute feature information and item flow feature information set comprises: screening item flow information meeting a screening condition from the item flow information set as candidate item flow information to obtain a candidate item flow information set; and determining a ratio of a number of candidate item flow information included in the candidate item flow information set to a number of item flow information included in the item flow information set as the non-zero demand interval feature information.

[0011] Optionally, the item flow feature information set comprises: demand coefficient of variation feature information; and the feature extraction on the item attribute information and the item flow information set respectively to generate item attribute feature information and item flow feature information set comprises: determining a standard deviation of the item flow information included in the item flow information set to generate a demand standard deviation feature; determining a mean value of the item flow information included in the item flow information set to generate a demand mean value feature; and determining a ratio of the demand standard deviation feature to the demand mean value feature as the demand coefficient of variation feature information.

[0012] Optionally, the generating demand posterior information according to the demand prior information and the item flow feature information set comprises: for each of the at least one classification category, performing the following processing steps: determining probability information of each item flow feature information in the item flow feature information set under the classification category to generate candidate probability information to obtain a candidate probability information set; determining candidate probability information corresponding to the probability information according to the probability information of the classification category in the demand prior information and the candidate probability information set; and determining a probability value corresponding to the classification category in the at least one classification category according to the joint probability information of the obtained multiple candidate probability information and the item flow feature information set to obtain the demand posterior information.

[0013] Optionally, the item flow feature information in the item flow feature information set is independent of each other under a classification category.

[0014] Optionally, the method further comprises: according to the demand posterior information, performing item allocation on the target item.

[0015] In a second aspect, some embodiments of the present disclosure provide a demand information generation apparatus, the apparatus comprising: an acquisition unit configured to acquire item attribute information and an item flow information set of a target item; a feature extraction unit configured to perform feature extraction on the item attribute information and the item flow information set respectively, to generate item attribute feature information and an item flow feature information set; a first generation unit configured to generate demand prior information according to the item attribute feature information; and a second generation unit configured to generate demand posterior information according to the demand prior information and the item flow feature information set.

[0016] Optionally, the item attribute information comprises a non-numeric item attribute information set, the item attribute feature information comprises a non-numeric item attribute feature set, and the feature extraction unit is further configured to: perform feature encoding on each non-numeric item attribute information in the non-numeric item attribute information set, to generate a non-numeric item attribute feature, to obtain the non-numeric item attribute feature set.

[0017] Optionally, the item attribute feature information further comprises a numeric item attribute feature set, and the first generation unit is further configured to: for each classification category in at least one classification category, perform the following processing steps: weighted sum processing on the item attribute features in the non-numeric item attribute feature set and the numeric item attribute feature set, to generate weighted feature information corresponding to the classification category; interval compression on the weighted feature information, to generate probability information corresponding to the classification category; and generation of the demand prior information according to the probability information corresponding to the classification category in the at least one classification category.

[0018] Optionally, the item flow feature information set comprises non-zero demand interval feature information, and the feature extraction unit is further configured to: filter out item flow information satisfying a filtering condition from the item flow information set as candidate item flow information, to obtain a candidate item flow information set; and determine a ratio of a number of candidate item flow information included in the candidate item flow information set to a number of item flow information included in the item flow information set as the non-zero demand interval feature information.

[0019] Optionally, the aforementioned set of goods circulation feature information includes: demand variation coefficient feature information; and the aforementioned feature extraction unit is further configured to: determine the standard deviation of the goods circulation information included in the aforementioned set of goods circulation information to generate demand standard deviation features; determine the mean of the goods circulation information included in the aforementioned set of goods circulation information to generate demand mean features; and determine the ratio of the aforementioned demand standard deviation features to the aforementioned demand mean features as the aforementioned demand variation coefficient feature information.

[0020] Optionally, the second generation unit is further configured to perform the following processing steps for each of the at least one category: determine the probability information of each item circulation feature information in the item circulation feature information set under the category, so as to generate candidate probability information and obtain a candidate probability information set; determine the candidate probability information corresponding to the probability information based on the probability information of the category in the prior information of the requirement and the candidate probability information set; determine the probability value corresponding to the category in the at least one category based on the obtained multiple candidate probability information and the joint probability information of the item circulation feature information set, so as to obtain the posterior information of the requirement.

[0021] Optionally, the item circulation feature information in the above-mentioned item circulation feature information set is independent of each other under the classification category.

[0022] Optionally, the above-mentioned device further includes: allocating the target items according to the above-mentioned demand post-hoc information.

[0023] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0024] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0025] The above embodiments of this disclosure have the following beneficial effects: the demand information generation method of some embodiments of this disclosure improves the accuracy of the obtained demand classification results. Specifically, the reason why the obtained demand classification results are not accurate enough is that the threshold setting method is highly subjective, resulting in an imprecise demand classification result. Based on this, some embodiments of this disclosure first obtain the item attribute information and item circulation information set of the target item. Second, feature extraction is performed on the item attribute information and the item circulation information set respectively to generate item attribute feature information and item circulation feature information set. By performing feature extraction, feature construction is realized. Next, demand prior information is generated based on the item attribute feature information. The prior demand information of the target item is determined through the inherent attributes of the item. Finally, demand posterior information is generated based on the demand prior information and the item circulation feature information set. Then, the demand posterior information is corrected based on the item circulation feature information set to generate demand posterior information. In this way, the item attribute features and circulation features are combined in the demand classification process, which greatly improves the accuracy of item demand classification. Attached Figure Description

[0026] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0027] Figure 1 This is a schematic diagram illustrating an application scenario of the requirement information generation method according to some embodiments of this disclosure;

[0028] Figure 2 These are flowcharts of some embodiments of the requirement information generation method according to this disclosure;

[0029] Figure 3 These are flowcharts of some other embodiments of the requirement information generation method according to this disclosure;

[0030] Figure 4 These are schematic diagrams illustrating the structure of some embodiments of the demand information generation apparatus according to this disclosure;

[0031] Figure 5 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0032] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0033] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0034] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0035] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0036] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0037] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0038] Figure 1 This is a schematic diagram illustrating an application scenario of the requirement information generation method according to some embodiments of this disclosure.

[0039] exist Figure 1In this application scenario, firstly, the computing device 101 can obtain the item attribute information 102 and the item circulation information set 103 of the target item. In this application scenario, the item attribute information 102 can be {item name: Name A; item brand: XX brand; item category: electronic product; item color: black; item weight: 500 grams}. The item circulation information set 103 can be {[time: 2021-4-8; item circulation volume: 20], [time: 2021-4-9; item circulation volume: 100], [time: 2021-4-10; item circulation volume: 0], [time: 2021-4-11; item circulation volume: 0], [time: 2021-4-12; item circulation volume: 20]}. Then, the computing device 101 can extract features from the aforementioned item attribute information 102 and the aforementioned item circulation information set 103 to generate item attribute feature information 104 and item circulation feature information set 105. In this application scenario, the item attribute feature information 104 can be {item name: 000001; item brand: 000010; item category: 100000; item color: 010000; item weight: 500}. The aforementioned item circulation feature information set 105 can be {mean of item circulation volume: 28; variance of item circulation volume: 1376; standard deviation of item circulation volume: 37.09}. Next, the computing device 101 can generate prior demand information 106 based on the aforementioned item attribute feature information 104. Finally, the computing device 101 can generate posterior demand information 107 based on the aforementioned prior demand information 106 and the aforementioned item circulation feature information set 105.

[0040] It should be noted that the aforementioned computing device 101 can be either hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed within the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0041] It should be understood that Figure 1 The number of computing devices shown is merely illustrative. Any number of computing devices can be used depending on implementation needs.

[0042] Continue to refer to Figure 2 The flowchart 200 illustrates some embodiments of a requirement information generation method according to the present disclosure. The requirement information generation method includes the following steps:

[0043] Step 201: Obtain the set of item attribute information and item transfer information of the target item.

[0044] In some embodiments, the execution entity of the demand information generation method (e.g. Figure 1 The computing device 101 shown can acquire item attribute information and item circulation information set of a target item via wired or wireless connection. The target item can be an item for which demand forecasting is to be performed. For example, the target item could be an "apple". The item attribute information can be information about the inherent attributes of the target item. For example, the item attribute information can include, but is not limited to, at least one of the following: item name, item brand, item category, item color, item weight, and item volume. The item circulation information set represents the circulation data of the target item within a target time period. For example, the target time period can be the period from the first time the target item began circulating to the present. The circulation data can be "sales volume".

[0045] As an example, the above item attribute information could be {Item Name: Name A; Item Brand: XX Brand; Item Category: Electronic Products; Item Color: Black; Item Weight: 500 grams; Item Volume: 0.05 cubic meters}. The above item circulation information set could be {[Time: 2021-4-8; Item Circulation Volume: 20], [Time: 2021-4-9; Item Circulation Volume: 100], [Time: 2021-4-10; Item Circulation Volume: 0], [Time: 2021-4-11; Item Circulation Volume: 0], [Time: 2021-4-12; Item Circulation Volume: 20]}.

[0046] Optionally, the aforementioned item attribute information may include: a set of non-numerical item attribute information and a set of numerical item attribute information. The non-numerical item attribute information represents the non-numerical type of item attributes of the target item. The numerical item attribute information represents the numerical type of item attributes of the target item. For example, the non-numerical item attribute information could be {Item Name: Name A}. The numerical item attribute information could be {Item Weight: 500 grams}.

[0047] Step 202: Extract features from the sets of item attribute information and item circulation information respectively to generate sets of item attribute feature information and item circulation feature information.

[0048] In some embodiments, the executing entity may extract features from the item attribute information and the item circulation information set respectively to generate item attribute feature information and item circulation feature information set. The item attribute feature information can represent features extracted from the item attribute information. The item circulation feature information in the item circulation feature information set can represent features extracted from the item circulation information set. For example, the item circulation feature information set may include: the mean of item circulation volume, the variance of item circulation volume, and the standard deviation of item circulation volume.

[0049] As an example, the aforementioned executing entity extracts features from the item attribute information to generate the aforementioned item attribute feature information, which may include the following steps:

[0050] The first step is to perform vector conversion on the above item attribute information to generate item attribute vectors.

[0051] For example, the aforementioned executing entity can perform word vector conversion on the aforementioned item attribute vector to generate the aforementioned item attribute vector. For instance, the aforementioned executing entity can use the Word2vec model to perform word vector conversion on the aforementioned item attribute vector.

[0052] The second step is to extract features from the above-mentioned item attribute vectors to generate the above-mentioned item attribute information.

[0053] For example, the aforementioned execution entity can use a feature extraction model to extract features from the aforementioned item attribute vector to generate the aforementioned item attribute feature information. For instance, the aforementioned feature extraction model could be, but is not limited to, VGGNet (VisualGeometry Group Network).

[0054] As another example, the aforementioned executing entity may extract features from the aforementioned set of goods circulation information to generate a set of goods circulation feature information, which may include the following steps:

[0055] The first step is to determine the mean of the above-mentioned set of goods circulation information in order to generate the mean of goods circulation volume in the above-mentioned set of goods circulation feature information.

[0056] The second step is to determine the variance of the above-mentioned item circulation information set in order to generate the variance of item circulation volume in the above-mentioned item circulation feature information set.

[0057] The third step is to determine the standard deviation of the above-mentioned set of goods circulation information in order to generate the standard deviation of the goods circulation volume in the above-mentioned set of goods circulation feature information.

[0058] Optionally, the aforementioned item attribute feature information includes: a set of non-numerical item attribute features and a set of numerical item attribute features. Specifically, non-numerical item attribute features are the attribute features of non-numerical item attribute information. Numerical item attribute features are the attribute features of numerical item attribute information.

[0059] Optionally, the set of goods circulation characteristic information may include: non-zero demand interval characteristic information, demand coefficient of variation characteristic information, demand length information, and demand mean information. Specifically, the non-zero demand interval characteristic information represents the average interval of non-zero demands in the goods circulation information set. The demand coefficient of variation characteristic information represents the demand coefficient of variation in the goods circulation information set. The demand length information represents the quantity of goods circulation information included in the aforementioned goods circulation information set. The demand mean information represents the mean of the goods circulation information in the goods circulation information set.

[0060] Step 203: Generate prior information of requirements based on the item attribute characteristics.

[0061] In some embodiments, the aforementioned executing entity can generate prior demand information based on the item attribute characteristics. This prior demand information represents the prior demand classification of the target item. For example, the demand classification corresponding to the prior demand information may include, but is not limited to, any of the following: high-selling, high-value, stable; high-selling, high-value, intermittent; high-selling, low-value, stable; low-selling, low-value, intermittent; medium-low-selling, high-value, stable; medium-low-selling, high-value, intermittent; medium-low-selling, low-value, stable; medium-low-selling, low-value, intermittent; and slow-moving. Specifically, high-selling, high-value, stable indicates that the target item is a high-turnover, high-value, and stable-turnover item. High-selling, high-value, intermittent indicates that the target item is a high-turnover, high-value, and intermittent-turnover item. High-selling, low-value, stable indicates that the target item is a high-turnover, low-value, and stable-turnover item. Low-selling, low-value, intermittent indicates that the target item is a low-turnover, low-value, and intermittent-turnover item. The "Low-to-Medium Sales, High Value, Stable" type indicates that the target item has a low to medium turnover, high value, and stable circulation. The "Low-to-Medium Sales, High Value, Discontinuous" type indicates that the target item has a low to medium turnover, high value, and discontinuous circulation. The "Low-to-Medium Sales, Low Value, Stable" type indicates that the target item has a low to medium turnover, low value, and stable circulation. The "Low-to-Medium Sales, Low Value, Discontinuous" type indicates that the target item has a low to medium turnover, low value, and discontinuous circulation. The "Slow-Moving" type indicates that the target item is slow-moving.

[0062] As an example, the aforementioned execution entity can input the aforementioned item attribute feature information into a pre-trained prior classification model to generate the aforementioned prior demand information. The aforementioned prior classification model can be a classification model trained using prior training samples. Prior training samples may include: item attribute feature information of non-target items, and pre-labeled prior demand classifications of non-target items. The aforementioned prior classification model may be, but is not limited to, a random forest model.

[0063] Step 204: Generate demand posterior information based on the set of prior demand information and item flow characteristic information.

[0064] In some embodiments, the aforementioned executing entity can generate posterior demand information based on a set of prior demand information and item circulation characteristic information. This posterior demand information can characterize the posterior demand classification of the target item. For example, the demand classification categories corresponding to the posterior demand information can include, but are not limited to, any of the following: high-selling, high-value, stable; high-selling, high-value, intermittent; high-selling, low-value, stable; low-selling, low-value, intermittent; medium-low-selling, high-value, stable; medium-low-selling, high-value, intermittent; medium-low-selling, low-value, stable; medium-low-selling, low-value, intermittent; and slow-moving. Specifically, the high-selling, high-value, stable type indicates that the target item is a high-volume, high-value, and stable-circulation item. The high-selling, high-value, intermittent type indicates that the target item is a high-volume, high-value, and intermittent-circulation item. The high-selling, low-value, stable type indicates that the target item is a high-volume, low-value, and stable-circulation item. The low-selling, low-value, intermittent type indicates that the target item is a low-volume, low-value, and intermittent-circulation item. The "Low-to-Medium Sales, High Value, Stable" type indicates that the target item has a low to medium turnover, high value, and stable circulation. The "Low-to-Medium Sales, High Value, Discontinuous" type indicates that the target item has a low to medium turnover, high value, and discontinuous circulation. The "Low-to-Medium Sales, Low Value, Stable" type indicates that the target item has a low to medium turnover, low value, and stable circulation. The "Low-to-Medium Sales, Low Value, Discontinuous" type indicates that the target item has a low to medium turnover, low value, and discontinuous circulation. The "Slow-Moving" type indicates that the target item is slow-moving.

[0065] As an example, the aforementioned implementing entity can generate subsequent demand information based on the set of prior demand information and goods flow characteristic information, which may include the following steps:

[0066] The first step is to perform vector transformation on the aforementioned demand prior information and the aforementioned set of item flow characteristic information to generate demand prior information vector and item flow characteristic information vector.

[0067] The second step is to concatenate the aforementioned demand prior information vector and the item flow feature information vector to generate candidate vectors.

[0068] The third step is to input the above candidate vectors into a pre-trained posterior classification model to generate posterior requirement information.

[0069] For example, the aforementioned posterior classification model can be a classification model trained using posterior training samples. Posterior training samples can include: candidate vectors of non-target items, and pre-labeled posterior classification categories of non-target items. The aforementioned candidate vectors of non-target items are vectors generated based on a set of prior demand information and item circulation feature information for non-target items. The aforementioned posterior classification model can be, but is not limited to: CNN (Convolutional Neural Networks) models and LSTM (Long Short-Term Memory) models.

[0070] The above embodiments of this disclosure have the following beneficial effects: the demand information generation method of some embodiments of this disclosure improves the accuracy of the obtained demand classification results. Specifically, the reason why the obtained demand classification results are not accurate enough is that the threshold setting method is highly subjective, resulting in an imprecise demand classification result. Based on this, some embodiments of this disclosure first obtain the item attribute information and item circulation information set of the target item. Second, feature extraction is performed on the item attribute information and the item circulation information set respectively to generate item attribute feature information and item circulation feature information set. By performing feature extraction, feature construction is realized. Next, demand prior information is generated based on the item attribute feature information. The prior demand information of the target item is determined through the inherent attributes of the item. Finally, demand posterior information is generated based on the demand prior information and the item circulation feature information set. Then, the demand posterior information is corrected based on the item circulation feature information set to generate demand posterior information. In this way, the item attribute features and circulation features are combined in the demand classification process, which greatly improves the accuracy of item demand classification.

[0071] Further reference Figure 3 This illustrates a flow 300 of another embodiment of the demand information generation method. Flow 300 of this demand information generation method includes the following steps:

[0072] Step 301: Obtain the set of item attribute information and item circulation information of the target item.

[0073] In some embodiments, the specific implementation of step 301 and its resulting technical effects can be found in [reference needed]. Figure 2 Step 201 in the corresponding embodiment will not be repeated here.

[0074] Step 302: Perform feature encoding on each non-numerical item attribute information in the non-numerical item attribute information set to generate non-numerical item attribute features, and obtain a non-numerical item attribute feature set.

[0075] In some embodiments, the execution entity of the demand information generation method (e.g. Figure 1 The computing device 101 shown can perform feature encoding on each non-numerical item attribute information in the non-numerical item attribute information set to generate non-numerical item attribute features, thereby obtaining a non-numerical item attribute feature set.

[0076] For example, the aforementioned executing entity can perform one-hot encoding on the aforementioned non-numerical item attribute information to generate non-numerical item attribute features corresponding to the aforementioned non-numerical item attribute information.

[0077] Optionally, the aforementioned executing entity may determine each numerical item attribute information in the numerical item attribute information set as the numerical item attribute feature corresponding to the numerical item attribute information, thereby obtaining a set of numerical item attribute features.

[0078] Step 303: Select the item flow information that meets the selection criteria from the item flow information set as candidate item flow information to obtain the candidate item flow information set.

[0079] In some embodiments, the aforementioned executing entity can filter out item flow information that meets the filtering criteria from the item flow information set, and use this as candidate item flow information to obtain a candidate item flow information set. The filtering criteria may be that the item flow information corresponds to a non-zero item flow volume, i.e., the item flow information corresponds to non-zero demand. Candidate item flow information refers to item flow information with a corresponding non-zero item flow volume.

[0080] As an example, the set of item circulation information could be {[Time: 2021-4-8; Item circulation volume: 20], [Time: 2021-4-9; Item circulation volume: 100], [Time: 2021-4-10; Item circulation volume: 0], [Time: 2021-4-11; Item circulation volume: 0], [Time: 2021-4-12; Item circulation volume: 20]}. Then, the candidate set of item circulation information could be {[Time: 2021-4-8; Item circulation volume: 20], [Time: 2021-4-9; Item circulation volume: 100], [Time: 2021-4-12; Item circulation volume: 20]}.

[0081] Step 304: The ratio of the number of candidate item circulation information included in the candidate item circulation information set to the number of item circulation information included in the item circulation information set is determined as the non-zero demand interval feature information.

[0082] In some embodiments, the aforementioned executing entity may determine the non-zero demand interval characteristic information as the ratio of the number of candidate item circulation information included in the candidate item circulation information set to the number of item circulation information included in the item circulation information set.

[0083] As an example, the set of goods circulation information could be {[Time: 2021-4-8; Item circulation volume: 20], [Time: 2021-4-9; Item circulation volume: 100], [Time: 2021-4-10; Item circulation volume: 0], [Time: 2021-4-11; Item circulation volume: 0], [Time: 2021-4-12; Item circulation volume: 20]}. The candidate set of goods circulation information could be {[Time: 2021-4-8; Item circulation volume: 20], [Time: 2021-4-9; Item circulation volume: 100], [Time: 2021-4-12; Item circulation volume: 20]}. The generated non-zero demand interval feature information could then be 1.67 (rounded to two decimal places).

[0084] Optionally, the aforementioned implementing entity may include the following steps in determining the demand variation coefficient feature information included in the set of goods circulation feature information:

[0085] The first step is to determine the standard deviation of the item circulation information included in the above-mentioned item circulation information set in order to generate the demand standard deviation feature.

[0086] As an example, the item circulation information set could include the circulation volumes of each item as [10, 20, 10, 10, 10, 0, 20, 10]. The demand standard deviation characteristic could then be 6.61 (rounded to two decimal places).

[0087] The second step is to determine the mean of the item circulation information included in the above item circulation information set in order to generate the demand mean feature.

[0088] As an example, the item circulation information set could include the circulation volume of each item as [10, 20, 10, 10, 10, 0, 20, 10]. Then the demand mean characteristic could be 12.5.

[0089] The third step is to determine the ratio of the aforementioned standard deviation of demand to the aforementioned mean of demand as the aforementioned coefficient of variation of demand characteristic information.

[0090] As an example, the standard deviation of demand could be 6.61. The mean of demand could be 12.5. The coefficient of variation of demand could be 0.53 (rounded to two decimal places).

[0091] Step 305: Generate prior information of requirements based on the item attribute characteristics.

[0092] In some embodiments, the execution entity can generate prior demand information based on item attribute feature information. Specifically, for each of at least one classification category, the execution entity can perform the following processing steps:

[0093] The first step involves performing a weighted summation of the item attribute features in both the non-numerical and numerical item attribute feature sets to generate weighted feature information corresponding to the aforementioned classification categories. The weights include constant term weights and non-constant term weights. The number of non-constant term weights is equal to the sum of the first and second quantities. The first quantity is the number of non-numerical item attribute features included in the non-numerical item attribute feature set. The second quantity is the number of numerical item attribute features included in the numerical item attribute feature set. The constant term weight is 1. Both the constant term weights and the non-constant term weights are to be solved.

[0094] The second step is to perform interval compression on the weighted feature information to generate probability information corresponding to the above classification categories.

[0095] The aforementioned executing entity can use the Sigmoid function to perform interval compression on the weighted feature information. The probability information corresponding to the classification category contains both constant and non-constant weights that have been solved.

[0096] The third step is to generate the aforementioned prior information for the requirements based on the probability information corresponding to at least one of the above classification categories.

[0097] The aforementioned prior information includes probability information corresponding to at least one of the classification categories.

[0098] Step 306: Generate demand posterior information based on the set of demand prior information and item flow characteristic information.

[0099] In some embodiments, the aforementioned executing entity generates demand post-demand information based on a set of demand prior information and item flow characteristic information.

[0100] As an example, the aforementioned implementing entity generates subsequent demand information based on the set of prior demand information and goods flow characteristic information, which may include the following steps:

[0101] First, for each of at least one classification category, the aforementioned implementing entity may perform the following processing steps:

[0102] The first sub-step involves determining the probability information of each item circulation feature in the above item circulation feature information set under the above classification category, in order to generate candidate probability information and obtain a candidate probability information set.

[0103] For example, firstly, the executing entity can determine the statistical distribution of the item circulation characteristic information under the aforementioned classification category. For instance, the statistical distribution could be a normal distribution. Next, the executing entity can convert the probability formula of the item circulation characteristic information under the aforementioned classification category into a probability density function to generate the aforementioned candidate probability information.

[0104] The second sub-step involves determining the candidate probability information corresponding to the aforementioned probability information based on the probability information of the aforementioned classification categories and the aforementioned candidate probability information set in the aforementioned prior information of requirements.

[0105] For example, the statistical distribution of the aforementioned probability information can be a normal distribution, and the executing entity can convert the probability information of the classification category into a probability density function. Specifically, the executing entity can determine the candidate probability information as the product of the first target probability density function and the second target probability density function. The first target probability density function is the product of the probability density functions corresponding to the candidate probability information in the candidate probability information set. The second target probability density function is the probability density function of the classification category probability information.

[0106] The second step is to determine the probability value corresponding to the classification category in at least one classification category based on the joint probability information of multiple candidate probability information and the above-mentioned item circulation feature information set, and obtain the above-mentioned demand posterior information.

[0107] The joint probability information is equal to the sum of the candidate probability information included in the multiple candidate probability information. For example, the joint probability information can be obtained by, but is not limited to, any of the following methods: gradient descent, stochastic gradient descent, Newton's method, and quasi-Newton's method. The sum of the candidate probability information included in the multiple candidate probability information can be obtained by the maximum likelihood estimation method. The specific values ​​of the weights in the probability information corresponding to each classification category can be obtained, and thus the probability value of each classification analogy can be obtained to generate the above-mentioned demand posterior information. For example, the above demand posterior information could be {[Category: High-selling, high-value, stable type, probability value: 0.92;], [Category: High-selling, high-value, discontinuous type, probability value: 0.32], [Category: High-selling, low-value, stable type, probability value: 0.11]}. Then the classification category of the target item is "High-selling, high-value, stable type".

[0108] Optionally, the item circulation feature information in the item circulation feature information set is independent of each other under the classification category.

[0109] Step 307: Based on the post-request information, allocate the target items.

[0110] In some embodiments, the aforementioned executing entity allocates the target items based on the demand post-hoc information.

[0111] As an example, when the aforementioned demand posterior information indicates that the target item is "high-selling, high-value, and stable," the current replenishment strategy for the target item remains unchanged, and the target item is allocated accordingly.

[0112] As another example, when the aforementioned demand posterior information indicates that the target item is of the "high-selling, high-value, stable type", the replenishment quantity of the target item is reduced, and the target item is allocated according to the reduced replenishment quantity.

[0113] from Figure 3 It can be seen that, with Figure 2 Compared to the descriptions of some corresponding embodiments, this disclosure firstly refines the feature construction methods for item attribute characteristic information and item circulation characteristic information. Secondly, it refines the generation methods for demand prior information and demand posterior information, and eliminates the need to construct complex neural network models, thereby reducing the generation cost of demand posterior information.

[0114] Further reference Figure 4 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a demand information generation apparatus, which are similar to... Figure 2 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0115] like Figure 4 As shown, the demand information generation device 400 in some embodiments includes: an acquisition unit 401, a feature extraction unit 402, a first generation unit 403, and a second generation unit 404. The acquisition unit 401 is configured to acquire a set of item attribute information and item circulation information of a target item; the feature extraction unit 402 is configured to extract features from the item attribute information and the item circulation information set respectively to generate item attribute feature information and item circulation feature information set; the first generation unit 403 is configured to generate prior demand information based on the item attribute feature information; and the second generation unit 404 is configured to generate posterior demand information based on the prior demand information and the item circulation feature information set.

[0116] Optionally, the above-mentioned item attribute information includes: a set of non-numerical item attribute information, and the item attribute feature information includes: a set of non-numerical item attribute features; and the above-mentioned feature extraction unit 402 is further configured to: perform feature encoding on each non-numerical item attribute information in the above-mentioned non-numerical item attribute information set to generate non-numerical item attribute features, thereby obtaining the above-mentioned non-numerical item attribute feature set.

[0117] Optionally, the aforementioned item attribute feature information further includes: a set of numerical item attribute features; and the aforementioned first generation unit 403 is further configured to: for each of the at least one classification category, perform the following processing steps: perform weighted summation processing on the item attribute features in the aforementioned non-numerical item attribute feature set and the numerical item attribute feature set to generate weighted feature information corresponding to the aforementioned classification category; perform interval compression on the aforementioned weighted feature information to generate probability information corresponding to the aforementioned classification category; and generate the aforementioned requirement prior information based on the probability information corresponding to the aforementioned at least one classification category.

[0118] Optionally, the aforementioned set of goods circulation feature information includes: non-zero demand interval feature information; and the aforementioned feature extraction unit 402 is further configured to: filter out goods circulation information that meets the filtering conditions from the aforementioned set of goods circulation information as candidate goods circulation information, thereby obtaining a set of candidate goods circulation information; and determine the non-zero demand interval feature information as the ratio of the number of candidate goods circulation information included in the aforementioned set of candidate goods circulation information to the number of goods circulation information included in the aforementioned set of goods circulation information.

[0119] Optionally, the aforementioned set of goods circulation feature information includes: demand variation coefficient feature information; and the aforementioned feature extraction unit 402 is further configured to: determine the standard deviation of the goods circulation information included in the aforementioned set of goods circulation information to generate demand standard deviation features; determine the mean of the goods circulation information included in the aforementioned set of goods circulation information to generate demand mean features; and determine the ratio of the aforementioned demand standard deviation features to the aforementioned demand mean features as the aforementioned demand variation coefficient feature information.

[0120] Optionally, the second generation unit 404 is further configured to perform the following processing steps for each of the at least one category: determine the probability information of each item circulation feature information in the item circulation feature information set under the category, so as to generate candidate probability information and obtain a candidate probability information set; determine the candidate probability information corresponding to the probability information based on the probability information of the category in the prior information of the requirement and the candidate probability information set; determine the probability value corresponding to the category in the at least one category based on the obtained multiple candidate probability information and the joint probability information of the item circulation feature information set, so as to obtain the posterior information of the requirement.

[0121] Optionally, the item circulation feature information in the above-mentioned item circulation feature information set is independent of each other under the classification category.

[0122] Optionally, the device 400 further includes an allocation unit, wherein the allocation unit can be configured to allocate the target item according to the a posteriori information of the demand.

[0123] It is understandable that the units described in the device 400 are related to the reference. Figure 2 The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method also apply to device 400 and the units contained therein, and will not be repeated here.

[0124] The following is for reference. Figure 5 It illustrates electronic devices suitable for implementing some embodiments of the present disclosure (such as...). Figure 1 The diagram shows the structure of the computing device 101)500. Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0125] like Figure 5 As shown, the electronic device 500 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. The RAM 503 also stores various programs and data required for the operation of the electronic device 500. The processing unit 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0126] Typically, the following devices can be connected to I / O interface 505: input devices 506 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 507 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 508 including, for example, magnetic tapes, hard disks, etc.; and communication devices 509. Communication device 509 allows electronic device 500 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5 An electronic device 500 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 5 Each box shown can represent a device or multiple devices as needed.

[0127] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 509, or installed from storage device 508, or installed from ROM 502. When the computer program is executed by processing device 501, it performs the functions defined in the methods of some embodiments of this disclosure.

[0128] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0129] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0130] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: acquire a set of item attribute information and item flow information of the target item; extract features from the aforementioned item attribute information and the aforementioned item flow information set to generate item attribute feature information and an item flow feature information set; generate prior demand information based on the aforementioned item attribute feature information; and generate posterior demand information based on the aforementioned prior demand information and the aforementioned item flow feature information set.

[0131] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0132] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions 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, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0133] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including an acquisition unit, a feature extraction unit, a first generation unit, and a second generation unit. The names of these units do not necessarily limit the specific unit; for example, the acquisition unit may also be described as "a unit that acquires a set of item attribute information and item flow information of a target item."

[0134] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0135] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for generating demand information, comprising: Obtain a set of item attribute information and item circulation information for the target item; Feature extraction is performed on the item attribute information and the item circulation information set respectively to generate item attribute feature information and item circulation feature information set, wherein the item circulation feature information in the item circulation feature information set represents the features extracted from the item circulation information set; Based on the item attribute feature information, demand prior information is generated, wherein the item attribute feature information is input into a pre-trained prior classification model to generate the demand prior information, and the demand prior information represents the prior demand classification of the target item. Based on the prior information of demand and the set of characteristics of item circulation, posterior information of demand is generated, and the posterior information of demand represents the posterior demand classification of the target item.

2. The method according to claim 1, wherein, The item attribute information includes: a set of non-numerical item attribute information; the item attribute feature information includes: a set of non-numerical item attribute features; and The step of extracting features from the item attribute information and the item circulation information set to generate item attribute feature information and item circulation feature information set respectively includes: Each non-numerical item attribute in the non-numerical item attribute information set is feature-encoded to generate non-numerical item attribute features, thus obtaining the non-numerical item attribute feature set.

3. The method according to claim 2, wherein, The item attribute feature information also includes: a set of numerical item attribute features; and The step of generating prior demand information based on the item attribute feature information includes: For each of at least one classification category, perform the following processing steps: The item attribute features in the non-numerical item attribute feature set and the numerical item attribute feature set are weighted and summed to generate the weighted feature information corresponding to the classification category; The weighted feature information is compressed into intervals to generate probability information corresponding to the classification category; The requirement prior information is generated based on the probability information corresponding to the classification category in the at least one classification category.

4. The method according to claim 1, wherein, The set of goods circulation characteristic information includes: non-zero demand interval characteristic information; and The step of extracting features from the item attribute information and the item circulation information set to generate item attribute feature information and item circulation feature information set respectively includes: Item flow information that meets the filtering criteria is selected from the item flow information set and used as candidate item flow information to obtain a candidate item flow information set; The ratio of the number of candidate item circulation information included in the candidate item circulation information set to the number of item circulation information included in the item circulation information set is determined as the non-zero demand interval feature information.

5. The method according to claim 1, wherein, The set of goods circulation characteristic information includes: demand variation coefficient characteristic information; and The step of extracting features from the item attribute information and the item circulation information set to generate item attribute feature information and item circulation feature information set respectively includes: Determine the standard deviation of the item circulation information included in the item circulation information set to generate the demand standard deviation feature; Determine the mean value of the item circulation information included in the item circulation information set to generate the demand mean feature; The ratio of the demand standard deviation feature to the demand mean feature is determined as the demand variation coefficient feature information.

6. The method according to any one of claims 1-5, wherein, The step of generating post-demand information based on the prior demand information and the set of item flow characteristic information includes: For each of at least one classification category, perform the following processing steps: Determine the probability information of each item circulation feature information in the item circulation feature information set under the classification category to generate candidate probability information and obtain a candidate probability information set; Based on the probability information of the classification category in the prior information of the requirements and the set of candidate probability information, determine the candidate probability information corresponding to the probability information; Based on the joint probability information of multiple candidate probability information and the set of item circulation feature information, the probability value corresponding to the classification category in at least one classification category is determined, and the demand posterior information is obtained.

7. The method according to claim 6, wherein, The item circulation feature information in the set of item circulation feature information is independent of each other under the classification category.

8. The method according to claim 1, wherein, The method further includes: Based on the required post-hoc information, the target item is allocated.

9. A demand information generation device, comprising: The acquisition unit is configured to acquire a set of item attribute information and item circulation information of the target item; The feature extraction unit is configured to extract features from the item attribute information and the item circulation information set, respectively, to generate item attribute feature information and item circulation feature information set, wherein the item circulation feature information in the item circulation feature information set represents the features extracted from the item circulation information set. The first generation unit is configured to generate demand prior information based on the item attribute feature information, wherein the item attribute feature information is input into a pre-trained prior classification model to generate the demand prior information, and the demand prior information represents the prior demand classification of the target item. The second generation unit is configured to generate demand posterior information based on the demand prior information and the set of item flow feature information, wherein the demand posterior information represents the demand classification of the target item in the posterior sense.

10. An electronic device, comprising: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 8.

11. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 8.

Images