CN113807905A

CN113807905A - Article recommendation method and device, computer storage medium and electronic equipment

Info

Publication number: CN113807905A
Application number: CN202011225139.8A
Authority: CN
Inventors: 赵森; 董亚楠
Original assignee: Beijing Jingdong Century Trading Co Ltd; Beijing Wodong Tianjun Information Technology Co Ltd
Current assignee: Beijing Jingdong Century Trading Co Ltd; Beijing Wodong Tianjun Information Technology Co Ltd
Priority date: 2020-11-05
Filing date: 2020-11-05
Publication date: 2021-12-17

Abstract

The disclosure relates to the technical field of data processing, and provides an article recommendation method, an article recommendation device, a computer storage medium and an electronic device, wherein the article recommendation method comprises the following steps: determining a target category to which a target user belongs under preset M characteristic dimensions; matching the score values corresponding to the feature dimensions according to the corresponding relation between the feature dimensions and the score values stored in the storage space in advance; determining a weighting parameter corresponding to each target category according to a corresponding relation between the target categories pre-stored in a storage space and the weighting parameters; determining a consumption potential evaluation value of the target user according to the score value and the weighting parameter; recommending the target item corresponding to the consumption potential evaluation value for the target user. The method and the device can recommend the articles matched with the consumption capacity of the newly registered user, so that the recommended articles can accurately cover the user requirements, and the article searching time of the user is saved.

Description

Article recommendation method and device, computer storage medium and electronic equipment

Technical Field

The present disclosure relates to the field of data processing technologies, and in particular, to a method and an apparatus for recommending an article, a computer storage medium, and an electronic device.

Background

With the rapid development of computer and internet technologies, the related data processing field is also rapidly developing. For the related goods display platform, if the goods matched with the consumption capability can be recommended for different types of users, the time for the users to search the goods can be greatly saved, and the user satisfaction degree is improved.

In the related art, the consumption potential of the consumer is generally evaluated by matching with models such as a decision tree and the like based on existing behaviors of the user, such as purchasing behavior, browsing behavior, purchased commodity attributes and the like, and thus, the consumption potential of a newly registered user cannot be evaluated and the newly registered user cannot be accurately recommended in the related art.

In view of the above, there is a need in the art to develop a method and apparatus for recommending new items.

It is to be noted that the information disclosed in the background section above is only used to enhance understanding of the background of the present disclosure.

Disclosure of Invention

The object of the present disclosure is to provide an article recommendation method, an article recommendation apparatus, a computer storage medium and an electronic device, so as to avoid, at least to a certain extent, a defect that a newly registered user cannot be accurately recommended in the related art.

Additional features and advantages of the disclosure will be set forth in the detailed description which follows, or in part will be obvious from the description, or may be learned by practice of the disclosure.

According to a first aspect of the present disclosure, there is provided a recommendation method of an item, including: determining a target category to which a target user belongs under preset M characteristic dimensions; the target users comprise new registered users, and M is a positive integer; matching the score values corresponding to the feature dimensions according to the corresponding relation between the feature dimensions and the score values stored in a storage space in advance; determining a weighting parameter corresponding to each target category according to a corresponding relation between the target categories pre-stored in the storage space and the weighting parameters; determining a consumption potential evaluation value of the target user according to the scoring value and the weighting parameter; recommending the target item corresponding to the consumption potential evaluation value for the target user.

In an exemplary embodiment of the present disclosure, the determining a target category to which the target user belongs under the preset M feature dimensions includes: acquiring user characteristic values of a target user under the M characteristic dimensions; and determining the target category of the target user under each feature dimension according to the user feature value.

In an exemplary embodiment of the present disclosure, before determining a target category to which a target user belongs under preset M feature dimensions, the method further includes: dividing the obtained sample users into positive sample users and negative sample users; the sample user is a user who purchased a specified type of article in a historical time period; determining a target class to which each sample user belongs under the M characteristic dimensions; determining a chi-square value corresponding to each characteristic dimension according to the number of positive sample users in each target category, the number of negative sample users in each target category, the total number of positive sample users, the total number of negative sample users, the number of sample users contained in each target category and the total number of sample users; determining a scoring value corresponding to each characteristic dimension according to the chi-square value and the total number of the sample users; and correspondingly storing the feature dimension and the score value into the storage space.

In an exemplary embodiment of the present disclosure, the determining, according to the number of positive sample users in each target category, the number of negative sample users in each target category, the total number of positive sample users, the total number of negative sample users, the number of sample users included in each target category, and the total number of sample users, a chi-square value corresponding to each feature dimension includes: determining a first statistical value according to the number of positive sample users, the total number of positive sample users, the number of sample users contained in each target category and the total number of sample users in each target category; determining a second statistical value according to the number of negative sample users, the total number of the negative sample users, the number of sample users contained in each target category and the total number of the sample users in each target category; and determining an accumulated value of the first statistical value and the second statistical value as the chi-square value corresponding to each feature dimension.

In an exemplary embodiment of the present disclosure, the dividing the obtained sample users into positive sample users and negative sample users includes: dividing the specified type of articles into low-price articles and high-price articles; determining a sample user who purchased the high-priced item as the positive sample user; and determining a sample user who purchased the low-priced item as the negative sample user.

In an exemplary embodiment of the present disclosure, the dividing the specified category of items into low-value items and high-value items includes: acquiring first scores of prices of a plurality of articles corresponding to the specified articles; determining an item having an item price greater than the first score as the high-value item; determining an item having an item price less than or equal to the first quantile as the low price item.

In an exemplary embodiment of the present disclosure, the method further comprises: acquiring a first ratio of the number of positive sample users in each target category to the total number of the positive sample users; acquiring a second ratio of the number of negative sample users in each target category to the total number of the negative sample users; acquiring a third ratio of the first ratio to the second ratio; determining the natural logarithm of the third ratio as a weighting parameter corresponding to each target category; and correspondingly storing the target category and the weighting parameter into the storage space.

In an exemplary embodiment of the present disclosure, the determining the consumption potential assessment value of the target user according to the score value and the weighting parameter includes: obtaining the product of the score value and the weighting parameter under each characteristic dimension; determining a sum of the M products as a consumption potential assessment value of the target user.

In an exemplary embodiment of the present disclosure, the recommending, for the target user, the target item corresponding to the consumption potential evaluation value includes: acquiring second scores of a plurality of consumption potential evaluation values corresponding to a plurality of target users; determining the consumption potential grade of the target user according to the size relation between the consumption potential evaluation value of the target user and the second score; and recommending and specifying the target item for the target user according to the pre-stored corresponding relation between the consumption potential grade and the target item.

According to a second aspect of the present disclosure, there is provided an article recommendation apparatus comprising: the determining module is used for determining the target category to which the target user belongs under the preset M characteristic dimensions; the target users comprise new registered users, and M is a positive integer; the evaluation module is used for matching the score values corresponding to the feature dimensions according to the corresponding relation between the feature dimensions and the score values stored in the storage space in advance; determining a weighting parameter corresponding to each target category according to a corresponding relation between the target categories pre-stored in the storage space and the weighting parameters; determining a consumption potential evaluation value of the target user according to the scoring value and the weighting parameter; and the item recommending module is used for recommending the target item corresponding to the consumption potential evaluation value for the target user.

According to a third aspect of the present disclosure, there is provided a computer storage medium having stored thereon a computer program which, when executed by a processor, implements the method of recommending items of the first aspect described above.

According to a fourth aspect of the present disclosure, there is provided an electronic device comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the method of recommending items according to the first aspect described above via execution of the executable instructions.

As can be seen from the foregoing technical solutions, the method for recommending an article, the apparatus for recommending an article, the computer storage medium, and the electronic device in the exemplary embodiment of the disclosure have at least the following advantages and positive effects:

in the technical solutions provided by some embodiments of the present disclosure, on one hand, the target category to which the target user belongs under the preset M feature dimensions is determined, the target user can be divided into different types, and the subsequent recommendation accuracy is improved. Furthermore, according to the corresponding relation between the characteristic dimension and the score value stored in the storage space in advance, the score value corresponding to each characteristic dimension is matched, and according to the corresponding relation between the target category stored in the storage space in advance and the weighting parameter, the weighting parameter corresponding to each target category is determined, so that the consumption potential evaluation value of the target user is determined according to the score value and the weighting parameter, a new thought for evaluating the consumption potential of the user from the characteristics of the user is provided, the target user can be a historical user or a new registered user without historical behaviors, and the problem that in the related technology, the evaluation of the historical user is only carried out from the existing behaviors of the user, and the evaluation of the new registered user cannot be carried out is solved. On the other hand, the target item corresponding to the consumption potential evaluation value is recommended for the target user, the item database and the user database can be communicated, the time for the user to search for the item is saved, the user satisfaction degree is improved, and the browsing efficiency and the item transaction rate of the user are improved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and together with the description, serve to explain the principles of the disclosure. It is to be understood that the drawings in the following description are merely exemplary of the disclosure, and that other drawings may be derived from those drawings by one of ordinary skill in the art without the exercise of inventive faculty.

FIG. 1 shows a flow diagram of a method of recommending items in an exemplary embodiment of the present disclosure;

FIG. 2 illustrates a sub-flow diagram of a method of recommending items in an exemplary embodiment of the present disclosure;

FIG. 3 illustrates a sub-flow diagram of a method of recommending items in an exemplary embodiment of the present disclosure;

FIG. 4 illustrates a sub-flow diagram of a method of recommending items in an exemplary embodiment of the present disclosure;

FIG. 5 is a sub-flow diagram illustrating a method of recommending items in an exemplary embodiment of the present disclosure;

FIG. 6 is a general flow chart diagram illustrating a method for recommending items according to an exemplary embodiment of the present disclosure;

fig. 7 is a schematic structural diagram of an article recommendation device in an exemplary embodiment of the present disclosure;

FIG. 8 shows a schematic diagram of a structure of a computer storage medium in an exemplary embodiment of the disclosure;

fig. 9 shows a schematic structural diagram of an electronic device in an exemplary embodiment of the present disclosure.

Detailed Description

Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the disclosure. One skilled in the relevant art will recognize, however, that the subject matter of the present disclosure can be practiced without one or more of the specific details, or with other methods, components, devices, steps, and the like. In other instances, well-known technical solutions have not been shown or described in detail to avoid obscuring aspects of the present disclosure.

The terms "a," "an," "the," and "said" are used in this specification to denote the presence of one or more elements/components/parts/etc.; the terms "comprising" and "having" are intended to be inclusive and mean that there may be additional elements/components/etc. other than the listed elements/components/etc.; the terms "first" and "second", etc. are used merely as labels, and are not limiting on the number of their objects.

Furthermore, the drawings are merely schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repetitive description will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities.

In the related art, a statistical learning-based method is generally used for predicting the consumption potential of related users according to models such as historical purchasing behaviors (for example, order amount, order times and order time), browsing behaviors (for example, PV (total browsing amount of all pages in a store), UV (number of visitors to each page of the store, and number of visitors when a user visits the store Multiple times in one day), purchased commodity attributes (for example, gear of the commodity and price of the commodity), and the like which have occurred to the user, and a Decision Tree, a random forest, GBDT (Multiple subscription Decision Tree, which is an iterative Decision Tree algorithm) and the like are matched.

However, the above method has the following problems: firstly, the current technology is too dependent on the behavior characteristics of the user, and cannot be used for a newly registered user without the behavior characteristics; secondly, the decision tree algorithm processes relatively small amount of data, and cannot perform complex calculation and iteration for a large amount of data.

In the embodiment of the disclosure, firstly, a method for recommending an article is provided, which overcomes the defect that accurate recommendation cannot be performed on a newly registered user in the related art at least to a certain extent.

Fig. 1 is a flowchart illustrating a recommendation method for an item in an exemplary embodiment of the present disclosure, where an execution subject of the recommendation method for an item may be a server that performs recommendation processing on the item.

Referring to fig. 1, a recommendation method of an item according to an embodiment of the present disclosure includes the steps of:

step S110, determining a target category to which a target user belongs under preset M characteristic dimensions;

step S120, matching the score values corresponding to the feature dimensions according to the corresponding relation between the feature dimensions and the score values stored in the storage space in advance;

step S130, determining a weighting parameter corresponding to each object category according to the corresponding relation between the object categories pre-stored in the storage space and the weighting parameters;

step S140, determining a consumption potential evaluation value of the target user according to the score value and the weighting parameter;

and step S150, recommending the target item corresponding to the consumption potential evaluation value for the target user.

In the technical scheme provided by the embodiment shown in fig. 1, the target category to which the target user belongs under the preset M characteristic dimensions is determined, the target user can be divided into different types, and the subsequent recommendation accuracy is improved. Furthermore, according to the corresponding relation between the characteristic dimension and the score value stored in the storage space in advance, the score value corresponding to each characteristic dimension is matched, and according to the corresponding relation between the target category stored in the storage space in advance and the weighting parameter, the weighting parameter corresponding to each target category is determined, so that the consumption potential evaluation value of the target user is determined according to the score value and the weighting parameter, a new thought for evaluating the consumption potential of the user from the characteristics of the user is provided, the target user can be a historical user or a new registered user without historical behaviors, and the problem that in the related technology, the evaluation of the historical user is only carried out from the existing behaviors of the user, and the evaluation of the new registered user cannot be carried out is solved. On the other hand, the target item corresponding to the consumption potential evaluation value is recommended for the target user, the item database and the user database can be communicated, the time for the user to search for the item is saved, the user satisfaction degree is improved, and the browsing efficiency and the item transaction rate of the user are improved.

The following describes the specific implementation of the steps in fig. 1 in detail:

in an exemplary embodiment of the present disclosure, sample users may be screened from historical users in advance, and a score value corresponding to each feature dimension may be determined by combining relevant feature information of the sample users. Specifically, referring to fig. 2, fig. 2 shows a sub-flow diagram of a recommendation method for an article in an exemplary embodiment of the present disclosure, and specifically shows a sub-flow diagram for calculating a user feature value corresponding to each feature dimension according to related information of a sample user, including steps S201 to S205, and a specific implementation is explained below with reference to fig. 2.

In step S201, the acquired sample users are divided into positive sample users and negative sample users.

In this step, a sample user may be obtained, where the sample user may be a user who purchased a specific type of article in a historical period of time. The specified articles can be mobile phone articles, household appliances, daily necessities and the like. For example, when the designated type of object is a mobile phone type of object, the designated type of object may be a mobile phone under the same brand, or a mobile phone under different brands. For example, the multiple mobile phones under the same brand may be mobile phones of the millet series, such as millet 6, millet 7, millet 8, and the like. For example, the plurality of mobile phones under different brands may be millet 8, iphone11, hua be P40, etc., which may be set by themselves according to actual situations, and belong to the protection scope of the present disclosure.

Further, the sample users may be divided into positive sample users and negative sample users according to the prices of the items purchased by the sample users. Specifically, referring to fig. 3, fig. 3 shows a sub-flow diagram of a recommendation method for an item in an exemplary embodiment of the present disclosure, and in particular, shows a sub-flow diagram of how to divide a sample user into a positive sample user and a negative sample user, including steps S301 to S302, and a specific implementation is explained below with reference to fig. 3.

In step S301, the specified category is divided into low-value items and high-value items.

In this step, the specific type of articles including article a and article B … …, article Z is taken as an example, the price of article a is priceA, the price of article B is priceB … …, and article Z is priceZ, and the priceA and priceB … … priceZ may be sorted in descending order to form a price list. Further, 0.9 quantiles of the price series (0.9 quantile is also referred to as 0.9 quantile point, which is obtained by taking a group of numerical values from small to large, and the numerical value number from the minimum value to 0.9 quantile accounts for 90% of all the numerical value numbers) are calculated, and further, the item whose item price is less than the 0.9 quantile is determined as a low-priced item, and the item whose item price is greater than or equal to the 0.9 quantile is determined as a high-priced item.

Illustratively, 0.7 quantiles (0.7 quantile is also referred to as 0.7 quantile, which is a set of numbers taken from small to large, and the number of numbers from the minimum to 0.7 quantile accounts for 70% of all the numbers) and 0.9 quantiles of the price series may also be calculated, and further, items whose price is less than or equal to 0.7 quantile are determined as low-priced items, items whose price is greater than 0.7 quantile and less than 0.9 quantile are determined as medium-priced items, and items whose price is greater than or equal to 0.9 quantile are determined as high-priced items.

In step S302, a sample user who purchases an item of high price is determined as a positive sample user, and a sample user who purchases an item of low price is determined as a negative sample user.

Accordingly, referring to the related explanation of step S301, the sample user who purchases the high-priced items may be determined as a positive sample user, and the sample user who purchases the low-priced items may be determined as a negative sample user. Alternatively, the sample user who purchases the high-price item may be determined as a positive sample user, and the user who purchases the low-price item and the medium-price item may be determined as a negative sample user.

With continued reference to fig. 2, in step S202, the target category to which each sample user belongs under M feature dimensions is determined.

In this step, the user feature values of the sample user in the M feature dimensions may be determined. For example, the M feature dimensions may include: the method comprises the following steps of setting user income (such as monthly income), user age, city level of a user, user science and technology preference, user occupation, user marital status, user automobile brand gear, possibility of children in the user, user mobile phone brand preference, user snack type preference, user child age and the like according to actual conditions, and belongs to the protection range of the method. Therefore, the related evaluation indexes are diversified, and the subsequent evaluation accuracy is guaranteed.

For example, when the feature dimension is User input, it may be determined that the User feature value corresponding to the User1 is: 60000, the User characteristic value corresponding to the User2 is 20000, and the User characteristic value corresponding to the User3 is: 3000. when the feature dimension is age, it may be determined that the User feature value corresponding to the User1 is 55, the User feature value corresponding to the User2 is 30, and the User feature value corresponding to the User3 is 21.

Furthermore, the sample users can be classified into different target categories of feature dimensions according to the user feature values. Illustratively, when the feature dimension is "user revenue", the target category may include "high revenue, medium revenue, low revenue", and the like.

For example, when the user income is greater than or equal to 50000, the target category to which the user income belongs may be set as "high income"; when the user income is more than 10000 and less than 50000, the target category to which the user income belongs is 'medium income'; when the user income is less than or equal to 10000, the target category to which the user belongs is 'low income'. As can be seen from the above explanation of step S203, the target category to which the User1 belongs in the feature dimension "User income" is "high income", the target category to which the User2 belongs in the feature dimension "User income" is "medium income", and the target category to which the User3 belongs in the feature dimension "User income" is "low income".

For example, it can also be set that when the age of the user is greater than or equal to 60 years, the target category to which the user belongs is "old"; when the age of the user is more than 45 years and less than 59 years, the target category to which the user belongs is 'middle-aged'; when the age of the user is less than or equal to 45 years, the target category to which the user belongs is "young". As can be seen from the above explanation of step S203, the target category to which the User1 belongs in the characteristic dimension "User age" is "middle age", and the target categories to which the User2 and the User3 belong in the characteristic dimension "User age" are "young age".

In step S203, a chi-squared value corresponding to each feature dimension is determined according to the number of positive sample users in each target category, the number of negative sample users in each target category, the total number of positive sample users, the total number of negative sample users, the number of sample users included in each target category, and the total number of sample users.

In this step, for example, referring to fig. 4, fig. 4 shows a sub-flow diagram of a consumption potential evaluation method in an exemplary embodiment of the present disclosure, specifically shows a sub-flow diagram of determining a chi-squared value corresponding to a feature dimension, which includes steps S401 to S403, and a specific implementation is explained below with reference to fig. 4.

In step S401, a first statistical value is determined according to the number of positive sample users, the total number of positive sample users, the number of sample users included in each target category, and the total number of sample users in each target category.

For example, referring to table 1 below, taking the target category as "high income", the number of included positive sample users is a11, and the number of included sample users is P1; taking the target category as "medium income", the number of the included positive sample users is a12, and the number of the included sample users is P2; taking the target category as "low income", the number of the included positive sample users is a13, and the number of the included sample users is P3; the total number of positive sample users is R1, and the total number of sample users is N.

TABLE 1

Thus, for example, for a positive sample user whose target category is "high income", the number of positive sample users belonging to "high income" may be taken as the observation frequency f₀Then f is₀A 11. Further, the expectation frequency f may be calculated based on the following formula 1_e：

Where R1 is the number of positive sample users, P1 is the number of sample users belonging to "high income", and N is the total number of sample users.

At the determination of f₀And f_eThen, the statistical value T11 may be calculated based on the following formula 2:

similarly, a statistical value may be calculated for a user with a target category of "medium income

For users with a target category of "low income

A statistical value T13 may be calculated, and the accumulated values of T11, T12, and T13 may be determined as the first statistical value.

In step S402, a second statistical value is determined according to the number of negative-sample users, the total number of negative-sample users, the number of sample users included in each target category, and the total number of sample users in each target category.

For example, referring to table 1 in step S401, taking the target category as "high income", the number of negative sample users included is a21, and the number of sample users included is P1; taking the target category as "medium income", the number of the included negative sample users is a22, and the number of the included sample users is P2; taking the target category as "low income", the number of the included positive sample users is a23, and the number of the included sample users is P3; the total number of negative sample users is R2, and the total number of sample users is N.

For example, for the negative sample users with the target category of "high income", the number of negative sample users belonging to "high income" can be used as the observation frequency f₀Then f is₀A 21. Further, the frequency f may be calculated based on the following formula 3_e：

Where R2 is the number of negative sample users, P1 is the number of sample users belonging to "high income", and N is the total number of sample users.

At the determination of f₀And f_eThen, the statistical value T21 may be calculated based on the above equation 4:

A statistical value may be calculated for a user with a target category of "low income

Further, the accumulated value of T21, T22, and T23 may be determined as the second statistical value.

In step S403, an accumulated value of the first statistical value and the second statistical value is determined as a chi-squared value corresponding to each feature dimension.

After the first statistical value and the second statistical value are determined, the accumulated sum of the first statistical value and the second statistical value may be determined as a chi-square value X corresponding to the characteristic dimension "user income²。

In step S204, a score value corresponding to each feature dimension is determined according to the chi-square value and the total number of the sample users.

In this step, the score value C1 corresponding to the characteristic dimension "user income" can be calculated based on the following formula 5:

wherein, X²N is the total number of the sample users.

In step S205, the feature dimensions and the score values are stored in a storage space.

In this step, the corresponding score values of the feature dimensions can be calculated with reference to the relevant explanations of the above steps, and are stored in the storage space correspondingly. For example, the feature dimension "user income" and the corresponding score value C1 may be stored in the storage space correspondingly, the feature dimension "user age" and the corresponding score value C2 may be stored in the storage space correspondingly, and the feature dimension "user city level" and the corresponding score value C3 may be stored in the storage space correspondingly. The storage space may be set by itself, for example, the storage space may be a user database, an optical disc drive, a mobile storage device, a certain disc (e.g., an E-disc) of a computer, a specific folder, an internal storage space of a mobile phone, an external Secure Digital Card (SD Card) of a mobile phone, or the like.

For example, weighting parameters corresponding to different object categories under each feature dimension may also be calculated, specifically, referring to fig. 5, fig. 5 shows a sub-flow diagram of a recommendation method for an article in an exemplary embodiment of the present disclosure, and specifically shows a sub-flow diagram of calculating weighting parameters corresponding to different object categories, including steps S501 to S505, and the following explains a specific implementation manner with reference to fig. 5.

In step S501, a first ratio of the number of positive sample users to the total number of positive sample users in each target category is obtained.

In this step, still taking the characteristic dimension as "user income" as an example for explanation, when the target category is "high income", a first ratio a11/R1 of the number of positive sample users a11 to the total number of positive sample users R1 may be obtained.

In step S502, a second ratio of the number of negative-sample users to the total number of negative-sample users in each target category is obtained.

In this step, when the target category is "high income", a second ratio a12/R2 of the number a12 of negative sample users to the total number R2 of negative sample users may be obtained.

In step S503, a third ratio of the first ratio to the second ratio is acquired.

In this step, a third ratio of the first ratio to the second ratio, that is, the first ratio and the second ratio may be obtained

In step S504, the natural logarithm of the third ratio is determined as the weighting parameter corresponding to each target class.

In this step, the natural logarithm of the third ratio may be determined as the weighting parameter W11 corresponding to the target category "high income", that is, the target category "high income" is

In step S505, the target category and the weighting parameter are stored in a storage space in association with each other.

Illustratively, a weighting parameter corresponding to a target category of "medium income" may also be calculated

Weighting parameter W13 corresponding to target class "Low revenue

Further, the "high income" and the corresponding weighting parameter W11 may be stored in the storage space, the "medium income" and the corresponding weighting parameter W12 may be stored in the storage space, and the "low income" and the corresponding weighting parameter W13 may be stored in the storage space.

Similarly, when the feature dimension is "user age", the weighting parameter W21 corresponding to "old age" may be correspondingly stored in the storage space, the weighting parameter W22 corresponding to "middle age" may be correspondingly stored in the storage space, and the weighting parameter W23 corresponding to "young age" may be correspondingly stored in the storage space.

With continued reference to fig. 1, in step S110, a target category to which the target user belongs under the preset M feature dimensions is determined.

In this step, user feature values of the target user in M feature dimensions may be obtained, and the target category to which the target user belongs in each feature dimension may be determined according to the user feature values. Specifically, reference may be made to the related explanation of the step S202, and the disclosure will not be repeated here. Therefore, the target users can be divided into different types, and the subsequent recommendation accuracy is improved.

It should be noted that the target user may be a historical user or a newly registered user. Therefore, a new idea for evaluating the consumption potential of the user based on the user characteristics is provided, and the problem that the evaluation of the historical user is only performed based on the existing behavior of the user, but the evaluation of the newly registered user cannot be performed in the related technology is solved.

When the target user is a historical user, user characteristic values of the target user under preset M characteristic dimensions can be determined through the user portrait, and then the target category of the target user under each characteristic dimension is determined according to the user characteristic values.

When the target user is a history user, a questionnaire can be issued to the history user, the user characteristic values corresponding to the characteristic dimensions of the history user are determined according to questionnaire results fed back by the target user, and the target category of the target user under the characteristic dimensions is determined according to the user characteristic values.

When the target user is a new registered user, the registration information of the user can be acquired, the user characteristic value corresponding to each characteristic dimension is determined according to the registration information, and the target category of the target user under each characteristic dimension is determined according to the user characteristic value.

In step S120, the score values corresponding to the feature dimensions are matched according to the corresponding relationship between the feature dimensions and the score values stored in the storage space in advance.

In this step, the scoring values corresponding to the feature dimensions may be read from the storage space (e.g., the user database) with reference to the related explanation in step S205. For example, when the feature dimension is the same as the M feature dimensions, score values C1, C2, and C3 … … CM corresponding to the feature dimension may be read.

In step S130, a weighting parameter corresponding to the target category is determined based on the correspondence between the target category and the weighting parameter stored in advance.

In this step, the weighting parameters corresponding to each target category may be read from the storage space (e.g., the user database) with reference to the relevant explanation in step S505, for example, when the target category to which the target user belongs under the characteristic dimension "user income" is "high income", the weighting parameters corresponding to the characteristic dimension "user income" may be read as W11; when the target category to which the age of the user belongs is "middle age", the weighting parameters corresponding to the characteristic dimension "user age" may be read as W22, … …, and similarly, the weighting parameters corresponding to the target categories to which other characteristic dimensions belong may be read.

In step S140, a consumption potential evaluation value of the target user is determined according to the score value and the weighting parameter.

After determining the above-mentioned score values and the corresponding weighting parameters, the consumption potential assessment value S of the target user can be calculated based on the following formula 6:

S-C1W 11+ C2W 22+ … … + CM WMj formula 6

Wherein S represents the consumption potential evaluation value, C1 represents the score value corresponding to the 1 st characteristic dimension, and W11 represents the weighting parameter corresponding to the target category 1 to which the target user belongs under the 1 st characteristic dimension; c2 represents the score value corresponding to the 2 nd characteristic dimension, and W22 represents the weighting parameter corresponding to the target class 2 to which the target user belongs under the 2 nd characteristic dimension; CM represents a score value corresponding to the mth characteristic dimension, and WMj represents a weighting parameter corresponding to the target category j to which the target user belongs under the mth characteristic dimension.

Furthermore, the consumption potential evaluation value of the target user can be written into the user database to avoid data loss.

In step S150, a target item corresponding to the consumption potential evaluation value is recommended for the target user.

In this step, a target item corresponding to the consumption potential evaluation value may be recommended to the target user. Illustratively, the presentation of the specified item in the current recommendation list of the target user may also be reversed based on the consumption potential assessment to avoid interface occupation of invalid items (i.e., items that do not match the user's consumption potential, such as top-level luxury items recommended for users with low consumption potential or discounted items recommended for target users with high consumption potential), and optimize the user experience.

For example, when the target users add up to a plurality of target users, the consumption potential evaluation values of the target users may be sorted from small to large to form a numerical sequence. Further, 0.9 quantiles corresponding to the numerical sequence may be calculated, and a target user having a consumption potential evaluation value greater than 0.9 quantile may be determined as a "high consumption potential user", and a target user having a consumption potential evaluation value less than or equal to the above-mentioned 0.9 quantile may be determined as a "low consumption potential user".

For example, the 0.7 quantile and the 0.9 quantile corresponding to the numerical sequence may be calculated, and further, the target user with the consumption potential evaluation value greater than or equal to the 0.9 quantile is determined as a "high consumption potential user", the target user with the consumption potential evaluation value greater than the 0.7 quantile and less than the 0.9 quantile is determined as a "medium consumption potential user", and the target user with the consumption potential evaluation value less than or equal to the 0.7 quantile is determined as a "low consumption potential user".

Furthermore, the target item can be recommended and specified for the target user according to the pre-stored corresponding relation between the consumption potential grade and the target item. For example, if the target user is the "high consumption potential user", the "high-end configuration mobile phone of a certain brand, the high-end configuration computer of a certain brand" and the like may be recommended to the target user to match the consumption capability of the target user. Therefore, the interest degree of the user for the target object can be improved, the time for the user to search the object is saved, the user satisfaction degree is improved, and the browsing efficiency and the object transaction rate of the user are improved.

Illustratively, the information of the target object can be read from the object database, and the association relationship between the target object and the target user is established, so that the object database and the user database are opened, different target objects are recommended to the target user at different time periods, the repeated determination of the consumption potential evaluation value of the target user within a preset time period (for example, one year) is avoided, the computing resources of the system are saved, and the processing efficiency of the system on other services is improved.

It should be noted that, in the present disclosure, the characteristic dimension of the sample user and the user characteristic value of the sample user in the M characteristic dimensions may also be input into a machine learning model (for example, a decision tree, a random forest, a reinforcement learning, and the like), and parameters are adjusted multiple times to train the machine learning model, so that a loss function of the machine learning model tends to converge, and a consumption potential prediction model is obtained. And then, the user characteristic values of the target user in each characteristic dimension can be input into the trained consumption potential evaluation model, and the consumption potential evaluation value corresponding to the target user is obtained according to the output of the consumption potential evaluation model. Further, the target item matched with the consumption capability of the target user can be recommended to the target user according to the consumption potential evaluation value.

For example, referring to fig. 6, fig. 6 shows an overall flowchart of a recommendation method for an item in an exemplary embodiment of the disclosure, which includes steps S601-S605, and a specific implementation is explained below with reference to fig. 6.

In step S601, screening M feature dimensions according to user attributes of sample users;

in step S602, the sample users are divided into positive sample users and negative sample users; respectively calculating the grading value (C coefficient, also called a series coefficient) corresponding to each characteristic dimension of different sample users and the weighting parameter (WOE code, Weight of Evidence, for short: WOE) corresponding to different target classes of each characteristic dimension; correspondingly storing the data into a storage space;

in step S603, according to the user feature value of the target user in each feature dimension, matching the score value corresponding to each feature dimension, and matching the weighting parameter corresponding to the target category;

in step S604, determining a consumption potential evaluation value of the target user according to the score value and the weighting parameter;

in step S605, a consumption potential grade of the target user is determined according to the consumption potential evaluation value.

Based on the technical scheme, the method and the device for recommending the articles can solve the technical problem that the articles can only be recommended to the user based on the existing behaviors (such as purchasing behaviors and browsing behaviors) of the user but cannot be recommended to a newly registered user in the related technology, and can also recommend the articles matched with the consumption capacity of the user, so that the recommended articles can accurately cover the user requirements, the time for the user to search the articles is saved, and the interest degree of the user on the target articles, the user satisfaction degree, the browsing efficiency of the user and the article transaction rate are improved.

The present disclosure also provides a recommendation apparatus for an article, and fig. 7 shows a schematic structural diagram of the recommendation apparatus for an article in an exemplary embodiment of the present disclosure; as shown in fig. 7, the recommendation apparatus 700 for an item may include a determination module 701, an evaluation module 702, and an item recommendation module 703. Wherein:

a determining module 701, configured to determine a target category to which a target user belongs under preset M feature dimensions.

In an exemplary embodiment of the present disclosure, the determining module is configured to obtain user feature values of a target user in M feature dimensions; and determining the target category of the target user under each feature dimension according to the user feature value.

The evaluation module 702 is configured to match score values corresponding to the feature dimensions according to a correspondence between the feature dimensions and the score values stored in the storage space in advance; determining a weighting parameter corresponding to each target category according to a corresponding relation between the target categories pre-stored in a storage space and the weighting parameters; and determining the consumption potential evaluation value of the target user according to the score value and the weighting parameter.

In an exemplary embodiment of the present disclosure, the evaluation module is configured to divide the obtained sample users into positive sample users and negative sample users; the sample user is a user who purchased a specified type of article in a historical time period; determining the target category of each sample user under M characteristic dimensions; determining chi-square values corresponding to the characteristic dimensions according to the number of positive sample users in each target category, the number of negative sample users in each target category, the total number of positive sample users, the total number of negative sample users, the number of sample users contained in each target category and the total number of sample users; determining the score value corresponding to each characteristic dimension according to the chi-square value and the total number of the sample users; and correspondingly storing the feature dimension and the score value into a storage space.

In an exemplary embodiment of the present disclosure, the evaluation module is configured to determine a first statistical value according to the number of positive sample users, the total number of positive sample users, the number of sample users included in each target category, and the total number of sample users in each target category; determining a second statistical value according to the number of negative sample users, the total number of the negative sample users, the number of sample users contained in each target category and the total number of the sample users in each target category; and determining the accumulated value of the first statistical value and the second statistical value as a chi-square value corresponding to each feature dimension.

In an exemplary embodiment of the present disclosure, the evaluation module is configured to classify the specified category of items into low-value items and high-value items; determining a sample user who purchases an over-priced item as a positive sample user; and determining a sample user who purchased the low value item as a negative sample user.

In an exemplary embodiment of the disclosure, the evaluation module is configured to obtain a first score of prices of a plurality of items corresponding to a specified type of item; determining the articles with the price of the articles larger than the first score as high-price articles; and determining the items with the price less than or equal to the first quantile as low-price items.

In an exemplary embodiment of the present disclosure, the evaluation module is configured to obtain a first ratio of the number of positive sample users to the total number of positive sample users in each target category; acquiring a second ratio of the number of the negative sample users in each target category to the total number of the negative sample users; acquiring a third ratio of the first ratio to the second ratio; determining the natural logarithm of the third ratio as a weighting parameter corresponding to each target category; and correspondingly storing the target category and the weighting parameter into a storage space.

And an item recommending module 703, configured to recommend a target item corresponding to the consumption potential evaluation value for the target user.

In an exemplary embodiment of the present disclosure, the item recommendation module is configured to obtain a product of the score value and the weighting parameter under each feature dimension; and determining the sum of the M products as the consumption potential evaluation value of the target user.

In an exemplary embodiment of the present disclosure, the item recommendation module is configured to obtain a second score of a plurality of consumption potential evaluation values corresponding to a plurality of target users; determining the consumption potential grade of the target user according to the size relation between the consumption potential evaluation value of the target user and the second score; and recommending and specifying the target item for the target user according to the pre-stored corresponding relation between the consumption potential grade and the target item.

The details of each module in the recommendation apparatus for an article have been described in detail in the recommendation method for a corresponding article, and therefore are not described herein again.

It should be noted that although in the above detailed description several modules or units of the device for action execution are mentioned, such a division is not mandatory. Indeed, the features and functionality of two or more modules or units described above may be embodied in one module or unit, according to embodiments of the present disclosure. Conversely, the features and functions of one module or unit described above may be further divided into embodiments by a plurality of modules or units.

Moreover, although the steps of the methods of the present disclosure are depicted in the drawings in a particular order, this does not require or imply that the steps must be performed in this particular order, or that all of the depicted steps must be performed, to achieve desirable results. Additionally or alternatively, certain steps may be omitted, multiple steps combined into one step execution, and/or one step broken down into multiple step executions, etc.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which may be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

In an exemplary embodiment of the present disclosure, there is also provided a computer storage medium capable of implementing the above method. On which a program product capable of implementing the above-described method of the present specification is stored. In some possible embodiments, aspects of the disclosure may also be implemented in the form of a program product comprising program code for causing a terminal device to perform the steps according to various exemplary embodiments of the disclosure described in the "exemplary methods" section above of this specification, when the program product is run on the terminal device.

Referring to fig. 8, a program product 800 for implementing the above method according to an embodiment of the present disclosure is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto, and in this document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

A computer readable signal medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable signal medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

Program code embodied on a readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Program code for carrying out operations for the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).

In addition, in an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided.

As will be appreciated by one skilled in the art, aspects of the present disclosure may be embodied as a system, method or program product. Accordingly, aspects of the present disclosure may be embodied in the form of: an entirely hardware embodiment, an entirely software embodiment (including firmware, microcode, etc.) or an embodiment combining hardware and software aspects that may all generally be referred to herein as a "circuit," module "or" system.

An electronic device 900 according to this embodiment of the disclosure is described below with reference to fig. 9. The electronic device 900 shown in fig. 9 is only an example and should not bring any limitations to the functionality or scope of use of the embodiments of the present disclosure.

As shown in fig. 9, the electronic device 900 is embodied in the form of a general purpose computing device. Components of electronic device 900 may include, but are not limited to: the at least one processing unit 910, the at least one storage unit 920, a bus 930 connecting different system components (including the storage unit 920 and the processing unit 910), and a display unit 940.

Wherein the storage unit stores program code that is executable by the processing unit 910 to cause the processing unit 910 to perform steps according to various exemplary embodiments of the present disclosure described in the above section "exemplary method" of the present specification. For example, the processing unit 910 may perform the following as shown in fig. 1: step S110, determining a target category to which a target user belongs under preset M characteristic dimensions; step S120, matching the score values corresponding to the feature dimensions according to the corresponding relation between the feature dimensions and the score values stored in the storage space in advance; step S130, determining a weighting parameter corresponding to each object category according to the corresponding relation between the object categories pre-stored in the storage space and the weighting parameters; step S140, determining a consumption potential evaluation value of the target user according to the score value and the weighting parameter; and step S150, recommending the target item corresponding to the consumption potential evaluation value for the target user.

The storage unit 920 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM)9201 and/or a cache memory unit 9202, and may further include a read only memory unit (ROM) 9203.

Storage unit 920 may also include a program/utility 9204 having a set (at least one) of program modules 9205, such program modules 9205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

Bus 930 can be any of several types of bus structures including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

The electronic device 900 may also communicate with one or more external devices 1000 (e.g., keyboard, pointing device, bluetooth device, etc.), with one or more devices that enable a user to interact with the electronic device 900, and/or with any devices (e.g., router, modem, etc.) that enable the electronic device 900 to communicate with one or more other computing devices. Such communication may occur via input/output (I/O) interface 950. Also, the electronic device 900 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet) via the network adapter 960. As shown, the network adapter 960 communicates with the other modules of the electronic device 900 via the bus 930. It should be appreciated that although not shown, other hardware and/or software modules may be used in conjunction with the electronic device 900, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which may be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

Furthermore, the above-described figures are merely schematic illustrations of processes included in methods according to exemplary embodiments of the present disclosure, and are not intended to be limiting. It will be readily understood that the processes shown in the above figures are not intended to indicate or limit the chronological order of the processes. In addition, it is also readily understood that these processes may be performed synchronously or asynchronously, e.g., in multiple modules.

Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

Claims

1. A method for recommending items, comprising:

determining a target category to which a target user belongs under preset M characteristic dimensions; the target users comprise new registered users, and M is a positive integer;

matching the score values corresponding to the feature dimensions according to the corresponding relation between the feature dimensions and the score values stored in a storage space in advance;

determining a weighting parameter corresponding to each target category according to a corresponding relation between the target categories pre-stored in the storage space and the weighting parameters;

determining a consumption potential evaluation value of the target user according to the scoring value and the weighting parameter;

recommending the target item corresponding to the consumption potential evaluation value for the target user.

2. The method according to claim 1, wherein the determining the target category to which the target user belongs under the preset M feature dimensions comprises:

acquiring user characteristic values of a target user under the M characteristic dimensions;

and determining the target category of the target user under each feature dimension according to the user feature value.

3. The method according to claim 1 or 2, wherein before determining the target category to which the target user belongs under the preset M feature dimensions, the method further comprises:

dividing the obtained sample users into positive sample users and negative sample users; the sample user is a user who purchased a specified type of article in a historical time period;

determining a target class to which each sample user belongs under the M characteristic dimensions;

determining a chi-square value corresponding to each characteristic dimension according to the number of positive sample users in each target category, the number of negative sample users in each target category, the total number of positive sample users, the total number of negative sample users, the number of sample users contained in each target category and the total number of sample users;

determining a scoring value corresponding to each characteristic dimension according to the chi-square value and the total number of the sample users;

and correspondingly storing the feature dimension and the score value into the storage space.

4. The method according to claim 3, wherein the determining a chi-squared value corresponding to each feature dimension according to the number of positive sample users, the number of negative sample users, the total number of positive sample users, the total number of negative sample users, the number of sample users included in each target category, and the total number of sample users in each target category includes:

determining a first statistical value according to the number of positive sample users, the total number of positive sample users, the number of sample users contained in each target category and the total number of sample users in each target category;

determining a second statistical value according to the number of negative sample users, the total number of the negative sample users, the number of sample users contained in each target category and the total number of the sample users in each target category;

and determining an accumulated value of the first statistical value and the second statistical value as the chi-square value corresponding to each feature dimension.

5. The method according to claim 3, wherein the dividing the obtained sample users into positive sample users and negative sample users comprises:

dividing the specified type of articles into low-price articles and high-price articles;

determining a sample user who purchased the high-priced item as the positive sample user; and the number of the first and second groups,

determining a sample user who purchased the low-priced item as the negative sample user.

6. The method of claim 5, wherein the classifying the specified categories of items into low-value items and high-value items comprises:

acquiring first scores of prices of a plurality of articles corresponding to the specified articles;

determining an item having an item price greater than the first score as the high-value item;

determining an item having an item price less than or equal to the first quantile as the low price item.

7. The method of claim 3, further comprising:

acquiring a first ratio of the number of positive sample users in each target category to the total number of the positive sample users;

acquiring a second ratio of the number of negative sample users in each target category to the total number of the negative sample users;

acquiring a third ratio of the first ratio to the second ratio;

determining the natural logarithm of the third ratio as a weighting parameter corresponding to each target category;

and correspondingly storing the target category and the weighting parameter into the storage space.

8. The method of claim 1, wherein determining the target user's consumption potential assessment value according to the score value and the weighting parameter comprises:

obtaining the product of the score value and the weighting parameter under each characteristic dimension;

determining a sum of the M products as a consumption potential assessment value of the target user.

9. The method of claim 1, wherein recommending the target item corresponding to the consumer potential assessment value for the target user comprises:

acquiring second scores of a plurality of consumption potential evaluation values corresponding to a plurality of target users;

determining the consumption potential grade of the target user according to the size relation between the consumption potential evaluation value of the target user and the second score;

and recommending and specifying the target item for the target user according to the pre-stored corresponding relation between the consumption potential grade and the target item.

10. An article recommendation device, comprising:

the determining module is used for determining the target category to which the target user belongs under the preset M characteristic dimensions; the target users comprise new registered users, and M is a positive integer;

the evaluation module is used for matching the score values corresponding to the feature dimensions according to the corresponding relation between the feature dimensions and the score values stored in the storage space in advance; determining a weighting parameter corresponding to each target category according to a corresponding relation between the target categories pre-stored in the storage space and the weighting parameters; determining a consumption potential evaluation value of the target user according to the scoring value and the weighting parameter;

and the item recommending module is used for recommending the target item corresponding to the consumption potential evaluation value for the target user.

11. A computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements a method of recommending items according to any of claims 1-9.

12. An electronic device, comprising:

a processor; and

a memory for storing executable instructions of the processor;

wherein the processor is configured to execute the recommendation method for an item of any of claims 1-9 via execution of the executable instructions.