WO2023284434A1 - Target information recommendation method and apparatus, and electronic device and storage medium - Google Patents

Abstract

The embodiments of the present disclosure relate to a target information recommendation method and apparatus, and an electronic device and a storage medium. The method comprises: acquiring a reference parameter set of at least one type of dimensional information and a plurality of alternative parameter sets of the at least one type of dimensional information of alternative information to be recommended, wherein the reference parameter set comprises a reference weight subset of the at least one type of dimensional information and a reference mapping value subset corresponding to the reference weight subset, and the alternative parameter sets comprise alternative weight subsets of the at least one type of dimensional information and alternative mapping value subsets corresponding to the alternative weight subsets; determining a target mapping value subset according to the difference between the sum of elements in any one alternative mapping value subset and the sum of elements in the reference mapping value subset, wherein each element in the alternative mapping value subset has a proportion coefficient, which is used for representing the degree of effect of the dimensional information on the determination of target information to be recommended from the alternative information; determining a target weight subset according to the target mapping value subset; and selecting said target information from the alternative information according to the target weight subset.

Description

Target information recommendation method and device, electronic device and storage medium

This disclosure claims the priority of the Chinese application with application number 202110807187.6 and the filing date is July 16, 2021; all the content of the Chinese application involved in the priority is the content of this disclosure.

technical field

The present disclosure relates to the field of information processing but is not limited to the field of information processing, and in particular relates to a method and device for recommending target information, electronic equipment, and a storage medium.

Background technique

With the development of science and technology, many emerging application technologies and application scenarios have emerged, and these application technologies can be adapted to be configured in relevant application scenarios. Information corresponding to the application scenario will be generated in different application scenarios, and by processing the information, the information can be better applied in related application scenarios and better application effects can be achieved.

For example, when recommending information, after processing relevant information through information processing technology, more matching information can be recommended to users.

Contents of the invention

Embodiments of the present disclosure provide a method and device for recommending target information, electronic equipment, and a storage medium.

The first aspect of the embodiments of the present disclosure provides a method for recommending target information, including: acquiring a reference parameter set of at least one dimension information of the candidate information to be recommended and multiple candidate parameter sets of the at least one dimension information ; Wherein, the reference parameter set includes: a reference weight subset of at least one dimension information and a reference mapping value subset corresponding to the reference weight subset; the candidate parameter set includes: the at least one dimension information The candidate weight subsets and the candidate mapping value subsets corresponding to each of the candidate weight subsets;

Determine the target mapping value subset from the candidate mapping value subset according to the difference between the first sum value and the second sum value; wherein, the first sum value is: any one of the candidate mapping value The sum of the elements in the subset; the second sum is the sum of the elements in the reference mapping value subset, and each element in the candidate mapping value subset has a proportional coefficient, and the proportional coefficient is configured to represent the The degree of influence of the dimension information on determining the recommended target information from the candidate information;

Determine a target weight subset according to the target mapping value subset; the target weight subset includes: a target weight affecting at least one dimension information of the candidate information;

Selecting the recommended target information from the candidate information according to the target weight subset.

The second aspect of the embodiments of the present disclosure provides an apparatus for recommending target information, the apparatus including:

The obtaining module is configured to obtain a reference parameter set of at least one dimension information of the candidate information to be recommended and a plurality of candidate parameter sets of the at least one dimension information; wherein the reference parameter set includes: at least one A reference weight subset of one dimension information and a reference mapping value subset corresponding to the reference weight subset; the candidate parameter set includes: the candidate weight subset of the at least one dimension information and each of the candidate the subset of alternative mapping values corresponding to the subset of weights;

A target mapping value subset determining module configured to determine a target mapping value subset from the candidate mapping value subset according to the difference between the first sum and the second sum; wherein the first and The value is: the sum of elements in any one of the candidate mapping value subsets; the second sum value is the sum of elements in the reference mapping value subset, and each element in the candidate mapping value subset has a proportional coefficient , the proportional coefficient is configured to represent the degree of influence of the dimensional information on determining recommended target information from the candidate information;

The target weight subset determination module is configured to determine a target weight subset according to the target mapping value subset; the target weight subset includes: target weights affecting at least one dimension information of the candidate information;

The recommendation module is configured to select the recommended target information from the candidate information according to the target weight.

A third aspect of the embodiments of the present disclosure provides an electronic device, including:

memory configured to store processor-executable instructions;

a processor connected to the memory;

Wherein, the processor is configured to execute the method for recommending target information as provided in any technical solution of the aforementioned first aspect.

The fourth aspect of the embodiments of the present disclosure provides a non-transitory computer-readable storage medium, wherein computer-executable instructions are stored in the computer-readable storage medium, and when the computer-executable instructions are executed by a processor, the aforementioned first A recommended method for target information provided by any technical solution.

In the embodiment of the present disclosure, when determining the first sum value, by adding a proportional coefficient to each element in the candidate mapping value subset, that is, before each candidate mapping value, the method in this embodiment can improve the accuracy of the target information. Recommendation accuracy, reducing the mutual influence and restriction between the weights corresponding to multiple dimension information, reducing the influence of the weight corresponding to one part of the dimension information on the weight corresponding to the other part of the dimension information, and reducing the weight corresponding to the other part of the dimension information The degree of sacrifice of the accuracy of determining the target information further improves the effect of the weight corresponding to another part of the dimension information on the accuracy of determining the target information, thus improving the weight corresponding to the multiple dimension information in determining the target information. Balanced, it is convenient to improve the accuracy of determining the target information.

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 present disclosure.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description serve to explain the principles of the invention.

Fig. 1 is a schematic flowchart of a method for recommending target information according to an exemplary embodiment;

Fig. 2 is a schematic flowchart of determining a subset of target mapping values according to an exemplary embodiment;

Fig. 3 is a schematic flow diagram of obtaining a set of candidate parameters according to an exemplary embodiment;

Fig. 4 is a schematic flow chart of updating a proportional coefficient according to an exemplary embodiment;

Fig. 5 is a schematic diagram showing a change trend of alternative mapping values in an application scenario according to an exemplary embodiment;

Fig. 6 is a schematic diagram showing a change trend of alternative mapping values in an application scenario according to an exemplary embodiment;

Fig. 7 is a schematic flowchart of another method for recommending target information according to an exemplary embodiment;

Fig. 8 is a block diagram of a terminal device according to an exemplary embodiment.

detailed description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with aspects of the invention as recited in the appended claims.

Usually, when recommending information, since there are many reference factors, that is, there are many factors affecting information recommendation results, multiple reference factors are combined to recommend the final information when recommending information. However, under normal circumstances, multiple reference factors will restrict and influence each other. When recommending information, it will be affected by one or several reference factors, which may cause the final recommended information to be affected by one or several of them. The influence of factors is large, and it is easy to fall into local optimum.

In this case, if there are reference factors that play a key role in the final recommended information among other reference factors, the degree of determination of the final recommended information by these key reference factors will be reduced, resulting in The determination degree of other reference information to the final recommended information is far less than that of this one or several reference factors, so that the relationship between multiple reference factors cannot be balanced according to multiple reference factors to achieve an overall optimal effect. As a result, the accuracy of the final recommended information is reduced, and the most suitable or matching recommendation information cannot be recommended, thereby reducing the recommendation effect and user experience.

Referring to FIG. 1 , it is a schematic flowchart of a method for recommending target information provided by an embodiment of the present disclosure. The data processing method includes the following steps:

Step S100, obtaining a reference parameter set of at least one dimension information of the candidate information to be recommended and multiple candidate parameter sets of the at least one dimension information; wherein, the reference parameter set includes: at least one dimension information The reference weight subset of the reference weight subset and the reference mapping value subset corresponding to the reference weight subset; the candidate parameter set includes: the candidate weight subset of the at least one dimension information and each of the candidate weight subsets The corresponding subset of alternative mapping values.

Step S200, according to the difference between the first sum value and the second sum value, determine the target mapping value subset from the candidate mapping value subset; wherein, the first sum value is: any one of the candidate mapping value subsets The sum of the elements in the selected mapping value subset; the second sum value is the sum of the elements in the reference mapping value subset, and each element in the candidate mapping value subset has a proportional coefficient, and the proportional coefficient is used to represent The degree of influence of the dimension information on determining the recommended target information from the candidate information.

Step S300, according to the subset of target mapping values, determine a subset of target weights; the subset of target weights includes: target weights affecting at least one dimension information of the candidate information.

Step S400, selecting the recommended target information from the candidate information according to the target weight subset.

The method can at least be executed in the mobile terminal, that is, the execution body of the method can at least include the mobile terminal. Mobile terminals can include mobile phones, tablet computers, vehicle-mounted central control devices, wearable devices, smart devices, etc., and smart devices can include smart office equipment and smart home devices.

The method of this embodiment can be executed by the execution terminal in various recommendation scenarios, and can be applied in a recommendation scenario in which a recommendation result is determined by one dimension information, or can be applied in a recommendation scenario in which a recommendation result is determined by multiple dimension information. In these recommendation scenarios, the candidate information to be recommended has multiple dimension information that can be referred to, and the final recommendation information can be determined according to these dimension information. Different dimensions of information have different proportions in determining whether a certain candidate information is the final recommendation information, so different dimensions of information have different reference values for determining the final recommendation information.

Of course, this method can also be executed by the execution terminal in various recommendation scenarios. For example, some candidate targets have multiple dimension information, as long as it is determined according to these different dimension information whether the candidate target is the final target scenario, all Within the protection scope of this embodiment.

Through this method, the optimal target weight corresponding to each dimension information can be determined, so as to determine the recommended target information according to the optimal target weight.

In this embodiment, the candidate information to be recommended includes at least one of the following: video, official account, article and/or product description information, and the like. Products here include, but are not limited to: physical goods and/or services. Of course, other types of content may also be included, which will not be listed one by one here, as long as the content including multiple dimensions of information can be used as candidate information to be recommended, all within the scope of protection of this embodiment.

In one embodiment, the dimensional information of the candidate information to be recommended includes at least one of the following: downloads, favorites, likes, number of users who follow the author of the candidate information, average browsing time, number of comments, and forwarding Quantity etc. Of course, different candidate information may have at least one dimension information matching the candidate information. In this embodiment, the dimension information of the candidate information to be recommended may be: generated according to historical operations on the candidate information by the user who has received the information.

In another embodiment, the dimensional information of the candidate information to be recommended includes at least one of the following: the difference between the amount of information of the information to be recommended and the amount of information preferred by the recommended user, the difference between the information content of the information to be recommended and the amount of information to be recommended The similarity between the recommended user's preferred content, the strength of the information to be recommended, etc.

In specific application scenarios, for example, when watching or browsing videos, official accounts and articles, videos, official accounts or articles may be downloaded, favorited, liked, commented and/or forwarded, etc. The author of the target message such as an account or an article may also be followed and so on. Downloads, favorites, likes, the number of users followed by the author of the alternative information, the average browsing time, the number of comments, and the number of reposts, etc., can reflect the user's awareness of the videos, official accounts and The degree of liking for content such as articles.

Different dimension information occupies different weights in the candidate information to be recommended. The weight of these dimension information in the candidate information can determine whether the candidate information is the target information, that is, the candidate information can be ranked according to these dimension information. The weight occupied by the target information is determined from the alternative information. When recommending follow-up videos, official accounts or articles to users, it can be based on information such as downloads, favorites, likes, comments, and/or forwarding of videos, official accounts, articles or other forms of content to be recommended. The target information that needs to be recommended is determined in the alternative information.

Therefore, the weights of different dimension information of the candidate information to be recommended play a key role in determining the target information. This embodiment is to determine the weight of the target information from the candidate information according to the weights of different dimension information of the candidate information illustrate.

For step S100, obtain a reference parameter set of at least one dimension information of the candidate information to be recommended, the reference parameter set is used as a basic parameter set for determining the target information, combine the information in the candidate parameter set with the reference parameter set In the information, determine the target information.

The reference parameter set includes: a reference weight subset of at least one dimension information and a reference mapping value subset corresponding to the reference weight subset. The reference weight subset is a collection of reference weights of at least one dimension information, that is, the reference weight subset includes reference weights of at least one dimension information. The reference mapping value subset is a set of mapping values corresponding to the reference weights of at least one dimension information, that is, the reference mapping value subset includes the mapping values of the reference weights of at least one dimension information. The set of reference parameters may be a predetermined set.

The reference weight subset can be expressed as Xbase={xbase1, xbase2, xbase3, ..., xbasen}, that is, the candidate information to be recommended has n dimension information, each dimension information has a corresponding reference weight, and n can be based on the The alternative information is determined. In this embodiment, each weight in the reference weight subset is referred to as an element of the reference weight subset.

For example, the candidate information to be recommended is a video, which includes seven dimensions of information: downloads, collections, likes, number of users who follow the author of the candidate information, average browsing time, number of comments, and number of reposts. Each dimension information has its own reference weight. For example, the reference weight subset is Xbase={xbase1, xbase2, xbase3, xbase4, xbase5, xbase6, xbase7}. The weight value corresponding to each dimension information in the reference weight subset, that is, the value of each element may be predetermined, that is, a reference value matching the candidate information to be recommended, or a randomly initialized value.

The reference mapping value subset may be expressed as Ybase={ybase1, ybase2, ybase3...,ybasen}, that is, the mapping values corresponding to the weights of the n dimension information. Each mapping value in the reference mapping value subset is referred to as an element of the reference mapping value subset, and the number of elements in the subset is equal to the number of elements in the reference weight subset. There is a corresponding mapping relationship between the reference mapping value and the reference weight, and the mapping relationship may be determined according to actual business requirements, or may be a preset mapping relationship.

For example, through a preset model or functional relationship, the corresponding mapping value can be determined according to the reference weight. For example, according to the preset model or functional relationship, the mapping value ybase1 of xbase1 can be determined, and the mapping value ybase2 of xbase2 can be determined. Each reference The weight corresponds to a reference mapping value, so that the mapping value of each element in the reference weight subset can be determined, thereby determining the reference mapping value subset. For example, Ybase={ybase1, ybase2, ybase3, ybase4, ybase5, ybase6, ybase7}.

In step S100, it also includes: obtaining a plurality of candidate parameter sets of at least one dimension information, that is, acquiring a plurality of candidate reference sets, and each candidate parameter set includes: candidate weights of at least one dimension information set and a subset of alternative mapping values corresponding to each subset of alternative weights.

The number of candidate weights included in the candidate weight subset is the same as the number of reference weights of the dimension information included in the reference weight subset, and each candidate weight in the candidate weight subset is also used as each element of the candidate weight subset. That is, the number of elements included in the candidate weight subset is the same as the number of elements in the reference weight subset.

In this embodiment, the number of candidate weight subsets may be determined according to actual requirements, for example, three or five. The candidate weight subsets may be expressed as X1, X2, X3...Xt, that is, t candidate weight subsets are included. Each candidate weight subset includes the same number of elements as the number of elements in the reference map value subset, for example, 7 elements.

For example, to obtain 5 candidate weight subsets X1, X2, X3, X4 and X5, X1={x11, x12, x13, x14, x15, x16, x17}, X2={x21, x22, x23, x24, x25 , x26, x27}, x3 = {x31, x32, x33, x34, x35, x36, x37}, x4 = {x41, x42, x43, x44, x45, x46, x47}, x5 = {x51, x52, x53 , x54, x55, x56, x57}.

For any two candidate weight subsets, when the candidate weight values corresponding to at least one dimension information in the candidate weight subsets are different, the two candidate weight subsets are different. For information of the same dimension, the weights included in any two different candidate weight subsets may be the same or may not be the same. In this embodiment, for the same dimensional information, the weights included in any two different candidate weight subsets are different as an example. When different candidate weight subsets are used as target weight subsets, the determined target information is different. By determining the target weight subsets from these different candidate weight subsets, the optimal target weight subset can be determined, thereby determining the recommended The target information with the highest accuracy improves the accuracy of recommendation.

Each candidate weight in the candidate weight subset can be predetermined, or can be a candidate weight matched from the database and matched with the candidate information to be recommended, of course, it can also be determined by other methods, and will not be limited here .

The candidate mapping value subsets corresponding to the candidate weight subsets can be expressed as Y1, Y2, Y3...Yt, and the number of candidate mapping value subsets is the same as the number of candidate weight subsets. If there are 5 candidate weight subsets, then there are 5 candidate mapping value subsets, there is a mapping relationship between Y1 and X1, there is a mapping relationship between Y2 and X2, ... there is a mapping relationship between Y5 and X5. The number of mapping values included in each candidate mapping value subset is the same as the number of weights in the candidate weight subset, that is, the number of elements in the candidate mapping value subset is the same as the number of elements in the candidate weight subset.

For example, Y1={y11, y12, y13, y14, y15, y16, y17}, Y2={y21, y22, y23, y24, y25, y26, y27}, Y3={y31, y32, y33, y34, y35 , y36, y37} and so on, Y4 and Y5 and so on. There is a mapping relationship between y11 and x11, there is a mapping relationship between y12 and x12, there is a mapping relationship between y13 and x13, ... there is a mapping relationship between y17 and x17. There is a mapping relationship between y21 and x21, there is a mapping relationship between y22 and x22, there is a mapping relationship between y23 and x23, ... there is a mapping relationship between y27 and x27, and so on.

In this embodiment, the candidate mapping value subset may be determined according to the candidate weight subset, or the candidate mapping value subset may be determined according to the candidate weight subset by using a preset model or functional relationship. The preset model or functional relationship here may be the same as the preset model or functional relationship used to determine the reference mapping value subset according to the reference weight subset.

The subset of candidate weights and the subset of candidate mapping values may be used to determine the subset of target weights, thereby determining target information.

For step S200, the target mapping value subset can be determined through this step.

The first sum value is: the sum of elements in any candidate mapping value subset. When determining the first sum value, each element in each candidate mapping value subset has a proportional coefficient, which is used to represent the corresponding The degree of influence of dimension information on determining recommended target information from alternative information. The first sum value is determined according to each element in the candidate mapping value subset and the proportionality factor of each element.

For example, the alternative mapping value subset Y1={y11, y12, y13, y14, y15, y16, y17}, when determining the first sum value, each element in the alternative mapping value subset Y1 has a proportional coefficient, such as Y1 ={a*y11, b*y12, c*y13, d*y14, e*y15, f*y16, g*y17}, a, b, c, d, e, f and g are proportional coefficients. The first sum value corresponding to Y1 can be expressed as sum(a*y11, b*y12, c*y13, d*y14, e*y15, f*y16, g*y17). Each element in other candidate mapping value subsets is the same, and also has a proportional coefficient, such as Y2={a*y21, b*y22, c*y23, d*y24, e*y25, f*y26, g*y27} , the first sum corresponding to Y2 is sum(a*y21, b*y22, c*y23, d*y24, e*y25, f*y26, g*y27).

These proportional coefficients are used to represent the dimensional information corresponding to the weights that have a mapping relationship with the mapping values, and the weights occupied when determining the target information from the candidate information to be recommended. The larger the proportional coefficient, the greater the weight. , the greater the degree of influence.

The second sum value is: the sum of the elements in the reference map value subset. When determining the first sum value, the elements in the reference mapping value subset can be directly added. For example, sum(ybase1, ybase2, ybase3, ybase4, ybase5, ybase6, ybase7).

Then, according to the difference between the first sum and the second sum, a target subset of mapped values is determined from the subset of candidate mapped values. The specific determination method is not limited, for example, reference may be made to subsequent embodiments, for example, the candidate mapping value subset with the largest difference may be used as the target mapping value subset, and so on.

For example, the first sum value sum(a*y51, b*y52, c*y53, d*y54 , e*y55, f*y56, g*y57) and the second sum value sum (ybase1, ybase2, ybase3, ybase4, ybase5, ybase6, ybase7), determine the target mapping value subset as Y5={y51, y52, y53, y54, y55, y56, y57}.

For step S300,

After the target mapping value subset is determined, a target weight subset is determined according to the determined target mapping value subset. Since the target mapping value subset is selected from the candidate mapping value subsets, and each candidate mapping value subset corresponds to an alternative weight subset that has a mapping relationship, the corresponding existence of the target mapping value subset has a mapping A subset of alternative weights for the relationship, as a subset of target weights. That is, the candidate weight subset corresponding to the target mapping value subset is determined as the target weight subset.

The target weight subset includes: a target weight affecting at least one dimension information of the candidate information. The target weight subset includes target weights corresponding to at least one dimension information, and the target weights can affect the determination of target information from candidate information, that is, influence whether a certain candidate information is enough target information.

For example, the target mapping value subset is Y5={y51, y52, y53, y54, y55, y56, y57}, and the alternative weight subset due to the mapping relationship with Y5 is X5={x51, x52, x53, x54, x55, x56, x57}, so X5 is the target weight subset.

When determining the first sum, each element in the candidate mapping value subset, that is, adding a proportional coefficient before each candidate mapping value, on the one hand retains the Bayesian algorithm’s ability to solve the waste of computing time and space, and on the other hand On the one hand, it also reduces the problem that the Bayesian algorithm is easy to fall into local optimum, and the weights of multi-dimensional information are mutually restricted and influenced, resulting in inaccurate determined target information, thereby improving the accuracy of determined target information.

For step S400,

After the target weight subset is determined, recommended target information can be selected from candidate information according to the target weight subset. Since the target weight subset includes the weight of at least one dimension information, the target information can be determined from candidate information to be recommended according to the weight of at least one dimension information.

For example, if the target weight subset is X5={x51, x52, x53, x54, x55, x56, x57}, then according to each element in the target weight subset, that is, according to the weights corresponding to the information of the seven dimensions, from the recommended Select the video to determine the target video.

Through the method in this embodiment, the recommendation accuracy of target information can be improved, the mutual influence and restriction between the weights corresponding to multiple dimension information can be reduced, and the weight corresponding to some dimension information can be reduced to the weight corresponding to another part of dimension information. , which reduces the sacrifice degree of the weight corresponding to another part of the dimension information on the accuracy of determining the target information, and then improves the effect of the weight corresponding to the other part of the dimension information on the accuracy of determining the target information, thereby improving The balance of the weights corresponding to the multi-dimensional information when determining the target information facilitates the improvement of the accuracy of determining the encoded information.

In another embodiment, when determining the first sum value, the proportional coefficients of each element in the candidate mapping value subset may be a preset value, and the preset value may be fixed. The specific value of the proportional coefficient is determined according to actual business requirements, that is, according to the candidate information to be recommended. For example, according to the recommendation requirements of the candidate information to be recommended, if the degree of influence of some of the dimension information on the determination of the target information is greater than that of the other part of the dimension information on the determination of the target information, the weight corresponding to a part of the dimension information will be adjusted to a certain part of the dimension The scaling factor of the mapping value corresponding to the information is set to be larger than the scaling factor of the mapping value corresponding to another part of dimension information.

Referring to FIG. 2, it is a schematic flow chart for determining the target mapping value subset, step S200, according to the difference between the first sum value and the second sum value, determining the target mapping value subset from the candidate mapping value subset, including:

Step S201, determining a first sum value according to the proportional coefficients of each element in the candidate mapping value subset.

The first sum value is: the sum of elements in any candidate mapping value subset. When determining the first sum value, each element in each candidate mapping value subset has a proportional coefficient, which is used to represent the corresponding The degree of influence of dimension information on determining recommended target information from alternative information. The first sum value is determined according to each element in the candidate mapping value subset and the proportionality factor of each element.

For example, the alternative mapping value subset Y1={y11, y12, y13, y14, y15, y16, y17}, when determining the first sum value, each element in the alternative mapping value subset Y1 has a proportional coefficient, such as Y1 ={a*y11, b*y12, c*y13, d*y14, e*y15, f*y16, g*y17}, a, b, c, d, e, f and g are proportional coefficients. The first sum value corresponding to Y1 can be expressed as sum(a*y11, b*y12, c*y13, d*y14, e*y15, f*y16, g*y17). Each element in other candidate mapping value subsets is the same, and also has a proportional coefficient, such as Y2={a*y21, b*y22, c*y23, d*y24, e*y25, f*y26, g*y27} , the first sum corresponding to Y2 is sum(a*y21, b*y22, c*y23, d*y24, e*y25, f*y26, g*y27).

These proportional coefficients are used to represent the dimensional information corresponding to the weights that have a mapping relationship with the mapping values, and the weights occupied when determining the target information from the candidate information to be recommended. The larger the proportional coefficient, the greater the weight. , the greater the degree of influence.

Step S202, according to the difference between the first sum and the second sum, determine the candidate mapping value subset with the largest difference, or the candidate mapping value subset whose difference is greater than the preset difference, as the target mapping value Subset.

After determining the difference between the first sum and the second sum, since the selection of the optimal target mapping subset is determined by maximizing the difference, in this embodiment, the candidate mapping with the largest difference The value subset is determined as the target mapping value subset, so that an optimal mapping value subset among all candidate mapping value subsets can be determined, and the optimal mapping value subset is used as the target mapping value subset.

It is also possible to use the candidate mapping value subset whose difference is greater than the preset difference as the target mapping value subset. The preset difference can be determined according to the candidate information and match the preset difference with the candidate information. When the difference between the first sum value and the second sum value is greater than the preset difference value, the optimal mapping value subset can be determined among the candidate mapping value subsets, and the optimal mapping value subset can be determined as the target mapping value Subset.

In this embodiment, the target information may be determined according to the target weight subset corresponding to the optimal mapping value subset.

In another embodiment, referring to FIG. 3 , which is a schematic flowchart of obtaining a set of candidate parameters, the step S100 of obtaining multiple set of candidate parameters of the at least one dimension information includes:

Step S101, preset T initialization parameter sets, the initialization parameter sets include: an initialization weight subset of at least one dimension information and an initialization mapping value subset corresponding to the initialization weight subset.

In this step, T initialization parameter sets can be preset in a preset manner, and each initialization parameter set includes an initialization weight subset of at least one dimension information and an initialization mapping value subset corresponding to the initialization weight subset Set, that is, there is a mapping relationship between the initialized weight subset and the initialized map value subset.

Each initialization weight subset includes weights of at least one dimension information, and the number of initialization weights in the subset is the same as the number of weights in the reference weight subset. Each initialization mapping value subset includes mapping values corresponding to weights of at least one dimension information, and the number of mapping values in the subset is the same as the number of initialization weights in the initialization weight subset.

The number of T can be set according to actual needs, for example, it can be a positive integer such as 3 or 5.

For example, the initialization weight subsets are X6, X7, X8, X9 and X10, the initialization mapping value subsets are Y6 to Y10, there is a mapping relationship between X6 and Y6, there is a mapping relationship between X7 and Y7, and so on. X6={x61, x62, x63, x64, x65, x66, x67}, X7={x71, x72, x73, x74, x75, x76, x77}, X3, X4 and X5 and so on. Y6={y61, y62, y63, y64, y65, y66, y67}, Y7={y71, y72, y73, y74, y75, y76, y77} and so on, Y8, Y9 and Y10 and so on. There is a mapping relationship between Y61 and x61, there is a mapping relationship between y62 and x62, there is a mapping relationship between y63 and x73, ... there is a mapping relationship between y67 and x67. There is a mapping relationship between Y71 and x71, there is a mapping relationship between y72 and x72, there is a mapping relationship between y73 and x73, ... there is a mapping relationship between y77 and x77, and so on.

Step S102, according to the third sum of each element in the initialization weight subset in each initialization parameter set, and the fourth sum of each element in each initialization mapping value subset, determine the T+1th parameter set; the T+1th The parameter set includes: a T+1th weight subset and a T+1th mapping value subset, where T is a positive integer greater than 1.

In this step, according to each element in each initialization weight subset, that is, the initialization weight, the sum of all initialization weights in each initialization weight subset is determined as the third sum value. According to each element in each subset of initialization mapping values, that is, the initialization mapping value, the sum of all initialization mapping values in each subset of initialization mapping values is determined as the fourth sum value.

Since there is a mapping relationship between the initialization weight subset and the initialization mapping value subset, each initialization weight subset corresponds to an initialization mapping value subset, and the initialization weight subset and the initialization weight subset with mapping relationship can be used as a set parameter. Different initialization parameter sets include a subset of initialization weights and a subset of initialization mapping values, so the third sum corresponding to a subset of initialization weights and the fourth sum corresponding to a subset of initialization mapping values can form The coordinates of an initialization parameter in the coordinate system. The third sum corresponds to the X-axis, and the fourth sum corresponds to the Y-axis.

Predict the coordinates of the T+1th parameter set according to the development trend of the coordinate positions from the first initial parameter set to the Tth initial parameter set in the known T initialization parameter sets. The T+1th parameter set includes a T+1th weight subset and a T+1th mapping value subset, and there is a mapping relationship between the T+1th mapping value subset and the T+1th weighting subset.

The specific prediction method can be determined according to the Gaussian distribution and the conditional probability distribution of each subset of mapping values. For example:

By adjusting x in X, the optimal target information can be determined from the alternative information, and whether the target information is optimal is evaluated by Y corresponding to X, that is, the subset of mapping values, such as Y takes an extreme value (in a certain interval maximum or minimum value). The recommendation method of target information can be realized through the recommendation model. In the recommendation scene, auc and gauc are generally used to evaluate the accuracy of the recommendation model of target information. Through the equation f of auc and gauc, f(X)=Y can be used to evaluate The recommendation accuracy of the model. Therefore, when f is determined, it is necessary to determine a set of X to maximize Y.

By constructing a surrogate function of the f objective function, the surrogate function can obtain the optimal solution by evaluating the score of the surrogate function on X. Assuming that X and Y obey the Gaussian distribution, the posterior probability distribution of (X,Y) can be fitted through the Gaussian distribution, then we can predict the value of Y based on X. Every time Y obtained according to X can correct the prediction of the posterior probability distribution. Of course, the more samples are taken, the closer the obtained distribution is to the real one, but the cost of calculation is huge. If any infinite sampling is performed, time-consuming and traversal The time-consuming of X to find Y is close.

So in this embodiment, a sampling limit is set, called the harvest function. , which can be max(mean+var). mean represents the mean value, and the higher the mean value, the region sampling where the global optimal solution is most likely to appear. var represents the variance, and the higher the variance, the more likely the global optimal solution is in the unsampled area, that is, to obtain sampling points in the unsampled area. Blindly pursuing the global optimal solution, the optimal X will linger at the points around X, which is not conducive to quickly fitting the real distribution. Therefore, it is also necessary to add variance to expand the variance and find some unsampled areas. According to the harvest function max(mean+var) plus the constraint relationship between mean and var, the optimal X under the current distribution can be determined, and Y can be obtained through f, and (X, Y) is the predicted information. Use To update the data distribution, continue to iterate.

Similarly, through the function Y=f(X), the function between y and x is also satisfied, which is equivalent to optimizing y1+y2+y3.....yn=f(x1)+f(x2)+f(x3)... ..f(xn), the x included in X can be determined by using the function f, and after x is determined, y can be determined according to the function f. The specific process will not be described in detail here, refer to the Bayesian optimization algorithm.

The sum of each weight in the T+1th weight subset of the T+1th parameter set is a local maximum value, and the sum of each mapping value in the T+1th mapping value subset is a local maximum value.

Step S103, according to the T+n parameter sets, determine the T+n+1th parameter set; n is a positive integer, and the sum of the elements in the T+nth mapping value subset is a local maximum value.

The T+1th parameter set can be determined according to the known T initialization parameter sets, that is, according to the third sum corresponding to the initialization weight subset in the known T initialization parameter sets and the first corresponding to the initialization mapping value subset The four sum values can determine the positions in the coordinate system of the third sum value and the fourth sum value respectively corresponding to the first initialization parameter set to the Tth initialization parameter set. According to the Gaussian distribution, the position of the sum value corresponding to the weight subset in the T+1th parameter set and the sum value corresponding to the mapping value subset in the coordinate system can be predicted, so as to determine the weight subset in the T+1th parameter set The corresponding sum value and the sum value corresponding to the mapping value subset, and then determine the weight subset and mapping value subset corresponding to the T+1th parameter set.

According to the third sum value and the fourth sum value corresponding to the T initialization parameter set and the T+1th parameter set respectively, determine the sum value of the weight subset and the sum value of the mapping value subset in the T+2th parameter set , so as to determine the weight subset and mapping value subset corresponding to the T+2th parameter set.

By analogy, the sum value corresponding to the weight subset in the T+nth parameter set and the sum value corresponding to the mapping value subset can be based on the sum value and mapping corresponding to the weight subset in the first T+n-1 parameter set The sum value corresponding to the subset of values is determined. According to the sum value corresponding to the weight subset in the first T+n parameter sets and the sum value corresponding to the mapping value subset, determine the corresponding sum value corresponding to the weight subset in the T+n+1 parameter set and the mapping value subset correspondence , so as to determine the T+n+1th parameter set.

In this step, n may be determined according to actual requirements, for example, may be a positive integer ranging from 1 to 10000. When determining the T+nth parameter set, the sum of the weights of the weight subsets in each parameter set is a local maximum value, and the sum of the mapping values in the mapping value subset is also a local maximum value.

Step S104, determining T+n+1 parameter sets as candidate parameter sets.

After determining T+n+1 parameter sets, determine these T+n+1 parameter sets as alternative parameter sets, and determine the target parameter set from these T+n+1 parameter sets, thereby determining the target mapping value Subset.

In another embodiment, T+n parameter sets may also be determined as candidate parameter sets.

In another embodiment, step S101, preset T sets of initialization parameters, including:

According to the historical parameter set of the candidate information, determine T initialization parameter sets that match the candidate information; wherein, the historical parameter set includes: a historical weight subset of at least one dimension information and a historical mapping value sub-set corresponding to the historical weight subset set.

The number of elements in the historical weight subset and the historical mapping value subset included in the historical parameter set may be equal to the number of elements in the reference weight subset and the reference mapping value subset.

By determining T initialization parameter sets in the historical parameter sets, the degree of matching between the initialization parameter sets and candidate information can be improved, thereby improving the accuracy of the determined target information.

or,

Randomly generate T sets of initialization parameters according to the candidate information to be recommended.

In another embodiment, the method further includes: updating the scaling factor according to the subset of candidate mapping values.

In this embodiment, the proportional coefficient can be changed dynamically. By dynamically adjusting the proportional coefficient, according to the change of the sum of the mapped values in the candidate mapped value subset, the proportional coefficient of each element in the candidate mapped value subset can be dynamically updated. In this way, the degree of influence of at least one dimension information on the target information is balanced, and the mutual constraints between the target weights corresponding to multiple dimension information are reduced, resulting in a situation where the accuracy of the determined target information is low, thereby improving the accuracy of determining the target information Spend.

In another embodiment, refer to FIG. 4 , which is a schematic flow chart of updating the proportional coefficient. Update the scale factor based on a subset of alternative mapping values, including:

Step S10, according to the plurality of candidate mapping values corresponding to each dimension information, determine the change trend of the plurality of candidate mapping values respectively corresponding to each dimension information.

After determining the candidate mapping value subsets in the plurality of candidate parameter sets corresponding to at least one dimension information, each mapping value in the candidate mapping value subsets in each candidate parameter set can be determined, and each candidate mapping value subset Both include mapping values corresponding to each dimension information, so that multiple mapping values corresponding to each dimension information can be determined, and thus the change trend of each mapping value corresponding to each dimension information can be obtained. That is, each dimension information corresponds to a change trend of multiple mapping values corresponding to the dimension information.

Step S20, adjust the proportional coefficient according to each change trend until each change trend reaches the target range; wherein, the change trends corresponding to different dimension information have corresponding target ranges respectively. The target range is: a threshold range determined according to a proportion of dimension information in selecting target information from candidate information. That is, each change trend corresponds to a threshold range, and the threshold range is determined by the proportion of information in different dimensions when selecting target information from candidate information.

After the change trend corresponding to each dimension information is determined, the proportional coefficient of the candidate mapping value corresponding to each dimension information can be adjusted according to the change trend, and the different change trends can be adjusted to the corresponding target range. In this way, the degree of influence of at least one dimensional information on the target information can be balanced, and the mutual constraints between the target weights corresponding to multiple dimensional information can be reduced, resulting in a situation where the accuracy of the determined target information is low, thereby improving the accuracy of determining the target information. Accuracy.

For example, the seven dimension information of downloads, favorites, likes, number of users followed by the author of alternative information, average browsing time, number of comments, and number of retweets correspond to thresholds ranging from 0 to 0.02% and 0.1% respectively to 0.3%, 0.2% to 0.5%, 0.3% to 0.7%, 0.1% to 0.6%, -0.9% to 0 and 0.05% to 0.08%, etc. When the change rate corresponding to the change trend reaches the corresponding threshold range, the adjustment of the proportional coefficient can be realized. When the rate of change is outside the corresponding threshold range, the corresponding rate of change is adjusted to be within the corresponding threshold range.

In another embodiment, it may also be: according to the change trend, when the multiple candidate mapping values corresponding to each dimension information exceed the corresponding threshold range, adjust the proportional coefficients of the candidate mapping values corresponding to each dimension information, and The candidate mapping values respectively corresponding to the dimension information are adjusted within a threshold range, and the threshold range in this embodiment is the numerical range of the candidate mapping values.

In another embodiment, referring to FIG. 5 , it is a schematic diagram of an apparatus for recommending target information, which includes:

The acquisition module 1 is configured to acquire a reference parameter set of at least one dimension information of the candidate information to be recommended and a plurality of candidate parameter sets of the at least one dimension information; wherein, the reference parameter set includes: at least A reference weight subset of dimension information and a reference mapping value subset corresponding to the reference weight subset; the candidate parameter set includes: the candidate weight subset of the at least one dimension information and each of the candidate parameters Select a subset of alternative mapping values corresponding to the subset of weights;

The target mapping value subset determination module 2 is configured to determine the target mapping value subset from the candidate mapping value subset according to the difference between the first sum value and the second sum value; wherein, the first The sum value is: the sum of elements in any one of the candidate mapping value subsets; the second sum value is the sum of elements in the reference mapping value subset, and each element in the candidate mapping value subset has a ratio A coefficient, the proportional coefficient is used to represent the degree of influence of the dimension information on determining the recommended target information from the candidate information;

The target weight subset determining module 3 is configured to determine a target weight subset according to the target mapping value subset; the target weight subset includes: a target weight affecting at least one dimension information of the candidate information;

The recommendation module 4 is configured to select the recommended target information from the candidate information according to the target weight.

In another embodiment, the proportional coefficient is a preset value;

Target mapping value subset determination module 2, including:

A first sum value determination unit configured to determine the first sum value according to the proportionality coefficient of each element in the candidate mapping value subset;

The target mapping value subset determining unit is configured to, according to the difference between the first sum and the second sum, select the candidate mapping value subset with the largest difference, or the difference The candidate mapping value subset greater than a preset difference is determined as the target mapping value subset.

In another embodiment, the device also includes:

A scaling factor updating module configured to update the scaling factor according to the subset of candidate mapping values.

The scale factor update module includes:

The change trend determination unit is configured to determine the change trend of the plurality of candidate mapping values corresponding to each dimension information respectively according to the plurality of candidate mapping values corresponding to each dimension information;

The adjustment unit is configured to adjust the proportional coefficient according to each change trend until each change trend reaches a target range; wherein the change trends corresponding to different dimensional information have corresponding target ranges respectively.

In another embodiment, the target range is: a threshold range determined according to a proportion of the dimension information when selecting the target information from the candidate information.

Acquisition Module 1 includes:

The initialization parameter set acquisition unit is configured to preset T initialization parameter sets, the initialization parameter sets include: the initialization weight subset of the at least one dimension information and the initialization mapping value subset corresponding to the initialization weight subset ;

The T+1th parameter set determining unit is configured to, according to the third sum of each element in the initialization weight subset in each initialization parameter set, and the fourth sum of each element in each of the initialization mapping value subsets, Determine the T+1th parameter set; wherein, the T+1th parameter set includes: the T+1th weight subset and the T+1th mapping value subset, and T is a positive integer greater than 1;

The T+n+1th parameter set determining unit is configured to determine the T+n+1th parameter set according to the T+n parameter set; n is a positive integer, and in the T+nth mapping value subset The sum of each element is a local maximum;

The candidate parameter set determining unit is configured to determine the T+n+1 parameter sets as the candidate parameter sets.

The initialization parameter set acquisition unit is also configured as:

According to the historical parameter set of the candidate information, determine T initialization parameter sets matching the candidate information; wherein, the historical parameter set includes: a historical weight subset of at least one dimension information and the historical weight A subset of historical mapping values corresponding to the subset;

or,

Randomly generate T initialization parameter sets according to the candidate information to be recommended.

In another embodiment, the target weight subset determination module 3 is further configured to:

A candidate weight subset corresponding to the target mapping value subset is determined as the target weight subset.

In another embodiment, the candidate information to be recommended includes at least one of the following:

video;

the public;

article;

The dimension information includes at least one of the following:

The number of downloads, favorites, likes, the number of users who follow the author of the alternative information, the average browsing time, the number of comments, and the number of reposts.

An embodiment of the present disclosure provides an electronic device, including:

memory for storing processor-executable instructions;

Processor, memory connection respectively;

Wherein, the processor is configured to execute the video processing method provided by any of the foregoing technical solutions by executing the computer-executable instructions stored in the memory.

The processor may include various types of storage media, which are non-transitory computer storage media, and can continue to memorize and store information thereon after the mobile terminal is powered off.

The processor may be connected to the memory through a bus, etc., and configured to read the executable program stored on the memory, for example, at least one of the methods shown in any one of Figures 1 to 4 and the method shown in Figure 7 one.

In another embodiment, another method for recommending target information is provided.

With the rapid development of big data technology and machine learning technology, the application of machine learning has become more abundant and complex, resulting in a large number of parameter adjustments that not only require more machine learning background and business knowledge of algorithms, especially now that the model update is changing with each passing day. Rapid product iteration. The demand for automatic parameter adjustment is becoming more and more intense, but traditional automatic parameter adjustment cannot completely solve the problem of simultaneously adjusting multiple parameters, and the adjustment results of multiple interactive parameter adjustments are often unsatisfactory. Therefore, the present invention was conceived to solve Under the influence of multiple parameters, reasonable hyperparameters can also be effectively selected.

Usually, the automatic parameter tuning methods used include grid search, random search, and Bayesian optimization. However, there are various problems in these methods: grid search is often not suitable for rapid product iteration due to the waste of a lot of time and space; random search is likely to miss the optimal point; traditional Bayesian optimization, Although a hyperparameter black box is provided, it is easy to fall into a local optimum under multiple interacting parameters, or it is difficult to find a balance to optimize the overall effect, especially in multi-parameter scenarios, multiple If the parameters restrict each other and affect each other, the optimal hyperparameters selected by the Bayesian algorithm are often not effective.

In this process, Bayesian optimization depends on the objective function y=f(x), that is, Max(sum(y1,y2…..yt)-sum(ybase1,ybase2,…ybaset)), but In the adjustment of multiple parameters of the traditional objective function, Bayesian often seeks the optimal point, which is not the optimal result of the effect in many scenarios. It can be seen that the above expression seeks to maximize the total value. In the scenario where multiple parameters are independent of each other, the ideal effect can be achieved. However, in many scenarios, the parameters are mutually restricted. Under the influence of this goal, often The effect brought by some parameters will be sacrificed. In some cases, the effect brought by these sacrificed parameters will play a crucial role in the overall improvement. Therefore, a strong constraint algorithm is defined in combination with Bayesian optimization, to The blindness of the improved objective equation avoids obtaining local optimum or effect deviation. In this process, the interaction of parameters can be strongly restricted, and effective business experience can be effectively transformed into solvable equations for repeated use.

First look at the optimization of a single parameter. Generally, the objective equation Ysingle=f(Xsingle) is defined. By finding the optimal parameter xt, the result yt is the optimal solution, that is, max(yn)n=1, 2, 3 is obtained. …..t., for multiple parameters, the maximum of the sum of multiple Ysingles will be found, that is, a set of parameters Xt={x1,x2…..xn} will be found to maximize the result, and Max(Yt) will be found . Max(Yt)=Max(y1+y2+y3...yt). Here the two problems mentioned before are exposed. One is that the rate at which each parameter yt changes with the change of xt is different, which will cause more weights to be assigned to the faster-changing yt, that is, for a change Faster (equation with greater slope g(x)) g(x+1) will increase more than the value of f(x+1) with smaller slope, so Bayesian will be more willing to adjust the larger slope g(x). The second is that it is difficult for business experience to fall into the Bayesian equation, so no matter what kind of automatic parameter tuning, it is impossible to learn business experience.

This embodiment provides a combination method based on Bayesian principle and strong constraint algorithm. On the one hand, it retains Bayesian ability to solve the waste of calculation time and space. On the other hand, the strong constraint algorithm of parameters solves the problem of Bayesian It is easy to fall into local optimum, and finally it is difficult to achieve the ideal income. The overall effect of the optimized hyperparameters caused by multiple parameters due to mutual constraints and influences is not increased but decreased. At the same time, the usable business experience is transformed into usable The solved equation is convenient for repeated use.

In order to achieve the above object, the present invention adopts following technical scheme, and concrete steps are as follows:

Step 1: Manually select a set of hyperparameters X, or randomly initialize a set of hyperparameters X, and calculate the corresponding score of each target for the hyperparameter set Xbase={xbase1,xbase2,xbase3,...,xbasen} Ybase={ybase1, ybase2, ybase3...,ybasen}. This constitutes the basic group (Xbase, Ybase), which is used for comparison with the experimental group.

Step 2: Define the optimization goal. Every time the hyperparameter X selected in the optimization process will get the target score Y={y1,y2,y3...,yn}, it is necessary to expand the sum of the target score Y and the sum of the basic group The distance between them, that is, Max(sum(y1,y2…..yn)-sum(ybase1,ybase2,…ybasen)), choose the one greater than 0, (select the experimental group) This is the traditional business scenario selection and delivery target equation The form is mainly improved here, see the detailed explanation in the notes of the next steps for details. (Not necessary, can be used to determine step 5, stop iteration)

Step 3: In the initialization stage, Bayesian will randomly generate several sets of hyperparameters X1, X2, X3...Xt, and their corresponding target score results combined with Y1, Y2, Y3, Y4...Yt. Bayesian believes that it obeys the Gaussian distribution, so the data set D={(X1,Y1),(X2,Y2)...(Xt,Yt)} and the conditional probability distribution of Y after updating D can be constructed.

Step 4: Bayesian uses the Gaussian distribution initialized in Step 3, and the objective equation defined in Step 2 to find a balance between development and exploration, and find the next set of parameters that maximize Y (the sum of y) Dt = (Xt, Yt), and then add Dt=(Xt, Yt) to the total set D. It can stop when the number of iterations t reaches a certain number of times.

Step 5: The maximum parameter aggregation Dt=(Xt,Yt) found in each iteration will update the original data set D, the change of the original data set brings the new conditional probability distribution Y, and the new distribution can be used The next optimal solution needs to be found. Then iterate step four until the final convergence, so that the posterior distribution is close to the real distribution. Get the best hyperparameters.

In another embodiment, refer to FIG. 6 , which is a schematic diagram of a change trend of alternative mapping values in an application scenario.

Among them, p1, p2, p3, p4, p5, and p6 are 6 different dimensions of information, the abscissa indicates the number of alternative mapping values, and the ordinate indicates the rate of change of the alternative mapping values as the number of alternative mapping values increases . As the value of the abscissa parameter increases, the change of the alternative mapping value corresponding to the ordinate coordinate includes the change of the parameter auc.

The traditional Bayesian iterative calculation will solve Y=MAX(X*AUC), which will eventually cause the value of p6 to be too large because its growth rate is very fast, and p3 may gradually disappear because it is faster decrease. But in fact this is not what you want. Because it is possible that the online income brought by the 1% growth of p1 will be more than the online income brought by the 10% growth of p6.

Therefore, f(X)=Max(a*y1+b*y2+c*y3+d*y4+e*y5) appears as a constraint condition.

It is possible to add 10 times more weight on p1 than p6 to balance the gains. In this way, all the constraints and constraints can be reasonably combined with the business to form a solvable equation.

Referring to FIG. 7 , it is a schematic flowchart of another method for recommending target information.

The business experience analyzer in this figure is used to determine the candidate information to be recommended in combination with actual business needs, and the equation builder in the target equation builder can be used to determine the first sum value, the second sum value, and the first sum value sum The difference of the second sum, etc. Numeric converters can be used to convert between weights and map values.

Fig. 8 is a block diagram of an electronic device 800 according to an exemplary embodiment. For example, the electronic device 800 may be a mobile phone, a mobile computer, and the like. The electronic device may be a terminal for executing the foregoing method.

8, electronic device 800 may include one or more of the following components: processing component 802, memory 804, power supply component 806, multimedia component 808, audio component 810, input/output (I/O) interface 812, sensor component 814 , and the communication component 816.

The processing component 802 generally controls the overall operations of the electronic device 800, such as those associated with display, telephone calls, data communications, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above method. Additionally, processing component 802 may include one or more modules that facilitate interaction between processing component 802 and other components. For example, processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802 .

The memory 804 is configured to store various types of data to support operations at the device 800 . Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phonebook data, messages, pictures, videos, and the like. The memory 804 can be implemented by any type of volatile or non-volatile storage device or their combination, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.

The power supply component 806 provides power to various components of the electronic device 800 . Power components 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for electronic device 800 .

The multimedia component 808 includes a screen providing an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may not only sense a boundary of a touch or a swipe action, but also detect duration and pressure associated with the touch or swipe operation. In some embodiments, the multimedia component 808 includes a front camera and/or a rear camera. When the device 800 is in an operating state, such as a shooting state or a video state, the front camera and/or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have focal length and optical zoom capability.

The audio component 810 is configured to output and/or input audio signals. For example, the audio component 810 includes a microphone (MIC), which is configured to receive an external audio signal when the electronic device 800 is in an operating state, such as a calling state, a recording state, and a speech recognition state. Received audio signals may be further stored in memory 804 or sent via communication component 816 . In some embodiments, the audio component 810 also includes a speaker for outputting audio signals.

The I/O interface 812 provides an interface between the processing component 802 and a peripheral interface module, which may be a keyboard, a click wheel, a button, and the like. These buttons may include, but are not limited to: a home button, volume buttons, start button, and lock button.

Sensor assembly 814 includes one or more sensors for providing status assessments of various aspects of electronic device 800 . For example, the sensor component 814 can detect the open/closed state of the device 800, the relative positioning of components, such as the display and the keypad of the electronic device 800, and the sensor component 814 can also detect the position of the electronic device 800 or a component of the electronic device 800 changes, the presence or absence of user contact with the electronic device 800 , the orientation or acceleration/deceleration of the electronic device 800 and the temperature change of the electronic device 800 . Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact. Sensor assembly 814 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor component 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor or a temperature sensor.

The communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices. The electronic device 800 can access a wireless network based on communication standards, such as Wi-Fi, 4G or 5G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra Wide Band (UWB) technology, Bluetooth (BT) technology and other technologies.

In an exemplary embodiment, electronic device 800 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A programmable gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation for performing the methods described above.

In an exemplary embodiment, there is also provided a non-transitory computer-readable storage medium including instructions, such as the memory 804 including instructions, which can be executed by the processor 820 of the electronic device 800 to complete the above method. For example, the non-transitory computer readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like.

An embodiment of the present disclosure provides a non-transitory computer-readable storage medium. When the instructions in the storage medium are executed by the processor of the mobile terminal, the mobile terminal can execute the prompt method for image acquisition provided by any of the foregoing embodiments, and can execute At least one of the methods as shown in any one of Fig. 1 , Fig. 3 to Fig. 6 .

Other embodiments of the invention will be readily apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any modification, use or adaptation of the present invention, these modifications, uses or adaptations follow the general principles of the present invention and include common knowledge or conventional technical means in the technical field not disclosed in this disclosure . The specification and examples are to be considered exemplary only, with a true scope and spirit of the invention being indicated by the following claims.

It should be understood that the present invention is not limited to the precise constructions which have been described above and shown in the accompanying drawings, and various modifications and changes may be made without departing from the scope thereof. The scope of the invention is limited only by the appended claims.

Claims (12)

A method for recommending target information, including: Obtaining a reference parameter set of at least one dimension information of the candidate information to be recommended and a plurality of candidate parameter sets of the at least one dimension information; wherein, the reference parameter set includes: a reference weight of at least one dimension information The subset and the reference mapping value subset corresponding to the reference weight subset; the candidate parameter set includes: the candidate weight subsets of the at least one dimension information and the candidate weight subsets corresponding to each candidate weight subset Select a subset of mapped values; Determine the target mapping value subset from the candidate mapping value subset according to the difference between the first sum value and the second sum value; wherein, the first sum value is: any one of the candidate mapping value The sum of the elements in the subset; the second sum is the sum of the elements in the reference mapping value subset, and each element in the candidate mapping value subset has a proportional coefficient, and the proportional coefficient is used to represent the dimension The degree of influence of the information on determining the recommended target information from the candidate information; Determine a target weight subset according to the target mapping value subset; the target weight subset includes: a target weight affecting at least one dimension information of the candidate information; Selecting the recommended target information from the candidate information according to the target weight subset. The method according to claim 1, wherein the proportional coefficient is a preset value; The determining the target mapping value subset from the candidate mapping value subset according to the difference between the first sum value and the second sum value includes: determining the first sum value according to the scaling factor of each element in the candidate mapping value subset; According to the difference between the first sum and the second sum, subset the candidate mapping value with the largest difference, or the candidate mapping with the difference greater than a preset difference A subset of values, identified as a subset of values for the target map. The method according to claim 1, wherein the method further comprises: The scaling factor is updated based on the subset of candidate mapping values. The method according to claim 3, wherein said updating said scaling factor according to said subset of candidate mapping values comprises: According to the plurality of candidate mapping values respectively corresponding to each dimension information, determining the change trend of the plurality of candidate mapping values respectively corresponding to each dimension information; According to each change trend, the proportional coefficient is adjusted until each change trend reaches a target range; wherein, change trends corresponding to different dimension information have corresponding target ranges respectively. The method according to claim 4, wherein the target range is: a threshold range determined according to a proportion of the dimension information when selecting the target information from the candidate information. The method according to claim 1, wherein said obtaining a plurality of candidate parameter sets of said at least one dimension information comprises: Presetting T initialization parameter sets, the initialization parameter sets include: an initialization weight subset of the at least one dimension information and an initialization mapping value subset corresponding to the initialization weight subset; According to the third sum value of each element in the initialization weight subset in each initialization parameter set, and the fourth sum value of each element in each initialization mapping value subset, determine the T+1th parameter set; wherein, the The T+1th parameter set includes: the T+1th weight subset and the T+1th mapping value subset, where T is a positive integer greater than 1; According to the T+n parameter set, determine the T+n+1 parameter set; n is a positive integer, and the sum of each element in the T+n mapping value subset is a local maximum value; The T+n+1 parameter sets are determined as the candidate parameter sets. The method according to claim 6, wherein the preset T initialization parameter sets include: According to the historical parameter set of the candidate information, determine T initialization parameter sets matching the candidate information; wherein, the historical parameter set includes: a historical weight subset of at least one dimension information and the historical weight A subset of historical mapping values corresponding to the subset; or, Randomly generate T initialization parameter sets according to the candidate information to be recommended. The method according to claim 1, wherein said determining a subset of target weights according to said subset of target mapping values comprises: A candidate weight subset corresponding to the target mapping value subset is determined as the target weight subset. The method according to any one of claims 1 to 8, wherein the candidate information to be recommended includes at least one of the following: video; the public; article; The dimension information includes at least one of the following: The number of downloads, favorites, likes, the number of users who follow the author of the alternative information, the average browsing time, the number of comments, and the number of reposts. A device for recommending target information, including: The obtaining module is configured to obtain a reference parameter set of at least one dimension information of the candidate information to be recommended and a plurality of candidate parameter sets of the at least one dimension information; wherein the reference parameter set includes: at least one A reference weight subset of one dimension information and a reference mapping value subset corresponding to the reference weight subset; the candidate parameter set includes: the candidate weight subset of the at least one dimension information and each of the candidate the subset of alternative mapping values corresponding to the subset of weights; A target mapping value subset determining module configured to determine a target mapping value subset from the candidate mapping value subset according to the difference between the first sum and the second sum; wherein the first and The value is: the sum of elements in any one of the candidate mapping value subsets; the second sum value is the sum of elements in the reference mapping value subset, and each element in the candidate mapping value subset has a proportional coefficient , the proportional coefficient is used to represent the degree of influence of the dimension information on determining the recommended target information from the candidate information; The target weight subset determination module is configured to determine a target weight subset according to the target mapping value subset; the target weight subset includes: target weights affecting at least one dimension information of the candidate information; The recommendation module is configured to select the recommended target information from the candidate information according to the target weight. An electronic device, comprising: memory for storing processor-executable instructions; a processor connected to the memory; Wherein, the processor is configured to execute the method for recommending target information as provided in any one of claims 1-9. A non-transitory computer-readable storage medium, wherein computer-executable instructions are stored in the computer-readable storage medium, and when the computer-executable instructions are executed by a processor, the method provided by any one of claims 1 to 9 is implemented. Recommended method for target information.

Images