CN113763027B

CN113763027B - Recommendation information processing method, recommendation information generation method and device

Info

Publication number: CN113763027B
Application number: CN202110503431.XA
Authority: CN
Inventors: 刘凯; 杨秀金
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2021-05-10
Filing date: 2021-05-10
Publication date: 2023-08-01
Anticipated expiration: 2041-05-10
Also published as: CN113763027A

Abstract

The disclosure provides a recommendation information processing method, a recommendation information generating method and a recommendation information generating device. The recommended information processing method comprises the following steps: acquiring a recommendation information set containing a plurality of recommendation information; extracting characteristic elements of a plurality of categories of each recommendation information in the recommendation information set to obtain a characteristic element set; feature elements for each category: dividing the recommendation information in the recommendation information set into at least two groups, wherein the attribute of the characteristic element of the category of the recommendation information in each group of the at least two groups is mutually exclusive with the attribute of the characteristic element of the category of the recommendation information in other groups of the at least two groups; acquiring at least one result data corresponding to the recommended information in each group; for each type of result data, a degree of difference between any two groups of result data is calculated, and in the event that the degree of difference is greater than a first predetermined threshold, it is determined that the characteristic element of the category has significance to the result data.

Description

Recommendation information processing method, recommendation information generation method and device

Technical Field

The present disclosure relates to the field of multimedia, and more particularly, to a recommendation information processing method, recommendation information generation method, and apparatus.

Background

Advertisements have become an important element in the commodity market, requiring effective analysis and evaluation of advertisements in order to learn or predict the effectiveness of advertising, or to produce quality advertisements. With the rapid development of internet technology, the number of advertisements is greatly increased, and compared with the traditional advertisements which mainly comprise graphic advertisements, the video advertisements have higher and higher proportion nowadays, and the receiving degree of the audience to the video advertisements is gradually increased. However, the conventional advertisement analysis method is mostly suitable for the graphic advertisement, but cannot distinguish the characteristics of the video advertisement from the graphic advertisement, or cannot be suitable for the video advertisement at all, so that it is difficult to accurately evaluate and predict the video advertisement.

Disclosure of Invention

In order to solve the above-described problems, the present disclosure provides a recommendation information processing method, a recommendation information generating method, a recommendation information processing apparatus, a recommendation information generating apparatus, a device, a computer-readable storage medium, and a computer program product.

According to an aspect of the embodiments of the present disclosure, there is provided a recommendation information processing method including: acquiring a recommendation information set containing a plurality of recommendation information; extracting characteristic elements of a plurality of categories of each piece of recommended information in the recommended information set to obtain a characteristic element set; feature elements for each category in the feature element set: dividing a plurality of recommendation information in the recommendation information set into at least two groups, wherein the attribute of the characteristic element of the category of recommendation information in each group of the at least two groups is mutually exclusive with the attribute of the characteristic element of the category of recommendation information in other groups of the at least two groups; acquiring at least one result data corresponding to the recommendation information in each of the at least two groups; for each of the at least one result data, calculating a degree of difference between the result data of any two of the at least two groupings, and determining that the category of feature elements has significance to the result data if the degree of difference is greater than a first predetermined threshold.

According to an example of an embodiment of the present disclosure, the recommendation information processing method further includes: in the case where it is determined that the characteristic elements of the category have significance to the result data, outputting information about the characteristic elements of the category, wherein the information includes evaluation information indicating an influence of the characteristic elements of the category on the result data, the method further including generating a recommendation information analysis model for providing recommendation information production guidance based on the evaluation information.

According to an example of an embodiment of the present disclosure, the information further includes viewer feature information indicating a common feature of viewers of the recommendation information in the packet in which the result data is better among the two packets, and the recommendation information processing method further includes: generating recommendation information-oriented crowd recommendation based on the viewer characteristic information.

According to an example of an embodiment of the present disclosure, the recommendation information processing method further includes: based on the outputted information about the feature elements of different categories, the delivering effect of the new recommendation information is predicted by extracting and analyzing the feature elements of a plurality of categories in the new recommendation information.

According to an example of an embodiment of the present disclosure, wherein the feature elements of the plurality of categories include feature elements of a plurality of base attribute categories and/or a plurality of content attribute categories, wherein the feature elements of the plurality of base attribute categories include at least one of frame rate, code rate, size, resolution, saturation, sharpness, number of lenses of the recommendation information; the feature elements of the plurality of content attribute categories include at least one of shots, scenes, characters, objects, and presentations of the recommendation information.

According to an example of an embodiment of the present disclosure, wherein the at least two groupings comprise a first grouping and a second grouping, wherein dividing the plurality of recommendation information in the recommendation information set into at least two groupings for each category of feature element in the feature element set comprises: for each of the plurality of recommendation information, dividing the recommendation information into a first group in a case where a value or a number of feature elements of the category of the recommendation information satisfies a predetermined condition; otherwise, the recommendation information is divided into a second group.

According to an example of an embodiment of the present disclosure, extracting, for the feature elements of the plurality of content attribute categories, feature elements of a plurality of categories of each recommendation information in the recommendation information set includes: extracting feature elements of a plurality of content attribute categories in the recommendation information set by using a deep learning model; and for each recommendation information, generating a confidence that the recommendation information contains the extracted feature elements of each content attribute category, and a confidence that the feature elements of the content attribute category appear in the recommendation information.

According to an example of an embodiment of the present disclosure, the recommendation information processing method further includes: by adjusting the confidence, the outputted information about the feature elements of the different content attribute categories is adjusted.

According to an example of an embodiment of the present disclosure, the recommendation information processing method further includes: extracting a time stamp of each feature element in the feature element set; and generating a recommendation information analysis model for providing recommendation information production guidance having a time partition based on the time stamp and the outputted information about the feature elements of the different categories, wherein the time stamp is attached to each feature element of the recommendation information when each recommendation information in the recommendation information set is generated.

According to an example of an embodiment of the present disclosure, the at least one result data includes at least one part of a delivery effect data, a viewer behavior data, and an industry distribution data, wherein the delivery effect data includes at least one of a recommended information click rate and a recommended information conversion rate, the viewer behavior data includes at least one of a viewing duration, a number of times of viewing, a number of times of daily repeated viewing, a completion rate, a jump rate, a conversion time point, and a conversion rate of the recommended information viewer, and the industry distribution data indicates a distribution characteristic of an industry to which the recommended information viewer belongs.

According to another aspect of the embodiments of the present disclosure, there is provided a recommendation information generation method, including: extracting characteristic elements of a plurality of categories in the recommendation information to be optimized; feature elements for each of the plurality of categories of feature elements: dividing the recommendation information to be optimized into groups in at least two groups, wherein the attribute of the characteristic element of the category of the recommendation information in each group in the at least two groups is mutually exclusive with the attribute of the characteristic element of the category of the recommendation information in other groups in the at least two groups; predicting at least one result data of the recommendation information to be optimized according to predetermined evaluation information about the grouping of feature elements for the category based on the grouping in which the recommendation information to be optimized is located; and optimizing the characteristic elements of the category of the recommended information to be optimized by utilizing the predicted at least one result data.

According to another aspect of the embodiments of the present disclosure, there is provided a recommended information processing apparatus including: a recommendation information acquisition unit configured to acquire a recommendation information set including a plurality of recommendation information; a feature extraction unit configured to extract feature elements of a plurality of categories of each of the recommendation information in the recommendation information set to obtain a feature element set; and a recommendation information processing unit configured to, for each category of feature elements in the feature element set: dividing a plurality of recommendation information in the recommendation information set into at least two groups, wherein the attribute of the characteristic element of the category of recommendation information in each of the at least two groups is mutually exclusive with the attribute of the characteristic element of the category of recommendation information in other groups of the at least two groups, acquiring at least one result data corresponding to the recommendation information in each of the at least two groups, calculating a degree of difference between the result data of any two groups of the at least two groups for each of the at least one result data, and determining that the characteristic element of the category has significance to the result data if the degree of difference is greater than a first predetermined threshold.

According to an example of an embodiment of the present disclosure, the recommendation information processing unit is further configured to: in a case where it is determined that the characteristic element of the category has significance to the result data, outputting information about the characteristic element of the category, wherein the information includes evaluation information indicating an influence of the characteristic element of the category on the result data, the recommendation information processing unit is further configured to generate a recommendation information analysis model for providing recommendation information production guidance based on the evaluation information.

According to an example of an embodiment of the present disclosure, wherein the information further includes viewer feature information indicating a common feature of viewers of the recommendation information in the group in which the result data is better among the two groups, the recommendation information processing unit is further configured to: generating recommendation information-oriented crowd recommendation based on the viewer characteristic information.

According to an example of an embodiment of the present disclosure, the recommendation information processing unit is further configured to: based on the outputted information on the feature elements of different categories, the delivery effect of the new recommendation information is predicted by analyzing the feature elements of a plurality of categories in the new recommendation information extracted by the feature extraction unit.

According to an example of an embodiment of the present disclosure, wherein the at least two packets include a first packet and a second packet, wherein the recommendation information processing unit is further configured to: for each of the plurality of recommendation information, dividing the recommendation information into a first group in a case where a value or a number of feature elements of the category of the recommendation information satisfies a predetermined condition; otherwise, the recommendation information is divided into a second group.

According to an example of an embodiment of the present disclosure, the feature extraction unit is further configured to: extracting characteristic elements of the plurality of content attribute categories in the recommendation information set by using a deep learning model for the characteristic elements of the plurality of content attribute categories; and for each recommendation information, generating a confidence that the recommendation information contains the extracted feature elements of each content attribute category, and a confidence that the feature elements of the content attribute category appear in the recommendation information.

According to an example of an embodiment of the present disclosure, the recommendation information processing unit is further configured to: by adjusting the confidence, the outputted information about the feature elements of the different content attribute categories is adjusted.

According to an example of an embodiment of the disclosure, the feature extraction unit is further configured to: extracting a time stamp of each feature element in the feature element set; and, the recommendation information processing unit is further configured to generate a recommendation information analysis model for providing recommendation information production guidance having a time partition based on the time stamp and the outputted information about the different categories of feature elements, wherein the time stamp is attached to each feature element of the recommendation information when each recommendation information in the recommendation information set is generated.

According to another aspect of the embodiments of the present disclosure, there is provided a recommendation information generating apparatus including: a feature extraction unit configured to extract feature elements of a plurality of categories in the recommendation information to be optimized; a recommendation information generation unit configured to, for each of the feature elements of the plurality of categories: dividing the recommendation information to be optimized into groups in at least two groups, wherein the attribute of the characteristic element of the category of the recommendation information in each group in the at least two groups is mutually exclusive with the attribute of the characteristic element of the category of the recommendation information in other groups in the at least two groups, predicting at least one result data of the recommendation information to be optimized according to the preset evaluation information of the group of the characteristic element of the category based on the group in which the recommendation information to be optimized is located, and optimizing the characteristic element of the category of the recommendation information to be optimized by utilizing the predicted at least one result data.

According to another aspect of the embodiments of the present disclosure, there is provided a recommended information processing apparatus including: one or more processors; and one or more memories, wherein the memories have stored therein computer readable code, which when executed by the one or more processors, causes the one or more processors to perform the methods described in the various aspects above.

According to another aspect of the disclosed embodiments, there is provided a computer readable storage medium having stored thereon computer readable instructions, which when executed by a processor, cause the processor to perform the method described in the above aspects.

According to another aspect of the disclosed embodiments, there is provided a computer program product comprising computer readable instructions which, when executed by a processor, cause the processor to perform the method described in the above aspects.

By using the recommended information processing method, the recommended information generating device, the recommended information generating equipment, the computer readable storage medium and the computer program product according to the aspects, the influence of different types of characteristic elements on the recommended information throwing result can be accurately analyzed and obtained; the method can provide recommendation information production guidance and targeted crowd recommendation of recommendation information delivery; the release effect of the recommended information can be predicted; and optimizing and adjusting the recommended information based on the predicted putting effect so as to obtain a better recommended information putting effect.

Drawings

The above and other objects, features and advantages of the presently disclosed embodiments will become more apparent from the more detailed description of the presently disclosed embodiments when taken in conjunction with the accompanying drawings. The accompanying drawings are included to provide a further understanding of embodiments of the disclosure, and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description serve to explain the disclosure, without limitation to the disclosure. In the drawings, like reference numerals generally refer to like parts or steps.

FIG. 1 illustrates a flowchart of a recommendation information processing method according to an embodiment of the present disclosure;

FIG. 2 shows a schematic diagram of an example extracted feature element according to an embodiment of the present disclosure;

FIG. 3 illustrates a schematic diagram of grouping advertisements in an advertisement collection, according to an example of an embodiment of the disclosure;

FIG. 4 illustrates a schematic diagram of associating result data for various packets, according to an example of an embodiment of the disclosure;

FIG. 5 illustrates a schematic diagram of calculating variance of deviations of result data within respective packets, according to an example of an embodiment of the disclosure;

FIG. 6 illustrates a flowchart of a recommendation information generation method according to an embodiment of the present disclosure;

fig. 7 illustrates a schematic configuration diagram of a recommended information processing apparatus according to an embodiment of the present disclosure;

fig. 8 illustrates a schematic configuration diagram of a recommendation information generating device according to an embodiment of the present disclosure;

fig. 9 shows a schematic diagram of an architecture of an exemplary computing device, according to an embodiment of the present disclosure.

Detailed Description

The technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the drawings in the embodiments of the present disclosure. It will be apparent that the described embodiments are merely embodiments of a portion, but not all, of the present disclosure. All other embodiments, which can be made by one of ordinary skill in the art without the need for inventive faculty, are intended to be within the scope of the present disclosure, based on the embodiments in this disclosure.

In the embodiment of the present disclosure, the recommendation information may be, for example, information pushed to the terminal device for display in a manner of pictures, text, video, or any combination thereof. For example, the recommendation information may be any content such as advertisement, push news, etc., which is not particularly limited by the embodiments of the present disclosure. In the following embodiments of the present disclosure, an advertisement will be described as an example of recommendation information.

Today, the video advertisement is getting higher and higher in the advertisement. Video advertisements, for example, refer to advertisements presented in the form of videos that may be part or all of the video, e.g., video advertisements may be appended to the beginning of a particular video, end of a video, or any other location, or occupy the entire content of the video. Video advertisements have a number of features that are significantly different from teletext advertisements due to their rich content properties, complex production patterns, diversified presentation forms, etc. For example, for a graphic advertisement, the graphic advertisement can be analyzed only by the corresponding advertisement click rate, and it is difficult to know that the watching time length, the watching times, and the like of a viewer of the graphic advertisement can reflect more characteristics of the quality of the graphic advertisement; the video advertisement is different, and the watching time, watching times, jumping-out time point, conversion time point and the like of the video advertisement can be collected for analyzing the video advertisement, so that more and more complex advertisement putting result data of the video advertisement are endowed.

However, the conventional advertisement analysis method is mostly suitable for the graphic advertisement, but cannot accurately analyze and evaluate the video advertisement. The present disclosure provides a recommendation information processing method that can effectively analyze recommendation information such as video advertisements and can provide a recommendation information analysis model for recommendation information production guidance and directed crowd recommendation for recommendation information delivery. In the following embodiments of the present disclosure, a video advertisement is described as an example, but it should be understood that some or all of the steps of the recommendation information processing method according to the embodiments of the present disclosure may also be applied to any other recommendation information.

A recommended information processing method according to an embodiment of the present disclosure is described below with reference to fig. 1. Fig. 1 shows a flowchart of a recommendation information processing method 100 according to an embodiment of the present disclosure.

As shown in fig. 1, in step S110, a recommendation information set including a plurality of recommendation information is acquired. By way of example and not limitation, each recommendation information in the set of recommendation information may be an advertisement, such as a teletext advertisement or a video advertisement presented in the form of a video, or the like. The recommendation information set may be obtained from an open source database, for example, the recommendation information processing method according to the embodiment of the disclosure may utilize existing data to perform analysis, such as obtaining valuable advertisement information from historical advertisement data to perform advertisement analysis; alternatively, the recommendation information set may be recommendation information about to be delivered, for example, a part or all of advertisements of a specific advertiser about to be delivered, that is, the recommendation information processing method according to the embodiment of the present disclosure may perform tracking analysis on the recommendation information about to be delivered, so as to analyze and adjust the delivery effect of the recommendation information in real time.

In step S120, feature elements of a plurality of categories of each of the recommendation information in the recommendation information set are extracted to obtain a feature element set. Taking advertisements as an example, each advertisement is made up of a plurality of different categories of feature elements. For example, a teletext advertisement may include characteristic elements of the class of colors, shapes, fonts, patterns, etc. The video advertisement may include more abundant feature elements, for example, the feature elements of the video advertisement may be divided into a basic attribute category and a content attribute category, where the basic attribute category may include, for example, feature elements of various categories such as frame rate, code rate, size, resolution, saturation, definition, lens number, and the like of the advertisement; the content attribute categories may include, for example, feature elements of various categories of shots, scenes, people, objects, presentations, etc. of the advertisement. Here, for the video advertisement, the presentation form of the advertisement refers to a manner in which content included in the advertisement is presented, for example, the presentation form of the video advertisement may include at least one of a dialogue, a mouth cast, a screen recording, a presentation (PPT), etc., or any combination thereof.

The steps of extracting feature elements will be described below with reference to fig. 2 by taking a video advertisement as an example. Fig. 2 shows a schematic diagram of an example extracted feature element according to an embodiment of the present disclosure. As shown in fig. 2, for video advertisements, extracting feature elements of multiple categories of each recommendation information in the set of recommendation information may include extracting feature elements of multiple underlying attribute categories and feature elements of multiple content attribute categories for each recommendation information.

Specifically, for example, for feature elements of a plurality of basic attribute categories of a video advertisement, feature elements of a frame rate, a code rate, a size, a resolution, saturation, sharpness, and the like of the video advertisement may be extracted using a multimedia video processing tool such as FFMPEG, and the like, and the number of lenses of the video advertisement may be extracted using a lens boundary detection algorithm such as a pixel comparison method, a double threshold detection method, and the like. For feature elements of multiple content attribute categories of video advertisements, wherein feature elements such as shots, scenes, characters, etc. of the video advertisements may also be extracted using shot boundary detection algorithms such as pixel comparison, dual threshold detection, etc., but the disclosure is not limited thereto. According to examples of embodiments of the present disclosure, a deep learning model may be utilized to extract feature elements of multiple content attribute categories for advertisements in an advertisement collection. For example, taking the presentation form of a video advertisement as an example, different frames in the video advertisement may be first compared to extract key frames in the video advertisement, and then the key frames of the video advertisement are processed using a deep learning model to identify the presentation form included in the video advertisement. Here, the deep learning model in the embodiments of the present disclosure may be a trained and fixed model for video analysis; alternatively, a neural network may be used to build a deep learning model, such as a Convolutional Neural Network (CNN), a long-term short memory model (LSTM), etc., and train it with video samples from an open source video database, which is not particularly limited by the embodiments of the present disclosure.

In addition, when feature elements of a plurality of content attribute categories of each recommendation information in the recommendation information set are extracted using the deep learning model, a confidence that the recommendation information contains the extracted feature elements of each content attribute category and a confidence that the extracted feature elements of the plurality of content attribute categories appear in the recommendation information may also be generated. For example, for a presentation of a video advertisement, after determining that a video advertisement includes a conversation and a recording using a deep learning model, a frame position or a timeline position at which the conversation and recording occur in the video advertisement may also be obtained, and a confidence level for the video advertisement including the conversation and recording, and a confidence level for the conversation and recording at the frame position or the timeline position may be generated for subsequent adjustment and optimization of advertisement analysis results, as will be described further below.

The extracted feature elements of the plurality of categories of each recommendation information in the recommendation information set form a feature element set, and the feature element set can comprise values of various feature elements such as resolution, saturation, definition, lens number and the like of each advertisement in the plurality of advertisements. After extracting the feature element set from the recommendation information set, for each category of feature elements in the feature element set, the influence of the category of feature elements on the result data (which may be referred to as result data) after the recommendation information is put in may be determined through step S130. Step S130 may in turn include steps S131, S132 and S133.

In step S131, for each category of feature element in the feature element set, the plurality of recommendation information in the recommendation information set is divided into at least two groups, wherein the attribute of the category of feature element of the recommendation information within each of the at least two groups is mutually exclusive with the attribute of the category of feature element of the recommendation information within other of the at least two groups. Here, the mutual exclusion of the attributes of the two feature elements may mean that the values or the numbers of the two feature elements have diametrically opposite attributes or attributes that are not compatible with each other. For example, for video advertisements, the attributes of high and low resolution, high and low numbers of shots, recorded and non-recorded presentations, etc. are diametrically opposed, the advertisements in the advertisement set may be divided accordingly into two groupings such that the advertisements in the two groupings have diametrically opposed attributes. For example, for video advertisements, the high resolution, medium resolution, and low resolution, the large number of shots, medium and small number of shots, large number of recordings, medium and small number of recordings, etc. are incompatible with each other, and the plurality of advertisements in the advertisement collection may be divided into three groupings accordingly such that the advertisements in the three groupings have properties that are incompatible with each other. It should be understood that, only the advertisement is taken as an example, and an example of dividing advertisements in the advertisement set into two groups and three groups is schematically listed herein, but the embodiment of the present disclosure is not limited thereto, and recommendation information in the recommendation information set may be divided into more groups according to actual requirements.

Referring to fig. 3, description will be given taking recommendation information as an advertisement and dividing advertisements in an advertisement set into two groups as an example. FIG. 3 illustrates a schematic diagram of grouping advertisements in an advertisement collection, according to an example of an embodiment of the disclosure. According to an example of an embodiment of the present disclosure, for each category of feature element in a set of feature elements, for each advertisement in a set of advertisements, dividing the advertisement into a first group if a value or number of feature elements of the category of the advertisement meets a predetermined condition; otherwise, the advertisement is divided into a second group. Here, the predetermined condition may be determined according to a numerical value or a number distribution of feature elements, or an actual requirement, for example, the predetermined condition may be that the value or the number of feature elements of the category of the advertisement is greater than, less than or equal to a second predetermined threshold, or the predetermined condition may also be that the value or the number of feature elements of the category of the advertisement belongs to a specific category, which is not particularly limited in the embodiments of the present disclosure.

For example, as shown in FIG. 3, for feature element class A, advertisements in an advertisement collection may be divided into a first group A1 and a second group A2; for feature element class B, advertisements in the advertisement collection may be divided into a first group B1 and a second group B2, and so on. For example, the feature element category a may be a resolution of a video advertisement, and when the resolution of a video advertisement in the advertisement set is greater than a second predetermined threshold, the video advertisement may be divided into a first group A1; otherwise, dividing the video advertisement into a second packet A2 when the resolution of the video advertisement is less than or equal to a second predetermined threshold. For another example, the feature element category B may be a person in a video advertisement, and when the number of persons in a certain video advertisement in the advertisement set is greater than a second predetermined threshold, the video advertisement may be divided into a first group B1; otherwise, dividing the video advertisement into a second group B2 when the number of people in the video advertisement is less than or equal to a second predetermined threshold.

In step S132, at least one kind of result data corresponding to the recommendation information in each of the at least two packets is acquired. The at least one result data corresponding to the recommended information in each group refers to various result data generated after the recommended information of the group is put in, and may include at least a part of put effect data, viewer behavior data, industry distribution data, and the like, for example. Taking an advertisement as an example, the placement effect data may include an advertisement click rate, which may be a ratio of a number of people who point to click on a product link included in the advertisement to a total number of people who view the advertisement, an advertisement conversion rate, which may be a ratio of a number of people who acquire a product by clicking on a product link included in the advertisement to a total number of people who click on a product link included in the advertisement, and the like. The viewer behavior data may include a viewing duration, a number of views, a number of daily repeat views, a run-out rate, a jump rate, a conversion time point, a conversion rate, and the like of the advertisement viewer. The completion rate may refer to a ratio of the number of people who completely watch the video advertisement to the total number of people who open the video advertisement; the jump rate refers to the ratio of the number of people who jump out of the advertisement page to the total number of people watching the advertisement due to reasons such as clicking on a product link included in the advertisement, closing the advertisement page, etc.; the conversion time point refers to a time point when a viewer clicks a product link included in an advertisement in the process of viewing a video advertisement; conversion success refers to the ratio of the number of people purchasing a product by clicking on the product link contained in the advertisement to the total number of people viewing the video advertisement. The industry distribution data represents the distribution characteristics of the industries to which the advertisement viewers belong, for example, the industry distribution data may include the proportion of the number of people belonging to the financial industry, the proportion of the number of people belonging to the teacher industry, and the like, and further, the above-mentioned delivery effect data and the viewer behavior data are combined, so that the delivery effect data and the viewer behavior data in different industries may be respectively obtained through statistics.

According to an example of an embodiment of the present disclosure, after various result data of all recommendation information in a recommendation information set is acquired, these result data may be associated with respective groupings obtained for different categories of feature elements, i.e., with respective groupings for different categories of feature elements and their corresponding result data. Referring to fig. 4, the advertisement sets are described as being divided into two groups. Fig. 4 shows a schematic diagram of result data associated with various packets according to an example of an embodiment of the disclosure. As shown in fig. 4, for the first group A1 and the second group A2 of the advertisement set divided for the feature element category a, the first group A1 and the second group A2 are respectively associated with a plurality of kinds of result data including at least a part of the impression effect data, the viewer behavior data, the industry distribution data, and the like corresponding thereto; similarly, for the first group B1 and the second group B2 of the advertisement set divided for the feature element category B, the first group B1 and the second group B2 are respectively associated with a plurality of result data corresponding thereto including at least a part of the delivery effect data, the viewer behavior data, the industry distribution data, and the like, and so on, and will not be described in detail herein.

Thereafter, in step S133, for at least one kind of result data corresponding to the recommendation information within each of at least two groups of the acquired recommendation information set, for each of the result data, a degree of difference between the result data of any two groups is calculated, and in the case where the degree of difference is greater than a first predetermined threshold, it is determined that the feature element of the current category has significance to the result data.

Specifically, for each result data, a total average value of the result data corresponding to any two groups is calculated first, for example, the result data corresponding to any two groups may be accumulated and summed, and the resulting accumulated sum is divided by the total number of recommended information in the two groups to obtain the above total average value; then, in each of the two groups, a difference value (may be referred to as a back deviation) between the result data of the respective recommendation information in the group and the total average value is calculated, and a variance (may be referred to as a deviation variance) of the back deviation of the respective recommendation information is calculated, for example, the deviation variances of the two groups may be referred to as a first deviation variance and a second deviation variance, respectively; the degree of difference of the first deviation variance and the second deviation variance is calculated, for example, the degree of difference may be an absolute difference of the first deviation variance and the second deviation variance. It should be noted that, although the degree of difference between the result data of any two packets is calculated by the variance of the deviation of the result data of each packet is described above, the embodiment of the present disclosure is not limited thereto. The degree of difference between the result data of any two packets may also be calculated by an average value of the result data of each packet or any other statistical index, which is not particularly limited by the embodiments of the present disclosure.

Referring to fig. 5, a specific step of calculating the degree of difference will be further described taking still an example of dividing the advertisement sets into two groups, and in this example, the grouping of the advertisement sets divided for the feature element class a is taken as an example.

Fig. 5 illustrates a schematic diagram of calculating variance of deviations of result data within individual packets, according to an example of an embodiment of the disclosure. As shown in fig. 5, for feature element category a, advertisements in the advertisement collection are divided into a first group A1 and a second group A2, wherein the first group A1 and the second group A2 have corresponding multiple result data including at least a portion of impression effect data, viewer behavior data, industry distribution data, and the like, respectively. In this step, for each type of result data, first, a total average value of the result data corresponding to the recommendation information of the first packet A1 and the second packet A2 is calculated, and for example, a quotient of the sum of the result data corresponding to the recommendation information of the packets A1, A2 and the total number of the recommendation information of the packets A1, A2 may be taken as the total average value; then, respectively calculating a first deviation variance of the result data corresponding to the first group A1 and a second deviation variance of the result data corresponding to the second group A2, for example, respectively calculating a first deviation variance A1-1 of the advertisement click rate in the first group A1 and a second deviation variance A2-1 of the advertisement click rate in the second group A2; a first deviation variance A1-2 of advertisement conversion rate in the first group A1 and a second deviation variance A2-2 of advertisement conversion rate in the second group A2; a first deviation variance A1-3 of the viewer completion rate in the first group A1 and a second deviation variance A2-3 of the viewer completion rate in the second group A2, etc.

In the case that there is a degree of difference between the result data of two groups among the at least two groups greater than a first predetermined threshold, it is determined that the feature element of the current category has significance to the result data. In the presently disclosed embodiments, significance means that the value of a characteristic element has a significant impact on the result data, i.e. the value of the result data can be significantly changed, for example as described above, such that the degree of difference between the result data for any two groupings of the characteristic element is greater than a first predetermined threshold. Still referring to the example in FIG. 5 above, for example, if the degree of difference between the advertisement click-through rate within the first group A1 and the second deviation variance A2-1 within the second group A2 is greater than a first predetermined threshold, it may be determined that the characteristic element category A has significance to the advertisement conversion rate. For another example, if the degree of difference between the first deviation variance A1-3 in the first group A1 and the second deviation variance A2-3 in the second group A2 is greater than a first predetermined threshold, it may be determined that the feature element class a has significance to the viewer's completion rate. Here, the first predetermined threshold may be determined according to actual application requirements, which is not particularly limited by the embodiments of the present disclosure.

It should be noted that, although the advertisement sets are described herein as being divided into two groups, the embodiments of the present disclosure are not limited thereto, and as described above, the advertisement sets may be divided into more groups. For example, if the advertisement set is divided into a first group A1, a second group A2, and a second group A3 for the feature element category a, in step S133, the degree of difference in the result data between any two of the three groups may be calculated, respectively, and compared with a first predetermined threshold value, respectively; alternatively, two packets of the three packets having the best and worst deviation variance of the result data may be selected, the degree of difference in the deviation variance between the two packets may be calculated and compared with a first predetermined threshold, and so on.

According to an example of an embodiment of the present disclosure, in a case where it is determined that a feature element of a certain category has significance to certain result data, information about the feature element of the category, for example, evaluation information indicating an influence of the feature element of the category on the result data, may be output. For example, as described above, in the case where it is determined that the feature element category a has significance to the advertisement click rate, evaluation information indicating the influence of the feature element category a on the advertisement click rate may be further output. Taking the resolution of the video advertisement as the characteristic element category a as an example, assuming that the resolution of the video advertisement in the first group A1 is high, the resolution of the video advertisement in the second group A2 is low, and the first deviation variance A1-1 of the advertisement click rate within the first group A1 is much smaller than the second deviation variance A2-1 of the advertisement click rate within the second group A2, i.e., the degree of difference therebetween is greater than a first predetermined threshold, evaluation information such as "high resolution makes the deviation variance of the advertisement click rate smaller, performs better" or the like may be output. Alternatively, if the average value is employed instead of deviating from the variance to calculate the degree of difference in the above example, evaluation information such as "high resolution makes the average value of advertisement click rate larger" may be output. According to examples of embodiments of the present disclosure, the rating information may be used to produce a recommendation information analysis model, such as an advertisement analysis model, that provides a recommendation information production guideline, as will be described further below.

Further, according to examples of embodiments of the present disclosure, the outputted information about the feature elements of the category may also include viewer feature information indicating common features of viewers of the recommendation information in the better grouping of result data, for example, for generating recommendation information targeting crowd recommendations. The viewer characteristic information may be, for example, information indicating the characteristics of the viewer such as age, sex, industry, taste, region, etc., which may reflect the influence of the characteristic elements of the category on the viewing behavior of the crowd having these characteristics.

For example, in the example in which the above-described feature element category a is the resolution of a video advertisement, assuming that a high resolution has been determined such that the deviation variance of the advertisement click rate is smaller, for the first group A1 and the second group A2 obtained by grouping for the resolution, if the first deviation variance A1-1 of the advertisement click rate within the first group A1 is smaller, that is, the deviation of the advertisement click rate within the first group A1 from the total average value is smaller, the common feature of viewers of the advertisement within the first group A1 can be extracted at this time. For example, all features of the viewers of the advertisement within the first group A1 may be extracted, the probabilities of the individual features may be statistically analyzed, and one or more features with the highest probabilities may be taken as common features for those viewers. For example, assuming that the advertisement viewers within the first group A1 are all extracted to have a common characteristic of "male, 18-25 years old, gaming industry", viewer characteristic information indicating the common characteristic may be output, which may indicate that, for example, a high-resolution video advertisement has a large influence on a crowd having a characteristic of "male, 18-25 years old, gaming industry". Then, when video advertisements are placed, if an optimal advertisement click-through rate is desired, a selection of people with common characteristics of "men, 18-25 years old, gaming industry" may be recommended for targeted placement to obtain better advertisement click-through benefits.

In the example of fig. 5, the analysis for the feature element class a ends up here. Next, the analysis may be continued with respect to other feature element categories using step S130 as described above, for example, the analysis may be continued with respect to feature element category B to determine the influence of feature element category B on different result data, and output corresponding evaluation information and viewer feature information; then, the characteristic element class C may be analyzed to determine the influence of the characteristic element class C on different result data, and output corresponding evaluation information, viewer characteristic information, and the like, which are not described herein.

According to an example of an embodiment of the present disclosure, evaluation information about different categories of feature elements output using the recommendation information processing method according to an embodiment of the present disclosure may be used to generate a recommendation information analysis model, such as an advertisement analysis model, that provides recommendation information production guidance. The recommendation information analysis model contains evaluation information indicating the influence of different types of characteristic elements on different result data, such as the influence of resolution on advertisement click rate, the influence of advertisement presentation form on advertisement conversion rate and the like. For example, a map reflecting the mapping relation of each characteristic element and the delivery result data expression may be generated based on the obtained various evaluation information, and the map may be built in the advertisement analysis model. Thus, when generating a new advertisement, desired result data, such as desired advertisement click-through rate, advertisement conversion rate, etc., may be input into an advertisement analysis model, which uses the mapping table to guide the generation of the new advertisement, for example, to output the new advertisement with the best desired performance. For example, the advertisement analysis model may direct the user on which feature element categories and what values or numbers are most likely to be the desired result data for those feature element categories based on the mapping table, or may directly provide multiple advertisements with the best performance for the user to select, thereby helping to produce a better quality advertisement. For another example, a neural network may be utilized to train an advertisement analysis model capable of intelligently generating new advertisements based on the various evaluation information obtained, e.g., product attributes and advertisement requirements such as "high click-through rate" may be directly entered, new advertisements with optimal click-through rates may be generated, and so forth. In addition, with the outputted information on the feature elements of different categories, or the advertisement analysis model containing such information, for an unknown new advertisement, the impression of the new advertisement can be predicted and optimized by extracting and analyzing the feature elements of a plurality of categories contained in the new advertisement, as will be described in further detail below.

According to an example of an embodiment of the present disclosure, viewer feature information about different categories of feature elements output using a recommendation information processing method according to an embodiment of the present disclosure may be used to generate recommendation information targeting crowd recommendations. After the new recommendation information is generated by using the recommendation information analysis model as described above, the group of people most suitable for the new recommendation information can be targeted by using the viewer characteristic information about the characteristic elements of the categories based on the category of the characteristic elements adopted in the new recommendation information and the numerical value or the number thereof, thereby being beneficial to obtaining more ideal benefits.

In addition, the recommendation information processing method according to the embodiment of the present disclosure may also adjust the output information about the feature elements of different content attribute categories by adjusting the confidence of the feature elements of the content attribute categories. For characteristic elements of the content attribute category, such as video advertisement presentation forms, when the presentation form of the video advertisement is very complex, the presentation form of the video advertisement obtained through the deep learning model may have errors, thereby affecting the accuracy of the finally output information. At this time, as described above, when feature elements of a plurality of content attribute categories in an advertisement are extracted using a deep learning model, a confidence that the advertisement contains the extracted feature element of each content attribute category, and a confidence that the feature element of that content attribute category appears in the advertisement may be generated. For example, after outputting information (e.g., evaluation information) about the feature elements of a certain content attribute category, if the information is found to be inaccurate after verifying the information in connection with the actual situation, the information may be adjusted and optimized by adjusting the confidence level of the feature elements of the content attribute category to output the most accurate information about the feature elements of the content attribute category, thereby ensuring the accuracy of advertisement analysis.

In practical applications, the feature elements of the respective categories employed by the recommendation information may be updated frequently, or more specifically, different versions may exist for the feature elements of the same category. For example, in the recommendation information set acquired in step S110, different recommendation information may include feature elements of the same category of different versions. Taking the characteristic elements of the character class as an example, for example, the advertisement 1 and the advertisement 2 in the advertisement set both contain characters, but the advertisement 1 with earlier time contains the character a, and the advertisement 2 with later time contains the character b updated, and the influence of the character a and the character b on the result data may be distinct. If the advertisement analysis is performed only for the character, which is a characteristic element category, regardless of the update of the character version, key information of the influence of the character update on the result data may be omitted. Conventional advertisement analysis methods often can only handle static (i.e., fixed) feature elements, and it is difficult to provide solutions to the problems caused by such feature element updates.

The recommendation information processing method according to the embodiment of the present disclosure can effectively solve the above-mentioned problems, unlike the conventional advertisement analysis method which can only rely on the service log after advertisement delivery, the recommendation information processing method 100 according to the embodiment of the present disclosure can intervene in the delivery link. Specifically, for example, in the advertisement generation phase, a time stamp may be appended to each category of feature elements that the advertisement contains, which may indicate, for example, the generation time, version, etc. of the category of feature elements for subsequent advertisement analysis. For example, in the case where the feature elements of different categories included in the advertisements in the advertisement set acquired in step S110 all have corresponding time stamps, when the feature elements of different categories of each advertisement in the advertisement set are extracted in step S120, the time stamp of the feature element of each category may also be extracted. At this time, for different versions indicated by the time stamps of the feature elements of each category, it is possible to analyze and output corresponding information about the category of the feature element of the version for the different versions, respectively, and generate an advertisement analysis model having a time partition according thereto in step S130. That is, in the advertisement analysis model, for each category of feature elements, time partitioning may be performed based on the time stamp thereof, and each version of the category of feature elements may be associated with the corresponding output information thereof, so that the influence of the feature elements of different versions on the advertisement delivery result may be effectively obtained.

By using the recommendation information processing method according to the embodiment of the disclosure, the influence of the feature elements of different categories on the recommendation information throwing result can be accurately analyzed, the recommendation information production guidance and the directional crowd recommendation of the recommendation information throwing can be provided, and the throwing effect of the recommendation information can be predicted.

A recommendation information generation method according to an embodiment of the present disclosure is described below with reference to fig. 6. Fig. 6 shows a flowchart of a recommendation information generation method 600 according to an embodiment of the present disclosure, where the recommendation information generation method 600 can perform optimization adjustment on any recommendation information to be optimized. Since the details of part of the steps of the recommendation information generation method 600 are the same as those of the recommendation information processing method 100 described above with reference to fig. 1, repeated descriptions of the same contents are omitted here for simplicity.

As shown in fig. 6, in step S610, feature elements of a plurality of categories in the recommendation information to be optimized are extracted. Here, the recommendation information to be optimized may be any information such as advertisement, for example, may include a teletext advertisement, a video advertisement, and the like. The following description will take the recommendation information to be optimized as video advertisement as an example.

According to an example of an embodiment of the present disclosure, extracting feature elements of a plurality of categories of recommendation information to be optimized may include extracting feature elements of a plurality of base attribute categories and feature elements of a plurality of content attribute categories of recommendation information to be optimized. Specifically, taking an advertisement as an example, for feature elements of a plurality of basic attribute categories of the advertisement to be optimized, feature elements of frame rate, code rate, size, resolution, saturation, sharpness, and the like of the advertisement to be optimized may be extracted using a multimedia video processing tool such as FFMPEG, and the number of lenses of the advertisement to be optimized may be extracted using a lens boundary detection algorithm such as a pixel comparison method, a dual threshold detection method, and the like. For feature elements of a plurality of content attribute categories of the advertisement to be optimized, wherein feature elements such as shots, scenes, characters, etc. of the advertisement to be optimized may also be extracted using a shot boundary detection algorithm such as a pixel comparison method, a double threshold detection method, etc., but the present disclosure is not limited thereto.

According to examples of embodiments of the present disclosure, a deep learning model may be utilized to extract feature elements of multiple content attribute categories of recommendation information to be optimized. For example, taking the presentation form of the advertisement to be optimized as an example, different frames in the advertisement to be optimized may be compared first to extract key frames in the advertisement to be optimized, and then the key frames of the advertisement to be optimized are processed by using a deep learning model to identify the presentation form included in the advertisement to be optimized. Here, the deep learning model in the embodiments of the present disclosure may be a trained and fixed model for video analysis; alternatively, a neural network may be used to build a deep learning model, such as a Convolutional Neural Network (CNN), a long-term short memory model (LSTM), etc., and train it with video samples from an open source video database, which is not particularly limited by the embodiments of the present disclosure.

After extracting the feature elements of the plurality of categories from the information to be optimized, for each category of feature elements, at least one result data of the information to be optimized may be predicted by step S620, and the feature elements of the category of the information to be optimized may be optimized accordingly. Step S620 may in turn include steps S621, S622, and S623.

In step S621, for each category of feature elements, the recommendation information to be optimized is divided into a certain group of at least two groups, wherein the attribute of the category of feature elements of the recommendation information within each of the at least two groups is mutually exclusive with the attribute of the category of feature elements of the recommendation information within the other of the at least two groups. Here, the at least two packets may include two or more packets, which may be specifically determined according to actual requirements, which is not specifically limited in the embodiments of the present disclosure.

Taking the recommendation information to be optimized as the advertisement to be optimized and the at least two groups comprise two groups as an example, according to the example of the embodiment of the disclosure, for each of the feature elements of the plurality of categories, classifying the advertisement to be optimized into a first group if the value or the number of the feature elements of the category of the advertisement to be optimized satisfies a predetermined condition; otherwise, the advertisement to be optimized is divided into a second group. The predetermined condition may be determined according to a numerical value or a number distribution of feature elements, or an actual requirement, for example, the predetermined condition may be that the value or the number of feature elements of the category of the advertisement is greater than, less than or equal to a second predetermined threshold, or the predetermined condition may also be that the value or the number of feature elements of the category of the advertisement belongs to a specific category, which is not specifically limited in the embodiments of the present disclosure. For example, regarding the resolution of the advertisement to be optimized, if the resolution of the advertisement to be optimized is greater than a second predetermined threshold, the advertisement to be optimized may be divided into a first group A1; otherwise, dividing the advertisement to be optimized into a second packet A2 when the resolution of the advertisement to be optimized is less than or equal to a second predetermined threshold.

In step S622, at least one result data of the recommendation information to be optimized is predicted based on the group in which the recommendation information to be optimized is located, according to predetermined evaluation information on the group of feature elements for the category. Here, the predetermined evaluation information may be, for example, information indicating the influence of different groupings of feature elements for different categories on different result data, which may include at least a part of the delivery effect data, the viewer behavior data, and the industry distribution data, as described previously. For example, in the above example, with respect to the resolution of the advertisement to be optimized, assuming that the advertisement to be optimized is divided into the first group A1 of high resolution, and the predetermined evaluation information for the first group A1 of resolution indicates the influence of the first group A1 of resolution on different result data, for example, the high resolution makes the deviation variance of the click rate of the advertisement small, the average value high, etc., at least one result data of the advertisement to be optimized may be predicted accordingly. The predetermined evaluation information may be acquired by the recommended information processing method 100 as described above, may be, for example, evaluation information about characteristic elements of different categories contained in a recommended information analysis model generated by the recommended information processing method 100, or may be information generated in advance by other methods, which is not particularly limited in the embodiment of the present disclosure.

After predicting at least one result data of the recommendation information to be optimized for the feature element of the current category, in step S623, the feature element of the category of the recommendation information to be optimized may be optimized using the predicted at least one result data. For example, regarding the resolution of the advertisement to be optimized, assuming that the advertisement to be optimized is divided into the first group A1 of low resolution, and the result data predicted according to step S622 performs poorly, for example, the deviation variance of the predicted advertisement click rate is large and the average value is small, at this time, the resolution of the advertisement to be optimized may be optimized by increasing it.

By using the recommendation information generation method according to the embodiment, the feature elements of different categories contained in any recommendation information to be optimized can be optimized and adjusted, so that the recommendation information to be optimized can have a better throwing effect.

A recommended information processing apparatus according to an embodiment of the present disclosure is described below with reference to fig. 7. Fig. 7 illustrates a schematic configuration diagram of a recommendation information processing apparatus 700 according to an embodiment of the present disclosure. As shown in fig. 7, the recommended information processing device 700 includes a recommended information acquiring unit 710, a feature extracting unit 720, and a recommended information processing unit 730. In addition to these three units, the recommended information processing apparatus 700 may include other components, however, since these components are not related to the contents of the embodiments of the present disclosure, illustration and description thereof are omitted here.

The recommendation information obtaining unit 710 is configured to obtain a recommendation information set containing a plurality of recommendation information. By way of example and not limitation, each recommendation information in the set of recommendation information may be an advertisement, such as a video advertisement, i.e., an advertisement presented in the form of a video. The recommendation information set may be obtained from an open-source advertisement database, for example, the recommendation information processing apparatus according to the embodiment of the present disclosure may perform analysis using existing data, such as obtaining valuable advertisement information from historical advertisement data to perform advertisement analysis; alternatively, the recommendation information set may be recommendation information about to be delivered, for example, a part or all of advertisements of a specific advertiser about to be delivered, that is, the recommendation information processing device according to the embodiment of the present disclosure may perform tracking analysis on the recommendation information about to be delivered, so as to analyze and adjust the delivery effect of the recommendation information in real time.

The feature extraction unit 720 is configured to extract feature elements of a plurality of categories of each of the recommendation information in the recommendation information set to obtain a feature element set. Taking advertisements as an example, each advertisement is made up of a plurality of different categories of feature elements. For example, a teletext advertisement may include characteristic elements of the class of colors, shapes, fonts, patterns, etc. The video advertisement may include more abundant feature elements, for example, the feature elements of the video advertisement may be divided into a basic attribute category and a content attribute category, where the basic attribute category may include, for example, feature elements of various categories such as frame rate, code rate, size, resolution, saturation, definition, lens number, and the like of the advertisement; the content attribute categories may include, for example, feature elements of various categories of shots, scenes, people, objects, presentations, etc. of the advertisement. Here, for the video advertisement, the presentation form of the advertisement refers to a manner in which content included in the advertisement is presented, for example, the presentation form of the video advertisement may include at least one of a dialogue, a mouth cast, a screen recording, a presentation (PPT), etc., or any combination thereof.

The steps of feature extraction unit 720 to extract feature elements will be described with reference to fig. 2 by taking video advertisement as an example. As shown in fig. 2, for video advertisements, extracting feature elements of multiple categories of each recommendation information in the set of recommendation information may include extracting feature elements of multiple underlying attribute categories and feature elements of multiple content attribute categories for each recommendation information.

Specifically, for example, for feature elements of a plurality of basic attribute categories of a video advertisement, the feature extraction unit 720 may extract feature elements of a frame rate, a code rate, a size, a resolution, saturation, sharpness, and the like of the video advertisement using a multimedia video processing tool such as FFMPEG, and may extract the number of lenses of the video advertisement using a lens boundary detection algorithm such as a pixel comparison method, a double threshold detection method, and the like. For feature elements of multiple content attribute categories of video advertisements, wherein feature elements such as shots, scenes, characters, etc. of the video advertisements may also be extracted using shot boundary detection algorithms such as pixel comparison, dual threshold detection, etc., but the disclosure is not limited thereto. According to an example of an embodiment of the present disclosure, the feature extraction unit 720 may extract feature elements of a plurality of content attribute categories of advertisements in the advertisement set using a deep learning model. For example, taking the presentation form of the video advertisement as an example, the feature extraction unit 720 may first compare different frames in the video advertisement to extract key frames in the video advertisement, and then process the key frames of the video advertisement using a deep learning model to identify the presentation form included in the video advertisement. Here, the deep learning model in the embodiments of the present disclosure may be a trained and fixed model for video analysis; alternatively, a neural network may be used to build a deep learning model, such as a Convolutional Neural Network (CNN), a long-term short memory model (LSTM), etc., and train it with video samples from an open source video database, which is not particularly limited by the embodiments of the present disclosure.

Further, the feature extraction unit 720 may be further configured to generate, when feature elements of a plurality of content attribute categories of each recommendation information in the recommendation information set are extracted using the deep learning model, a confidence that the recommendation information contains the extracted feature elements of each content attribute category, and a confidence that the extracted feature elements of the plurality of content attribute categories appear in the recommendation information. For example, for the presentation form of a video advertisement, the feature extraction unit 720 may also obtain the frame position or the time axis position where the conversation and the screen appear in the video advertisement, and generate the confidence that the video advertisement includes the conversation and the screen, and the confidence that the conversation and the screen are at the frame position or the time axis position, after determining that the video advertisement includes the conversation and the screen using the deep learning model, for subsequent adjustment and optimization of the advertisement analysis result, as will be described further below.

The extracted feature elements of the plurality of categories of each recommendation information in the recommendation information set form a feature element set, and the feature element set can comprise values of various feature elements such as resolution, saturation, definition, lens number and the like of each advertisement in the plurality of advertisements. After extracting the feature element set from the recommendation information set, the recommendation information processing unit 730 may determine, for each category of feature elements in the feature element set, an effect of the category of feature elements on result data (which may be referred to as result data) after delivery of the recommendation information, as follows.

First, the recommendation information processing unit 730 is configured to divide a plurality of recommendation information in the recommendation information set into at least two groups for each class of feature elements in the feature element set, wherein an attribute of the class of feature elements of the recommendation information in each of the at least two groups is mutually exclusive with an attribute of the class of feature elements of the recommendation information in other of the at least two groups. Here, the mutual exclusion of the attributes of the two feature elements may mean that the values or the numbers of the two feature elements have diametrically opposite attributes or attributes that are not compatible with each other. For example, for video advertisements, the attributes of high and low resolution, high and low numbers of shots, recorded and non-recorded presentations, etc. are diametrically opposed, the advertisements in the advertisement set may be divided accordingly into two groupings such that the advertisements in the two groupings have diametrically opposed attributes. For example, for video advertisements, the high resolution, medium resolution, and low resolution, the large number of shots, medium and small number of shots, large number of recordings, medium and small number of recordings, etc. are incompatible with each other, and the plurality of advertisements in the advertisement collection may be divided into three groupings accordingly such that the advertisements in the three groupings have properties that are incompatible with each other. It should be understood that only the example of dividing the advertisements in the advertisement sets into two groups and three groups is schematically listed here by way of example only, but the embodiments of the present disclosure are not limited thereto and the advertisements in the advertisement sets may be divided into more groups according to actual needs.

Thereafter, the recommendation information processing unit 730 is configured to acquire at least one kind of result data corresponding to the recommendation information within each of the at least two packets. The at least one result data corresponding to the recommended information in each group refers to various result data generated after the recommended information of the group is put in, and may include at least a part of put effect data, viewer behavior data, industry distribution data, and the like, for example. Taking an advertisement as an example, the placement effect data may include an advertisement click rate, which may be a ratio of a number of people who point to click on a product link included in the advertisement to a total number of people who view the advertisement, an advertisement conversion rate, which may be a ratio of a number of people who acquire a product by clicking on a product link included in the advertisement to a total number of people who click on a product link included in the advertisement, and the like. The viewer behavior data may include a viewing duration, a number of views, a number of daily repeat views, a run-out rate, a jump rate, a conversion time point, a conversion rate, and the like of the advertisement viewer. The completion rate may refer to a ratio of the number of people who completely watch the video advertisement to the total number of people who open the video advertisement; the jump rate refers to the ratio of the number of people who jump out of the advertisement page to the total number of people watching the advertisement due to reasons such as clicking on a product link included in the advertisement, closing the advertisement page, etc.; the conversion time point refers to a time point when a viewer clicks a product link included in an advertisement in the process of viewing a video advertisement; conversion success refers to the ratio of the number of people purchasing a product by clicking on the product link contained in the advertisement to the total number of people viewing the video advertisement. The industry distribution data represents the distribution characteristics of the industries to which the advertisement viewers belong, for example, the industry distribution data may include the proportion of the number of people belonging to the financial industry, the proportion of the number of people belonging to the teacher industry, and the like, and further, the above-mentioned delivery effect data and the viewer behavior data are combined, so that the delivery effect data and the viewer behavior data in different industries may be respectively obtained through statistics.

According to an example of an embodiment of the present disclosure, after the recommendation information processing unit 730 acquires various result data of all recommendation information in the recommendation information set, these result data may be associated with respective groups obtained for different categories of feature elements, i.e., respective groups for different categories of feature elements are associated with their corresponding result data. Referring to fig. 4, the advertisement sets are described as being divided into two groups. Fig. 4 shows a schematic diagram of result data associated with various packets according to an example of an embodiment of the disclosure. As shown in fig. 4, for the first group A1 and the second group A2 of the advertisement set divided for the feature element category a, the first group A1 and the second group A2 are respectively associated with a plurality of kinds of result data including at least a part of the impression effect data, the viewer behavior data, the industry distribution data, and the like corresponding thereto; similarly, for the first group B1 and the second group B2 of the advertisement set divided for the feature element category B, the first group B1 and the second group B2 are respectively associated with a plurality of result data corresponding thereto including at least a part of the delivery effect data, the viewer behavior data, the industry distribution data, and the like, and so on, and will not be described in detail herein.

Thereafter, the recommendation information processing unit 730 is configured to calculate, for each of at least one kind of result data corresponding to each of at least two groups of the acquired recommendation information set, a degree of difference between the result data of any two groups, and determine that the feature element of the current category has significance to the result data in a case where the degree of difference is greater than a first predetermined threshold.

Specifically, for each result data, the recommendation information processing unit 730 may first calculate a total average value of the result data corresponding to any two groups, for example, may add up the result data corresponding to any two groups, and divide the resulting added up sum by the total number of recommendation information in the two groups to obtain the above total average value; then, in each of the two groups, a difference value (may be referred to as a back deviation) between the result data of the respective recommendation information in the group and the total average value is calculated, and a variance (may be referred to as a deviation variance) of the back deviation of the respective recommendation information is calculated, for example, the deviation variances of the two groups may be referred to as a first deviation variance and a second deviation variance, respectively; the degree of difference of the first deviation variance and the second deviation variance is calculated, for example, the degree of difference may be an absolute difference of the first deviation variance and the second deviation variance. It should be noted that, although the degree of difference between the result data of any two packets is calculated by the variance of the deviation of the result data of each packet is described above, the embodiment of the present disclosure is not limited thereto. The degree of difference between the result data of any two packets may also be calculated by an average value of the result data of each packet or any other statistical index, which is not particularly limited by the embodiments of the present disclosure.

Referring to fig. 5, a concrete step of calculating the degree of difference by the recommendation information processing unit 730 is further described taking still an example of dividing the advertisement sets into two groups, and in this example, the grouping of the advertisement sets divided for the feature element category a is taken as an example.

As shown in fig. 5, for feature element category a, advertisements in the advertisement collection are divided into a first group A1 and a second group A2, wherein the first group A1 and the second group A2 have corresponding multiple result data including at least a portion of impression effect data, viewer behavior data, industry distribution data, and the like, respectively. For each result data, the recommendation information processing unit 730 first calculates a total average value of the result data of the recommendation information of the first and second packets A1, A2, for example, a quotient of the sum of the result data of the recommendation information of the packets A1, A2 and the total number of the recommendation information of the packets A1, A2 may be taken as the above total average value; then, respectively calculating a first deviation variance of the result data corresponding to the first group A1 and a second deviation variance of the result data corresponding to the second group A2, for example, respectively calculating a first deviation variance A1-1 of the advertisement click rate in the first group A1 and a second deviation variance A2-1 of the advertisement click rate in the second group A2; a first deviation variance A1-2 of advertisement conversion rate in the first group A1 and a second deviation variance A2-2 of advertisement conversion rate in the second group A2; a first deviation variance A1-3 of the viewer completion rate in the first group A1 and a second deviation variance A2-3 of the viewer completion rate in the second group A2, etc.

In case that there are two packets having the degree of difference of the result data corresponding to the two packets greater than the first predetermined threshold, the recommendation information processing unit 730 may determine that the feature element of the current category has significance to the result data. In the presently disclosed embodiments, significance means that the value of a characteristic element has a significant impact on the result data, i.e. the value of the result data can be significantly changed, for example as described above, such that the degree of difference between the result data for any two groupings of the characteristic element is greater than a first predetermined threshold. Still referring to the example in FIG. 5 above, for example, if the degree of difference between the advertisement click-through rate within the first group A1 and the second deviation variance A2-1 within the second group A2 is greater than a first predetermined threshold, it may be determined that the characteristic element category A has significance to the advertisement conversion rate. For another example, if the degree of difference between the viewer completion rate in the first group A1 within the first group A1 and the second deviation variance A1-3 in the second group A2 is greater than a first predetermined threshold, it may be determined that the feature element class a has significance to the viewer completion rate. Here, the first predetermined threshold may be determined according to actual application requirements, which is not particularly limited by the embodiments of the present disclosure.

According to an example of an embodiment of the present disclosure, in a case where it is determined that a feature element of a certain category has significance to certain result data, the recommendation information processing unit 730 may output information regarding the feature element of the category, for example, evaluation information indicating an influence of the feature element of the category on the result data. For example, as described above, in the case where it is determined that the feature element category a has significance to the advertisement click rate, evaluation information indicating the influence of the feature element category a on the advertisement click rate may be further output. Taking the resolution of the video advertisement as the characteristic element category a as an example, assuming that the resolution of the video advertisement in the first group A1 is high, the resolution of the video advertisement in the second group A2 is low, and the first deviation variance A1-1 of the advertisement click rate within the first group A1 is much smaller than the second deviation variance A2-1 of the advertisement click rate within the second group A2, i.e., the degree of difference therebetween is greater than a first predetermined threshold, evaluation information such as "high resolution makes the deviation variance of the advertisement click rate smaller, performs better" or the like may be output. Alternatively, if the average value is employed instead of deviating from the variance to calculate the degree of difference in the above example, evaluation information such as "high resolution makes the average value of advertisement click rate larger" may be output. According to examples of embodiments of the present disclosure, the rating information may be used to produce a recommendation information analysis model that provides recommendation information production guidance, such as an advertisement analysis model, as will be described further below.

Further, according to an example of an embodiment of the present disclosure, the information about the feature elements of the category that the recommendation information processing unit 730 may output may further include viewer feature information indicating common features of viewers of the recommendation information in the group in which the result data is good, for example, for generating recommendation information-oriented crowd recommendations. The viewer characteristic information may be, for example, information indicating the characteristics of the viewer such as age, sex, industry, taste, region, etc., which may reflect the influence of the characteristic elements of the category on the viewing behavior of the crowd having these characteristics.

For example, in the example in which the above-described feature element category a is the resolution of a video advertisement, assuming that a high resolution has been determined such that the deviation variance of the advertisement click rate is smaller, for the first group A1 and the second group A2 obtained by grouping for the resolution, if the first deviation variance A1-1 of the advertisement click rate within the first group A1 is smaller, that is, the deviation of the advertisement click rate within the first group A1 from the total average value is smaller, the common feature of viewers of the advertisement within the first group A1 can be extracted at this time. For example, all features of viewers of the advertisement within the first group A1 may be extracted, and the probabilities of the respective features may be statistically analyzed, with one or more features having the highest probabilities being common features of those viewers. Assuming that the advertisement viewers within the first group A1 all have a common characteristic of "male, 18-25 years old, game industry", viewer characteristic information indicating the common characteristic may be output, which may indicate that, for example, a high-resolution video advertisement has a large influence on a crowd having a characteristic of "male, 18-25 years old, game industry". According to examples of embodiments of the present disclosure, viewer characteristic information may be used to generate ad targeting crowd recommendations, as will be described further below. Then, when video advertisements are placed, if an optimal advertisement click-through rate is desired, a selection of people with common characteristics of "men, 18-25 years old, gaming industry" may be recommended for targeted placement to obtain better advertisement click-through benefits.

In the example of fig. 5, up to this point, the recommendation information processing unit 730 may end the analysis for the feature element category a. Next, the recommended information processing unit 730 may continue to analyze the other feature element categories using step S130 as described above, for example, may continue to analyze the feature element category B to determine the influence of the feature element category B on different result data, and output corresponding evaluation information and viewer feature information; then, the characteristic element class C may be analyzed to determine the influence of the characteristic element class C on different result data, and output corresponding evaluation information, viewer characteristic information, and the like, which are not described herein.

According to an example of an embodiment of the present disclosure, evaluation information on different categories of feature elements output by the recommendation information processing apparatus according to an embodiment of the present disclosure may be used to generate a recommendation information analysis model, such as an advertisement analysis model, that provides recommendation information production guidance. The recommendation information analysis model contains evaluation information indicating the influence of different types of characteristic elements on different result data, such as the influence of resolution on advertisement click rate, the influence of advertisement presentation form on advertisement conversion rate and the like. For example, a map reflecting the mapping relation of each characteristic element and the delivery result data expression may be generated based on the obtained various evaluation information, and the map may be built in the advertisement analysis model. Thus, when generating a new advertisement, desired result data, such as desired advertisement click-through rate, advertisement conversion rate, etc., may be input into an advertisement analysis model, which uses the mapping table to guide the generation of the new advertisement, for example, to output the new advertisement with the best desired performance. For example, the advertisement analysis model may direct which feature element categories to choose and what values or numbers to take for these feature element categories are most likely to achieve the desired outcome data based on the mapping table, or may directly provide multiple advertisements with the best performance for selection by the user, thereby helping to produce better quality advertisements. For another example, a neural network may be utilized to train an advertisement analysis model capable of intelligently generating new advertisements based on the various evaluation information obtained, e.g., product attributes and advertisement requirements such as "high click-through rate" may be directly entered, new advertisements with optimal click-through rates may be generated, and so forth. In addition, with the outputted information on the feature elements of different categories, or the advertisement analysis model containing such information, for an unknown new advertisement, the impression of the new advertisement can be predicted by extracting and analyzing the feature elements of a plurality of categories contained in the new advertisement, as will be described in further detail below.

According to an example of an embodiment of the present disclosure, viewer feature information about feature elements of different categories output by a recommendation information processing apparatus according to an embodiment of the present disclosure may be used to generate recommendation information targeting crowd recommendations. After the new recommendation information is generated by using the recommendation information analysis model as described above, the group of people most suitable for the new recommendation information can be targeted by using the viewer characteristic information about the characteristic elements of the categories based on the category of the characteristic elements adopted in the new recommendation information and the numerical value or the number thereof, thereby being beneficial to obtaining more ideal benefits.

In addition, the recommended information processing apparatus according to the embodiment of the present disclosure may also adjust the output information on the feature elements of different content attribute categories by adjusting the confidence of the feature elements of the content attribute categories. For characteristic elements of the content attribute category, such as video advertisement presentation forms, when the presentation form of the video advertisement is very complex, the presentation form of the video advertisement obtained through the deep learning model may have errors, thereby affecting the accuracy of the finally output information. At this time, as described above, when feature elements of a plurality of content attribute categories in an advertisement are extracted using a deep learning model, a confidence that the advertisement contains the extracted feature element of each content attribute category, and a confidence that the feature element of that content attribute category appears in the advertisement may be generated. For example, after outputting information (e.g., evaluation information) about the feature elements of a certain content attribute category, if the information is found to be inaccurate after verifying the information in connection with the actual situation, the information may be adjusted and optimized by adjusting the confidence level of the feature elements of the content attribute category to output the most accurate information about the feature elements of the content attribute category, thereby ensuring the accuracy of advertisement analysis.

In practical applications, the feature elements of the respective categories employed by the recommendation information may be updated frequently, or more specifically, different versions may exist for the feature elements of the same category. For example, in the recommendation information collection acquired by the recommendation information acquiring unit 710, different recommendation information may contain feature elements of the same category of different versions. Taking the characteristic elements of the character class as an example, for example, the advertisement 1 and the advertisement 2 in the advertisement set both contain characters, but the advertisement 1 with earlier time contains the character a, and the advertisement 2 with later time contains the character b updated, and the influence of the character a and the character b on the result data may be distinct. If the advertisement analysis is performed only for the character, which is a characteristic element category, regardless of the update of the character version, key information of the influence of the character update on the result data may be omitted. Conventional advertisement analysis methods often can only handle static (i.e., fixed) feature elements, and it is difficult to provide solutions to the problems caused by such feature element updates.

The recommended information processing device according to the embodiment of the present disclosure can effectively solve the above-mentioned problems, unlike the conventional advertisement analysis method which can only rely on the service log after advertisement delivery, the recommended information processing device 700 according to the embodiment of the present disclosure can intervene in the delivery link. Specifically, for example, in the advertisement generation phase, a time stamp may be appended to each category of feature elements that the advertisement contains, which may indicate, for example, the generation time, version, etc. of the category of feature elements for subsequent advertisement analysis. For example, in the case where the feature elements of different categories included in the advertisements in the advertisement set acquired by the recommendation information acquiring unit 710 all have corresponding time stamps, when the feature extracting unit 720 extracts the feature elements of different categories of each advertisement in the advertisement set, the time stamp of the feature element of each category may also be extracted. At this time, for different versions indicated by the time stamps of the feature elements of each category, the recommendation information processing unit 730 may analyze and output corresponding information about the category of the feature element of the version for the different versions, respectively, and generate an advertisement analysis model having a time partition according thereto. That is, in the advertisement analysis model, for each category of feature elements, time partitioning may be performed based on the time stamp thereof, and each version of the category of feature elements may be associated with the corresponding output information thereof, so that the influence of the feature elements of different versions on the advertisement delivery result may be effectively obtained.

By using the recommended information processing device according to the embodiment of the disclosure, the influence of the characteristic elements of different categories on the recommended information delivery result can be accurately analyzed, the recommended information production guidance and the directional crowd recommendation of the recommended information delivery can be provided, and the delivery effect of the recommended information can be predicted.

A recommendation information generating apparatus according to an embodiment of the present disclosure is described below with reference to fig. 8. Fig. 8 illustrates a schematic configuration diagram of a recommendation information generating device 800 according to an embodiment of the present disclosure. As shown in fig. 8, the recommendation information generating device 800 includes a feature extraction unit 810 and a recommendation information generating unit 820. In addition to these two units, the recommendation information generation apparatus 800 may include other components, however, since these components are not related to the contents of the embodiments of the present disclosure, illustration and description thereof are omitted here.

As shown in fig. 8, the feature extraction unit 810 is configured to extract feature elements of a plurality of categories in the recommendation information to be optimized. Here, the recommendation information to be optimized may be any information such as advertisement, for example, may include a teletext advertisement, a video advertisement, and the like. The following description will take the recommendation information to be optimized as video advertisement as an example.

According to an example of an embodiment of the present disclosure, extracting feature elements of a plurality of categories of recommendation information to be optimized may include extracting feature elements of a plurality of base attribute categories and feature elements of a plurality of content attribute categories of recommendation information to be optimized. Specifically, taking an advertisement as an example, for feature elements of a plurality of basic attribute categories of the advertisement to be optimized, the feature extraction unit 810 may extract feature elements of a frame rate, a code rate, a size, a resolution, saturation, sharpness, etc. of the advertisement to be optimized using a multimedia video processing tool such as FFMPEG, etc., and may extract the number of lenses of the advertisement to be optimized using a lens boundary detection algorithm such as a pixel comparison method, a double threshold detection method, etc. For feature elements of a plurality of content attribute categories of the advertisement to be optimized, wherein feature elements such as shots, scenes, characters, etc. of the advertisement to be optimized may also be extracted using a shot boundary detection algorithm such as a pixel comparison method, a double threshold detection method, etc., but the present disclosure is not limited thereto.

According to an example of an embodiment of the present disclosure, the feature extraction unit 810 may extract feature elements of a plurality of content attribute categories of recommendation information to be optimized using a deep learning model. For example, taking the presentation form of the advertisement to be optimized as an example, the feature extraction unit 810 may first compare different frames in the advertisement to be optimized to extract key frames in the advertisement to be optimized, and then process the key frames of the advertisement to be optimized using the deep learning model to identify the presentation form included in the advertisement to be optimized. Here, the deep learning model in the embodiments of the present disclosure may be a trained and fixed model for video analysis; alternatively, a neural network may be used to build a deep learning model, such as a Convolutional Neural Network (CNN), a long-term short memory model (LSTM), etc., and train it with video samples from an open source video database, which is not particularly limited by the embodiments of the present disclosure.

After extracting the feature elements of the plurality of categories from the recommendation information to be optimized, for each category of feature elements, the recommendation information generating unit 820 is configured to predict at least one result data of the recommendation information to be optimized, and accordingly optimize the feature elements of the category of the recommendation information to be optimized, as follows.

First, the recommendation information generation unit 820 is configured to divide the recommendation information to be optimized into a certain group of at least two groups for each category of feature elements, wherein the attribute of the category of feature elements of the recommendation information within each of the at least two groups is mutually exclusive with the attribute of the category of feature elements of the recommendation information within other groups of the at least two groups. Here, the at least two packets may include two or more packets, which may be specifically determined according to actual requirements, which is not specifically limited in the embodiments of the present disclosure.

Thereafter, the recommendation information generation unit 820 is configured to predict at least one kind of result data of the recommendation information to be optimized based on the group in which the recommendation information to be optimized is located, according to predetermined evaluation information on the group for the feature elements of the category. Here, the predetermined evaluation information may be, for example, information indicating the influence of different groupings of feature elements for different categories on different result data, which may include at least a part of the delivery effect data, the viewer behavior data, and the industry distribution data, as described previously. For example, in the above example, with respect to the resolution of the advertisement to be optimized, assuming that the advertisement to be optimized is divided into the first group A1 of high resolution, and the predetermined evaluation information for the first group A1 of resolution indicates the influence of the first group A1 of resolution on different result data, for example, the high resolution makes the deviation variance of the click rate of the advertisement small, the average value high, etc., at least one result data of the advertisement to be optimized may be predicted accordingly. The predetermined evaluation information may be acquired by the recommended information processing method 100 as described above, may be, for example, evaluation information on feature elements of different categories contained in an advertisement analysis model generated by the recommended information processing method 100, or may be information generated in advance by other methods, which is not particularly limited in the embodiment of the present disclosure.

After predicting at least one result data of the advertisement to be optimized for the feature element of the current category, the recommendation information generation unit 820 is configured to optimize the feature element of the category of the recommendation information to be optimized using the predicted at least one result data. For example, regarding the resolution of the advertisement to be optimized, assuming that the advertisement to be optimized is divided into a first group A1 with low resolution and the predicted result data performs poorly, for example, the deviation variance of the predicted advertisement click rate is large and the average value is small, the advertisement to be optimized can be optimally adjusted by increasing the resolution thereof.

By using the recommendation information generation device according to the embodiment, the feature elements of different categories contained in any recommendation information to be optimized can be optimized and adjusted, so that the recommendation information to be optimized can have a better advertisement putting effect.

Further, devices (e.g., recommendation information processing devices, recommendation information generating devices, etc.) according to embodiments of the present disclosure may also be implemented by means of the architecture of the exemplary computing device shown in fig. 9. Fig. 9 shows a schematic diagram of an architecture of an exemplary computing device, according to an embodiment of the present disclosure. As shown in fig. 9, computing device 900 may include a bus 910, one or more CPUs 920, a Read Only Memory (ROM) 930, a Random Access Memory (RAM) 940, a communication port 950 connected to a network, an input/output component 960, a hard disk 970, and the like. A storage device in computing device 900, such as ROM 930 or hard disk 970, may store various data or files for computer processing and/or communication and program instructions for execution by the CPU. Computing device 900 may also include a user interface 980. Of course, the architecture shown in FIG. 9 is merely exemplary, and one or more components of the computing device shown in FIG. 9 may be omitted as may be practical in implementing different devices. The apparatus according to the embodiments of the present disclosure may be configured to perform the recommended information processing method or the recommended information generating method according to the above-described respective embodiments of the present disclosure, or to implement the recommended information processing device or the recommended information generating device according to the above-described respective embodiments of the present disclosure.

Embodiments of the present disclosure may also be implemented as a computer-readable storage medium. Computer readable storage media according to embodiments of the present disclosure have computer readable instructions stored thereon. The recommended information processing method or recommended information processing generating method according to the embodiments of the present disclosure described with reference to the above drawings may be performed when the computer readable instructions are executed by the processor. Computer-readable storage media include, but are not limited to, volatile memory and/or nonvolatile memory, for example. Volatile memory can include, for example, random Access Memory (RAM) and/or cache memory (cache) and the like. The non-volatile memory may include, for example, read Only Memory (ROM), hard disk, flash memory, and the like.

According to an embodiment of the present disclosure, there is also provided a computer program product or a computer program comprising computer readable instructions stored in a computer readable storage medium. The processor of the computer device may read the computer-readable instructions from the computer-readable storage medium, and execute the computer-readable instructions, so that the computer device performs the recommended information processing method or the recommended information generating method described in the respective embodiments described above.

Those skilled in the art will appreciate that various modifications and improvements can be made to the disclosure. For example, the various devices or components described above may be implemented in hardware, or may be implemented in software, firmware, or a combination of some or all of the three.

Furthermore, as shown in the present disclosure and claims, unless the context clearly indicates otherwise, the words "a," "an," "the," and/or "the" are not specific to the singular, but may include the plural. The terms "first," "second," and the like, as used in this disclosure, do not denote any order, quantity, or importance, but rather are used to distinguish one element from another. Likewise, the word "comprising" or "comprises", and the like, means that elements or items preceding the word are included in the element or item listed after the word and equivalents thereof, but does not exclude other elements or items. The terms "connected" or "connected," and the like, are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

Further, a flowchart is used in this disclosure to describe the operations performed by the system according to embodiments of the present disclosure. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously. Also, other operations may be added to the processes or a step or steps may be removed from the processes.

Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

While the present disclosure has been described in detail above, it will be apparent to those skilled in the art that the present disclosure is not limited to the embodiments described in the present specification. The present disclosure may be embodied as modifications and variations without departing from the spirit and scope of the disclosure, which is defined by the appended claims. Accordingly, the description herein is for the purpose of illustration and is not intended to be in any limiting sense with respect to the present disclosure.

Claims

1. A recommended information processing method, comprising:

acquiring a recommendation information set containing a plurality of recommendation information;

extracting characteristic elements of a plurality of categories of each piece of recommended information in the recommended information set to obtain a characteristic element set; and

Feature elements for each category in the feature element set:

dividing a plurality of recommendation information in the recommendation information set into at least two groups, wherein the attribute of the characteristic element of the category of recommendation information in each group of the at least two groups is mutually exclusive with the attribute of the characteristic element of the category of recommendation information in other groups of the at least two groups;

acquiring at least one result data corresponding to the recommendation information in each of the at least two groups;

for each of the at least one result data, calculating a degree of difference between the result data of any two of the at least two groupings, and determining that the category of feature elements has significance to the result data if the degree of difference is greater than a first predetermined threshold.

2. The recommended information processing method according to claim 1, further comprising:

in case it is determined that the characteristic elements of the category have significance to the result data, outputting information about the characteristic elements of the category,

wherein the information includes evaluation information indicating an influence of the characteristic element of the category on the result data,

The method further includes generating a recommendation information analysis model for providing recommendation information production guidance based on the evaluation information.

3. The recommended information processing method according to claim 2, wherein the information further includes viewer characteristic information indicating common characteristics of viewers of recommended information in a group in which result data is good in the two groups, the method further comprising:

generating recommendation information-oriented crowd recommendation based on the viewer characteristic information.

4. The recommended information processing method according to claim 2, further comprising:

based on the outputted information about the feature elements of different categories, the delivering effect of the new recommendation information is predicted by extracting and analyzing the feature elements of a plurality of categories in the new recommendation information.

5. The recommended information processing method according to claim 1, wherein the characteristic elements of the plurality of categories include characteristic elements of a plurality of basic attribute categories and/or a plurality of content attribute categories, wherein,

the characteristic elements of the plurality of basic attribute categories comprise at least one of frame rate, code rate, size, resolution, saturation, definition and lens number of recommended information;

The feature elements of the plurality of content attribute categories include at least one of shots, scenes, characters, objects, and presentations of the recommendation information.

6. The recommended information processing method according to claim 5, wherein the at least two packets include a first packet and a second packet,

wherein dividing the plurality of recommendation information in the recommendation information set into at least two groupings for each category of feature element in the feature element set comprises:

for each of the plurality of recommendation information, dividing the recommendation information into a first group in a case where a value or a number of feature elements of the category of the recommendation information satisfies a predetermined condition; otherwise, the recommendation information is divided into a second group.

7. The recommendation information processing method according to claim 5, wherein extracting feature elements of a plurality of categories of each recommendation information in the recommendation information set for feature elements of the plurality of content attribute categories comprises:

extracting feature elements of a plurality of content attribute categories in the recommendation information set by using a deep learning model; and

for each recommendation information, generating a confidence that the recommendation information contains the extracted feature elements of each content attribute category, and a confidence that the feature elements of the content attribute category appear in the recommendation information.

8. The recommended information processing method of claim 7, further comprising:

by adjusting the confidence, the outputted information about the feature elements of the different content attribute categories is adjusted.

9. The recommended information processing method according to claim 2, further comprising:

extracting a time stamp of each feature element in the feature element set; and

generating a recommended information analysis model for providing recommended information production guidance with time division based on the time stamp and the outputted information on the different categories of feature elements,

wherein the timestamp is appended to each feature element of the recommendation information when each recommendation information in the set of recommendation information is generated.

10. The recommended information processing method according to any one of claims 1 to 9, wherein the at least one kind of result data includes at least one part of delivery effect data, viewer behavior data, and industry distribution data, wherein,

the delivery effect data includes at least one of a recommended information click rate and a recommended information conversion rate,

the viewer behavior data includes at least one of a viewing duration, a number of times of viewing, a number of times of daily repeated viewing, a complete broadcast rate, a jump rate, a conversion time point, and a conversion success rate of a viewer of the recommended information,

The industry distribution data indicates a distribution characteristic of an industry to which the recommended information viewer belongs.

11. A recommendation information generation method, comprising:

extracting characteristic elements of a plurality of categories in the recommendation information to be optimized; and

feature elements for each of the plurality of categories of feature elements:

dividing the recommendation information to be optimized into groups in at least two groups, wherein the attribute of the characteristic element of the category of the recommendation information in each group in the at least two groups is mutually exclusive with the attribute of the characteristic element of the category of the recommendation information in other groups in the at least two groups;

predicting at least one result data of the recommendation information to be optimized according to predetermined evaluation information about the grouping of feature elements for the category based on the grouping in which the recommendation information to be optimized is located;

and optimizing the characteristic elements of the category of the recommended information to be optimized by utilizing the predicted at least one result data.

12. A recommended information processing apparatus comprising:

a recommendation information acquisition unit configured to acquire a recommendation information set including a plurality of recommendation information;

A feature extraction unit configured to extract feature elements of a plurality of categories of each of the recommendation information in the recommendation information set to obtain a feature element set; and

a recommendation information processing unit configured to, for each category of feature elements in the feature element set:

13. A recommended information processing apparatus comprising:

one or more processors; and

One or more memories having stored therein computer readable code which, when executed by the one or more processors, causes the one or more processors to perform the method of any of claims 1-11.

14. A computer readable storage medium having stored thereon computer readable instructions which, when executed by a processor, cause the processor to perform the method of any of claims 1-11.

15. A computer program product comprising computer readable instructions which, when executed by a processor, cause the processor to perform the method of any of claims 1-11.