Advertisement information delivery method and device applied to an advertisement system and electronic equipment

Abstract

This disclosure provides embodiments of an advertising information delivery method, apparatus, and electronic device applied to an advertising system. One specific implementation of the method includes: collecting page behavior information and page description information corresponding to a target page; for each piece of page content information in a set of page content information, determining the content attention level corresponding to the page content information based on the page behavior information and user attribute information corresponding to the target user; determining dwell tendency information for the target page based on the page description information and the content attention level corresponding to the page content information; determining whether to deliver advertising information based on the dwell tendency information; in response to delivering advertising information, determining advertising description information based on the dwell tendency information and the set of page content information; and delivering the advertising information to be delivered to the target page through the advertising system based on the advertising description information. This implementation improves the matching accuracy of advertising information delivery.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the field of computer technology, the field of information recommendation, and in particular, to an advertisement information putting method and device applied to an advertisement system and an electronic device. BACKGROUND

[0002] As a vertical subdivision field of the field of information recommendation, advertisement information recommendation is mainly used for accurately recommending advertisement information to users. At present, when performing advertisement information recommendation, an advertisement information screening and recommendation is mainly performed in a manner such as “recall + sorting”.

[0003] However, the above manner mainly takes the matching degree between advertisement information and user operation behaviors as a recommendation basis, fails to effectively pay attention to other factors affecting advertisement information recommendation, and thus it is difficult to further improve the recommendation matching degree. SUMMARY

[0004] The summary part of the present disclosure is used to introduce the concepts in a brief form, which will be described in detail in the specific embodiments part. The summary part of the present disclosure is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to be used to limit the scope of the claimed technical solutions.

[0005] Some embodiments of the present disclosure provide an advertisement information putting method and device applied to an advertisement system and an electronic device to solve the technical problems mentioned in the background part.

[0006] In a first aspect, some embodiments of the present disclosure provide an advertisement information putting method applied to an advertisement system, which comprises: collecting page behavior information and page description information corresponding to a target page, wherein the page behavior information represents a page behavior of a target user when browsing the target page, and the page description information comprises page style information, page layout information, and a page content information set; for each page content information in the page content information set, determining a content attention degree corresponding to the page content information according to the page behavior information and user attribute information corresponding to the target user, wherein the content attention degree comprises a direct attention degree and an indirect attention degree; determining page residence tendency information for the target page according to the page description information and the content attention degree corresponding to the page content information, wherein the page residence tendency information comprises a page residence tendency value and a content residence tendency value; determining whether to perform advertisement information putting according to the page residence tendency information; in response to performing advertisement information putting, determining advertisement description information according to the page residence tendency information and the page content information set, wherein the advertisement description information comprises advertisement attribute information and putting style information; and putting, by the advertisement system, advertisement information corresponding to the advertisement description information to the target page according to the advertisement description information.

[0007] Secondly, some embodiments of this disclosure provide an advertising information delivery device applied to an advertising system. The device includes: a collection unit configured to collect page behavior information and page description information corresponding to a target page, wherein the page behavior information characterizes the page behavior of a target user when browsing the target page, and the page description information includes: page style information, page layout information, and a set of page content information; a first determining unit configured to, for each piece of page content information in the set of page content information, determine the content attention level corresponding to the page content information based on the page behavior information and the user attribute information corresponding to the target user, wherein the content attention level includes: direct attention level and indirect attention level; and a second determining unit configured to... The system is configured to determine the dwell tendency information for the target page based on the content attention corresponding to the page description information and page content information, wherein the dwell tendency information includes: page dwell tendency value and content dwell tendency value; the third determining unit is configured to determine whether to deliver advertising information based on the dwell tendency information; the fourth determining unit is configured to determine advertising description information based on the dwell tendency information and the set of page content information in response to delivering advertising information, wherein the advertising description information includes: advertising attribute information and delivery style information; and the delivery unit is configured to deliver the advertising information to be delivered to the target page through the advertising system based on the advertising description information.

[0008] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0009] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0010] The above embodiments of this disclosure have the following beneficial effects: the advertising information delivery method applied to an advertising system through some embodiments of this disclosure improves the matching degree of advertising information delivery. Specifically, firstly, page behavior information and page description information corresponding to the target page are collected. Page behavior information characterizes the page behavior of the target user when browsing the target page, and page description information includes: page style information, page layout information, and a set of page content information. In practice, this disclosure has found that page style and layout affect user retention. For example, aesthetically pleasing and / or efficiently displayed pages can reduce user operations and promote the conversion (e.g., purchase) of products related to the advertising information. Therefore, this disclosure collects relevant data from two dimensions: page behavior and page description. Secondly, for each piece of page content information in the above-mentioned set of page content information, the content attention corresponding to the above-mentioned page content information is determined based on the above-mentioned page behavior information and the user attribute information corresponding to the target user. Content attention includes: direct attention and indirect attention. In practice, by focusing on the user's page behavior towards page content, the attention of the displayed page content to the user can be mapped. Meanwhile, considering that page-related operations are often optimized to reduce user complexity, and to ensure the effectiveness of targeted content attention, this disclosure further incorporates user attribute information to generate indirect attention levels. Next, based on the content attention levels corresponding to the aforementioned page description and content information, the dwell tendency information for the target page is determined. This dwell tendency information includes page dwell tendency value and content dwell tendency value. By combining page description information and content attention levels, user dwell tendency is judged from both page and content perspectives. Furthermore, based on the aforementioned dwell tendency information, it is determined whether to deliver advertising information. This achieves effective advertising information delivery under conditions of high user dwell tendency, significantly improving the effectiveness of information delivery compared to conventional indiscriminate delivery and avoiding waste of related traffic. Then, in response to delivering advertising information, based on the aforementioned dwell tendency information and page content information set, advertising description information is determined. This advertising description information includes advertising attribute information and delivery style information. When advertising needs to be delivered, the system further combines dwell time patterns and page content information to determine the ad content to be recommended and optimize the delivery style. Finally, based on the ad description information, the system delivers the corresponding ad content to the target page. In summary, this improves the matching accuracy of ad delivery. Attached Figure Description

[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0012] Figure 1 This is a flowchart of some embodiments of an advertising information delivery method applied to an advertising system according to the present disclosure;

[0013] Figure 2 This is a schematic diagram showing the style effect of the target page;

[0014] Figure 3 This is an illustration of another style effect for the target page;

[0015] Figure 4 This is a schematic diagram of the structure of some embodiments of an advertising information delivery device applied to an advertising system according to the present disclosure;

[0016] Figure 5 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0017] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0018] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0019] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0020] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0021] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0022] Before performing any of the operations involving the information disclosed herein (such as page behavior information and user attribute information), the relevant organizations or individuals shall fulfill their obligations, including conducting personal information security impact assessments, informing personal information subjects, and obtaining prior authorization and consent from personal information subjects.

[0023] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] Continue to refer to Figure 1 The flowchart 100 illustrates some embodiments of an advertising information delivery method applied to an advertising system according to the present disclosure. This advertising information delivery method applied to an advertising system includes the following steps:

[0025] Step 101: Collect page behavior information and page description information corresponding to the target page.

[0026] In some embodiments, the execution entity (e.g., a computing device) of the advertising information delivery method applied to the advertising system can collect page behavior information and page description information corresponding to the target page.

[0027] Page behavior information represents the page actions of a target user while browsing a target page. Specifically, page behavior information can be composed of a sequence of triples, including: page behavior type, behavior execution time, and behavior execution location. Page behavior type can include, but is not limited to: swiping, clicking, zooming in, zooming out, and double-clicking. Behavior execution time represents the time when the target user performs the page behavior. Behavior execution location represents the position of the page behavior within the target page.

[0028] The target page is the information display page that the target user is currently browsing. The target user is the user to whom the advertising information is to be delivered. The page description information includes: page style information, page layout information, and a set of page content information. Page style information describes the style of the target page. Page layout information describes the layout of the elements within the target page. The set of page content information describes the page content displayed on the target page (e.g., item information).

[0029] Specifically, page content information may include, but is not limited to: item type, item number, item summary description, and page content location. The page content location indicates the position of the corresponding page content within the target page.

[0030] In practice, firstly, page behavior information can be collected through event tracking. Then, the HTML (Hypertext Markup Language), JavaScript (JS), and CSS (Cascading Stylesheets) code corresponding to the target page are collected. Next, the page style information, page layout information, and page content information are determined by filtering the code using keywords.

[0031] As an example, first, see Figure 2 The image shows a schematic of the target page's style. The target page displays page content A1, A2, A3, A4, A5, A6, A7, and A8 in a two-row, side-by-side format. Next, see... Figure 3 The illustration shows another style of the target page, in which the target page displays page content A1, page content A2, page content A3, page content A4, page content A5, and page content A6 in a single row.

[0032] It should be noted that the aforementioned computing device can be either hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or terminal device. When the computing device is software, it can be installed within the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here. In practice, the aforementioned computing device can be used for an advertising system.

[0033] In some optional implementations of certain embodiments, the aforementioned execution entity collects page behavior information and page description information corresponding to the target page, including:

[0034] Step S1: Determine whether to obtain the information processing permissions authorized by the target user mentioned above.

[0035] Among them, information processing permissions represent whether the executing entity is allowed to collect, use, and store behavioral information and page description information for the target page.

[0036] In practice, permission queries can be used to determine whether the information processing permissions authorized by the target user have been granted. In particular, to ensure data security, relevant data is encrypted during transmission and anonymized as needed during use and storage.

[0037] Step S2: In response to the lack of the above information processing permission, an authorization prompt for the above information processing permission will pop up on the above target page.

[0038] The authorization prompt mentioned above is used to prompt the target user to authorize the above information processing permissions.

[0039] In practice, authorization prompts will include a detailed list of the permissions granted, as well as notifications of the corresponding legal risks.

[0040] Step S3: In response to having obtained the above information processing permissions, request a trust token.

[0041] In practice, a trust token can be requested from a dedicated trust token management server. This server is used for unified trust token management across different advertising platforms (including but not limited to trust token issuance). Specifically, since the advertising system disclosed herein can deliver advertising information through different advertising platforms, to ensure security during the information acquisition process, trust tokens are set up to guarantee identity verification for different advertising platforms when collecting information.

[0042] Step S4: In response to obtaining the aforementioned authorization token, activate the data collector within the target page and create a data transmission channel.

[0043] The data collector is embedded within the target page via data entry points. The data transmission channel is used to transmit page behavior and description information. For example, the data transmission channel can be a socket connection. The data transmission channel is controlled by the validity period of a trust token. For example, when the trust token expires, the data transmission channel is deregistered, thus avoiding memory consumption caused by the prolonged existence of the data transmission channel.

[0044] In practice, when a trust token is obtained and is valid, the data collector can be activated to collect data. Specifically, the data transmission channel uses national cryptographic algorithms for data encryption during transmission.

[0045] Step S5: Collect the page behavior information and page description information using the data collector within the target page.

[0046] In practice, firstly, when a target user engages in user behavior, a data collector captures this behavior to obtain page behavior information. Secondly, for the page content displayed on the target page, the data collector gathers page content information, page style parameters, and page layout parameters related to the page content to obtain page description information.

[0047] Step 102: For each piece of page content information in the set of page content information, determine the content attention level corresponding to the page content information based on the page behavior information and the user attribute information corresponding to the target user.

[0048] In some embodiments, the aforementioned execution entity may determine the content attention level corresponding to each page content information in the page content information set based on page behavior information and user attribute information corresponding to the target user.

[0049] Content attention includes direct attention and indirect attention. Content attention represents the level of attention a target user pays to the content displayed on the target page. User attribute information represents user attributes in the form of a user profile generated based on the target user's historical page behavior. The values ​​for both direct and indirect attention range from [0,1].

[0050] In practice, firstly, by combining page behavior information, we can statistically analyze the relevant page behaviors of target users towards the page content and map them to direct attention. Secondly, based on the similarity between user attribute information and page content information in the page content information set, we can generate indirect attention. By determining direct and indirect attention, we can quantify the target user's attention to the target page from the perspectives of user behavior and page content.

[0051] In some optional implementations of certain embodiments, the execution entity determines the content attention level corresponding to each page content information in the page content information set, based on the page behavior information and the user attribute information corresponding to the target user, including:

[0052] Step S1: Filter the page behaviors corresponding to the above page behavior information to obtain a page behavior sequence.

[0053] Among them, the page behaviors in the page behavior sequence represent the behavioral operations related to the above page content information, and the page behaviors in the page behavior sequence are arranged in order according to the corresponding behavior execution time.

[0054] In practice, because data collectors continuously capture page behavior, they will collect multiple page behaviors from the target user. Some of these behaviors may not be related to the page content. Therefore, to ensure the accuracy of the subsequently determined direct engagement, it is necessary to filter page behaviors unrelated to the page content. For example, if the target user clicks on page content A1, this behavior can be retained. Conversely, if the target user swipes on the target page, this behavior can be filtered.

[0055] Step S2: Generate a page feature map corresponding to the target page mentioned above.

[0056] The ratio of the feature size of the page feature map to the page size of the target page is K, where K is a positive integer. In practice, the feature values ​​in the page feature map are initially set to 0.

[0057] Step S2: For each page behavior in the above page behavior sequence, perform the following feature processing steps:

[0058] Step S21: Extract the page behavior features corresponding to the above page behaviors.

[0059] In practice, since the triples corresponding to page behaviors are stored in the form of text data, the Word2Vec model can be used to encode the behavior of the triples corresponding to page behaviors to obtain page behavior features.

[0060] Step S22: Project the above page behavior features onto the above page feature map to obtain the projected page behavior features.

[0061] In practice, based on the execution location of the behavior included in the triple corresponding to the page behavior, the page behavior features are projected onto the above page feature map to obtain the projected page behavior features.

[0062] Step S3: Generate the above direct attention based on the obtained post-projection page behavior feature sequence.

[0063] In practice, firstly, for each projected page behavior feature in the projected page behavior feature sequence, a downsampling network can be used to downsample the projected page behavior feature to obtain the downsampled page behavior feature. Then, the obtained downsampled page behavior feature sequences are superimposed to obtain the superimposed page behavior feature. Finally, a fully connected layer maps the superimposed page behavior features to direct attention.

[0064] Step S4: Encode the above page content information to obtain page content features.

[0065] In practice, the Word2Vec model can be used to encode the above page content information to obtain page content features.

[0066] Step S5: Extract user behavior attribute information from the above user attribute information.

[0067] In practice, user attribute information represents user attributes in the form of a user profile generated based on the target user's historical page behavior. For example, user attribute information can include static user attributes (such as user name, user gender, user age, user education, etc.). User behavior attributes can represent dynamic user attributes (such as user attributes related to page behavior). Specifically, user behavior attribute information can be extracted from user attribute information through key-value pair extraction.

[0068] Step S6: Encode the above user behavior attribute information to obtain user behavior attribute features.

[0069] In this context, page content features and user behavior attribute features reside in the same feature space.

[0070] In practice, the Word2Vec model can be used to encode the aforementioned user behavior attribute information, resulting in user behavior attribute features. By using the same information encoding method, page content features and user behavior attribute features are mapped to the same feature space, facilitating subsequent calculation of indirect attention. In particular, to ensure consistency in feature dimensions, when the feature dimensions of the encoded user behavior attribute features and page content features are inconsistent, feature downsampling of the features with larger feature dimensions is required before calculating indirect attention.

[0071] Step S7: Generate the indirect attention based on the above page content features and user behavior attribute features.

[0072] In practice, the feature similarity between page content features and the aforementioned user behavior attributes can be determined by calculating cosine similarity, which can then be used as an indirect measure of attention.

[0073] Step 103: Determine the dwell tendency information for the target page based on the content attention corresponding to the page description information and page content information.

[0074] In some embodiments, the aforementioned executing entity may determine dwell tendency information for a target page based on the content attention corresponding to the page description information and page content information.

[0075] The dwell tendency information includes: page dwell tendency value and content dwell tendency value. The page dwell tendency value represents the attractiveness of the target page to the target user. The content dwell tendency value represents the attractiveness of the page content displayed on the target page to the target user.

[0076] In practice, firstly, features can be extracted from the page description information, including page style information, page layout information, and page content information, and mapped to page dwell tendency values. Secondly, the content attention corresponding to the page content information, including direct attention and indirect attention, is weighted and summed to obtain the content dwell tendency value.

[0077] In some optional implementations of certain embodiments, the execution entity determines dwell tendency information for the target page based on the content attention corresponding to the page description information and page content information, including:

[0078] Step S1: Extract page style features from the above page style information using the style feature extraction module to obtain page style features.

[0079] The aforementioned page style features include: page color features, page texture features, and page animation features.

[0080] In practice, firstly, page code related to color, page code related to page texture, and page code related to page animation are extracted from the page style information. Then, using the LightLLM model, code effect descriptions are generated for each of these three types of page code. Next, using the RoBERTa model, features are extracted from the three text-based code effect descriptions to obtain page color features, page texture features, and page animation features.

[0081] Step S2: Generate a page layer sequence based on the page layout information above.

[0082] The page layer includes at least one page layout block that is located at the same display level and whose content is visible. Each page layer has a corresponding page layer weight, which is controlled by the layer occlusion of the corresponding page layer.

[0083] In practice, complex web pages often use multiple block elements, each with its own style attributes. Therefore, the first step is to generate the DOM (Document Object Model) tree for the target page. Then, using a breadth-first search, page layout blocks at the same depth within the DOM tree are filtered to obtain a set of page layout block groups. Each page layout block within a group resides on the same page layer. Next, for each group, page layout blocks whose corresponding page attributes indicate content visibility are filtered out, resulting in page layers. Finally, different page layer weights are assigned based on the vertical order of the page layers. Specifically, the page layer weight corresponding to an upper-level page image is greater than that corresponding to a lower-level page image. The page layer weight ranges from [0,1].

[0084] Step S3: Using the layout feature extraction module, extract layout features for each page layer in the above page layer sequence to generate candidate page layout features and obtain a candidate page layout feature sequence.

[0085] The layout feature extraction module can employ the Tiny-CNN model. For each page layer, since the page layer contains side-by-side block elements, it can be treated as an image for feature extraction. Considering that the page layer sequence does not have a temporal relationship, feature extraction can be performed on a single page layer to obtain the corresponding candidate page layout features.

[0086] Step S4: Based on the page layer weights corresponding to the page layers, perform feature fusion on the candidate page layout features in the obtained candidate page layout feature sequence to obtain the page layout features.

[0087] In practice, candidate page layout features in the candidate page layout feature sequence can be superimposed, and the page layer weight corresponding to the page layer is used as the weight when superimposing features.

[0088] Step S5: Based on the above page style features and page layout features, generate the above page dwell tendency value.

[0089] The value of the page dwell tendency value ranges from [0,1].

[0090] In practice, firstly, a dual-path feature extraction network with two convolutional neurons is used to perform feature shaping on the page style features and the aforementioned page layout features in parallel. Then, a stacking layer and a fully connected layer are added after the dual-path feature extraction network. The stacking layer is used to stack the feature-shaped page style features and the feature-shaped page layout features. The fully connected layer is used to perform feature mapping on the stacked features to obtain the page dwell tendency value.

[0091] Step S6: Based on the content position of the page content information on the target page and the content attention corresponding to the page content information, update the page layout features to obtain the updated page layout features.

[0092] In practice, for each piece of page content information in the set of page content information, the local features corresponding to the content position of the page content information on the target page are updated by using the content attention corresponding to the page content information as weight. Specifically, since the values ​​of direct and indirect attention are both in the range [0,1], when using content attention as weight, the direct and indirect attention included in the content attention are first weighted and summed. At this point, the value range of the weighted summed content attention is still [0,1]. If the weighted summed content attention is directly used as the weight for feature update, it will weaken the feature. Therefore, the weighted summed content attention needs to be incremented by 1 before feature update, thereby highlighting the features at the corresponding position of the page content.

[0093] Step S7: Based on the updated page layout features described above, generate the content dwell tendency value.

[0094] The content dwell tendency value ranges from [0,1].

[0095] In practice, by setting a fully connected layer, the updated page layout features can be mapped to the aforementioned content dwell tendency values.

[0096] Step 104: Based on the residence tendency information, determine whether to place advertising information.

[0097] In some embodiments, the aforementioned implementing entity may determine whether to deliver advertising information based on residency preference information.

[0098] In practice, an ad placement threshold can be set. When the page dwell tendency or content dwell tendency value is greater than the ad placement threshold, it indicates that ad placement is necessary. When both the page dwell tendency and content dwell tendency values ​​are less than or equal to the ad placement threshold, it indicates that ad placement is not necessary.

[0099] Step 105: In response to the delivery of advertising information, determine the advertising description information based on the set of dwell tendency information and page content information.

[0100] In some embodiments, the aforementioned implementing entity may determine the advertising description information based on a set of dwell tendency information and page content information in response to delivering advertising information.

[0101] The ad description information includes ad attribute information and ad placement style information. Ad attribute information represents the attributes of the ad to be placed. For example, ad attribute information may include, but is not limited to, ad type, ad placement source identifier, and ad content. The ad placement source identifier represents the unique identifier of the ad placement platform corresponding to the ad description information. The placement style represents the style parameters of the ad when embedded in the target page.

[0102] In some optional implementations of certain embodiments, the execution entity determines the advertisement description information based on the aforementioned dwell tendency information and the aforementioned page content information set, including:

[0103] Step S1: In response to the above page dwell tendency value being greater than the first threshold and the above content dwell tendency value being less than the second threshold, the following first processing step is executed:

[0104] Step S11: Extract the advertising information to be placed from the advertising information database.

[0105] The advertising information database is a database used to store advertising information.

[0106] In practice, when the dwell tendency value is greater than the first threshold and the dwell tendency value of the above content is less than the second threshold, the page style of the target page has a significant impact on the retention of target users. Therefore, advertising information to be delivered can be selected from the advertising information database according to the extraction rules.

[0107] Step S12: Extract the advertising attributes corresponding to the above-mentioned advertising information to be delivered, and use them as the advertising attribute information included in the above-mentioned advertising description information.

[0108] In practice, preset regular expressions can be used to extract the advertising attributes corresponding to the advertising information to be delivered, which are then used as the advertising attribute information included in the aforementioned advertising description information.

[0109] Step S13: Based on the above page style information, generate the placement style information included in the above advertisement description information.

[0110] In practice, since the page style of the target page has a significant impact on the retention rate of target users, the page style corresponding to the page content can be included as the targeting style information in the ad description. Specifically, the HTML, CSS, and JS code of the target page can be read to obtain the page style parameters related to the page content, which can then be included as the targeting style information in the ad description. This ensures that the ad information to be delivered has the same style as the page content, and under the influence of a high page dwell rate, the page style of the target page can promote the conversion of the corresponding product (e.g., sales) through the targeting page.

[0111] Step S2: In response to the above page dwell tendency value being less than or equal to the first threshold and the above content dwell tendency value being greater than or equal to the second threshold, the following second processing step is executed:

[0112] Step S21: Based on the page content characteristics corresponding to the page content information in the above page content information set, recall the advertising information to be delivered from the above advertising information database.

[0113] In practice, vector recall can be used to retrieve advertising information from the aforementioned advertising information database based on the page content features corresponding to the page content information in the aforementioned page content information set. Specifically, when the page dwell tendency value is less than or equal to a first threshold and the aforementioned content dwell tendency value is greater than or equal to a second threshold, it indicates that the page content (displayed items) on the target page has a significant impact on the retention rate of the target users. Therefore, advertising information for items similar to the page content can be retrieved as advertising information to be delivered.

[0114] Step S22: Extract the advertising attributes corresponding to the above-mentioned advertising information to be delivered, and use them as the advertising attribute information included in the above-mentioned advertising description information.

[0115] In practice, preset regular expressions can be used to extract the advertising attributes corresponding to the advertising information to be delivered, which are then used as the advertising attribute information included in the aforementioned advertising description information.

[0116] Step S23: Determine the advertising style that matches the above-mentioned advertising information to be placed, and obtain the placement style information included in the above-mentioned advertising description information.

[0117] In practice, multiple advertising styles (page style parameters) can be preset for different items corresponding to the advertised information. The aforementioned execution entity can randomly select one of the preset advertising styles as the advertising style information. Since page content has a significant impact on the retention rate of target users, recalling similar items to be advertised can promote the conversion of the items corresponding to the advertised information. At the same time, considering that the page style of the target page has a relatively small impact on the retention rate of target users in this case, adjusting the advertising style corresponding to the advertised information can increase the attention of target users.

[0118] Step 106: Based on the advertisement description information, the advertisement information to be delivered to the target page through the advertising system.

[0119] In some embodiments, the aforementioned implementing entity may, based on the advertising description information, deliver the advertising information to be delivered to the target page through the advertising system.

[0120] In practice, when a target user refreshes the target page, the ad information to be delivered is placed on the target page according to the ad description information, and the target page is re-rendered according to the ad attribute information and delivery style information included in the ad description information.

[0121] In some optional implementations of some embodiments, the above method further includes:

[0122] Step S1: Collect the delivery data for the above-mentioned advertising information to be delivered.

[0123] The aforementioned campaign data includes: clicks, impressions, and conversion rate.

[0124] In practice, advertising information to be delivered can be delivered to multiple different clients, thereby allowing for the collection of corresponding delivery data for the advertising information to be delivered.

[0125] Step S2: Perform data preprocessing on the delivery data to obtain preprocessed delivery data.

[0126] In practice, dirty or missing data can be filtered from the delivery data to obtain pre-processed delivery data. Alternatively, time periods can be set to filter the delivery data, resulting in pre-processed delivery data.

[0127] Step S3: Generate a delivery report corresponding to the preprocessed delivery data described above.

[0128] In practice, the pre-processed delivery data can be filled into the delivery report template to obtain the delivery report.

[0129] Step S4: Synchronize the above deployment report to the data visualization terminal.

[0130] Among them, the data visualization terminal can be an advertising platform corresponding to the advertising information to be advertised.

[0131] The above embodiments of this disclosure have the following beneficial effects: the advertising information delivery method applied to an advertising system through some embodiments of this disclosure improves the matching degree of advertising information delivery. Specifically, firstly, page behavior information and page description information corresponding to the target page are collected. Page behavior information characterizes the page behavior of the target user when browsing the target page, and page description information includes: page style information, page layout information, and a set of page content information. In practice, this disclosure has found that page style and layout affect user retention. For example, aesthetically pleasing and / or efficiently displayed pages can reduce user operations and promote the conversion (e.g., purchase) of products related to the advertising information. Therefore, this disclosure collects relevant data from two dimensions: page behavior and page description. Secondly, for each piece of page content information in the above-mentioned set of page content information, the content attention corresponding to the above-mentioned page content information is determined based on the above-mentioned page behavior information and the user attribute information corresponding to the target user. Content attention includes: direct attention and indirect attention. In practice, by focusing on the user's page behavior towards page content, the attention of the displayed page content to the user can be mapped. Meanwhile, considering that page-related operations are often optimized to reduce user complexity, and to ensure the effectiveness of targeted content attention, this disclosure further incorporates user attribute information to generate indirect attention levels. Next, based on the content attention levels corresponding to the aforementioned page description and content information, the dwell tendency information for the target page is determined. This dwell tendency information includes page dwell tendency value and content dwell tendency value. By combining page description information and content attention levels, user dwell tendency is judged from both page and content perspectives. Furthermore, based on the aforementioned dwell tendency information, it is determined whether to deliver advertising information. This achieves effective advertising information delivery under conditions of high user dwell tendency, significantly improving the effectiveness of information delivery compared to conventional indiscriminate delivery and avoiding waste of related traffic. Then, in response to delivering advertising information, based on the aforementioned dwell tendency information and page content information set, advertising description information is determined. This advertising description information includes advertising attribute information and delivery style information. When advertising needs to be delivered, the system further combines dwell time patterns and page content information to determine the ad content to be recommended and optimize the delivery style. Finally, based on the ad description information, the system delivers the corresponding ad content to the target page. In summary, this improves the matching accuracy of ad delivery.

[0132] Further reference Figure 4 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of an advertising information delivery device applied to an advertising system. These device embodiments are similar to... Figure 1Corresponding to the method embodiments shown, the advertising information delivery device applied to the advertising system can be specifically applied to various electronic devices.

[0133] like Figure 4 As shown, an advertising information delivery device 400 applied to an advertising system in some embodiments includes: a collection unit 401, a first determination unit 402, a second determination unit 403, a third determination unit 404, a fourth determination unit 405, and a delivery unit 406. The collection unit 401 is configured to collect page behavior information and page description information corresponding to the target page. The page behavior information characterizes the page behavior of the target user when browsing the target page, and the page description information includes: page style information, page layout information, and a set of page content information. The first determination unit 402 is configured to, for each piece of page content information in the set of page content information, determine the content attention level corresponding to the page content information based on the page behavior information and the user attribute information corresponding to the target user. The content attention level includes: direct attention level and indirect attention level. The second determination unit 403 is configured to, based on the page description information and the page content information... The system determines the dwell tendency information for the target page based on the content attention corresponding to the information, wherein the dwell tendency information includes: page dwell tendency value and content dwell tendency value; the third determining unit 404 is configured to determine whether to deliver advertising information based on the dwell tendency information; the fourth determining unit 405 is configured to determine advertising description information based on the dwell tendency information and the set of page content information in response to delivering advertising information, wherein the advertising description information includes: advertising attribute information and delivery style information; the delivery unit 406 is configured to deliver the advertising information to be delivered to the target page through the advertising system based on the advertising description information. It is understood that the units described in the advertising information delivery device 400 applied to the advertising system are similar to those in the reference system. Figure 1 The steps in the described method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method are also applicable to the advertising information delivery device 400 and the units contained therein applied to the advertising system, and will not be repeated here.

[0134] The following is for reference. Figure 5 It illustrates a schematic diagram of the structure of an electronic device (e.g., a computing device) suitable for implementing some embodiments of the present disclosure. Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality or scope of the embodiments of this disclosure. Figure 5As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The memory may include a non-volatile storage medium and internal memory. The non-volatile storage medium may store an operating system and a computer program. The computer program includes program instructions that, when executed, cause the processor to perform any of the methods described above. The processor provides computational and control capabilities to support the operation of the entire computer device. The internal memory provides an environment for the execution of the computer program in the non-volatile storage medium; when executed by the processor, the computer program causes the processor to perform any of the methods described above. The network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present disclosure and does not constitute a limitation on the computer device to which the present disclosure is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0135] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0136] In one embodiment, the processor is configured to run a computer program stored in a memory to perform the following steps: collecting page behavior information and page description information corresponding to the target page, wherein the page behavior information characterizes the page behavior of the target user when browsing the target page, and the page description information includes: page style information, page layout information, and a set of page content information; for each piece of page content information in the set of page content information, determining the content attention level corresponding to the page content information based on the page behavior information and the user attribute information corresponding to the target user, wherein the content attention level includes: direct attention level and indirect attention level; determining the dwell tendency information for the target page based on the content attention level corresponding to the page description information and the page content information, wherein the dwell tendency information includes: page dwell tendency value and content dwell tendency value; determining whether to deliver advertising information based on the dwell tendency information; in response to delivering advertising information, determining advertising description information based on the dwell tendency information and the set of page content information, wherein the advertising description information includes: advertising attribute information and delivery style information; and delivering the advertising information to be delivered to the target page through an advertising system based on the advertising description information.

[0137] This disclosure also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, and the method implemented when the program instructions are executed can be referred to the various embodiments of the methods described above.

[0138] The aforementioned computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. Alternatively, the aforementioned computer-readable storage medium may be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0139] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0140] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for delivering advertising information in an advertising system, characterized in that, include: Collect page behavior information and page description information corresponding to the target page. The page behavior information represents the page behavior of the target user when browsing the target page, and the page description information includes: page style information, page layout information and page content information. For each piece of page content information in the set of page content information, the content attention level corresponding to the page content information is determined based on the page behavior information and the user attribute information corresponding to the target user. The content attention level includes direct attention level and indirect attention level. Based on the content attention corresponding to the page description information and page content information, the dwell tendency information for the target page is determined, wherein the dwell tendency information includes: page dwell tendency value and content dwell tendency value; Based on the aforementioned dwelling tendency information, determine whether to deliver advertising information; In response to the delivery of advertising information, the advertising description information is determined based on the dwell tendency information and the set of page content information, wherein the advertising description information includes: advertising attribute information and delivery style information; Based on the advertisement description information, the advertisement information to be delivered, corresponding to the advertisement description information, is delivered to the target page through the advertising system. The step of determining dwell tendency information for the target page based on the content attention corresponding to the page description information and page content information includes: The page style information is extracted using the style feature extraction module to obtain page style features, which include: page color features, page texture features, and page animation features. Based on the page layout information, a page layer sequence is generated, wherein each page layer includes at least one page layout block that is located at the same display level and whose content is visible, and each page layer has a corresponding page layer weight, which is controlled by the layer occlusion degree of the corresponding page layer. The layout feature extraction module extracts layout features from each page layer in the page layer sequence to generate candidate page layout features, thus obtaining a candidate page layout feature sequence. Based on the page layer weights corresponding to the page layers, feature fusion is performed on the candidate page layout features in the obtained candidate page layout feature sequence to obtain the page layout features. Based on the page style features and the page layout features, the page dwell tendency value is generated; Based on the content position of the page content information on the target page and the content attention corresponding to the page content information, the page layout features are updated to obtain the updated page layout features. Based on the updated page layout features, the content dwell tendency value is generated.

2. The method according to claim 1, characterized in that, The collected page behavior information and the page description information corresponding to the target page include: Determine whether to obtain the target user's authorized information processing permissions; In response to the failure to obtain the information processing permission, an authorization prompt for the information processing permission is displayed on the target page, wherein the authorization prompt is used to prompt the target user to authorize the information processing permission; In response to having obtained the information processing permission, request a trust token; In response to obtaining the authorization token, the data collector within the target page is activated, and a data transmission channel is created, wherein the data transmission channel is used to transmit page behavior information and page description information, and the data transmission channel is controlled by the token validity period of the authorization token; The page behavior information and page description information are collected by a data collector within the target page.

3. The method according to claim 2, characterized in that, For each piece of page content information in the set of page content information, determining the content attention level corresponding to the page content information based on the page behavior information and the user attribute information corresponding to the target user includes: Page behavior filtering is performed on the page behaviors corresponding to the page behavior information to obtain a page behavior sequence. The page behaviors in the page behavior sequence represent behavioral operations related to the page content information. The page behaviors in the page behavior sequence are arranged in order according to their corresponding execution times. Generate a page feature map corresponding to the target page; For each page behavior in the page behavior sequence, perform the following feature processing steps: Extract the page behavior features corresponding to the page behavior; The page behavior features are projected onto the page feature map to obtain the projected page behavior features. The direct attention level is generated based on the obtained sequence of page behavior features after projection; The page content information is encoded to obtain page content features; Extract user behavior attribute information from the user attribute information; The user behavior attribute information is encoded to obtain user behavior attribute features, wherein the page content features and user behavior attribute features are located in the same feature space; The indirect attention level is generated based on the page content features and the user behavior attribute features.

4. The method according to claim 3, characterized in that, The step of determining the advertisement description information based on the dwell tendency information and the page content information set includes: In response to the page dwell tendency value being greater than a first threshold and the content dwell tendency value being less than a second threshold, the following first processing step is performed: Extract advertising information to be placed from the advertising information database; Extract the advertising attributes corresponding to the advertising information to be delivered, and use them as the advertising attribute information included in the advertising description information; Based on the page style information, the ad description information, including the placement style information, is generated; In response to the page dwell tendency value being less than or equal to a first threshold and the content dwell tendency value being greater than or equal to a second threshold, the following second processing step is performed: Based on the page content characteristics corresponding to the page content information in the page content information set, recall the advertising information to be delivered from the advertising information database; Extract the advertising attributes corresponding to the advertising information to be delivered, and use them as the advertising attribute information included in the advertising description information; Determine the advertising style that matches the advertising information to be delivered, and obtain the delivery style information included in the advertising description information.

5. The method according to claim 4, characterized in that, The step of delivering the advertisement information corresponding to the advertisement description information to the target page through the advertising system, based on the advertisement description information, includes: The advertising information to be delivered corresponding to the advertising description information is added to the target message queue, wherein the lifecycle of the target message queue is controlled by the page refresh frequency corresponding to the target page; In response to detecting a page refresh operation for the target page, the advertising system delivers the advertising information to be delivered to the target page through the target message queue.

6. The method according to claim 5, characterized in that, The method further includes: Collect delivery data for the advertisement information to be delivered, wherein the delivery data includes: clicks, impressions, and conversion rate; The delivery data is preprocessed to obtain preprocessed delivery data; Generate a delivery report corresponding to the preprocessed delivery data; The delivery report will be synchronized to the data visualization terminal.

7. An advertising information delivery device applied to an advertising system, characterized in that, include: The collection unit is configured to collect page behavior information and page description information corresponding to the target page. The page behavior information represents the page behavior of the target user when browsing the target page, and the page description information includes: page style information, page layout information and page content information. The first determining unit is configured to, for each page content information in the set of page content information, determine the content attention corresponding to the page content information based on the page behavior information and the user attribute information corresponding to the target user, wherein the content attention includes: direct attention and indirect attention; The second determining unit is configured to determine dwell tendency information for the target page based on the content attention corresponding to the page description information and page content information, wherein the dwell tendency information includes: page dwell tendency value and content dwell tendency value. The third determining unit is configured to determine whether to deliver advertising information based on the dwell tendency information; The fourth determining unit is configured to determine advertising description information based on the dwell tendency information and the set of page content information in response to advertising information delivery, wherein the advertising description information includes: advertising attribute information and delivery style information; The delivery unit is configured to deliver the advertisement information to the target page corresponding to the advertisement description information through the advertising system, based on the advertisement description information. The step of determining dwell tendency information for the target page based on the content attention corresponding to the page description information and page content information includes: The page style information is extracted using the style feature extraction module to obtain page style features, which include: page color features, page texture features, and page animation features. Based on the page layout information, a page layer sequence is generated, wherein each page layer includes at least one page layout block that is located at the same display level and whose content is visible, and each page layer has a corresponding page layer weight, which is controlled by the layer occlusion degree of the corresponding page layer. The layout feature extraction module extracts layout features from each page layer in the page layer sequence to generate candidate page layout features, thus obtaining a candidate page layout feature sequence. Based on the page layer weights corresponding to the page layers, feature fusion is performed on the candidate page layout features in the obtained candidate page layout feature sequence to obtain the page layout features. Based on the page style features and the page layout features, the page dwell tendency value is generated; Based on the content position of the page content information on the target page and the content attention corresponding to the page content information, the page layout features are updated to obtain the updated page layout features. Based on the updated page layout features, the content dwell tendency value is generated.

8. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 6.

9. A computer-readable medium, characterized in that, It stores a computer program thereon, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1 to 6.

Images