Advertisement putting system and method based on Android set top box

Abstract

The invention discloses an advertisement putting system and method based on an Android set top box, and relates to the technical field of smart televisions. Collecting data: collecting user watching behavior data by means of an Android set top box, and establishing an interest preference model: based on the collected behavior data, introducing a machine learning algorithm to construct the user interest preference model, and carrying out advertisement filtering: screening and filtering an advertisement resource pool according to the interest preference model, eliminating advertisements which are not interested in by the user, reducing the advertisement insertion frequency, and improving the advertisement insertion efficiency; according to an interest preference model, selecting advertisement categories according with user interests, determining advertisement putting opportunity and frequency in combination with the user watching rhythm, dynamically optimizing a putting strategy according to real-time feedback, and obtaining user feedback: establishing a convenient feedback channel, encouraging the user to feed back advertisement interaction, and improving user experience; feedback data is collected and analyzed, an advertisement putting strategy is continuously optimized according to the feedback data, and the satisfaction degree of the user and the advertiser is improved.

Description

Technical Field

[0001] The present invention discloses an advertising delivery system and method based on an Android set-top box, which relates to the technical field of smart TVs. Background Art

[0002] The popularization of smart TVs and set-top boxes has led to a huge change in viewing habits. The previous advertising insertion mode based on time periods seems out of place in today's environment where personalized needs are on the rise. On the one hand, the number of advertisements is overwhelming, and users are frequently interrupted by advertisements when watching their favorite programs, destroying the coherence of viewing and the immersive feeling of watching movies. On the other hand, the content of the advertisements deviates from the interests of the users, and a large number of irrelevant advertisements flood the screen, making it difficult for the advertisements to attract the attention of the users and greatly reducing the advertising effect. Moreover, users are often in a passive state of receiving these advertisements, which greatly reduces the user satisfaction. Summary of the Invention

[0003] In view of the problems of the prior art, the present invention provides an advertising delivery system and method based on an Android set-top box. By deeply analyzing user behavior data and accurately understanding interest preferences, unnecessary advertising insertions are reduced, and accurate advertising reach is achieved, comprehensively improving the user experience and advertising efficiency.

[0004] The specific solution proposed by the present invention is as follows:

[0005] The present invention provides an advertising delivery method based on an Android set-top box. Collect data: With the help of the Android set-top box, collect user viewing behavior data, including viewing duration, viewing content type, and ad skipping frequency, for subsequent analysis.

[0006] Build an interest preference model: Based on the collected behavior data, introduce machine learning algorithms to build a user interest preference model, where preprocess the behavior data and extract features, and use the extracted features for model training. The trained model is used to estimate the user's preference for different types of advertisements.

[0007] Perform advertisement filtering: According to the interest preference model, screen and filter the advertisement resource pool, exclude advertisements that the user is not interested in, reduce the advertisement insertion frequency, and purify the viewing experience.

[0008] Perform advertisement delivery: According to the interest preference model, select advertisement categories that meet the user's interests, determine the advertisement delivery timing and frequency in combination with the user's viewing rhythm, and dynamically optimize the delivery strategy according to real-time feedback.

[0009] Obtain user feedback: Build a convenient feedback channel, encourage users to give feedback on advertisement interactions, collect and analyze the feedback data, and continuously optimize the advertisement delivery strategy according to the feedback data to improve the satisfaction of both users and advertisers.

[0010] Further, the data collection in the advertising delivery method based on an Android set-top box includes:

[0011] Collecting viewing duration: Recording the time a user stays on each video, which is used to deeply explore the user's preference trends for various types of video content.

[0012] Collecting viewing content types: Collecting viewing content type data based on the metadata carried by the video. The metadata includes classification labels and theme descriptions.

[0013] Collecting the frequency of skipping ads: Tracking each operation of the user skipping ads, including the specific time nodes of skipping.

[0014] Further, the establishment of the interest preference model in the advertising delivery method based on an Android set-top box includes:

[0015] Performing preprocessing on behavioral data: Using data cleaning tools and normalization algorithms to eliminate noisy data with mixed data, ensuring the purity and reliability of the data input into the model.

[0016] Performing feature extraction: Extracting key features from the user's behavioral data and converting the key features into feature vectors recognizable by the model.

[0017] Using the extracted features for model training: Selecting a machine learning algorithm to repeatedly iterate and train the interest preference model using the extracted key features, enabling the interest preference model to have the ability to predict the user's interest level in different types of ads.

[0018] Further, the advertisement filtering in the advertising delivery method based on an Android set-top box includes:

[0019] Classifying ads according to the interest preference model: Classifying ads into different categories according to the core content, industry, and promotional selling points of the ads. The ad categories include fast-moving consumer goods, digital electronics, and cultural and tourism services.

[0020] Performing user interest matching according to the interest preference model: Invoking the evaluation results of the interest preference model and calculating the interest heat value of the user for each ad category.

[0021] Performing ad filtering: Setting a filtering threshold. Once the interest level corresponding to a certain ad category is lower than the filtering threshold, the ad category below the filtering threshold will be intercepted immediately.

[0022] The present invention also provides an advertising delivery system based on an Android set-top box, including a data collection module, an interest preference modeling module, an ad filtering module, a precise delivery module, and a user feedback module.

[0023] Data collection module collects data: With the help of Android set-top boxes, user viewing behavior data is collected. User viewing behavior data includes viewing time, viewing content type, and ad skipping frequency for subsequent analysis.

[0024] The interest preference modeling module builds an interest preference model: Based on the collected behavioral data, a machine learning algorithm is introduced to build a user interest preference model, which includes behavioral data preprocessing and feature extraction. The extracted features are used to train the model. The trained model is used to estimate the user's preference for different types of advertisements.

[0025] Ad filtering module performs ad filtering: based on the interest preference model, it filters the ad resource pool, excludes ads that users are not interested in, reduces the frequency of ad insertion, and purifies the viewing experience.

[0026] The precise delivery module delivers advertisements: Based on the interest preference model, it selects advertisement categories that match the user's interests, determines the timing and frequency of advertisement delivery based on the user's viewing rhythm, and dynamically optimizes the delivery strategy based on real-time feedback.

[0027] The user feedback module obtains user feedback: builds a convenient feedback channel, encourages users to provide feedback on advertising interactions, collects and analyzes feedback data, and continuously optimizes advertising delivery strategies based on feedback data to improve the satisfaction of both users and advertisers.

[0028] Furthermore, the data collection module of the Android set-top box-based advertising delivery system collects viewing time: records the user's stay time on each video, which is used to deeply explore the user's preference for various video contents.

[0029] Collect viewing content types: Collect viewing content type data based on the metadata carried by the video. The metadata includes classification tags, subject descriptions,

[0030] Collect the frequency of ad skipping: Track every time a user skips an ad, including the specific time of the skip.

[0031] Furthermore, the interest preference modeling module of the advertising delivery system based on Android set-top box establishes an interest preference model, including:

[0032] Preprocess behavioral data: Use data cleaning tools and normalization algorithms to remove noise data and ensure that the data input to the model is pure and reliable.

[0033] Perform feature extraction: extract key features from user behavior data and convert them into feature vectors that can be recognized by the model.

[0034] Model training using the extracted features: Select a machine learning algorithm to repeatedly iterate and train the interest preference model using the extracted key features, enabling the interest preference model to have the ability to predict the user's interest level in different types of advertisements.

[0035] Further, the advertisement filtering module of the described advertisement delivery system based on an Android set-top box performs advertisement filtering, including:

[0036] Classify advertisements according to the interest preference model: According to the core content, industry, and promotional selling points of the advertisements, regularize and divide the advertisements into different categories. The advertisement categories include fast-moving consumer goods, digital electronics, and cultural and tourism services.

[0037] Perform user interest matching according to the interest preference model: Invoke the evaluation results of the interest preference model to calculate the user's interest heat value for each advertisement category.

[0038] Perform advertisement filtering: Set a filtering threshold. Once the interest level corresponding to a certain advertisement category is lower than the filtering threshold, immediately intercept the advertisement category with an interest level lower than the filtering threshold.

[0039] The beneficial effects are:

[0040] With the in-depth insight into user behavior data and the accurate grasp of interest preferences, the present invention has achieved a major breakthrough in the field of Android set-top box advertisement delivery, bringing many significant benefits:

[0041] Improve the user experience: By intelligently screening and filtering advertisements that the user is not interested in, it greatly restores a pure viewing environment, allowing the user to immerse themselves in their favorite programs, and significantly improving the viewing fluency and immersion.

[0042] Improve the advertisement effect: Precisely push advertisements that match the user's interests, directly hit the user's pain points of demand, stimulate the user's desire for interaction, and significantly improve the advertisement click-through rate and conversion rate, helping advertisers achieve efficient marketing.

[0043] Enhance user control: Build a user feedback loop, endow the user with the right to judge and choose advertisements, enabling them to change from passive acceptance to active participation, enhancing the user's sense of control and satisfaction in the advertisement link, and injecting new vitality into the intelligent TV advertisement ecosystem. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0045] Figure 1 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION

[0046] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0047] Example 1

[0048] The present invention provides an advertisement delivery method based on an Android set-top box, and collects data: by means of an Android set-top box, user viewing behavior data is collected, and the user viewing behavior data includes viewing time, viewing content type, and skipping advertisement frequency, which are used for subsequent analysis. The data may include:

[0049] Collect viewing time: record the time users spend on each video, and use this to deeply explore users’ preferences for various video content. For example, focusing on documentary channels for a long time reflects users’ preference for knowledge-based content.

[0050] Collect viewing content types: Collect viewing content type data based on the metadata carried by the video. The metadata includes classification labels and subject descriptions, which can analyze the user's interest areas. It is easy to see whether the user is fond of comedy movies and TV shows, or is keen on science and popular science programs.

[0051] Collect the frequency of ad skipping: Track every time a user skips an ad, including the specific time of the skip, to accurately assess the user's tolerance for ads.

[0052] Establishing interest preference model: Based on the collected behavior data, introduce machine learning algorithms to build user interest preference models, which includes preprocessing behavior data and extracting features, using the extracted features to train the model, and using the trained model to estimate the user's preference for different types of advertisements. This may include:

[0053] Preprocess behavioral data: Use data cleaning tools and normalization algorithms to remove noise data and ensure that the data input into the model is pure and reliable, laying a solid foundation for accurate modeling.

[0054] Feature extraction: Extract key features from user behavior data, such as the degree of concentration reflected by viewing time, the interest direction reflected by content type, and the degree of preference implied by the frequency of skipping ads, etc., and convert key features into feature vectors that can be recognized by the model.

[0055] Use the extracted features for model training: Select a machine learning algorithm to use the extracted key features to repeatedly train the interest preference model, so that the interest preference model has the ability to predict the user's interest in different types of advertisements.

[0056] Perform ad filtering: Based on the interest preference model, screen and filter the ad resource pool, exclude ads that the user is not interested in, reduce the ad insertion frequency, and purify the viewing experience. This may include:

[0057] Classify ads according to the interest preference model: Classify ads into different categories according to the core content, industry, and promotional selling points of the ads. The ad categories include fast-moving consumer goods, digital electronics, and cultural and tourism services.

[0058] Match user interests based on the interest preference model: Invoke the evaluation results of the interest preference model and calculate the interest heat value of the user for each ad category.

[0059] Perform ad filtering: Set a filtering threshold. Once the interest level corresponding to a certain ad category is lower than the filtering threshold, immediately intercept the ad category with an interest level lower than the filtering threshold and prevent it from appearing in the user's view.

[0060] Perform ad placement: Based on the interest preference model, select ad categories that match the user's interests, determine the ad placement timing and frequency in combination with the user's viewing rhythm, and dynamically optimize the placement strategy based on real-time feedback. This may include:

[0061] Perform ad selection: Refer to the interest scores given by the model, lock in the category with the peak user interest among numerous ad categories, and use it as the preferred placement option.

[0062] Perform the placement strategy: Combine the user's daily viewing habits, such as the usual viewing time period, the switching rhythm of the viewing program type, etc., and skillfully select the best timing and appropriate frequency for ad placement to ensure that the ad appears neither abruptly nor fails to catch the user's attention.

[0063] Perform dynamic adjustment: Pay close attention to the user's immediate feedback on the placed ads, such as click, stay, skip, etc. behaviors, and quickly optimize and adjust the subsequent ad placement strategy based on this to achieve dynamic optimization of the placement.

[0064] Obtain user feedback: Build a convenient feedback channel, encourage users to provide feedback on ad interactions, collect and analyze the feedback data, and continuously optimize the ad placement strategy based on the feedback data to improve the satisfaction of both users and advertisers. This may include:

[0065] Feedback collection: Open a convenient feedback entry on the interactive interface of the Android set-top box. Users can easily perform operations such as skipping, liking, and reporting on the ad, and the system will immediately collect this feedback information.

[0066] Feedback analysis: Use data analysis algorithms to deeply interpret the user feedback data, accurately measure the ad placement effect, and gain insights into the root causes of user preferences and dislikes.

[0067] Strategy optimization: Based on the feedback analysis conclusions, optimize the advertising delivery strategy in a targeted manner, fine-tune parameters such as advertising category and delivery frequency, and improve overall advertising effectiveness.

[0068] Example 2

[0069] The present invention also provides an advertisement delivery system based on an Android set-top box, comprising a data collection module, an interest preference modeling module, an advertisement filtering module, a precision delivery module and a user feedback module.

[0070] Data collection module collects data: With the help of Android set-top boxes, user viewing behavior data is collected. User viewing behavior data includes viewing time, viewing content type, and ad skipping frequency for subsequent analysis.

[0071] The interest preference modeling module builds an interest preference model: Based on the collected behavioral data, a machine learning algorithm is introduced to build a user interest preference model, which includes behavioral data preprocessing and feature extraction. The extracted features are used to train the model. The trained model is used to estimate the user's preference for different types of advertisements.

[0072] Ad filtering module performs ad filtering: based on the interest preference model, it filters the ad resource pool, excludes ads that users are not interested in, reduces the frequency of ad insertion, and purifies the viewing experience.

[0073] The precise delivery module delivers advertisements: Based on the interest preference model, it selects advertisement categories that match the user's interests, determines the timing and frequency of advertisement delivery based on the user's viewing rhythm, and dynamically optimizes the delivery strategy based on real-time feedback.

[0074] The user feedback module obtains user feedback: builds a convenient feedback channel, encourages users to provide feedback on advertising interactions, collects and analyzes feedback data, and continuously optimizes advertising delivery strategies based on feedback data to improve the satisfaction of both users and advertisers.

[0075] As the information interaction and execution process between the modules in the above-mentioned system are based on the same concept as the embodiment of the method of the present invention, the specific contents can be found in the description of the embodiment of the method of the present invention and will not be repeated here.

[0076] Similarly, the system of the present invention has achieved a major breakthrough in the field of Android set-top box advertising by virtue of its deep insight into user behavior data and accurate grasp of interest preferences, bringing many significant benefits:

[0077] Improve user experience: By intelligently identifying and filtering advertisements that users are not interested in, the pure viewing environment can be restored to the greatest extent, allowing users to immerse themselves in their favorite programs, and the viewing fluency and immersion are greatly improved.

[0078] Improve advertising effectiveness: Precisely push ads that match users' interests, directly address the pain points of users' needs, stimulate users' desire for interaction, significantly improve the click-through rate and conversion rate of ads, and help advertisers achieve efficient marketing.

[0079] Enhance user control: Build a user feedback loop, endow users with the right to judge and choose ads, enable them to transform from passive acceptance to active participation, enhance users' sense of control and satisfaction in the advertising process, and inject new vitality into the smart TV advertising ecosystem.

[0080] It should be noted that not all steps and modules in the above-mentioned processes and system structures are necessary, and some steps or modules can be ignored according to actual needs. The execution order of each step is not fixed and can be adjusted according to needs. The system structures described in the above-mentioned embodiments can be physical structures or logical structures. That is, some modules may be implemented by the same physical entity, or some modules may be implemented separately by multiple physical entities, or some components in multiple independent devices may jointly implement them.

[0081] The above-described embodiments are merely preferred embodiments given to fully illustrate the present invention, and the protection scope of the present invention is not limited thereto. Equivalent substitutions or transformations made by those skilled in the art on the basis of the present invention are all within the protection scope of the present invention. The protection scope of the present invention shall be subject to the claims.

Claims

1. An advertising delivery method based on an Android set-top box, characterized by collecting data: using an Android set-top box to collect user viewing behavior data, the user viewing behavior data including viewing time, viewing content type, and skipping frequency of advertisements, for subsequent analysis, Establish interest preference model: Based on the collected behavior data, introduce machine learning algorithms to build user interest preference models, which includes behavior data preprocessing and feature extraction, and use the extracted features to train the model. The trained model is used to estimate the user's preference for different types of advertisements. Ad filtering: Based on the interest preference model, the ad resource pool is screened and filtered to exclude ads that users are not interested in, reduce the frequency of ad insertion, and purify the viewing experience. Advertisement delivery: Based on the interest preference model, select the ad categories that match the user's interests, determine the timing and frequency of ad delivery based on the user's viewing rhythm, and dynamically optimize the delivery strategy based on real-time feedback. Obtain user feedback: Build convenient feedback channels, encourage users to provide feedback on advertising interactions, collect and analyze feedback data, and continuously optimize advertising delivery strategies based on feedback data to improve the satisfaction of both users and advertisers.

2. The method for placing advertisements based on an Android set-top box according to claim 1, characterized in that The collected data include: Collect viewing time: record the time users spend on each video, and use it to deeply explore users' preferences for various video content. Collect viewing content types: Collect viewing content type data based on the metadata carried by the video. The metadata includes classification tags, subject descriptions, Collect the frequency of ad skipping: Track every time a user skips an ad, including the specific time of the skip.

3. The method for placing advertisements based on an Android set-top box according to claim 1, characterized in that The establishing of the interest preference model comprises: Preprocess behavioral data: Use data cleaning tools and normalization algorithms to remove noise data and ensure that the data input to the model is pure and reliable. Perform feature extraction: extract key features from user behavior data and convert them into feature vectors that can be recognized by the model. Use the extracted features for model training: Select a machine learning algorithm to use the extracted key features to repeatedly train the interest preference model, so that the interest preference model has the ability to predict the user's interest in different types of advertisements.

4. The method for placing advertisements based on an Android set-top box according to claim 1, characterized in that The advertisement filtering comprises: Advertisements are classified based on the interest preference model: Advertisements are divided into different categories according to their core content, industry, and selling points. The categories include fast-moving consumer goods, digital electronics, and cultural and travel services. According to the interest preference model, user interest matching is performed: the evaluation results of the interest preference model are called to calculate the user's interest heat value for each advertising category. Ad filtering: Set a filtering threshold. Once the interest level of a certain ad category is lower than the filtering threshold, the ad category below the filtering threshold will be blocked immediately.

5. An advertising delivery system based on Android set-top box, characterized by It includes data collection module, interest preference modeling module, advertisement filtering module, precision delivery module and user feedback module. Data collection module collects data: With the help of Android set-top boxes, user viewing behavior data is collected. User viewing behavior data includes viewing time, viewing content type, and ad skipping frequency for subsequent analysis. The interest preference modeling module builds an interest preference model: Based on the collected behavioral data, a machine learning algorithm is introduced to build a user interest preference model, which includes behavioral data preprocessing and feature extraction. The extracted features are used to train the model. The trained model is used to estimate the user's preference for different types of advertisements. Ad filtering module performs ad filtering: based on the interest preference model, it filters the ad resource pool, excludes ads that users are not interested in, reduces the frequency of ad insertion, and purifies the viewing experience. The precise delivery module delivers advertisements: Based on the interest preference model, it selects advertisement categories that match the user's interests, determines the timing and frequency of advertisement delivery based on the user's viewing rhythm, and dynamically optimizes the delivery strategy based on real-time feedback. The user feedback module obtains user feedback: builds a convenient feedback channel, encourages users to provide feedback on advertising interactions, collects and analyzes feedback data, and continuously optimizes advertising delivery strategies based on feedback data to improve the satisfaction of both users and advertisers.

6. The advertising delivery system based on Android set-top box according to claim 5 is characterized in that The data collection module collects viewing time: records the time users spend on each video, and is used to deeply explore users' preferences for various video contents. Collect viewing content types: Collect viewing content type data based on the metadata carried by the video. The metadata includes classification tags, subject descriptions, Collect the frequency of ad skipping: Track every time a user skips an ad, including the specific time of the skip.

7. The advertising delivery system based on Android set-top box according to claim 5 is characterized in that The interest preference modeling module establishes an interest preference model, including: Preprocess behavioral data: Use data cleaning tools and normalization algorithms to remove noise data and ensure that the data input to the model is pure and reliable. Perform feature extraction: extract key features from user behavior data and convert them into feature vectors that can be recognized by the model. Use the extracted features for model training: Select a machine learning algorithm to use the extracted key features to repeatedly train the interest preference model, so that the interest preference model has the ability to predict the user's interest in different types of advertisements.

8. The advertising delivery system based on Android set-top box according to claim 5 is characterized in that The advertisement filtering module performs advertisement filtering, including: Advertisements are classified based on the interest preference model: Advertisements are divided into different categories according to their core content, industry, and selling points. The categories include fast-moving consumer goods, digital electronics, and cultural and travel services. According to the interest preference model, user interest matching is performed: the evaluation results of the interest preference model are called to calculate the user's interest heat value for each advertising category. Ad filtering: Set a filtering threshold. Once the interest level of a certain ad category is lower than the filtering threshold, the ad category below the filtering threshold will be blocked immediately.

Images