Insurance advertisement dynamic loading method based on Internet

Abstract

The invention relates to the technical field of insurance advertisement dynamic loading based on the Internet, and discloses an insurance advertisement dynamic loading method based on the Internet. By acquiring the watching data of the user on the social media platform in recent three months, the watching preference and behavior mode of the user can be comprehensively known, and the advertisement system is helped to grasp the interest of the user. The videos are classified according to types, so that the performance of different types of videos can be analyzed, and the most interested content type of the user can be identified. The time point of video lagging is recorded and the waiting duration of the user is calculated, so that support is provided for understanding the tolerance of the user, and an advertisement display strategy is optimized. By counting the number of jamming times accepted by the user, the tolerance threshold value of the user can be revealed. The jamming tolerance of the user in different video types is known, and an advertiser is helped to formulate a precise putting strategy. The design of dynamically loading the advertisement can ensure that the advertisement resource is smoothly played under the condition that the video is not stuck.

Description

Technical Field

[0001] This invention relates to the field of dynamic loading technology for insurance advertisements based on the Internet, specifically to a method for dynamic loading of insurance advertisements based on the Internet. Background Technology

[0002] With the rapid development of the internet, video content consumption has become an important part of users' daily lives. Especially on social media platforms, users are watching videos more frequently, providing advertisers with more marketing opportunities. However, traditional advertising methods often face numerous challenges, particularly in the loading and display of ads, where user experience is severely impacted.

[0003] Currently, ad loading strategies primarily rely on fixed delivery rules, failing to dynamically adjust based on real-time user behavior and environmental changes. This approach not only leads to inefficient ad delivery but can also cause user dissatisfaction, ultimately impacting brand image. For example, if ads are slow to load or experience buffering while a user is watching a video, they may choose to close the video or switch to other content, resulting in a significant drop in ad exposure.

[0004] In existing technologies, many advertising systems still rely on static resource configuration and prioritization, failing to effectively analyze user behavior data. The limitation of this approach is its inability to respond in real-time to users' viewing habits and content preferences. For example, advertising systems may fail to update and optimize ad content based on users' browsing history on social media, resulting in ads that do not match users' interests. Furthermore, users' tolerance for buffering varies from person to person, making it difficult for existing technologies to accurately grasp each user's acceptable range.

[0005] To address the aforementioned issues, a method for loading insurance ads based on user behavior analysis and dynamic adjustment is urgently needed. This method should be able to acquire real-time user behavior data on social media, analyze user acceptance of different video types, and dynamically load ads accordingly, thereby improving the user's viewing experience and the effectiveness of ad delivery. Summary of the Invention

[0006] This invention provides a method for dynamically loading insurance advertisements based on the Internet, which helps to solve the problems mentioned in the background art above.

[0007] This invention provides the following technical solution: a method for dynamically loading insurance advertisements based on the Internet, comprising: Get all videos a user has viewed on social media platforms over the past three months; The acquired videos are categorized according to video type; Get the percentage of each type of video in all videos, and sort them in descending order of percentage. Then, label the sorted video types as type 1, type 2, ..., type N. When a user browses a video; If the video is choppy, obtain the time point when the chop starts and the time point when the chop ends for all videos. Calculate the time point when the video ends and the time point when the video starts to stop, and record the result as the user's waiting time. Get all the user's waiting time in a video and compare them, then record the longest user waiting time as the longest waiting time accepted by the user for this video. When a user switches to or closes a video during a video stutter, the system counts all video stutters that occurred before this stutter and records the result as the number of stutters the user has experienced. Calculate the average duration of buffering that users can tolerate in each type of video. The average duration of buffering that users can tolerate = the sum of the waiting time of all users in each type of video / the total number of buffering occurrences in each type of video. Get the longest duration of buffering or stuttering experienced by the user in each video category; Get the shortest duration of buffering that users can tolerate in each type of video; Calculate the average number of stutters that users can tolerate in each video category. The average number of stutters that users can tolerate = the sum of the number of stutters in all videos in each video category / the total number of videos in each video category. Get the maximum number of stutters that users can tolerate in each type of video; Find the minimum number of stutters that users can tolerate in each type of video; Set up a set of user video behaviors named after the video type. The elements in the set include the average duration of buffering that users can tolerate, the longest duration of buffering that users can tolerate, the shortest duration of buffering that users can tolerate, the average number of buffering instances that users can tolerate, the minimum number of buffering instances that users can tolerate, and the maximum number of buffering instances that users can tolerate. Design a dynamic loading strategy for insurance advertisements; If the video plays smoothly without buffering, the set insurance advertisement will be played when the video reaches the advertisement slot.

[0008] Optionally, the start time of the pause corresponding to the end time of the video pause specifically includes: Set a timeline in the order from the start to the end of the video playback. The beginning of the timeline is the time when the video starts playing, and the end of the timeline is the time when the video ends playing. Map all the stuttering times onto the timeline. The first time a lag occurs is recorded as the first time a lag occurs, the second time a lag occurs is recorded as the second time a lag occurs, and so on until the Mth time a lag occurs is recorded as the Mth time a lag occurs. The first time the lag occurred is recorded as the first time the lag occurred, the second time the lag occurred is recorded as the second time the lag occurred, and so on until the Mth time the lag occurred is recorded as the Mth time the lag occurred. The first start time of the lag corresponds to the first end time of the lag, the second start time of the lag corresponds to the second end time of the lag, and so on, the Mth start time of the lag corresponds to the Mth end time of the lag.

[0009] Optionally, the design of the dynamic loading strategy for insurance advertisements specifically includes: Set up a general resource set for insurance advertisements; Obtain the keywords of the top 30% of the video types watched by users, use the keywords to match the corresponding insurance advertising resources in the total insurance advertising resource set, and combine the matched insurance advertising resources into a user insurance advertising resource set; The elements in the total insurance advertising resource set and the elements in the user insurance advertising resource set are classified separately. Insurance advertisements are divided into priority category and secondary category. The priority category includes video insurance advertisements, and the secondary category includes image insurance advertisements. Within the priority and secondary categories, insurance advertisements are further classified into first-level resources, second-level resources, ... Q-level resources in descending order of resolution.

[0010] Optionally, obtaining the keywords representing the top 30% of the video types watched by users specifically includes: Get the sorted order of video categories; Get the top 30% of video categories after sorting; We analyzed the top 30% of video categories and extracted keywords.

[0011] Optionally, the design of the dynamic loading strategy for insurance advertisements specifically includes: Get the fixed and custom ad slots in the videos viewed by users; Get the type of video the user is currently viewing; Analyze the keywords in the video between the custom ad placement and the previous fixed ad placement or custom ad placement; Get the resolution set by the user when watching the video; In the user's insurance ad resource set, randomly match an insurance ad resource that matches the keywords obtained from the analysis and the resolution set by the user. If no match is found, match from the total insurance ad resource set. If no match is found from the total insurance ad resource set, then take any resource from the total insurance ad resource set as the ad in the custom ad slot.

[0012] Optionally, the design of the dynamic loading strategy for insurance advertisements specifically includes: When lag occurs in a fixed ad slot; A set of user video behaviors that match the type of video the user is currently viewing; Start timing when the lag begins; when the lag duration reaches the minimum acceptable lag duration for the user, obtain the type and resolution of the insurance ad in the current fixed ad slot. Load the current insurance advertisement resource at a lower resolution by one level. If the current insurance ad resource resolution is reduced by one level, and the loading time of the insurance ad after the resolution reduction is less than or equal to the average time that the user can tolerate the lag, then the insurance ad resource at the current resolution will be played. If the loading time of the insurance ad after reducing the resolution by one level is greater than the average time that users can tolerate with lag, then the resolution of the current insurance ad will be reduced by two levels. If the resolution of the current insurance advertisement resource is reduced by two levels, and the loading time of the insurance advertisement after the resolution is reduced by two levels is less than or equal to the longest time that the user can tolerate the lag, then the insurance advertisement resource at the current resolution will be played. If the loading time of the insurance ad after reducing the resolution by two levels is greater than the maximum duration of buffering that the user can tolerate, then the ad playback for that ad slot will be skipped.

[0013] Optionally, the design of the dynamic loading strategy for insurance advertisements specifically includes: When lag occurs in a custom ad slot; A set of user video behaviors that match the type of video the user is currently viewing; When the lag starts, a timer is set. When the lag duration reaches the minimum acceptable lag duration for the user, the system matches the user's insurance ad resource set with the same keywords as the insurance ad in the current custom ad slot. If no match is found, the system matches the total insurance ad resource set and loads an insurance ad resource with a resolution one level lower than the current resolution. If no match is found in the total insurance ad resource set, the system loads an insurance ad resource with a resolution one level lower than the current resolution from the user's insurance ad resource set. If the loading time of the first rematched insurance ad resource is less than or equal to the average duration of buffering that the user can tolerate, then the rematched insurance ad resource will be played. If, after the first rematch of insurance ad resources, the loading time of the first rematch of insurance ad resources is greater than the average time that the user can tolerate from buffering, then the system will match ad resources in the user's insurance ad resource set that have the same keywords as the insurance ad in the current custom ad slot. If no match is found, the system will match in the total set of insurance ad resources and select an insurance ad resource with a resolution two levels lower than the current resolution for loading. If no match is found in the total set of insurance ad resources either, then the system will select an insurance ad resource in the user's insurance ad resource set with a resolution two levels lower than the current resolution for loading. If, after the second rematch of insurance ad resources, the loading time of the second rematch of insurance ad resources is less than or equal to the maximum duration during which the user can tolerate buffering, then the insurance ad resources at the current resolution will be played. If, after a second rematch of the insurance ad resource, the loading time of the second rematch exceeds the maximum acceptable buffering time for the user, then the ad playback for that ad slot will be skipped.

[0014] Optionally, the design of the dynamic loading strategy for insurance advertisements specifically includes: A set of user video behaviors that match the type of video the user is currently viewing; Get the number of times the user experiences buffering while watching a video; If the current number of stutters is less than the minimum number of stutters that the user can tolerate, then no action will be taken; If the minimum number of stutters that users can tolerate is less than the current number of stutters and the average number of stutters that users can tolerate, then reduce the resolution of the remaining insurance ad resources by one level. If the average number of stutters that users can tolerate is less than the current number of stutters and the maximum number of stutters that users can tolerate, then reduce the resolution of the remaining insurance ad resources by one level. If the current number of stutters equals the maximum number of stutters the user can tolerate, then the remaining insurance advertisement resources will not be played.

[0015] Optionally, the design of the dynamic loading strategy for insurance advertisements specifically includes: Obtain the encoding of videos and images compatible with the user's currently used device; Obtain the code of the insurance ad resource in the fixed ad slot. If the code of the insurance ad resource set initially is not a code compatible with the current device, convert the insurance ad resource that is not compatible with the current device to a compatible code. Obtain the current user's network resources; Get the network resources remaining while the user is loading a video; A set of user video behaviors that match the type of video the user is currently viewing; Use 50% of the remaining network resources to load insurance advertisement resources in sync with the video; Count the number of times the game has stopped running; If the minimum number of stutters that the user can tolerate is less than the current number of stutters and the average number of stutters that the user can tolerate, 25% of the remaining network resources will be used to load insurance advertisement resources synchronously with the video while the video is playing normally. If the current number of buffering instances equals the maximum number of buffering instances the user can tolerate, stop the synchronous loading of insurance advertising resources and videos.

[0016] The present invention has the following beneficial effects: 1. By collecting users' viewing data over the past three months, a comprehensive understanding of users' viewing preferences and behavioral patterns can be achieved. This data provides a foundation for subsequent analysis, helping the advertising system accurately grasp users' interests and ensure the relevance and effectiveness of advertisements. Categorizing videos by type helps to better analyze the performance of different types of videos. By analyzing the proportion of video types, the types of content that users are most interested in can be identified. Recording the time points of video buffering helps to analyze the impact of buffering on the user's viewing experience. Calculating the user's waiting time can provide data support for understanding the user's tolerance. This data analysis enables the advertising system to more accurately predict user reactions and make adjustments when ads load.

[0017] 2. Determining the longest acceptable waiting time for users can help optimize ad display strategies. Statistics on the number of buffering instances users tolerate reveal their patience threshold. This data provides a basis for optimizing video content smoothness and ad display timing. By understanding user tolerance, the ad system can avoid displaying ads during high-frequency buffering, thereby improving the overall user viewing experience. Calculating the average buffering time that users can tolerate for each video type provides important reference for ad loading strategies. This analysis allows the ad system to set reasonable loading times for different video types, reducing user dissatisfaction and churn. Understanding the longest buffering time users can tolerate across different video types provides advertisers with precise targeting strategies. Through this analysis, advertisers can design more targeted ad content and playback timing, ensuring smooth and effective ad display. Obtaining the shortest acceptable buffering time for users identifies user sensitivity to ad interference. Calculating the average number of buffering instances users can tolerate helps the ad system design more reasonable ad loading strategies. This analysis enables advertisers to avoid high-frequency buffering interference when placing ads.

[0018] 3. Understanding the maximum number of buffering instances users can tolerate across different video types provides crucial data for developing advertising strategies. By mastering this data, advertisers can more effectively control ad display frequency, ensuring that excessive interruptions don't lead to user churn. Conversely, obtaining the minimum number of buffering instances users can tolerate allows the advertising system to better understand users' tolerance limits. Creating user video behavior datasets provides comprehensive data support for advertising strategies. This dataset aggregates user tolerance for buffering across different video types, enabling the advertising system to develop more precise targeting strategies. Designing dynamic loading strategies effectively improves ad relevance and user experience. By adjusting ad display timing based on user behavior and preferences, advertisers can ensure ad content is more attractive when users watch videos. Playing ad resources without video buffering guarantees smooth ad display and enhances the user's viewing experience.

[0019] 4. By creating a timeline and mapping stuttering times, the user's viewing experience during video playback can be visually displayed. Numbering the start stuttering time helps to systematically record and analyze each stuttering event. This method provides a convenient index for subsequent analysis, enabling quick location of the specific time each stuttering occurs. By analyzing each end time, developers can assess the duration of the stuttering and its impact on user experience. This data will provide a basis for optimizing video loading and smoothness, thereby reducing the risk of user churn, improving ad view rates, and increasing overall user satisfaction. By mapping the start and end stuttering times one-to-one, the duration of each stutter can be effectively analyzed. Establishing this correspondence not only facilitates data analysis but also reveals user tolerance and behavioral patterns.

[0020] 5. By establishing a comprehensive insurance advertising resource set, all available advertising resources can be centrally managed and categorized. This centralized management not only improves the accessibility and retrieval efficiency of advertising resources but also ensures that all possible advertising materials can be quickly obtained during ad delivery. By analyzing the keywords users use to watch videos, a deeper understanding of user interests and preferences can be gained. This process not only makes the matching of advertising resources more personalized and targeted but also improves the relevance of the ads. Combining the matched advertising resources into a user insurance advertising resource set helps ensure that users see advertising content highly relevant to their interests when watching videos, thereby increasing user engagement and advertising effectiveness, effectively improving ad conversion rates and user satisfaction. By categorizing advertising resources, the advertising system can more effectively manage and allocate advertising materials. This prioritization ensures that the highest quality advertising resources are displayed first during ad delivery, improving the user's viewing experience. Video ads are generally more attractive than image ads, so prioritizing video ads can effectively increase user engagement and attention. Through resolution sorting, it also ensures that users can always see the best quality ads under different network conditions. This strategy helps enhance user acceptance of advertising content, thereby achieving better advertising results.

[0021] 6. Obtaining the sorting order of video categories helps to clarify the content types that users prefer. Focusing on the top 30% of video categories can help the system better meet user preferences. This strategy ensures that advertisers concentrate their resources on the most promising content, thereby maximizing the impact of advertising. By identifying and analyzing keywords, the advertising system can better understand user needs and thus optimize ad matching strategies.

[0022] 7. Identifying fixed and custom ad placements within a user's viewed video helps the advertising system understand the specific location of ad display. This information allows advertising strategies to be adjusted based on the characteristics of different ad placements, thereby improving ad display effectiveness. Fixed ad placements typically have higher exposure rates, while custom ad placements can be personalized based on user viewing habits. Confirming the type of video a user is currently viewing helps the advertising system understand the user's immediate interests and preferences. Analyzing video keywords between custom ad placements and previous ad placements helps understand the user's content flow and changes in interest during viewing. This analysis enables the advertising system to provide more relevant ads based on context, thereby improving ad targeting and effectiveness.

[0023] 8. Determining the resolution of the video being viewed by the user is a crucial step in optimizing ad delivery. Different resolutions affect ad loading speed and visual effects. By understanding the user's resolution settings, the ad system can select appropriate ad resources to ensure optimal display quality, thus avoiding ad stuttering or blurring due to resolution mismatch. If an ad matching the user's resolution and keywords can be found, user engagement and satisfaction will be enhanced. If no match is found, a search is performed in the overall resource set to ensure the ad still has a certain relevance and quality, preventing users from seeing completely irrelevant ads. Even if the final matched ad is not the most ideal, it still ensures uninterrupted ad display, enhancing the user's viewing experience.

[0024] 9. When video stuttering occurs in fixed ad slots, the advertising system can promptly identify this situation and take corresponding measures. By matching the user's current video viewing behavior set with the types of videos the user is currently viewing, the advertising system can obtain the user's preference for specific video types. By timing the occurrence of stuttering, the advertising system can assess the relationship between the stuttering duration and the user's acceptance threshold. This mechanism allows the system to adjust the advertising strategy at appropriate times, obtaining the relevant ad types and resolutions, which helps optimize ad loading strategies. Reducing the resolution of ad resources is an effective measure to deal with stuttering, which can speed up ad loading. This strategy can ensure that ad content is displayed as quickly as possible even under less than ideal technical conditions, reducing [the impact of stuttering]. By reducing user waiting time and minimizing the negative experience caused by lag, continuing to play the current ad when the ad loading time after reducing resolution is within the user's acceptable range can effectively improve user satisfaction. If reducing the resolution by one level still fails to meet the user's acceptance standard, further downgrading the ad resolution can be an effective strategy. This operation can quickly reduce ad loading time and avoid negative emotions caused by long waiting times. Through this dynamic adjustment, the advertising system demonstrates its flexibility and adaptability, ensuring consistent ad display under different network and device conditions. If the ad loading time exceeds the user's acceptable threshold, skipping the ad slot is an effective measure to prevent a deterioration in user experience. This decision reflects the importance placed on user satisfaction.

[0025] 10. When a custom ad slot experiences buffering during video playback, the system can quickly identify and initiate the corresponding processing flow. Custom ad slots are usually customized based on user interests, so their importance cannot be ignored. By matching the types of videos a user is currently browsing with their video behavior set, the system can obtain the user's viewing preferences and behavioral characteristics. Combined with the user's viewing habits, ads can be targeted more accurately, improving relevance and attractiveness. In the event of buffering, the system can provide ad content that matches the user's interests, thereby enhancing the viewing experience. Timing is performed when buffering occurs, allowing for real-time assessment of the relationship between buffering time and user acceptance.

[0026] 11. When the lag time reaches the minimum acceptable duration for the user, keyword matching helps to quickly find alternative ad resources. This strategy ensures that users see relevant and suitable ads during the lag period, reducing potential user dissatisfaction due to waiting. If no matching ad is found in the user's insurance ad resource set, continuing to search for lower-resolution ads in the total ad resource set can effectively speed up ad loading. Lowering the resolution maintains ad display continuity and improves the user viewing experience. This not only increases ad display opportunities but also reduces the risk of users leaving the page due to lag, improving overall ad effectiveness. If no suitable ad is found in all available resources, the system can select a lower-resolution ad from the user's insurance ad resource set. This flexibility ensures continuous ad display. This strategy prevents ad playback interruption due to lack of available resources, maintaining smooth user viewing. When the ad loading time after the first rematch is within the user's acceptable range, continuing to play the ad maximizes user satisfaction.

[0027] 12. If the loading time after the first rematch exceeds the user's acceptable threshold, continuing to search for resources that match the current ad keywords is an effective response. When no suitable ad resources are found, selecting an ad with a lower resolution from the total ad resource set is a pragmatic response. If all available resources fail to match, the system can still select an ad with a lower resolution from the user's insured ad resource set to ensure the flexibility of ad display. When the ad loading time after the second rematch is within the user's acceptable range, continuing to play the ad at the current resolution can effectively maintain the user's viewing experience. If the loading time after the second rematch exceeds the user's acceptable limit, skipping the ad slot can effectively prevent user dissatisfaction due to excessive waiting.

[0028] 13. By matching the types of videos a user is currently browsing with their video viewing behavior, the system can gain a deeper understanding of the user's viewing habits and preferences. Statistics on the number of stutters during video viewing help the system monitor the smoothness of video playback in real time. This data is crucial for evaluating user experience. By understanding the frequency of stutters, the system can determine the stability of the current playback environment and the feasibility of ad display. When the current number of stutters is below the user's acceptable threshold, the system decides not to take any action. Between the minimum and average number of stutters acceptable to the user, the system chooses to reduce the resolution of ad resources. This is an effective adjustment strategy. This measure aims to maintain ad playback while minimizing the impact on user experience. Reducing ad resolution can speed up loading, thereby improving the continuity of ad display. When the current number of stutters reaches the user's acceptable average but does not exceed the maximum value, the resolution of ad resources is further reduced. In this way, the system aims to maintain continuous ad playback while reducing user waiting time. When the current number of stutters reaches the user's acceptable limit, choosing not to play the remaining ad resources is an effective way to protect user experience. This strategy avoids forcibly playing ads when users are already experiencing discomfort, thus preventing user dissatisfaction and churn.

[0029] 14. By acquiring video and image encodings compatible with the user's device, the system can ensure that the playback content matches the user's device performance. This process not only improves the success rate of content loading but also avoids playback failures due to encoding incompatibility. When the encoding of insurance advertising resources does not meet the user's device compatibility standards, transcoding can effectively reduce playback obstacles caused by encoding mismatch. This flexible resource processing method can ensure the normal delivery of advertisements.

[0030] 15. Monitoring a user's current network resources helps the system assess available bandwidth, thereby optimizing content loading strategies. This information is crucial for the simultaneous loading of video and advertising resources. By understanding the user's network status, the system can dynamically adjust resource loading methods based on actual bandwidth conditions. By monitoring the remaining network bandwidth in real time, the system can determine whether it can support the simultaneous loading of advertising resources. This strategy not only improves the success rate of ad loading but also minimizes the buffering experience for users watching videos. Matching the types of videos a user is currently browsing with their behavioral patterns can provide profound insights, helping the system better understand the user's viewing habits.

[0031] 16. By using 50% of the remaining network resources for the synchronous loading of insurance advertisements, the system can effectively push advertisements to users while ensuring smooth video playback. This strategy maximizes the use of available bandwidth, improves the success rate of advertisement loading, and avoids playback interruptions due to insufficient network resources. Counting the number of stutters can provide important performance feedback for the system and help evaluate the smoothness of current video playback. Within the range of stutters that users can accept, choosing to use 25% of the remaining network resources for the synchronous loading of advertisements is a flexible response strategy. This strategy can moderately increase the loading of advertisements while ensuring normal video playback, ensuring that advertisements can be displayed smoothly. When the number of stutters reaches the user's maximum acceptable threshold, stopping the synchronous loading of advertisements is a strategy to protect the user experience. This measure avoids further burdening users when they are already feeling uncomfortable and helps maintain the user's viewing experience. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Example, refer to Figure 1 A method for dynamically loading insurance advertisements based on the Internet, comprising: Get all videos a user has viewed on social media platforms over the past three months; The acquired videos are categorized according to video type; Get the percentage of each type of video in all videos, and sort them in descending order of percentage. Then, label the sorted video types as type 1, type 2, ..., type N. When a user browses a video; If the video is choppy, obtain the time point when the chop starts and the time point when the chop ends for all videos. Calculate the time point when the video ends and the time point when the video starts to stop, and record the result as the user's waiting time. Get all the user's waiting time in a video and compare them, then record the longest user waiting time as the longest waiting time accepted by the user for this video. When a user switches to or closes a video during a video stutter, the system counts all video stutters that occurred before this stutter and records the result as the number of stutters the user has experienced. Calculate the average duration of buffering that users can tolerate in each type of video. The average duration of buffering that users can tolerate = the sum of the waiting time of all users in each type of video / the total number of buffering occurrences in each type of video. Get the longest duration of buffering or stuttering experienced by the user in each video category; Get the shortest duration of buffering that users can tolerate in each type of video; Calculate the average number of stutters that users can tolerate in each video category. The average number of stutters that users can tolerate = the sum of the number of stutters in all videos in each video category / the total number of videos in each video category. Get the maximum number of stutters that users can tolerate in each type of video; Find the minimum number of stutters that users can tolerate in each type of video; Set up a set of user video behaviors named after the video type. The elements in the set include the average duration of buffering that users can tolerate, the longest duration of buffering that users can tolerate, the shortest duration of buffering that users can tolerate, the average number of buffering instances that users can tolerate, the minimum number of buffering instances that users can tolerate, and the maximum number of buffering instances that users can tolerate. Design a dynamic loading strategy for insurance advertisements; If the video plays smoothly without buffering, the set insurance advertisement will be played when the video reaches the advertisement slot.

[0035] By collecting user viewing data over the past three months, a comprehensive understanding of user viewing preferences and behavioral patterns can be achieved. This data provides a foundation for subsequent analysis, helping the advertising system accurately grasp user interests and ensure the relevance and effectiveness of advertisements. Categorizing videos by type helps in better analyzing the performance of different video types. By analyzing the proportion of video types, the types of content users are most interested in can be identified. Recording the time points of video buffering helps analyze the impact of buffering on the user viewing experience. Calculating user waiting time provides data support for understanding user tolerance. This data analysis enables the advertising system to more accurately predict user reactions and make adjustments during ad loading. Determining the longest acceptable waiting time for users can help optimize ad display strategies. Statistics on the number of buffering instances users tolerate reveal their patience threshold. This data provides a basis for optimizing video content smoothness and ad display timing. By understanding user tolerance, ad systems can avoid displaying ads during periods of frequent buffering, thereby improving the overall user viewing experience. Calculating the average buffering time that users can tolerate for each video type provides important reference for ad loading strategies. This analysis allows ad systems to set reasonable loading times for different video types, thereby reducing user dissatisfaction and churn. Understanding the longest buffering time users can tolerate across different video types can provide valuable insights for ad loading strategies. This analysis provides advertisers with precise insights for ad placement strategies. Through this analysis, advertisers can design more targeted ad content and timing, ensuring smooth and effective ad display. By obtaining the shortest acceptable buffering time for users, the sensitivity to ad interruptions can be identified. Calculating the average number of buffering instances users can tolerate helps design more reasonable ad loading strategies. This analysis allows advertisers to avoid high-frequency buffering interference when placing ads, understanding the maximum number of buffering instances users can tolerate across different video types. This provides crucial data for developing ad placement strategies. By mastering this data, advertisers can more effectively control ad display frequency, ensuring that ads do not... Excessive interruptions can lead to user churn. By identifying the minimum number of buffering instances users can tolerate, advertising systems can better understand users' tolerance limits. Creating a set of user video behavior patterns can provide comprehensive data support for ad delivery strategies. This set aggregates users' tolerance for buffering across different video types, enabling advertising systems to develop more precise delivery strategies. Designing dynamic loading strategies can effectively improve ad relevance and user experience. By adjusting ad display timing based on user behavior and preferences, advertisers can ensure that ad content is more attractive when users watch videos. Playing ad resources without video buffering ensures smooth ad display and enhances the user's viewing experience.

[0036] The specific time points at which the video pause begins, corresponding to the time point at which the video ends and pauses, include: Set a timeline in the order from the start to the end of the video playback. The beginning of the timeline is the time when the video starts playing, and the end of the timeline is the time when the video ends playing. Map all the stuttering times onto the timeline. The first time a lag occurs is recorded as the first time a lag occurs, the second time a lag occurs is recorded as the second time a lag occurs, and so on until the Mth time a lag occurs is recorded as the Mth time a lag occurs. The first time the lag occurred is recorded as the first time the lag occurred, the second time the lag occurred is recorded as the second time the lag occurred, and so on until the Mth time the lag occurred is recorded as the Mth time the lag occurred. The first start time of the lag corresponds to the first end time of the lag, the second start time of the lag corresponds to the second end time of the lag, and so on, the Mth start time of the lag corresponds to the Mth end time of the lag.

[0037] By creating a timeline and mapping stuttering points, the user's viewing experience during video playback can be visually displayed. Numbering the start stuttering point helps to systematically record and analyze each stuttering event. This method provides a convenient index for subsequent analysis, enabling quick location of the specific time each stuttering occurs. By analyzing each end stuttering point, developers can assess the duration of the stuttering and its impact on user experience. This data will provide a basis for optimizing video loading and smoothness, thereby reducing the risk of user churn, improving ad view rates, and increasing overall user satisfaction. By mapping the start and end stuttering points one-to-one, the duration of each stuttering can be effectively analyzed. Establishing this correspondence not only facilitates data analysis but also reveals user tolerance and behavioral patterns.

[0038] The design of the dynamic loading strategy for insurance advertisements specifically includes: Set up a general resource set for insurance advertisements; Obtain the keywords of the top 30% of the video types watched by users, use the keywords to match the corresponding insurance advertising resources in the total insurance advertising resource set, and combine the matched insurance advertising resources into a user insurance advertising resource set; The elements in the total insurance advertising resource set and the elements in the user insurance advertising resource set are classified separately. Insurance advertisements are divided into priority category and secondary category. The priority category includes video insurance advertisements, and the secondary category includes image insurance advertisements. Within the priority and secondary categories, insurance advertisements are further classified into first-level resources, second-level resources, ... Q-level resources in descending order of resolution.

[0039] By establishing a centralized collection of insurance advertising resources, all available advertising resources can be managed and categorized in a centralized manner. This centralized management not only improves the accessibility and retrieval efficiency of advertising resources but also ensures that all possible advertising materials can be quickly obtained during ad delivery. By analyzing the keywords users use to watch videos, a deeper understanding of user interests and preferences can be gained. This process not only makes the matching of advertising resources more personalized and targeted but also improves the relevance of the ads. Combining the matched advertising resources into a user insurance advertising resource collection helps ensure that users see advertising content highly relevant to their interests when watching videos, thereby increasing user engagement and advertising effectiveness, effectively improving ad conversion rates and user satisfaction. By categorizing advertising resources, the advertising system can manage and allocate advertising materials more effectively. This prioritization ensures that the highest quality advertising resources are displayed first during ad delivery, improving the user's viewing experience. Video ads are generally more attractive than image ads, so prioritizing video ads can effectively increase user engagement and attention. Through resolution sorting, it also ensures that users can always see the best quality ads under different network conditions. This strategy helps enhance user acceptance of advertising content, thereby achieving better advertising results.

[0040] The keywords used to obtain the top 30% of the video types watched by users specifically include: Get the sorted order of video categories; The top 30% of video categories after sorting are obtained. These top 30% are representative and better reflect user preferences. The remaining 30% can be modified adaptively according to actual needs. We analyzed the top 30% of video categories and extracted keywords.

[0041] Obtaining the sort order of video categories helps to clarify the content types that users prefer. Focusing on the top 30% of video categories can help the system better meet user preferences. This strategy ensures that advertisers concentrate their resources on the most promising content, thereby maximizing the impact of their ads. By identifying and analyzing keywords, the advertising system can better understand user needs and thus optimize ad matching strategies.

[0042] The design of the dynamic loading strategy for insurance advertisements specifically includes: Get the fixed and custom ad slots in the videos viewed by users; Get the type of video the user is currently viewing; Analyze the keywords in the video between the custom ad placement and the previous fixed ad placement or custom ad placement; Get the resolution set by the user when watching the video; In the user's insurance ad resource set, randomly match an insurance ad resource that matches the keywords obtained from the analysis and the resolution set by the user. If no match is found, match from the total insurance ad resource set. If no match is found from the total insurance ad resource set, then take any resource from the total insurance ad resource set as the ad in the custom ad slot.

[0043] Identifying fixed and custom ad placements within a user's viewed video helps the advertising system understand the specific location of ad display. This information allows ad delivery strategies to be adjusted based on the characteristics of different ad placements, thereby improving ad performance. Fixed ad placements typically have higher exposure rates, while custom ad placements can be personalized based on user viewing habits. Confirming the type of video a user is currently viewing helps the advertising system understand the user's immediate interests and preferences. Analyzing video keywords between custom ad placements and previous ad placements helps understand the user's content flow and changes in interest during viewing. This analysis enables the advertising system to deliver more relevant ads based on context, thereby improving ad targeting and effectiveness. Determining the resolution of the video a user is currently watching is a crucial step in optimizing ad delivery. Different resolutions affect ad loading speed and visual effects. By understanding the user's resolution settings, the ad system can select appropriate ad resources to ensure optimal display quality, thus avoiding ad stuttering or blurring caused by resolution mismatch. If an ad matching the user's set resolution and keywords can be found, user engagement and satisfaction will be enhanced. If no match is found, a search is performed in the overall resource set to ensure that the ad still has a certain degree of relevance and quality, preventing users from seeing completely irrelevant ads. Even if the final matched ad is not the most ideal, it still ensures that ad display is uninterrupted, enhancing the user's viewing experience.

[0044] The design of the dynamic loading strategy for insurance advertisements specifically includes: When lag occurs in a fixed ad slot; A set of user video behaviors that match the type of video the user is currently viewing; Start timing when the lag begins; when the lag duration reaches the minimum acceptable lag duration for the user, obtain the type and resolution of the insurance ad in the current fixed ad slot. Load the current insurance advertisement resource at a lower resolution by one level. If the current insurance ad resource resolution is reduced by one level, and the loading time of the insurance ad after the resolution reduction is less than or equal to the average time that the user can tolerate the lag, then the insurance ad resource at the current resolution will be played. If the loading time of the insurance ad after reducing the resolution by one level is greater than the average time that users can tolerate with lag, then the resolution of the current insurance ad will be reduced by two levels. If the resolution of the current insurance advertisement resource is reduced by two levels, and the loading time of the insurance advertisement after the resolution is reduced by two levels is less than or equal to the longest time that the user can tolerate the lag, then the insurance advertisement resource at the current resolution will be played. If the loading time of the insurance ad after reducing the resolution by two levels is greater than the maximum duration of buffering that the user can tolerate, then the ad playback for that ad slot will be skipped.

[0045] When video stuttering occurs in fixed ad slots, the advertising system can promptly identify the situation and take corresponding measures. By matching the user's current video viewing behavior set with the types of videos they are currently viewing, the advertising system can obtain the user's preference for specific video types. By timing the occurrence of stuttering, the advertising system can assess the relationship between the stuttering duration and the user's acceptance threshold. This mechanism allows the system to adjust advertising strategies at appropriate times. Obtaining the relevant ad types and resolutions helps optimize ad loading strategies. Reducing the resolution of ad resources is an effective measure to deal with stuttering, which can speed up ad loading. This strategy can ensure that ad content is displayed as quickly as possible even under less than ideal technical conditions, reducing user fatigue. The system aims to reduce user waiting time and minimize the negative experience caused by buffering. If the ad loading time after reducing resolution is within the user's acceptable range, continuing to play the current ad effectively improves user satisfaction. If reducing the resolution by one level still fails to meet the user's acceptance standard, further downgrading the ad resolution is an effective strategy. This action quickly reduces ad loading time, preventing negative user emotions caused by prolonged waiting. Through this dynamic adjustment, the advertising system demonstrates its flexibility and adaptability, ensuring consistent ad display under different network and device conditions. If the ad loading time exceeds the user's acceptable threshold, skipping the ad slot is an effective measure to prevent a deterioration in user experience. This decision reflects a strong emphasis on user satisfaction.

[0046] The design of the dynamic loading strategy for insurance advertisements specifically includes: When lag occurs in a custom ad slot; A set of user video behaviors that match the type of video the user is currently viewing; When the lag starts, a timer is set. When the lag duration reaches the minimum acceptable lag duration for the user, the system matches the user's insurance ad resource set with the same keywords as the insurance ad in the current custom ad slot. If no match is found, the system matches the total insurance ad resource set and loads an insurance ad resource with a resolution one level lower than the current resolution. If no match is found in the total insurance ad resource set, the system loads an insurance ad resource with a resolution one level lower than the current resolution from the user's insurance ad resource set. If the loading time of the first rematched insurance ad resource is less than or equal to the average duration of buffering that the user can tolerate, then the rematched insurance ad resource will be played. If, after the first rematch of insurance ad resources, the loading time of the first rematch of insurance ad resources is greater than the average time that the user can tolerate from buffering, then the system will match ad resources in the user's insurance ad resource set that have the same keywords as the insurance ad in the current custom ad slot. If no match is found, the system will match in the total set of insurance ad resources and select an insurance ad resource with a resolution two levels lower than the current resolution for loading. If no match is found in the total set of insurance ad resources either, then the system will select an insurance ad resource in the user's insurance ad resource set with a resolution two levels lower than the current resolution for loading. If, after the second rematch of insurance ad resources, the loading time of the second rematch of insurance ad resources is less than or equal to the maximum duration during which the user can tolerate buffering, then the insurance ad resources at the current resolution will be played. If, after a second rematch of the insurance ad resource, the loading time of the second rematch exceeds the maximum acceptable buffering time for the user, then the ad playback for that ad slot will be skipped.

[0047] When custom ad slots experience buffering during video playback, the system can quickly identify and initiate corresponding processing. Custom ad slots are typically tailored to user interests, making their importance undeniable. By matching the user's current video viewing categories with their video behavior set, the system can acquire the user's viewing preferences and behavioral characteristics. Combined with user viewing habits, ads can be more precisely targeted, improving relevance and attractiveness. In the event of buffering, the system can provide ad content that matches the user's interests, thereby enhancing the viewing experience. Timing during buffering allows for real-time assessment of the relationship between buffering duration and user acceptance. When the buffering time reaches the user's minimum acceptable duration, keyword matching helps quickly find alternative ad resources. This strategy ensures users see relevant and suitable ads during buffering, reducing potential user dissatisfaction due to waiting. If no matching ad is found in the user's existing ad resource set, searching for lower-resolution ads in the overall ad resource set effectively speeds up ad loading. Lowering the resolution maintains ad display continuity, improving the user viewing experience. This not only increases ad display opportunities but also reduces the risk of users leaving the page due to buffering, improving overall performance. To ensure overall advertising effectiveness, if no suitable ad is found among all available resources, the system can select a lower-resolution ad from the user's insurance ad resource set. This flexibility ensures continuous ad display and prevents ad playback interruptions due to a lack of available resources, maintaining a smooth viewing experience for the user. When the ad loading time after the first rematch is within the user's acceptable range, continuing to play the ad maximizes user satisfaction. If the loading time after the first rematch exceeds the user's acceptable threshold, continuing to search for resources matching the current ad keywords is an effective response. When no suitable ad resource is found, selecting a lower-resolution ad from the total ad resource set is a pragmatic response. If all available resources fail to match, the system can still select a lower-resolution ad from the user's insurance ad resource set to load, ensuring ad display flexibility. When the ad loading time after the second rematch is within the user's acceptable range, continuing to play the ad at the current resolution effectively maintains the user's viewing experience. If the loading time after the second rematch exceeds the user's acceptable limit, skipping the ad slot effectively prevents user dissatisfaction due to excessive waiting time.

[0048] The design of the dynamic loading strategy for insurance advertisements specifically includes: A set of user video behaviors that match the type of video the user is currently viewing; Get the number of times the user experiences buffering while watching a video; If the current number of stutters is less than the minimum number of stutters that the user can tolerate, then no action will be taken; If the minimum number of stutters that users can tolerate is less than the current number of stutters and the average number of stutters that users can tolerate, then reduce the resolution of the remaining insurance ad resources by one level. If the average number of stutters that users can tolerate is less than the current number of stutters and the maximum number of stutters that users can tolerate, then reduce the resolution of the remaining insurance ad resources by one level. If the current number of stutters equals the maximum number of stutters the user can tolerate, then the remaining insurance advertisement resources will not be played.

[0049] By matching the types of videos a user is currently browsing with their video viewing behavior, the system can gain a deeper understanding of the user's viewing habits and preferences. Statistical analysis of the number of stutters during video viewing helps the system monitor video playback smoothness in real time. This data is crucial for evaluating user experience. By understanding the frequency of stutters, the system can determine the stability of the current playback environment and the feasibility of ad display. When the current number of stutters is below the user's acceptable threshold, the system decides not to take any action. Between the minimum and average number of stutters acceptable to the user, the system chooses to reduce the resolution of ad resources—an effective adjustment strategy. This measure aims to maintain ad playback while minimizing the impact on user experience. Lowering ad resolution can speed up loading, thereby improving the continuity of ad display. When the current number of stutters reaches the user's acceptable average but does not exceed the maximum, the resolution of ad resources is further reduced. In this way, the system aims to maintain continuous ad playback while reducing user waiting time. When the current number of stutters reaches the user's acceptable limit, choosing not to play the remaining ad resources is an effective way to protect user experience. This strategy avoids forcing ads to play when users are already experiencing discomfort, thus preventing user dissatisfaction and churn.

[0050] The design of the dynamic loading strategy for insurance advertisements specifically includes: Obtain the encoding of videos and images compatible with the user's currently used device; Obtain the code of the insurance ad resource in the fixed ad slot. If the code of the insurance ad resource set initially is not a code compatible with the current device, convert the insurance ad resource that is not compatible with the current device to a compatible code. Obtain the current user's network resources; Get the network resources remaining while the user is loading a video; A set of user video behaviors that match the type of video the user is currently viewing; Using 50% of the remaining network resources to load insurance advertisement resources synchronously with the video ensures that the insurance advertisement resources are loaded as quickly as possible without affecting the normal loading of the video, avoiding lag in the loading of both the video and the insurance advertisement resources. The 50% allocation can be adjusted adaptively. Count the number of times the game has stopped running; If the minimum number of stutters that the user can accept is less than the current number of stutters and the average number of stutters that the user can accept, then when the video is playing normally, 25% of the remaining network resources are used to load insurance advertisement resources synchronously with the video. Using 25% of the remaining network resources to load insurance advertisement resources synchronously with the video is an adaptive reduction of the remaining network usage after a certain number of stutters, in order to avoid frequent video stutters due to network occupation. The 25% can be adjusted adaptively. If the current number of buffering instances equals the maximum number of buffering instances the user can tolerate, stop the synchronous loading of insurance advertising resources and videos.

[0051] By acquiring video and image encodings compatible with the user's device, the system ensures that the playback content matches the user's device performance. This process not only improves the success rate of content loading but also avoids playback failures due to encoding incompatibility. When the encoding of insurance advertisement resources does not meet the user's device compatibility standards, transcoding can effectively reduce playback obstacles caused by encoding mismatch. This flexible resource handling method ensures normal ad delivery. Monitoring the user's current network resources helps the system assess available bandwidth, thereby optimizing content loading strategies. This information is crucial for the synchronous loading of video and ad resources. By understanding the user's network status, the system can dynamically adjust resource loading methods based on actual bandwidth conditions. By monitoring the remaining network bandwidth in real time, the system can determine whether it can support the synchronous loading of ad resources. This strategy not only improves the success rate of ad loading but also minimizes the user's buffering experience when watching videos, matching the user's current video browsing type with their behavior. The system can provide profound insights to better understand users' viewing habits. By using 50% of the remaining network resources for the simultaneous loading of ad resources, the system can effectively push ads to users while ensuring smooth video playback. This strategy maximizes the use of available bandwidth, improves the success rate of ad loading, and avoids playback interruptions due to insufficient network resources. Counting the number of stutters provides important performance feedback to the system, helping to assess the smoothness of current video playback. Within the range of stutters that users can tolerate, choosing to use 25% of the remaining network resources for the simultaneous loading of ads is a flexible strategy. This strategy can moderately increase ad loading while ensuring normal video playback, ensuring that ads can be displayed smoothly. When the number of stutters reaches the user's maximum acceptable threshold, stopping the simultaneous loading of ads is a strategy to protect the user experience. This measure avoids further burdening users when they are already experiencing discomfort, helping to maintain the user's viewing experience.

[0052] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0053] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for dynamically loading insurance advertisements based on the Internet, characterized in that: include: Get all videos that users have viewed on social media platforms in the past three months; Classify the acquired videos according to video types; Obtain the proportion of the total number of videos of each category in all videos, and sort the videos in descending order from the highest proportion, and record the sorted video categories as the first category video, the second category video, and so on, the Nth category video; When users browse videos; If the video freezes, get the time when the freeze starts and ends for all videos; Calculate the time point at which the video freezes and the time point at which the video freezes, respectively, and record the result as the user waiting time. Get all the waiting times of users in a video and compare them. The longest waiting time of a user is recorded as the longest waiting time accepted by the user in this video. When the user switches or closes the video when the video freezes, all video freezes that occurred before the current video freeze are counted, and the result is recorded as the number of freezes accepted by the user; Calculate the average duration of freezes that users can accept in each type of video. The average duration of freezes that users can accept = the sum of the waiting time of all users in each type of video / the total number of freezes in each type of video; Get the longest duration of video freeze that users can tolerate in each type of video; Get the shortest duration of freezes that users can tolerate in each type of video; Calculate the average number of freezes that users can accept in each type of video. The average number of freezes that users can accept = the sum of the number of freezes in all videos in each type of video / the total number of videos in each type of video; Get the maximum number of freezes that users can accept in each type of video; Get the minimum number of freezes that users can accept in each type of video; A user video behavior set is set based on the video category. The elements in the set include the average duration of freezes that users can accept for the video category, the longest duration of freezes that users can accept, the shortest duration of freezes that users can accept, the average number of freezes that users can accept, the minimum number of freezes that users can accept, and the maximum number of freezes that users can accept; Design a dynamic loading strategy for insurance ads; If the video is not stuck, the set insurance advertising resources will be played when the video reaches the advertising position.

2. The method for dynamically loading insurance advertisements based on the Internet according to claim 1, characterized in that: The time point at which the video freezes corresponding to the time point at which the video freezes ends specifically includes: In the sequence from the start of video playback to the end of video playback, a timeline is set, the start end of the timeline is the start time of the video playback, and the end end of the timeline is the end time of the video playback, and all the freeze time points are mapped to the timeline; The first time point of the start of the jam is recorded as the first time point of the start of the jam, the second time point of the start of the jam is recorded as the second time point of the start of the jam, and the Mth time point of the start of the jam is recorded as the Mth time point of the start of the jam; The first end jamming time point that appears is recorded as the first end jamming time point, the second end jamming time point that appears is recorded as the second end jamming time point, and the Mth end jamming time point that appears is recorded as the Mth end jamming time point; Then the first start-stuck time point corresponds to the first end-stuck time point, the second start-stuck time point corresponds to the second end-stuck time point... the Mth start-stuck time point corresponds to the Mth end-stuck time point.

3. The method for dynamically loading insurance advertisements based on the Internet according to claim 1, characterized in that: The design of the dynamic loading strategy for insurance advertisements specifically includes: Set the total insurance advertising resource collection; Obtain the top 30% keywords of the types of videos watched by the user, use the keywords to match the corresponding insurance advertising resources in the total insurance advertising resource set, and form the user's insurance advertising resource set with the matched insurance advertising resources; The elements in the total insurance advertisement resource set and the elements in the user insurance advertisement resource set are classified respectively, and the insurance advertisements are divided into a priority class and a secondary class. The priority class includes insurance advertisements in video form, and the secondary class includes insurance advertisements in picture form. In the priority class and the secondary class, the insurance advertisements are divided into primary resources, secondary resources...Q-level resources in descending order according to their resolutions.

4. The method for dynamically loading insurance advertisements based on the Internet according to claim 3, characterized in that: The keywords for obtaining the top 30% of the types of videos watched by users specifically include: Get the sorted order of video categories; Get the top 30% of video categories after sorting; Analyze the top 30% of video categories and extract keywords.

5. The method for dynamically loading insurance advertisements based on the Internet according to claim 3, characterized in that: The design of the dynamic loading strategy for insurance advertisements specifically includes: Get the fixed and custom ad slots in the videos that users browse; Get the type of video the user is currently browsing; Analyze the keywords of the video between the custom ad slot and the previous fixed ad slot or custom ad slot; Get the resolution set by the user when currently watching the video; In the user's insurance advertising resource set, any insurance advertising resource with the same keyword obtained by analysis and the resolution set by the user is matched. If no match is found, then match it in the total insurance advertising resource set. If no match is found in the total insurance advertising resource set, then obtain any resource in the total insurance advertising resource set as the advertisement in the custom ad slot.

6. The method for dynamically loading insurance advertisements based on the Internet according to claim 5, characterized in that: The design of the dynamic loading strategy for insurance advertisements specifically includes: When lag occurs in a fixed ad slot; A collection of user video behaviors that matches the type of video the user is currently browsing; Start counting at the beginning of the freeze, and when the freeze duration reaches the shortest freeze duration acceptable to users, obtain the type and resolution of the insurance ad in the current fixed ad slot; Reduce the resolution of the current insurance advertising resources by one level for loading; If the resolution of the current insurance ad resource is reduced by one level, and the loading time of the insurance ad after the resolution is reduced by one level is less than or equal to the average duration of lag that the user can accept, then the insurance ad resource with the current resolution will be played; If the insurance ad loading time after the resolution of the current insurance ad resource is reduced by one level is greater than the average loading time that users can accept, the resolution of the current insurance ad resource will be reduced by two levels; If the resolution of the current insurance ad resource is reduced by two levels, and the loading time of the insurance ad after the resolution is reduced by two levels is less than or equal to the longest duration of lag that the user can accept, then the insurance ad resource with the current resolution will be played; If the resolution of the current insurance advertising resources is reduced by two levels, and the loading time of the insurance advertisement after the resolution is reduced by two levels is greater than the longest duration of lag that the user can accept, the advertisement playback of this ad slot will be skipped.

7. The method for dynamically loading insurance advertisements based on the Internet according to claim 5, characterized in that: The design of the dynamic loading strategy for insurance advertisements specifically includes: When the freeze occurs in a custom ad slot; A collection of user video behaviors that matches the type of video the user is currently browsing; Start counting at the beginning of the freeze. When the freeze duration reaches the minimum freeze duration acceptable to the user, match the ad resource with the same keyword as the insurance ad in the current custom ad slot in the user insurance ad resource set. If no match is found, match it in the total insurance ad resource set. Select an insurance ad resource with a resolution one level lower than the current resolution to load. If no match is found in the total insurance ad resource set, select an insurance ad resource with a resolution one level lower than the current resolution in the user insurance ad resource set to load. If after the first re-matching of insurance advertising resources, the loading time of the first re-matched insurance advertising resources is ≤ the average duration of lag that users can accept, then the re-matched insurance advertising resources will be played; If after the first re-matching of insurance advertising resources, the loading time of the first re-matching of insurance advertising resources is greater than the average duration of lag that users can accept, then match the advertising resources with the same keywords as the insurance advertising on the current custom ad slot in the user insurance advertising resource set. If no match is found, match in the total insurance advertising resource set, and select an insurance advertising resource with a resolution two levels lower than the current resolution to load. If no match is found in the total insurance advertising resource set, select an insurance advertising resource with a resolution two levels lower than the current resolution in the user insurance advertising resource set to load. If after the second rematch of insurance ad resources, the loading time of the second rematch of insurance ad resources is less than or equal to the longest duration of freeze that the user can accept, then the insurance ad resources of the current resolution will be played; If after the second re-matching of insurance ad resources, the loading time of the second re-matching of insurance ad resources is greater than the longest duration of lag that users can accept, the ad playback of this ad slot will be skipped.

8. The method for dynamically loading insurance advertisements based on the Internet according to claim 5, characterized in that: The design of the dynamic loading strategy for insurance advertisements specifically includes: A collection of user video behaviors that matches the type of video the user is currently browsing; Get the number of times the user's video is currently stuck; If the current number of freezes is less than the minimum number of freezes that the user can accept, no processing will be done; If the minimum number of freezes that the user can accept is less than the current number of freezes and less than the average number of freezes that the user can accept, the resolution of the remaining insurance advertising resources will be reduced by one level; If the average number of freezes that users can accept is less than the current number of freezes and less than the maximum number of freezes that users can accept, the resolution of the remaining insurance advertising resources will be reduced by one level; If the current number of freezes is equal to the maximum number of freezes that the user can accept, the remaining insurance advertising resources will not be played.

9. The method for dynamically loading insurance advertisements based on the Internet according to claim 1, characterized in that: The design of the dynamic loading strategy for insurance advertisements specifically includes: Get the video and picture encodings that are compatible with the device currently used by the user; Obtain the code of the insurance advertising resources in the fixed advertising position. When the code of the insurance advertising resources initially set is not compatible with the current device, transcode the insurance advertising resources that are not compatible with the current device into a compatible code. Get the current user's network resources; Get the remaining network resources when the user loads the video; A set of user video behaviors that matches the type of video the user is currently browsing; Use 50% of the remaining network resources to load insurance advertising resources synchronously with the video; Count the current number of freezes; If the minimum number of freezes that the user can accept is less than or equal to the current number of freezes and less than the average number of freezes that the user can accept, when the video is playing normally, 25% of the remaining network resources are used to load the insurance ad resources synchronously with the video; If the current number of freezes is equal to the maximum number of freezes that the user can accept, stop the synchronous loading of insurance advertising resources and videos.

Images