A method and system for dynamic content carousel based on HarmonyOS distributed architecture

Abstract

This invention relates to the field of computer technology, specifically a dynamic content carousel method and system based on the HarmonyOS distributed architecture. The method includes: receiving a content carousel instruction; querying the HarmonyOS distributed soft bus based on the instruction to obtain multiple distributed devices; performing regional priority evaluation on multiple pieces of content to be played to obtain multiple content priority indices; analyzing user characteristics of the multiple distributed devices to obtain multiple user feature data; performing a genetic algorithm simulation on the multiple pieces of content to be played and the multiple distributed devices based on multiple distributed parsing rates, multiple user feature data, and multiple content priority indices to obtain an optimal carousel; and completing the dynamic content carousel based on the HarmonyOS distributed architecture using multiple target carousel devices. This invention can improve the content push effect and user experience of carousel across multiple distributed devices, and achieve intelligent allocation of carousel content among multiple HarmonyOS devices.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a dynamic content carousel method, system, electronic device, and computer-readable storage medium based on the HarmonyOS distributed architecture. Background Technology

[0002] With the advent of the Internet of Things era, the types and numbers of smart terminals have grown rapidly, from smartphones and tablets to smart TVs, in-vehicle screens, and smart advertising screens. Content display modes have also evolved from static playback on a single terminal to dynamic carousel modes involving multiple terminals. How to achieve smooth and personalized content delivery across multiple devices has become an important direction for improving user experience and information dissemination efficiency.

[0003] Currently, some content management platforms establish a central server to achieve multi-device content distribution and carousel control. In addition, some systems are trying to introduce big data analysis and cloud recommendation algorithms to optimize the content playback order, thereby improving content matching and playback efficiency to a certain extent.

[0004] However, most existing multi-device content carousel solutions still rely on centralized architectures, lacking the ability to dynamically perceive and collaboratively control the resource status among distributed devices. Furthermore, existing systems fail to comprehensively consider multi-dimensional data such as geographical factors, user access characteristics, and content priority during content allocation, resulting in insufficient targeting of content display and low carousel efficiency. Therefore, there is an urgent need for a dynamic content carousel allocation method based on the HarmonyOS distributed architecture, capable of achieving resource collaboration and intelligent content allocation in a multi-device environment. This method can achieve intelligent allocation of carousel content among multiple HarmonyOS devices by introducing optimization mechanisms such as genetic algorithms. Summary of the Invention

[0005] This invention provides a dynamic content carousel method based on the HarmonyOS distributed architecture and a computer-readable storage medium. Its main purpose is to improve the content push effect and user experience of carousel across multiple distributed devices and to achieve intelligent allocation of carousel content among multiple HarmonyOS devices.

[0006] To achieve the above objectives, the present invention provides a dynamic content carousel method based on the HarmonyOS distributed architecture, comprising:

[0007] Receive content carousel instructions, perform HarmonyOS distributed soft bus queries based on content carousel instructions to obtain multiple distributed devices, including: distributed CPUs, distributed GPUs and historical databases;

[0008] Get multiple pieces of content to be played;

[0009] Multiple pieces of content to be played are evaluated based on their geographic priority, resulting in multiple content priority indices. Each content priority index corresponds one-to-one with a piece of content to be played.

[0010] Network performance analysis is performed on multiple distributed devices to obtain multiple distributed resolution rates. User feature analysis is also performed on multiple distributed devices to obtain multiple user feature data. The distributed resolution rates and user feature data correspond one-to-one with the distributed devices. The user feature data includes: user usage frequency and multiple access word groups. The access word groups include: access words and access counts.

[0011] Based on multiple distributed parsing rates, multiple user feature data, and multiple content priority indices, a genetic algorithm simulation is performed on multiple content to be played and multiple distributed devices to obtain the optimal carousel.

[0012] Based on the optimal carousel body, content carousel settings are configured for multiple distributed devices to obtain multiple target carousel devices;

[0013] Dynamic content carousel based on HarmonyOS distributed architecture is achieved by using multiple target carousel devices.

[0014] Optionally, the process of performing a region-priority evaluation on multiple pieces of content to be played yields multiple content priority indices, including:

[0015] For each of the multiple pieces of content to be played, perform the following operation:

[0016] Confirm the origin coordinates, total number of clicks, and daily click growth rate of the content to be played. The origin coordinates include the origin longitude and origin latitude.

[0017] Obtain the coordinates of the user equipment, which include the equipment's longitude and latitude;

[0018] The content priority index is calculated based on user device coordinates, origin location coordinates, total clicks, and daily click growth rate.

[0019] By summing the content priority indices, multiple content priority indices are obtained.

[0020] Optionally, the step of performing network performance analysis on multiple distributed devices to obtain multiple distributed resolution rates includes:

[0021] Perform the following operation on each of the multiple distributed devices:

[0022] Read the CPU load of distributed CPUs in the distributed device, and read the GPU stream count, GPU memory capacity and GPU memory bandwidth of distributed GPUs in the distributed device.

[0023] Network status monitoring of distributed devices yields bandwidth utilization, packet loss rate, and network latency.

[0024] The distributed parsing rate is calculated based on CPU load, GPU stream count, GPU memory capacity, GPU memory bandwidth, bandwidth utilization, packet loss rate, and network latency. The calculation formula is as follows:

[0025]

[0026] in, For distributed analytical rate, , and These are GPU stream count, GPU memory capacity, and GPU memory bandwidth, respectively. CPU load, For bandwidth utilization, For packet loss rate, Due to network latency, It is the hyperbolic tangent function;

[0027] By summing the distributed resolution rates, multiple distributed resolution rates are obtained.

[0028] Optionally, the step of performing user feature analysis on multiple distributed devices to obtain multiple user feature data includes:

[0029] Perform the following operation on each of the multiple distributed devices:

[0030] Read multiple browsing titles from the historical database in the distributed device;

[0031] Perform the following operation for each of the multiple browse titles:

[0032] Extract Chinese characters from the browsing titles to obtain multiple title characters;

[0033] Based on a preset set of commonly used characters, common characters are removed from multiple titles to obtain a reference set of characters;

[0034] By compiling a collection of reference Chinese character sets, multiple reference Chinese character sets were obtained.

[0035] By merging multiple reference character sets, a total historical character set is obtained;

[0036] The total historical set of Chinese characters was categorized, resulting in multiple categorized character sets.

[0037] Perform the following operation on each of the multiple categorized Chinese character sets:

[0038] Confirm the accessed characters and their access counts in the categorized Chinese character set;

[0039] By combining the access words and the number of accesses, we obtain the access word group;

[0040] Summarize the access words to obtain multiple access words;

[0041] Confirm the user usage frequency of the distributed device, summarize the user usage frequency and multiple access word groups to obtain user feature data;

[0042] By aggregating user characteristic data, multiple user characteristic data points are obtained.

[0043] Optionally, the step of performing a genetic algorithm simulation on multiple content to be played and multiple distributed devices based on multiple distributed parsing rates, multiple user feature data, and multiple content priority indices to obtain the optimal carousel includes:

[0044] Randomly sort multiple pieces of content to be played to obtain a sequence of content to be played;

[0045] Confirm the number of distributed devices and the number of playbacks for multiple pieces of content to be played.

[0046] The number of carousels is calculated based on the number of distributions and the number of plays.

[0047] Multiple simulated chromosomes are constructed based on the number of carousels. Each simulated chromosome consists of multiple chromosome segments, and each chromosome segment includes a gene loci, where a is the number of carousels. The number of chromosome segments in the multiple chromosome segments is equal to the number of distributions, and each chromosome segment corresponds one-to-one with a distributed device.

[0048] Perform the following operation on each of the multiple simulated chromosomes:

[0049] Randomly assign gene values ​​to simulated chromosomes to obtain carousel chromosomes;

[0050] By summing the carousel chromosomes, multiple carousel chromosomes are obtained;

[0051] Perform the following operation on each of the multiple rotating chromosomes:

[0052] The carousel matching degree is calculated based on the playback content sequence, carousel chromosome, multiple distribution resolution rates, multiple user feature data, and multiple content priority indices.

[0053] Summarize the carousel matching scores to obtain multiple carousel matching scores. Sort the multiple carousel matching scores in descending order to obtain a carousel matching score sequence.

[0054] Confirm the number of chromosomes in multiple simulated chromosomes;

[0055] Based on the number of chromosomes, the carousel matching degree sequence is extracted by three parts to obtain n priority matching degrees, where n is equal to one-third of m, and m is the number of chromosomes;

[0056] Based on n priority matching degrees and multiple carousel chromosomes, n matching chromosomes were identified;

[0057] Randomly pair n matching chromosomes to obtain J chromosome pairs, where each chromosome pair includes a first chromosome and a second chromosome, and n=2J;

[0058] Gene exchange is performed on J chromosome pairs based on the distribution quantity to obtain n exchanged chromosomes;

[0059] Gene mutations are performed on n matching chromosomes to obtain n mutated chromosomes;

[0060] By summing up n matched chromosomes, n exchanged chromosomes, and n mutated chromosomes, multiple updated chromosomes are obtained, where the updated chromosomes are either matched chromosomes, exchanged chromosomes, or mutated chromosomes.

[0061] Confirm the number of times the following operation is performed on each of the multiple carousel chromosomes. Compare the number of executions with a preset threshold. If the number of executions is less than the threshold, treat the multiple updated chromosomes as multiple carousel chromosomes and return to the step of performing the following operation on each of the multiple carousel chromosomes until the number of executions equals the threshold. Then treat the multiple updated chromosomes as multiple target chromosomes.

[0062] Perform the following operation on each of the multiple target chromosomes:

[0063] The final carousel matching degree is calculated based on the playback content sequence, target chromosome, multiple distribution resolution rates, multiple user feature data, and multiple content priority indices.

[0064] The final carousel matching scores are aggregated to obtain multiple final carousel matching scores. The target chromosome corresponding to the highest final carousel matching score among the multiple final carousel matching scores is taken as the optimal carousel.

[0065] Optionally, the step of randomly assigning gene values ​​to the simulated chromosome to obtain a carousel chromosome includes:

[0066] Construct a playback tag set based on the number of plays;

[0067] Perform the following operation on each of the multiple chromosome segments in the simulated chromosome:

[0068] Gene loci were extracted sequentially from chromosome segments, and the following operations were performed on the extracted gene loci:

[0069] Randomly select playback tags from the playback tag set to obtain selected playback tags. Assign values ​​to gene loci based on the selected playback tags to obtain marker loci.

[0070] Remove selected playback tags from the playback tag set to obtain an updated tag set. Use the updated tag set as the playback tag set and return to the step of sequentially extracting gene loci from the chromosome segment until all a gene loci of the chromosome segment have been extracted.

[0071] By summing up the marker sites, we obtain a marker sites. Based on a marker sites and the chromosome segment, we identify the marker chromosomal segment.

[0072] Summarize the marked segments to obtain multiple marked segments;

[0073] Rotational chromosomes were identified based on multiple marker-stained segments and simulated chromosomes.

[0074] Optionally, the step of calculating the carousel matching degree based on the playback content sequence, carousel chromosome, multiple distribution resolution rates, multiple user feature data, and multiple content priority indices includes:

[0075] Perform the following operation on each marked chromatid segment in the rounded chromosomes:

[0076] The distributed device corresponding to the labeled segment is denoted as the evaluation device, and the distribution resolution rate corresponding to the evaluation device among multiple distribution resolution rates is taken as the evaluation resolution rate. The user feature data corresponding to the evaluation device among multiple user feature data is denoted as user evaluation data.

[0077] For each of the multiple marker sites in the labeled stained segment, the following operation is performed:

[0078] The evaluation playback content is identified in the playback content sequence based on the selected playback tag in the marker site, wherein the sequence number of the evaluation playback content in the playback content sequence is the selected playback tag;

[0079] The content priority index corresponding to the evaluation playback content in multiple content priority indices is recorded as the evaluation priority index.

[0080] Confirm the amount of evaluation data and the title of the evaluation content to be played;

[0081] The evaluation parsing time is obtained by dividing the amount of evaluation data by the evaluation parsing rate.

[0082] The Chinese characters in the evaluation content titles were extracted to obtain multiple evaluation Chinese characters;

[0083] User preference is calculated based on multiple Chinese characters used in the evaluation and user evaluation data.

[0084] The evaluation priority index, evaluation analysis time, and user preference were summarized separately to obtain multiple evaluation priority indices, multiple evaluation analysis times, and multiple user preferences;

[0085] The total priority index is calculated based on multiple evaluation priority indices, the total parsing time is calculated based on multiple evaluation parsing times, and the total preference score is calculated based on multiple user preferences. The total priority index is the sum of multiple evaluation priority indices, the total parsing time is the sum of multiple evaluation parsing times, and the total preference score is the sum of multiple user preferences.

[0086] The carousel distribution matching degree is calculated based on user usage frequency, total priority index, total parsing time, and total preference in user evaluation data. The calculation formula is as follows:

[0087]

[0088] in, For the matching degree of the carousel distribution, Based on user usage frequency, As the overall priority index, For overall preference, Total parsing time, It is a natural constant;

[0089] Summarize the carousel distribution matching scores to obtain multiple carousel distribution matching scores. Calculate the carousel matching score based on these multiple carousel distribution matching scores, where the carousel matching score is the sum of the multiple carousel distribution matching scores.

[0090] Optionally, the calculation of user preference based on multiple evaluation characters and user evaluation data includes:

[0091] Perform the following operation on each of the multiple evaluation characters:

[0092] Determine if there is an access character group among multiple access character groups in the user evaluation data that has the same access character as the evaluation Chinese character;

[0093] If there is a group of access characters in multiple access character groups that has the same access character as the Chinese character being evaluated, then the number of accesses in that access character group is recorded as the number of overlaps.

[0094] Otherwise, the preset zero value will be used as the number of overlaps;

[0095] The number of overlaps is summarized to obtain multiple overlap counts, and the user preference score is calculated based on the multiple overlap counts.

[0096] Optionally, the step of setting up content carousel for multiple distributed devices based on the optimal carousel body to obtain multiple target carousel devices includes:

[0097] For each labeled segment in the optimal carousel, perform the following operation:

[0098] The distributed devices corresponding to the marked segments in multiple distributed devices are used as the initial carousel devices;

[0099] Obtain the initial playlist of the initial carousel device, where the initial playlist is an empty set;

[0100] Extract the k-th marker site from the stained segment, where the initial value of k is 1;

[0101] Based on the selection of the playback tag at the k-th marker site, the target playback content is identified in the playback content sequence;

[0102] The target playback content is placed at the kth position in the initial playlist to obtain the updated playlist;

[0103] Let L = k + 1, take L as k, take the updated playlist as the target playlist, return to the step of extracting the kth marker site from the marked staining segment, until all a marker sites in the marked staining segment have been extracted, and take the updated playlist as the carousel playlist.

[0104] The target carousel device is identified based on the carousel playlist and the initial carousel device.

[0105] The target carousel devices are aggregated to obtain multiple target carousel devices.

[0106] To achieve the above objectives, the present invention also provides a dynamic content carousel system based on the HarmonyOS distributed architecture, comprising:

[0107] The distributed architecture confirmation module is used to receive content carousel instructions, perform HarmonyOS distributed soft bus queries based on the content carousel instructions, and obtain multiple distributed devices, including distributed CPUs, distributed GPUs, and historical databases.

[0108] The carousel content analysis module is used to acquire multiple pieces of content to be played, perform regional priority evaluation on these pieces of content to obtain multiple content priority indices, where each content priority index corresponds one-to-one with the content to be played. It also performs network performance analysis on multiple distributed devices to obtain multiple distributed resolution rates, and performs user characteristic analysis on these distributed devices to obtain multiple user characteristic data. Both the distributed resolution rates and user characteristic data correspond one-to-one with the distributed devices. The user characteristic data includes: user usage frequency and multiple access word groups, where each access word group includes: access word and access count.

[0109] The genetic algorithm simulation module is used to perform genetic algorithm simulation on multiple content to be played and multiple distributed devices based on multiple distributed parsing rates, multiple user feature data and multiple content priority indices to obtain the optimal carousel;

[0110] The optimal carousel setting module is used to set up content carousel for multiple distributed devices based on the optimal carousel body, obtain multiple target carousel devices, and complete dynamic content carousel based on the HarmonyOS distributed architecture based on multiple target carousel devices.

[0111] To address the above problems, the present invention also provides an electronic device, the electronic device comprising:

[0112] Memory, storing at least one instruction; and

[0113] The processor executes the instructions stored in the memory to implement the dynamic content carousel method based on the HarmonyOS distributed architecture described above.

[0114] To address the aforementioned issues, the present invention also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor in an electronic device to implement the aforementioned dynamic content carousel method based on the HarmonyOS distributed architecture.

[0115] To address the problems described in the background section, this invention receives content carousel commands and performs a HarmonyOS distributed soft bus query based on these commands to obtain multiple distributed devices. These distributed devices include distributed CPUs, distributed GPUs, and a historical database. This embodiment of the invention utilizes the HarmonyOS distributed soft bus to query devices, pre-identifying multiple distributed devices within the HarmonyOS distributed architecture. This provides the device foundation for subsequent content carousel setup. Multiple content items to be played are then acquired, and a regional priority evaluation is performed on these items to obtain multiple content priority indices. Each content priority index corresponds one-to-one with a specific content item. This embodiment of the invention provides foundational data support for subsequent content evaluation and allocation by acquiring multiple content items to be played. The regional priority evaluation allows the system to automatically identify content suitability in different regions, prioritizing the push of more regionally relevant content during carousel playback, thus improving the matching degree and user acceptance of the carousel content. Network performance analysis is performed on the multiple distributed devices to obtain multiple distributed parsing rates. User feature analysis is also performed on the multiple distributed devices to obtain multiple user feature data. Both the distributed parsing rates and user feature data correspond one-to-one with the distributed devices, and the user feature data... The data includes: user usage frequency and multiple access word groups, wherein the access word group includes: access word and access count. It is evident that this embodiment of the invention, through dual analysis of the network performance of distributed devices and user characteristics, can comprehensively grasp the real-time parsing capabilities of the device and user behavior characteristics, providing an analytical basis for subsequent genetic algorithm simulation, thereby improving the targeting and stability of carousel content push. Based on multiple distributed parsing rates, multiple user characteristic data, and multiple content priority indices, a genetic algorithm simulation is performed on multiple content to be played and multiple distributed devices to obtain the optimal carousel. It is evident that this embodiment of the invention, through genetic algorithm analysis of device performance, user characteristics, and user behavior characteristics, comprehensively grasps the real-time parsing capabilities of the device and user behavior characteristics, providing an analytical basis for subsequent genetic algorithm simulation, thereby improving the targeting and stability of carousel content push. By comprehensively simulating and iteratively optimizing features and content priorities, the optimal carousel scheme can be quickly converged, achieving intelligent, adaptive, and efficient content allocation across multiple terminals. This significantly improves the distribution accuracy of carousel content. Based on the optimal carousel body, content carousel settings are applied to multiple distributed devices to obtain multiple target carousel devices. Dynamic content carousel based on the HarmonyOS distributed architecture is then completed using these multiple target carousel devices. It is evident that this invention, by executing optimal content playback on multiple target carousel devices, achieves intelligent allocation of carousel content across multiple HarmonyOS devices, effectively improving content delivery performance and user experience, ultimately realizing efficient content delivery through multi-device collaboration. Therefore, this invention can improve the delivery performance and user experience of carousel content across multiple distributed devices, achieving intelligent allocation of carousel content across multiple HarmonyOS devices. Attached Figure Description

[0116] Figure 1 This is a flowchart illustrating a dynamic content carousel method based on the HarmonyOS distributed architecture provided in an embodiment of the present invention.

[0117] Figure 2 A functional block diagram of a dynamic content carousel system based on the HarmonyOS distributed architecture provided in an embodiment of the present invention;

[0118] Figure 3 This is a schematic diagram of the structure of an electronic device that implements the dynamic content carousel method based on the HarmonyOS distributed architecture, according to an embodiment of the present invention.

[0119] Explanation of reference numerals in the attached figures:

[0120] 10. Electronic device; 11. Processor; 12. Memory; 13. Bus.

[0121] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0122] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0123] This application provides a dynamic content carousel method based on the HarmonyOS distributed architecture. The executing entity of the dynamic content carousel method based on the HarmonyOS distributed architecture includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application embodiment: a server, a terminal, etc. In other words, the dynamic content carousel method based on the HarmonyOS distributed architecture can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0124] Reference Figure 1 The diagram shown is a flowchart illustrating a dynamic content carousel method based on the HarmonyOS distributed architecture according to an embodiment of the present invention. In this embodiment, the dynamic content carousel method based on the HarmonyOS distributed architecture includes:

[0125] S1. Receive the content carousel command, and perform HarmonyOS distributed soft bus query based on the content carousel command to obtain multiple distributed devices, including distributed CPUs, distributed GPUs, and historical databases.

[0126] It should be explained that the content carousel command is initiated by the HarmonyOS system's backend administrator. For example, Zhang, as the HarmonyOS system's backend administrator, is responsible for pushing or delivering various information content to system users through the HarmonyOS system backend. This information content includes news, weather forecasts, commercial advertisements, traffic alerts, and real-time event announcements. The aforementioned news, weather forecasts, commercial advertisements, traffic alerts, and real-time event announcements are all content to be played in subsequent embodiments. When it is necessary to push information to users and display the content to be played in a carousel format on multiple terminal devices with the HarmonyOS system built into the user's device, Zhang initiates the content carousel command.

[0127] It is understood that the querying of HarmonyOS distributed soft bus based on content carousel instructions to obtain multiple distributed devices refers to: querying all online devices connected via the HarmonyOS distributed soft bus within the user's local area network, based on the HarmonyOS distributed soft bus. These online devices connected via the HarmonyOS distributed soft bus are the distributed devices. For example, the distributed devices may include: smartphones, smart tablets, smart advertising screens, smart TVs, and smart display terminals all running HarmonyOS. The HarmonyOS distributed soft bus refers to the distributed soft bus technology of the HarmonyOS system, and this technology is existing technology and will not be elaborated upon here.

[0128] It should be explained that distributed CPU refers to the central processing unit of a distributed device, and distributed GPU refers to the graphics processing unit of a distributed device. The historical database is a database that stores the titles of carousel content viewed by a user in the past. For example, Xiao Wang is a user of a smartphone with the HarmonyOS operating system installed. In the past, Xiao Wang has clicked on and viewed titles of carousel content such as "New mobile phone coming soon" and "Metro Line 3 in this city opens for operation." These titles are the browsing titles in the following embodiments, and multiple historical browsing titles are stored in the historical database.

[0129] S2. Get multiple content to be played.

[0130] It should be explained that the multiple pieces of content to be played are news articles, weather forecasts, commercial advertisements, traffic tips, and real-time event announcements that have been manually selected by the backend administrator and need to be pushed to users. The specific content is provided by third-party users, media outlets, or advertisers.

[0131] S3. Perform a regional priority evaluation on multiple content to be played to obtain multiple content priority indices, where each content priority index corresponds one-to-one with the content to be played.

[0132] In detail, the method of performing a region-priority evaluation on multiple pieces of content to be played yields multiple content priority indices, including:

[0133] For each of the multiple pieces of content to be played, perform the following operation:

[0134] Confirm the origin coordinates, total number of clicks, and daily click growth rate of the content to be played. The origin coordinates include the origin longitude and origin latitude.

[0135] Obtain the coordinates of the user equipment, which include the equipment's longitude and latitude;

[0136] The content priority index is calculated based on user device coordinates, origin location coordinates, total clicks, and daily click growth rate. The calculation formula is shown below:

[0137]

[0138] in, As a content priority index, Total number of clicks This represents the daily click growth rate. and These are the origin longitude and origin latitude coordinates of the origin location, respectively. and These are the device longitude and device latitude coordinates of the user equipment, respectively. It is the natural logarithm;

[0139] By summing the content priority indices, multiple content priority indices are obtained.

[0140] It should be explained that the origin location coordinates refer to the latitude and longitude coordinates of the location where the content to be played is published by a third-party user, media, or advertiser. Origin longitude refers to the geographical longitude of the location where the content to be played is published, and origin latitude refers to the geographical latitude of the location where the content to be played is published. User device coordinates refer to the latitude and longitude coordinates of the location of the distributed device used by the user. Device longitude refers to the geographical longitude of the distributed device, and device latitude refers to the geographical latitude of the distributed device. Total clicks refer to the total number of times the content to be played has been viewed online, and the daily click growth rate refers to the number of times the content to be played has been viewed online today.

[0141] Understandably, the content priority index reflects the priority of the content to be played relative to the user on the distributed device. The higher the content priority index, the higher the priority of the content to be played relative to the user.

[0142] It should be understood that, by combining the coordinates of the user's device and the coordinates of the originating location, the embodiments of the present invention give higher push priority to the content to be played that is closer to the user's geographical location, so that the user can receive more valuable information from the vicinity during the carousel.

[0143] S4. Perform network performance analysis on multiple distributed devices to obtain multiple distributed resolution rates. Perform user characteristic analysis on multiple distributed devices to obtain multiple user characteristic data. The distributed resolution rates and user characteristic data correspond one-to-one with the distributed devices. The user characteristic data includes: user usage frequency and multiple access word groups. The access word groups include: access words and access counts.

[0144] In detail, the network performance analysis of multiple distributed devices to obtain multiple distributed resolution rates includes:

[0145] Perform the following operation on each of the multiple distributed devices:

[0146] Read the CPU load of distributed CPUs in the distributed device, and read the GPU stream count, GPU memory capacity and GPU memory bandwidth of distributed GPUs in the distributed device.

[0147] Network status monitoring of distributed devices yields bandwidth utilization, packet loss rate, and network latency.

[0148] The distributed parsing rate is calculated based on CPU load, GPU stream count, GPU memory capacity, GPU memory bandwidth, bandwidth utilization, packet loss rate, and network latency. The calculation formula is as follows:

[0149]

[0150] in, For distributed analytical rate, , and These are GPU stream count, GPU memory capacity, and GPU memory bandwidth, respectively. CPU load, For bandwidth utilization, For packet loss rate, Due to network latency, It is the hyperbolic tangent function;

[0151] By summing the distributed resolution rates, multiple distributed resolution rates are obtained.

[0152] It should be explained that CPU load refers to the utilization rate of the central processing unit in the distributed device. GPU stream count, GPU memory capacity, and GPU memory bandwidth refer to the number of stream processors in the graphics processing unit (GPU) of the distributed device, respectively. GPU memory capacity refers to the memory capacity of the GPU in the distributed device. GPU memory bandwidth refers to the memory bandwidth of the GPU in the distributed device. Bandwidth utilization rate refers to the proportion of bandwidth actually used during network transmission to the total available bandwidth. Packet loss rate refers to the proportion of data packets that fail to reach their destination address during network transmission during the period of network status monitoring. Network latency refers to the average time taken for multiple data packets to travel from the source address to the destination address during the period of network status monitoring. The technology for performing network status monitoring on distributed devices to obtain bandwidth utilization rate, packet loss rate, and network latency is existing technology and will not be elaborated here.

[0153] It should be explained that the distributed resolution rate reflects the processing speed of distributed devices from the moment they start receiving content to be played from the network to the moment they complete the parsing and display the content on the page. The higher the distributed resolution rate, the faster the distributed devices can receive content to be played from the network and complete the parsing and display the content on the page.

[0154] In detail, the user feature analysis of multiple distributed devices yields multiple user feature data, including:

[0155] Perform the following operation on each of the multiple distributed devices:

[0156] Read multiple browsing titles from the historical database in the distributed device;

[0157] Perform the following operation for each of the multiple browse titles:

[0158] Extract Chinese characters from the browsing titles to obtain multiple title characters;

[0159] Based on a preset set of commonly used characters, common characters are removed from multiple titles to obtain a reference set of characters;

[0160] By compiling a collection of reference Chinese character sets, multiple reference Chinese character sets were obtained.

[0161] By merging multiple reference character sets, a total historical character set is obtained;

[0162] The total historical set of Chinese characters was categorized, resulting in multiple categorized character sets.

[0163] Perform the following operation on each of the multiple categorized Chinese character sets:

[0164] Confirm the accessed characters and their access counts in the categorized Chinese character set;

[0165] Combine the access words and the number of access times to obtain an access word group;

[0166] Summarize the access word groups to obtain multiple access word groups;

[0167] Confirm the usage frequency of users of the distributed device, summarize the user usage frequency and multiple access word groups to obtain user feature data;

[0168] Summarize the user feature data to obtain multiple user feature data.

[0169] Exemplarily, the titles of the carousel content that the user has browsed or clicked on in a certain distributed device are: "New mobile phones are about to be launched", "The third subway line in this city has been opened for transportation", "The local intangible cultural heritage festival is held in the historical ancient street". Then the above "New mobile phones are about to be launched", "The third subway line in this city has been opened for transportation", and "The local intangible cultural heritage festival is held in the historical ancient street" are three browsing titles. Extract Chinese characters from the browsing title "New mobile phones are about to be launched" to obtain 8 title Chinese characters: new, model, mobile phone,..., up, market.

[0170] It should be explained that the common character set includes: multiple common characters. A common character refers to: numerals in Chinese characters (such as one, two), measure words (such as: piece, catty), prepositions (such as: at, from), conjunctions (such as: and, with) and pronouns (such as: I, you), and the specific common characters included in the common character set are artificially set by the background administrator. The elimination of common characters from multiple title Chinese characters based on the preset common character set means: eliminating the title Chinese characters that are the same as the common characters in the common character set from the multiple title Chinese characters, so that there are no Chinese characters without specific meanings in the multiple title Chinese characters, and the accuracy of subsequent evaluation of user browsing preferences is improved.

[0171] It can be understood that the merging of multiple reference Chinese character sets to obtain the total historical Chinese character set means: extracting all the reference Chinese characters in the multiple reference Chinese character sets and summarizing them into a set to obtain the total historical Chinese character set. The classification of the total historical Chinese character set to obtain multiple classified Chinese character sets means: classifying the same Chinese characters in the total historical Chinese character set into one category. Exemplarily, the total historical Chinese character set is {medical, insurance, benefit, medical, insurance, reform, hospital, reform, construction}, then the multiple classified Chinese character sets obtained by classification are {medical, medical, medical}, {insurance, insurance}, {reform, reform}, {benefit}, {reform}, {hospital}, {construction}. The confirmation of the access word and the number of access times of the classified Chinese character set means: taking the Chinese characters after classification in the classified Chinese character set as the access word, and taking the number of times the Chinese character appears in the classified Chinese character set as the number of access times. For example: the access word of {medical, medical, medical} is "medical", and the number of access times is 3, and the access word group is (medical, 3).

[0172] It should be explained that user usage frequency refers to the average number of times a user clicks on the carousel content on a distributed device per day over a period of time.

[0173] S5. Based on multiple distributed resolution rates, multiple user feature data, and multiple content priority indices, a genetic algorithm is used to simulate multiple content to be played and multiple distributed devices to obtain the optimal carousel.

[0174] In detail, the step of performing a genetic algorithm simulation on multiple content to be played and multiple distributed devices based on multiple distributed parsing rates, multiple user feature data, and multiple content priority indices to obtain the optimal carousel includes:

[0175] The number of carousels is determined based on multiple content to be played and multiple distributed devices;

[0176] Multiple simulated chromosomes are constructed based on the number of carousels. Each simulated chromosome consists of multiple chromosome segments, and each chromosome segment includes a gene loci, where a is the number of carousels. The number of chromosome segments in the multiple chromosome segments is equal to the number of distributions, and each chromosome segment corresponds one-to-one with a distributed device.

[0177] Perform the following operation on each of the multiple simulated chromosomes:

[0178] Randomly assign gene values ​​to simulated chromosomes to obtain carousel chromosomes;

[0179] By summing the carousel chromosomes, multiple carousel chromosomes are obtained;

[0180] Perform the following operation on each of the multiple rotating chromosomes:

[0181] The carousel matching degree is calculated based on the playback content sequence, carousel chromosome, multiple distribution resolution rates, multiple user feature data, and multiple content priority indices.

[0182] Summarize the carousel matching scores to obtain multiple carousel matching scores. Sort the multiple carousel matching scores in descending order to obtain a carousel matching score sequence.

[0183] Confirm the number of chromosomes in multiple simulated chromosomes;

[0184] Multiple updated chromosomes were identified based on multiple carousel chromosomes, carousel matching sequences, and chromosome numbers;

[0185] Confirm the number of times the following operation is performed on each of the multiple carousel chromosomes. Compare the number of executions with a preset threshold. If the number of executions is less than the threshold, treat the multiple updated chromosomes as multiple carousel chromosomes and return to the step of performing the following operation on each of the multiple carousel chromosomes until the number of executions equals the threshold. Then treat the multiple updated chromosomes as multiple target chromosomes.

[0186] Perform the following operation on each of the multiple target chromosomes:

[0187] The final carousel matching degree is calculated based on the playback content sequence, target chromosome, multiple distribution resolution rates, multiple user feature data, and multiple content priority indices.

[0188] The final carousel matching scores are aggregated to obtain multiple final carousel matching scores. The target chromosome corresponding to the highest final carousel matching score among the multiple final carousel matching scores is taken as the optimal carousel.

[0189] Specifically, determining the number of carousels based on multiple pieces of content to be played and multiple distributed devices includes:

[0190] Randomly sort multiple pieces of content to be played to obtain a sequence of content to be played;

[0191] Confirm the number of distributed devices and the number of playbacks for multiple pieces of content to be played.

[0192] The number of carousels is calculated based on the number of distributions and the number of plays, using the following formula:

[0193]

[0194] in, For the number of carousels, and These represent the number of distributions and the number of plays, respectively. This refers to rounding down to the nearest integer.

[0195] Specifically, the identification of multiple updated chromosomes based on multiple carousel chromosomes, carousel matching sequences, and chromosome numbers includes:

[0196] Based on the number of chromosomes, the carousel matching degree sequence is extracted by three parts to obtain n priority matching degrees, where n is equal to one-third of m, and m is the number of chromosomes;

[0197] Based on n priority matching degrees and multiple carousel chromosomes, n matching chromosomes were identified;

[0198] Randomly pair n matching chromosomes to obtain J chromosome pairs, where each chromosome pair includes a first chromosome and a second chromosome, and n=2J;

[0199] Gene exchange is performed on J chromosome pairs based on the distribution quantity to obtain n exchanged chromosomes;

[0200] Gene mutations are performed on n matching chromosomes to obtain n mutated chromosomes;

[0201] By summing up n matched chromosomes, n exchanged chromosomes, and n mutated chromosomes, multiple updated chromosomes are obtained, where the updated chromosomes are either matched chromosomes, exchanged chromosomes, or mutated chromosomes.

[0202] It should be explained that the number of distributed devices refers to the number of distributed devices among multiple distributed devices, while the number of playbacks refers to the number of content to be played among multiple content to be played.

[0203] It should be understood that the simulated chromosome is a virtual chromosome in a genetic algorithm, and it consists of multiple chromosome segments. Each chromosome segment includes *a* gene loci, and all gene loci initially have a value of 0. Subsequently, random gene values ​​will be assigned to the simulated chromosome through the playback sequence, causing changes in the values ​​at the gene loci. It is important to note that the position of each gene locus within the simulated chromosome is immutable; only its value can be changed. The number of chromosomes refers to the total number of simulated chromosomes in the multiple simulated chromosomes.

[0204] For example, since each simulated chromosome corresponds to one carousel match degree, if the number of chromosomes is 90, there are 90 carousel match degrees in the carousel match degree sequence. The three-part extraction based on the number of chromosomes means extracting the top one-third of the carousel match degrees in the sequence. Therefore, the top 30 carousel match degrees in the sequence are taken as 30 priority matches, and the 30 carousel chromosomes corresponding to these 30 priority matches are taken as 30 matching chromosomes. Then, two matching chromosomes are randomly extracted from these 30 matching chromosomes and paired to form a chromosome pair. This process is repeated until all chromosomes are extracted, resulting in 15 chromosome pairs. Any matching chromosome in a chromosome pair is the first chromosome, and the other matching chromosome in the pair is the second chromosome.

[0205] For example, when multiple updated chromosomes are obtained for the first time, i.e., the number of times the following operation is performed on each of the multiple carousel chromosomes is 1, if the number of times is less than the threshold of 80, then the multiple updated chromosomes are treated as multiple carousel chromosomes, and the step of performing the following operation on each of the multiple carousel chromosomes is returned, and multiple updated chromosomes are obtained again, this time the number of times is 2, and so on, until the number of times is reached 80. The multiple updated chromosomes in the last execution process are treated as multiple target chromosomes. The number of times threshold is set manually by the backend administrator. Preferably, the number of times threshold is 80, so as to obtain a better carousel content matching result while ensuring that the algorithm iteration time is reasonably controllable.

[0206] It should be understood that the method for calculating the final carousel matching degree based on the playback content sequence, target chromosome, multiple distribution resolution rates, multiple user feature data, and multiple content priority indices is the same as the method for calculating the carousel matching degree based on the playback content sequence, carousel chromosome, multiple distribution resolution rates, multiple user feature data, and multiple content priority indices, and will not be repeated here.

[0207] Specifically, the random gene assignment to the simulated chromosome to obtain a rotating chromosome includes:

[0208] Construct a playback tag set based on the number of plays;

[0209] Perform the following operation on each of the multiple chromosome segments in the simulated chromosome:

[0210] Gene loci were extracted sequentially from chromosome segments, and the following operations were performed on the extracted gene loci:

[0211] Randomly select playback tags from the playback tag set to obtain selected playback tags. Assign values ​​to gene loci based on the selected playback tags to obtain marker loci.

[0212] Remove selected playback tags from the playback tag set to obtain an updated tag set. Use the updated tag set as the playback tag set and return to the step of sequentially extracting gene loci from the chromosome segment until all a gene loci of the chromosome segment have been extracted.

[0213] By summing up the marker sites, we obtain a marker sites. Based on a marker sites and the chromosome segment, we identify the marker chromosomal segment.

[0214] Summarize the marked segments to obtain multiple marked segments;

[0215] Rotational chromosomes were identified based on multiple marker-stained segments and simulated chromosomes.

[0216] For example, if the number of plays is 10, then the play tag set is [1,2,3,…,10].

[0217] It should be explained that, "randomly selecting a playback tag from the playback tag set" means randomly selecting a value from the playback tag set as the playback tag. "Assigning a value to the gene locus based on the selected playback tag" means assigning the value within the gene locus to the value of the selected playback tag. "Identifying the marked chromosomal segment based on 'a' marker loci and the chromosome segment" means that when all 'a' gene loci in the chromosome segment are confirmed to have transformed into 'a' marker loci, the chromosome segment at this point is the marked chromosomal segment. "Identifying the revolving chromosome based on multiple marked chromosomal segments and the simulated chromosome" means that when multiple chromosome segments in the simulated chromosome are confirmed to have transformed into multiple marked chromosomal segments, the simulated chromosome at this point is the revolving chromosome.

[0218] In detail, the calculation of carousel matching degree based on the playback content sequence, carousel chromosome, multiple distribution resolution rates, multiple user feature data, and multiple content priority indices includes:

[0219] Perform the following operation on each marked chromatid segment in the rounded chromosomes:

[0220] The distributed device corresponding to the labeled segment is denoted as the evaluation device, and the distribution resolution rate corresponding to the evaluation device among multiple distribution resolution rates is taken as the evaluation resolution rate. The user feature data corresponding to the evaluation device among multiple user feature data is denoted as user evaluation data.

[0221] Based on the marked segments, the sequence of playback content, multiple content priority indices, and the evaluation parsing rate, the total priority index, total parsing time, and total preference were determined.

[0222] The carousel distribution matching degree is calculated based on user usage frequency, total priority index, total parsing time, and total preference in user evaluation data. The calculation formula is as follows:

[0223]

[0224] in, For the matching degree of the carousel distribution, Based on user usage frequency, As the overall priority index, For overall preference, Total parsing time, It is a natural constant;

[0225] Summarize the carousel distribution matching scores to obtain multiple carousel distribution matching scores. Calculate the carousel matching score based on these multiple carousel distribution matching scores, where the carousel matching score is the sum of the multiple carousel distribution matching scores.

[0226] In detail, the determination of the total priority index, total parsing time, and total preference based on the marked segments, the playback content sequence, multiple content priority indices, and the evaluation parsing rate includes:

[0227] For each of the multiple marker sites in the labeled stained segment, the following operation is performed:

[0228] The evaluation playback content is identified in the playback content sequence based on the selected playback tag in the marker site, wherein the sequence number of the evaluation playback content in the playback content sequence is the selected playback tag;

[0229] The content priority index corresponding to the evaluation playback content in multiple content priority indices is recorded as the evaluation priority index.

[0230] Confirm the amount of evaluation data and the title of the evaluation content to be played;

[0231] The evaluation parsing time is obtained by dividing the amount of evaluation data by the evaluation parsing rate.

[0232] The Chinese characters in the evaluation content titles were extracted to obtain multiple evaluation Chinese characters;

[0233] User preference is calculated based on multiple Chinese characters used in the evaluation and user evaluation data.

[0234] The evaluation priority index, evaluation analysis time, and user preference were summarized separately to obtain multiple evaluation priority indices, multiple evaluation analysis times, and multiple user preferences;

[0235] The total priority index is calculated based on multiple evaluation priority indices, the total parsing time is calculated based on multiple evaluation parsing times, and the total preference score is calculated based on multiple user preferences. The total priority index is the sum of multiple evaluation priority indices, the total parsing time is the sum of multiple evaluation parsing times, and the total preference score is the sum of multiple user preferences.

[0236] It should be explained that the evaluation data volume refers to the size of the data packet of the evaluation playback content, measured in bytes. The evaluation content title refers to the title of the evaluation playback content. The evaluation parsing time reflects the total time consumed when the evaluation playback content is transmitted to the evaluation device, parsed on the evaluation device, and displayed on the carousel interface. The method for extracting Chinese characters from the evaluation content title to obtain multiple evaluation Chinese characters is the same as the method for extracting Chinese characters from the browsing title to obtain multiple title Chinese characters, and will not be repeated here.

[0237] It is understood that the carousel distribution matching degree is a quantity used to characterize the playback effect when multiple evaluation playback contents corresponding to the distribution matching degree are distributed to the evaluation device for carousel playback. That is, the higher the carousel distribution matching degree, the faster the multiple evaluation contents are parsed and displayed on the page when they are carouseled on the evaluation device, and the easier they are to be noticed by users or meet users' preferences.

[0238] It should be understood that the embodiments of the present invention use a genetic algorithm to efficiently simulate the optimal carousel allocation scheme within a finite time without exhausting all possibilities. This scheme rationally allocates multiple content to be played to different distributed devices for carousel playback, thereby improving the overall compatibility between the carousel content and the distributed devices and thus optimizing the overall carousel effect.

[0239] In detail, the calculation of user preference based on multiple evaluation Chinese characters and user evaluation data includes:

[0240] Perform the following operation on each of the multiple evaluation characters:

[0241] Determine if there is an access character group among multiple access character groups in the user evaluation data that has the same access character as the evaluation Chinese character;

[0242] If there is a group of access characters in multiple access character groups that has the same access character as the Chinese character being evaluated, then the number of accesses in that access character group is recorded as the number of overlaps.

[0243] Otherwise, the preset zero value will be used as the number of overlaps;

[0244] The number of overlaps is summarized to obtain multiple overlap counts. User preference is then calculated based on these multiple overlap counts, using the following formula:

[0245]

[0246] in, For user preferences, The number of overlaps in multiple times Number of overlaps, This represents the number of overlaps out of multiple overlaps. This is the preset total number of reference attempts.

[0247] It should be noted that the value of zero is 0. The total reference count is a value manually set by the backend administrator; preferably, the total reference count is 10,000.

[0248] Specifically, the gene exchange of J chromosome pairs based on the distribution quantity to obtain n exchanged chromosomes includes:

[0249] Perform the following operation on each of the J chromosome pairs:

[0250] Construct a distribution label set based on the distribution quantity;

[0251] Exchange tags are obtained based on the distributed tag set, and the first target exchange segment is identified from the first chromosome of the chromosome pair based on the exchange tags.

[0252] The second target exchange segment was identified from the second chromosome of the chromosome pair based on the exchange tag;

[0253] The first target exchange segment of the first chromosome is replaced by the second target exchange segment to obtain the first updated chromosome; the second target exchange segment of the second chromosome is replaced by the first target exchange segment to obtain the second updated chromosome.

[0254] By combining the first and second updated chromosomes, we obtain the exchanged chromosome pairs;

[0255] By summing the exchanged chromosome pairs, J exchanged chromosome pairs are obtained;

[0256] Based on J pairs of exchanged chromosomes, n exchanged chromosomes were identified.

[0257] It is understood that the method of constructing a distribution tag set based on the number of distributions is the same as the method of constructing a playback tag set based on the number of playbacks, and the method of obtaining exchange tags based on the distribution tag set is the same as the method of randomly selecting playback tags from the playback tag set to obtain selected playback tags. These will not be described in detail here.

[0258] It should be understood that the first target exchange segment is the b-th marked chromatid segment on the first chromosome, where b is the value in the selected segment label. The method for identifying the second target exchange segment from the second chromosome of the chromosome pair based on the exchange label is the same as the method for identifying the first target exchange segment from the first chromosome of the chromosome pair based on the exchange label, and will not be described again here.

[0259] For example, the first exchange segment on the first chromosome is replaced by the second target exchange segment, and the replaced first chromosome is used as the first updated chromosome. Similarly, the second exchange segment on the second chromosome is replaced by the first target exchange segment, and the replaced second chromosome is used as the second updated chromosome. The step of identifying n exchange chromosomes based on J exchange chromosome pairs means: sequentially extracting the first and second updated chromosomes from each of the J exchange chromosome pairs and summing them all to obtain n exchange chromosomes, where each exchange chromosome is either the first or second updated chromosome.

[0260] Specifically, the process of performing gene mutations on n matching chromosomes to obtain n mutated chromosomes includes:

[0261] Extract matching chromosomes sequentially from n matching chromosomes, and perform the following operations on the extracted matching chromosomes:

[0262] Perform the following operation on each marked chromatid segment in the matched chromosomes:

[0263] Double random selection is performed on a marker sites on the stained segment to obtain the first site and the second site;

[0264] The first and second sites on the labeled stained segment are swapped to obtain the mutant stained segment;

[0265] By summarizing the mutated chromatogram segments, multiple mutated chromatogram segments were obtained. Based on the multiple mutated chromatogram segments and the matching chromosome, the mutated chromosome was identified.

[0266] By summing the mutated chromosomes, we obtain n mutated chromosomes.

[0267] It is understood that exchanging the first and second sites on the marked chromatid segment means swapping the value of the first site with the value of the second site on the marked chromatid segment. The method for identifying mutated chromosomes based on multiple mutated chromatid segments and matching chromosomes is the same as the method for identifying rotating chromosomes based on multiple marked chromatid segments and simulated chromosomes, and will not be described again here.

[0268] It should be explained that the double random selection of a marker sites on the stained segment to obtain the first site and the second site means that two marker sites are randomly selected from the a marker sites, and either one is used as the first site and the other is used as the second site.

[0269] S6. Based on the optimal carousel body, perform content carousel settings on multiple distributed devices to obtain multiple target carousel devices.

[0270] In detail, the step of setting up content carousel for multiple distributed devices based on the optimal carousel body to obtain multiple target carousel devices includes:

[0271] For each labeled segment in the optimal carousel, perform the following operation:

[0272] The distributed devices corresponding to the marked segments in multiple distributed devices are used as the initial carousel devices;

[0273] Obtain the initial playlist of the initial carousel device, where the initial playlist is an empty set;

[0274] Extract the k-th marker site from the stained segment, where the initial value of k is 1;

[0275] Based on the selection of the playback tag at the k-th marker site, the target playback content is identified in the playback content sequence;

[0276] The target playback content is placed at the kth position in the initial playlist to obtain the updated playlist;

[0277] Let L = k + 1, take L as k, take the updated playlist as the target playlist, return to the step of extracting the kth marker site from the marked staining segment, until all a marker sites in the marked staining segment have been extracted, and take the updated playlist as the carousel playlist.

[0278] The target carousel device is identified based on the carousel playlist and the initial carousel device.

[0279] The target carousel devices are aggregated to obtain multiple target carousel devices.

[0280] It should be explained that the initial playlist refers to the playlist that the initial carousel device will use for subsequent carousels, and the initial state of the playlist is an empty set. In this embodiment of the invention, the content to be played will be inserted into the playlist in the order of the marked sites corresponding to the content to be played in the marked coloring segments.

[0281] It is understood that the method of identifying the target playback content in the playback content sequence based on the selection of playback tags in the k-th marker site is the same as the method of identifying the evaluation playback content in the playback content sequence based on the selection of playback tags in the marker site, and will not be described again here.

[0282] It should be understood that the determination of the target carousel device based on the carousel playlist and the initial carousel device means that when it is confirmed that the initial playlist in the initial carousel device has been converted into a carousel playlist, the initial carousel device at this time is the target carousel device.

[0283] S7. Performs dynamic content carousel based on HarmonyOS distributed architecture using multiple target carousel devices.

[0284] For example, when multiple target carousel devices are obtained, the multiple target carousel devices are made to play the content to be played in sequence according to the order in their corresponding carousel playlists. The specific playback time is set manually by the backend administrator, and the carousel is started after all the content has been played. This completes the dynamic content carousel based on the HarmonyOS distributed architecture, improves the compatibility between the content to be played and the distributed devices and user satisfaction, and realizes the intelligent allocation and adaptive carousel of the content to be played.

[0285] To address the problems described in the background section, this invention receives content carousel commands and performs a HarmonyOS distributed soft bus query based on these commands to obtain multiple distributed devices. These distributed devices include distributed CPUs, distributed GPUs, and a historical database. This embodiment of the invention utilizes the HarmonyOS distributed soft bus to query devices, pre-identifying multiple distributed devices within the HarmonyOS distributed architecture. This provides the device foundation for subsequent content carousel setup. Multiple content items to be played are then acquired, and a regional priority evaluation is performed on these items to obtain multiple content priority indices. Each content priority index corresponds one-to-one with a specific content item. This embodiment of the invention provides foundational data support for subsequent content evaluation and allocation by acquiring multiple content items to be played. The regional priority evaluation allows the system to automatically identify content suitability in different regions, prioritizing the push of more regionally relevant content during carousel playback, thus improving the matching degree and user acceptance of the carousel content. Network performance analysis is performed on the multiple distributed devices to obtain multiple distributed parsing rates. User feature analysis is also performed on the multiple distributed devices to obtain multiple user feature data. Both the distributed parsing rates and user feature data correspond one-to-one with the distributed devices, and the user feature data... The data includes: user usage frequency and multiple access word groups, wherein the access word group includes: access word and access count. It is evident that this embodiment of the invention, through dual analysis of the network performance of distributed devices and user characteristics, can comprehensively grasp the real-time parsing capabilities of the device and user behavior characteristics, providing an analytical basis for subsequent genetic algorithm simulation, thereby improving the targeting and stability of carousel content push. Based on multiple distributed parsing rates, multiple user characteristic data, and multiple content priority indices, a genetic algorithm simulation is performed on multiple content to be played and multiple distributed devices to obtain the optimal carousel. It is evident that this embodiment of the invention, through genetic algorithm analysis of device performance, user characteristics, and user behavior characteristics, comprehensively grasps the real-time parsing capabilities of the device and user behavior characteristics, providing an analytical basis for subsequent genetic algorithm simulation, thereby improving the targeting and stability of carousel content push. By comprehensively simulating and iteratively optimizing features and content priorities, the optimal carousel scheme can be quickly converged, achieving intelligent, adaptive, and efficient content allocation across multiple terminals. This significantly improves the distribution accuracy of carousel content. Based on the optimal carousel body, content carousel settings are applied to multiple distributed devices to obtain multiple target carousel devices. Dynamic content carousel based on the HarmonyOS distributed architecture is then completed using these multiple target carousel devices. It is evident that this invention, by executing optimal content playback on multiple target carousel devices, achieves intelligent allocation of carousel content across multiple HarmonyOS devices, effectively improving content delivery performance and user experience, ultimately realizing efficient content delivery through multi-device collaboration. Therefore, this invention can improve the delivery performance and user experience of carousel content across multiple distributed devices, achieving intelligent allocation of carousel content across multiple HarmonyOS devices.

[0286] like Figure 2 The diagram shown is a functional block diagram of a dynamic content carousel system based on the HarmonyOS distributed architecture provided in an embodiment of the present invention.

[0287] The dynamic content carousel system 100 based on the HarmonyOS distributed architecture described in this invention can be installed in electronic devices. Depending on the functions implemented, the dynamic content carousel system 100 based on the HarmonyOS distributed architecture may include a distributed architecture confirmation module 101, a carousel content analysis module 102, a genetic algorithm simulation module 103, and an optimal carousel setting module 104. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device.

[0288] The distributed architecture confirmation module 101 is used to receive content carousel instructions, perform HarmonyOS distributed soft bus queries based on the content carousel instructions, and obtain multiple distributed devices, including: distributed CPUs, distributed GPUs and historical databases.

[0289] The carousel content analysis module 102 is used to acquire multiple content to be played, perform regional priority evaluation on the multiple content to be played to obtain multiple content priority indices, wherein the content priority indexes correspond one-to-one with the content to be played, perform network performance analysis on multiple distributed devices to obtain multiple distributed resolution rates, and perform user feature analysis on multiple distributed devices to obtain multiple user feature data, wherein the distributed resolution rate and user feature data correspond one-to-one with the distributed devices, and the user feature data includes: user usage frequency and multiple access word groups, wherein the access word group includes: access word and access count;

[0290] The genetic algorithm simulation module 103 is used to perform genetic algorithm simulation on multiple content to be played and multiple distributed devices based on multiple distributed parsing rates, multiple user feature data and multiple content priority indices to obtain the optimal carousel;

[0291] The optimal carousel setting module 104 is used to set the content carousel for multiple distributed devices based on the optimal carousel body, obtain multiple target carousel devices, and complete the dynamic content carousel based on the HarmonyOS distributed architecture based on the multiple target carousel devices.

[0292] In detail, the modules in the dynamic content carousel system 100 based on the HarmonyOS distributed architecture described in this embodiment of the invention adopt the same approach as described above when in use. Figure 1 The method used here is the same as the dynamic content carousel method based on the HarmonyOS distributed architecture, and can produce the same technical effect, so it will not be repeated here.

[0293] like Figure 3 The diagram shown is a structural schematic of an electronic device that implements a dynamic content carousel method based on the HarmonyOS distributed architecture, according to an embodiment of the present invention.

[0294] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and capable of running on the processor 10, such as a dynamic content carousel method program based on the HarmonyOS distributed architecture.

[0295] The memory 11 includes at least one type of readable storage medium, including flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as the portable hard drive of the electronic device 1. In other embodiments, the memory 11 can be an external storage device of the electronic device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device 1. Furthermore, the memory 11 includes both internal storage units and external storage devices of the electronic device 1. The memory 11 can be used not only to store application software and various types of data installed on the electronic device 1, such as the code of a dynamic content carousel method program based on the HarmonyOS distributed architecture, but also to temporarily store data that has been output or will be output.

[0296] In some embodiments, the processor 10 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control unit of the electronic device, connecting various components of the entire electronic device through various interfaces and lines. It executes programs or modules stored in the memory 11 (e.g., dynamic content carousel methods based on the HarmonyOS distributed architecture) and calls data stored in the memory 11 to perform various functions of the electronic device 1 and process data.

[0297] The bus 12 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 can be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to realize the connection and communication between the memory 11 and at least one processor 10, etc.

[0298] Figure 3 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0299] For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) to power the various components. Preferably, the power supply can be logically connected to the at least one processor 10 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0300] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the electronic device 1 and other electronic devices.

[0301] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired or wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device 1 and to display a visual user interface.

[0302] The dynamic content carousel method program based on the HarmonyOS distributed architecture stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When run in the processor 10, it can achieve the following:

[0303] Receive content carousel instructions, perform HarmonyOS distributed soft bus queries based on content carousel instructions to obtain multiple distributed devices, including: distributed CPUs, distributed GPUs and historical databases;

[0304] Get multiple pieces of content to be played;

[0305] Multiple pieces of content to be played are evaluated based on their geographic priority, resulting in multiple content priority indices. Each content priority index corresponds one-to-one with a piece of content to be played.

[0306] Network performance analysis is performed on multiple distributed devices to obtain multiple distributed resolution rates. User feature analysis is also performed on multiple distributed devices to obtain multiple user feature data. The distributed resolution rates and user feature data correspond one-to-one with the distributed devices. The user feature data includes: user usage frequency and multiple access word groups. The access word groups include: access words and access counts.

[0307] Based on multiple distributed parsing rates, multiple user feature data, and multiple content priority indices, a genetic algorithm simulation is performed on multiple content to be played and multiple distributed devices to obtain the optimal carousel.

[0308] Based on the optimal carousel body, content carousel settings are configured for multiple distributed devices to obtain multiple target carousel devices;

[0309] Dynamic content carousel based on HarmonyOS distributed architecture is achieved by using multiple target carousel devices.

[0310] Specifically, the processor 10's implementation method for the above instructions can be found in [reference needed]. Figures 1 to 3 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0311] Furthermore, if the modules / units integrated in the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0312] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following:

[0313] Receive content carousel instructions, perform HarmonyOS distributed soft bus queries based on content carousel instructions to obtain multiple distributed devices, including: distributed CPUs, distributed GPUs and historical databases;

[0314] Get multiple pieces of content to be played;

[0315] Multiple pieces of content to be played are evaluated based on their geographic priority, resulting in multiple content priority indices. Each content priority index corresponds one-to-one with a piece of content to be played.

[0316] Network performance analysis is performed on multiple distributed devices to obtain multiple distributed resolution rates. User feature analysis is also performed on multiple distributed devices to obtain multiple user feature data. The distributed resolution rates and user feature data correspond one-to-one with the distributed devices. The user feature data includes: user usage frequency and multiple access word groups. The access word groups include: access words and access counts.

[0317] Based on multiple distributed parsing rates, multiple user feature data, and multiple content priority indices, a genetic algorithm simulation is performed on multiple content to be played and multiple distributed devices to obtain the optimal carousel.

[0318] Based on the optimal carousel body, content carousel settings are configured for multiple distributed devices to obtain multiple target carousel devices;

[0319] Dynamic content carousel based on HarmonyOS distributed architecture is achieved by using multiple target carousel devices.

[0320] In the embodiments provided by this invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and actual implementations may have other classification methods.

[0321] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0322] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0323] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0324] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A dynamic content carousel method based on a distributed architecture of the Hong Meng, characterized in that, The method comprises: receiving a content carousel instruction, performing a HarmonyOS distributed bus query based on the content carousel instruction, and obtaining a plurality of distributed devices, wherein the distributed devices include a distributed CPU, a distributed GPU, and a historical database; obtaining a plurality of to-be-played contents; performing regional priority evaluation on the plurality of to-be-played contents to obtain a plurality of content priority indexes, wherein the content priority indexes correspond one-to-one to the to-be-played contents; performing network performance analysis on the plurality of distributed devices to obtain a plurality of distributed resolution rates, and performing user feature analysis on the plurality of distributed devices to obtain a plurality of user feature data, wherein the distributed resolution rates and the user feature data correspond one-to-one to the distributed devices, and the user feature data includes user usage frequency and a plurality of access word groups, wherein the access word group includes an access word and an access frequency; performing genetic algorithm simulation on the plurality of to-be-played contents and the plurality of distributed devices based on the plurality of distributed resolution rates, the plurality of user feature data, and the plurality of content priority indexes to obtain an optimal carousel body; based on the optimal carousel body, performing content carousel setting on the plurality of distributed devices to obtain a plurality of target carousel devices, wherein the content carousel setting based on the optimal carousel body on the plurality of distributed devices to obtain the plurality of target carousel devices comprises: performing the following operations on each marker dye segment in the optimal carousel body: taking the distributed device corresponding to the marker dye segment in the plurality of distributed devices as an initial carousel device; obtaining an initial play list of the initial carousel device, wherein the initial play list is an empty set; extracting a kth marker site from the marker dye segment, wherein the initial value of k is 1; confirming a target play content from a play content sequence based on a selected play tag of the kth marker site; placing the target play content at the kth position in the initial play list to obtain an updated play list; setting L=k+1 as k and the updated play list as a target play list, returning to the step of extracting the kth marker site from the marker dye segment until the a marker sites in the marker dye segment are all extracted, and taking the updated play list as a carousel play list; confirming a target carousel device based on the carousel play list and the initial carousel device; summarizing the target carousel devices to obtain the plurality of target carousel devices; completing dynamic content carousel based on the Harmony distributed architecture based on the plurality of target carousel devices.

2. The dynamic content wheelcasting method based on a distributed architecture of the HongMeng, as claimed in claim 1, characterized in that, The regional priority evaluation on the plurality of to-be-played contents to obtain a plurality of content priority indexes comprises: performing the following operations on each to-be-played content in the plurality of to-be-played contents: confirming a source location coordinate, a total click volume, and a single-day click growth rate of the to-be-played content, wherein the source location coordinate includes a source longitude and a source latitude; obtaining a user device coordinate, wherein the user device coordinate includes a device longitude and a device latitude; calculating a content priority index according to the user device coordinate, the source location coordinate, the total click volume, and the single-day click growth rate; summarizing the content priority indexes to obtain the plurality of content priority indexes.

3. The dynamic content wheelcasting method based on a distributed architecture of the HongMeng, as claimed in claim 2, wherein, The network performance analysis on the plurality of distributed devices to obtain a plurality of distributed resolution rates comprises: The following operations are performed on each of the plurality of distributed devices: Read the CPU load of the distributed CPU in the distributed device, and read the GPU stream number, GPU video memory capacity and GPU video memory bandwidth of the distributed GPU in the distributed device; Network state monitoring is performed on the distributed device to obtain bandwidth utilization, packet loss rate and network delay; The distributed resolution rate is calculated according to the CPU load, GPU stream number, GPU video memory capacity, GPU video memory bandwidth, bandwidth utilization, packet loss rate and network delay, and the calculation formula is as follows: ； wherein, is the distribution analysis rate, , and are the number of GPU streams, GPU memory capacity and GPU memory bandwidth, respectively, is the CPU load, is the bandwidth usage rate, is the packet loss rate, is the network delay, is the hyperbolic tangent function; The distributed resolution rates are summarized to obtain a plurality of distributed resolution rates.

4. The dynamic content carousel method based on a horcrumb distributed architecture as claimed in claim 3, wherein, The user feature analysis is performed on the plurality of distributed devices to obtain a plurality of user feature data, including: The following operations are performed on each of the plurality of distributed devices: Read a plurality of browsing titles in the historical database of the distributed device; The following operations are performed on each of the plurality of browsing titles: Chinese character extraction is performed on the browsing title to obtain a plurality of title Chinese characters; Based on the preset common character set, the common character elimination is performed on the plurality of title Chinese characters to obtain a reference Chinese character set; The reference Chinese character sets are summarized to obtain a plurality of reference Chinese character sets; The plurality of reference Chinese character sets are merged to obtain a total historical Chinese character set; The total historical Chinese character set is classified to obtain a plurality of classified Chinese character sets; The following operations are performed on each of the plurality of classified Chinese character sets: Confirm the access word and access frequency of the classified Chinese character set; Combine the access word and access frequency to obtain an access word group; Summarize the access word groups to obtain a plurality of access word groups; Confirm the user usage frequency of the distributed device, and summarize the user usage frequency and the plurality of access word groups to obtain user feature data; Summarize the user feature data to obtain a plurality of user feature data.

5. The dynamic content wheelcasting method based on a horcrn distributed architecture as claimed in claim 4, wherein, The genetic algorithm simulation is performed on the plurality of to-be-played contents and the plurality of distributed devices based on the plurality of distributed resolution rates, the plurality of user feature data and the plurality of content priority indexes to obtain an optimal carousel body, including: Confirm the carousel number based on the plurality of to-be-played contents and the plurality of distributed devices; Construct a plurality of simulation chromosomes based on the carousel number, wherein the simulation chromosomes are composed of a plurality of chromosome segments, and the chromosome segments include a gene locus, wherein a is the carousel number, the number of chromosome segments in the plurality of chromosome segments is equal to the distribution number, and the chromosome segments correspond to the distributed devices one by one; The following operations are performed on each of the plurality of simulation chromosomes: Random gene assignment is performed on the simulation chromosome to obtain a carousel chromosome; Summarize the carousel chromosomes to obtain a plurality of carousel chromosomes; The following operations are performed on each of the plurality of carousel chromosomes: Calculate the carousel matching degree according to the play content sequence, the carousel chromosome, the plurality of distributed resolution rates, the plurality of user feature data and the plurality of content priority indexes; Summarize the carousel matching degrees to obtain a plurality of carousel matching degrees, and sort the plurality of carousel matching degrees in descending order to obtain a carousel matching degree sequence; Confirm the number of chromosomes of the plurality of simulation chromosomes; Confirm a plurality of updated chromosomes based on the plurality of carousel chromosomes, the carousel matching degree sequence and the number of chromosomes; Confirming the number of times of performing the following operation on each of the plurality of carousel chromosomes, comparing the number of times of performing with a preset number threshold, if the number of times of performing is less than the number threshold, taking the plurality of updated chromosomes as the plurality of carousel chromosomes, returning to the step of performing the following operation on each of the plurality of carousel chromosomes until the number of times of performing is equal to the number threshold, taking the plurality of updated chromosomes as the plurality of target chromosomes; Performing the following operation on each of the plurality of target chromosomes: According to the sequence of playing content, the target chromosome, the plurality of distributed analysis rates, the plurality of user feature data and the plurality of content priority indexes, calculating the final carousel matching degree; Summarizing the final carousel matching degree to obtain a plurality of final carousel matching degrees, and taking the target chromosome corresponding to the maximum final carousel matching degree in the plurality of final carousel matching degrees as the optimal carousel body.

6. The dynamic content wheelcasting method based on a HongMeng distributed architecture of claim 5, wherein, The random gene assignment to the simulation chromosome to obtain the carousel chromosome comprises: Constructing a playing label set based on the number of plays; Performing the following operation on each of the plurality of chromosome segments of the simulation chromosome: Extracting gene sites from the chromosome segment in turn, and performing the following operation on the extracted gene sites: Randomly selecting a playing label from the playing label set to obtain a selected playing label, and assigning the selected playing label to the gene site to obtain a marked site; Removing the selected playing label from the playing label set to obtain an updated label set, taking the updated label set as the playing label set, and returning to the step of extracting the gene sites from the chromosome segment until the a gene sites of the chromosome segment are extracted; Summarizing the marked sites to obtain a plurality of marked sites, and confirming the marked chromosome segment based on the a marked sites and the chromosome segment; Summarizing the marked chromosome segment to obtain a plurality of marked chromosome segments; Confirming the carousel chromosome based on the plurality of marked chromosome segments and the simulation chromosome.

7. The dynamic content carousel method based on a Hmong distributed architecture as claimed in claim 6, characterized by, The carousel matching degree is calculated according to the sequence of playing content, the carousel chromosome, the plurality of distributed analysis rates, the plurality of user feature data and the plurality of content priority indexes, comprising: Performing the following operation on each of the plurality of carousel chromosomes: Taking the marked chromosome segment as an evaluation device, taking the distributed analysis rate corresponding to the evaluation device in the plurality of distributed analysis rates as an evaluation analysis rate, and taking the user feature data corresponding to the evaluation device in the plurality of user feature data as user evaluation data; Confirming the total priority index, the total analysis time and the total preference degree based on the marked chromosome segment, the sequence of playing content, the plurality of content priority indexes and the evaluation analysis rate; According to the user usage frequency in the user evaluation data, the total priority index, the total analysis time and the total preference degree, calculating the carousel distribution matching degree, and the calculation formula is as follows: ； wherein, is a wheelcast distribution match, is a user usage frequency, is a total priority index, is a total preference, is a total resolution time, is a natural constant; Summarizing the carousel distribution matching degree to obtain a plurality of carousel distribution matching degrees, and calculating the carousel matching degree according to the plurality of carousel distribution matching degrees, wherein the carousel matching degree is the sum of the plurality of carousel distribution matching degrees.

8. A dynamic content carousel system based on a distributed architecture of the HOMOGENEUS, characterized in that, The system comprises: The distribution architecture confirmation module is configured to receive a content carousel instruction, perform a HarmonyOS distributed soft bus query based on the content carousel instruction, and obtain a plurality of distributed devices, wherein the distributed devices include a distributed CPU, a distributed GPU, and a historical database. The carousel content analysis module is configured to obtain a plurality of to-be-played contents, perform regional priority evaluation on the plurality of to-be-played contents to obtain a plurality of content priority indexes, wherein the content priority indexes correspond to the to-be-played contents one by one, perform network performance analysis on the plurality of distributed devices to obtain a plurality of distributed resolution rates, perform user feature analysis on the plurality of distributed devices to obtain a plurality of user feature data, wherein the distributed resolution rates and the user feature data correspond to the distributed devices one by one, and the user feature data includes a user usage frequency and a plurality of access word groups, wherein the access word group includes an access word and an access frequency. The genetic algorithm simulation module is configured to perform genetic algorithm simulation on the plurality of to-be-played contents and the plurality of distributed devices based on the plurality of distributed resolution rates, the plurality of user feature data, and the plurality of content priority indexes, and obtain an optimal carousel body. The optimal carousel setting module is configured to perform content carousel setting on the plurality of distributed devices based on the optimal carousel body, and obtain a plurality of target carousel devices. The following operations are performed on each marker dye segment in the optimal carousel body: The corresponding distributed device of the marker dye segment in the plurality of distributed devices is taken as an initial carousel device; An initial play list of the initial carousel device is obtained, wherein the initial play list is an empty set; The kth marker site is extracted from the marker dye segment, wherein the initial value of k is 1; A target play content is confirmed from a play content sequence based on a selected play tag of the kth marker site; The target play content is placed in the kth position in the initial play list to obtain an updated play list; L is set to k+1, and L is taken as k, the updated play list is taken as a target play list, and the step of extracting the kth marker site from the marker dye segment is returned until the a marker sites in the marker dye segment are all extracted, and the updated play list is taken as a carousel play list; The target carousel device is confirmed based on the carousel play list and the initial carousel device; The target carousel devices are summarized to obtain a plurality of target carousel devices; The dynamic content carousel based on the Harmony distributed architecture is completed based on the plurality of target carousel devices.

Images