An intelligent broadcasting ecosystem and method based on artificial intelligence multi-terminal application

Abstract

The present invention relates to a smart broadcast ecosystem and method based on artificial intelligence multi-terminal applications, and relates to the field of audio and video playback management, including: obtaining a user interest tag matrix of a first online user, obtaining a user portrait of the first online user; obtaining an adjacent interest tag set based on a user basic portrait, a user behavior portrait, and a user browsing content portrait, updating a user interest tag matrix, obtaining a user interest tag update matrix, resetting the steady-state recording time to 0 to start timing; matching playback information based on the user interest tag update matrix, and pushing the matching playback information to a computer, a mobile phone, a speaker, and other multi-terminal applications for playback. The defects of the prior art of periodically recommending playback information to solve the poor diversity of interest-based recommendation information, as well as the low probability of matching with the user, which affects the user's browsing, playing, and listening experience are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of audio and video playback management, and in particular to an intelligent broadcasting ecosystem and method based on artificial intelligence multi-terminal applications. Background Art

[0002] The traditional audio playback mode is rigid, which makes it impossible to achieve a high degree of flexibility in playback. There are many defects, such as the poor fit between the audio playback content and the user, the fixed playback time of the audio playback content cannot be flexibly adjusted, such as the inability to flexibly play according to the user's switching of multi-terminal applications.

[0003] Among the above defects, the poor fit between audio playback content and users is the most difficult to solve. In traditional audio playback management systems, user playback information usually depends on the user's own portrait, which is often built based on the user's long-term behavior patterns and preferences. Although this method can provide a certain degree of personalized recommendations, over time, the recommended content will gradually tend to be single and lack diversity.

[0004] At present, in order to solve this defect, some random content will be recommended periodically to achieve diversity in the playback information. However, the disadvantage is that the probability of matching with the user is low, which affects the user's browsing experience. Summary of the invention

[0005] The present invention aims to solve the technical problems that the prior art uses periodic recommended playback information to solve the defects of poor diversity of interest-based recommended information, as well as the low probability of matching with users, which affects the user's browsing experience. An intelligent broadcast ecosystem and method based on artificial intelligence multi-terminal applications are provided to solve the problems.

[0006] The technical solution of the present invention to solve the above technical problems is as follows:

[0007] In a first aspect, the present invention provides an intelligent broadcast ecology method based on artificial intelligence multi-terminal applications, including: obtaining a user interest tag matrix of a first online user, wherein the user interest tag matrix has a steady-state recording duration, and the steady-state recording duration is the continuous duration that the user interest tag matrix is ​​not updated; when the steady-state recording duration is greater than or equal to a preset update period, obtaining a user portrait of the first online user, wherein the user portrait includes a user basic portrait, a user behavior portrait of a time zone covered by the steady-state recording duration, and a user browsing content portrait; performing adjacent interest retrieval based on the user basic portrait, the user behavior portrait, and the user browsing content portrait to obtain an adjacent interest tag set; updating the user interest tag matrix based on the adjacent interest tag set to obtain a user interest tag update matrix, and resetting the steady-state recording duration to 0 to start timing; matching playback information based on the user interest tag update matrix, and pushing the matching playback information to multi-terminal applications for playback, wherein, based on the user basic portrait, the user behavior portrait, and the user browsing content portrait, a user interest tag update matrix is ​​obtained, and the steady-state recording duration is reset to 0 to start timing; matching playback information based on the user interest tag update matrix, and pushing the matching playback information to multi-terminal applications for playback, Conducting adjacent user collection for the portrait and the user browsing content portrait to obtain a first-level adjacent user basic portrait, a first-level adjacent user behavior portrait, and a first-level adjacent user browsing content portrait; conducting adjacent user collection based on the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait, and the first-level adjacent user browsing content portrait to obtain a second-level adjacent user basic portrait, a second-level adjacent user behavior portrait, and a second-level adjacent user browsing content portrait; conducting adjacent user collection based on K-level adjacent user basic portraits, K-level adjacent user behavior portraits, and K-level adjacent user browsing content portraits to obtain K+1-level adjacent user basic portraits, K+1-level adjacent user behavior portraits, and K+1-level adjacent user browsing content portraits, 2≤K+1≤5; conducting adjacent interest frequency analysis based on the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, and up to K+1-level adjacent user behavior portraits and K+1-level adjacent user browsing content portraits to obtain the adjacent interest tag set.

[0008] In a second aspect, the present invention provides an intelligent broadcast ecosystem based on artificial intelligence multi-terminal applications, including: an interest tag matrix retrieval module, used to obtain a user interest tag matrix of a first online user, wherein the user interest tag matrix has a steady-state recording duration, and the steady-state recording duration is the continuous duration during which the user interest tag matrix is ​​not updated; a user portrait retrieval module, used to obtain a user portrait of the first online user when the steady-state recording duration is greater than or equal to a preset update period, wherein the user portrait includes a user basic portrait, a user behavior portrait of the time zone covered by the recording of the steady-state recording duration, and a user browsing content portrait. The invention relates to a method for performing adjacent interest retrieval based on the user basic profile, the user behavior profile and the user browsing content profile to obtain an adjacent interest tag set; an interest tag update module is used to update the user interest tag matrix based on the adjacent interest tag set to obtain a user interest tag update matrix, and at the same time reset the steady-state recording time to 0 to start timing; a smart broadcast execution module is used to match playback information based on the user interest tag update matrix, and push the matching playback information to multi-terminal applications for playback, wherein the adjacent interest retrieval module execution includes: The method further comprises: performing adjacent interest retrieval based on the user basic portrait, the user behavior portrait and the user browsing content portrait, and obtaining the adjacent interest tag set, including: performing adjacent user collection based on the user basic portrait, the user behavior portrait and the user browsing content portrait, and obtaining the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait; performing adjacent user collection based on the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait, and obtaining the second-level adjacent user basic portrait, the second-level adjacent user behavior portrait and the second-level adjacent user browsing content portrait; Until adjacent users are collected based on K-level adjacent user basic portraits, K-level adjacent user behavior portraits and K-level adjacent user browsing content portraits, K+1-level adjacent user basic portraits, K+1-level adjacent user behavior portraits and K+1-level adjacent user browsing content portraits are obtained, 2≤K+1≤5; adjacent interest frequency analysis is performed based on the first-level adjacent user behavior portraits, the first-level adjacent user browsing content portraits, the second-level adjacent user behavior portraits, the second-level adjacent user browsing content portraits, until K+1-level adjacent user behavior portraits and K+1-level adjacent user browsing content portraits, to obtain the adjacent interest tag set.

[0009] In a third aspect, the present invention provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby implementing the intelligent broadcast ecological method based on artificial intelligence multi-terminal applications described in the first aspect.

[0010] In a fourth aspect, the present invention provides a non-transitory computer-readable storage medium, in which a computer software program is stored. When the computer software program is executed by a processor, the smart broadcast ecological method based on artificial intelligence multi-terminal application described in the first aspect is implemented.

[0011] The beneficial effect of the present invention is as follows: by obtaining the user interest tag matrix of the first online user and recording its steady-state duration, that is, the continuous duration that the user interest tag matrix is ​​not updated. When the steady-state recording duration reaches or exceeds the preset update cycle, the user portrait of the user is obtained, and the user portrait includes the user basic portrait, the user behavior portrait and the user browsing content portrait. Adjacent interest retrieval is performed based on the user portrait to obtain a set of adjacent interest tags. The user interest tag matrix is ​​updated using the adjacent interest tag set to form a user interest tag update matrix, and the steady-state recording duration is reset. The technical solution of matching playback information according to the updated user interest tag matrix and pushing the matched playback information to multi-terminal applications for playback uses the user basic portrait, the user behavior portrait and the user browsing content portrait as the index basis of adjacent users, collects the interest tags of other users with similar user basic portraits, behaviors and browsing content, and uses them as indicators to enrich the user interest tag matrix. Compared with the traditional push method, it not only increases diversity, but also improves the probability of matching with users, achieving the technical effect of improving user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flowchart of an intelligent broadcast ecological method based on artificial intelligence multi-terminal application provided by the present invention;

[0013] Figure 2 A schematic diagram of the structure of an intelligent broadcast ecosystem based on artificial intelligence multi-terminal applications provided by the present invention;

[0014] Figure 3 A schematic diagram of the structure of an electronic device provided by the present invention;

[0015] Figure 4 A schematic diagram of the structure of a computer-readable storage medium provided by the present invention;

[0016] Figure 5 A schematic diagram of a possible structure of the audio playback management platform provided by the present invention.

[0017] In the accompanying drawings, the components represented by the reference numerals are listed as follows:

[0018] Electronic device 500 , memory 510 , processor 520 , computer program 511 , computer readable storage medium 600 , computer program 611 . DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0020] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0021] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0022] Embodiment 1:

[0023] like Figure 1 As shown, the embodiment of the present invention provides an intelligent broadcasting ecological method based on artificial intelligence multi-terminal application, including the steps of:

[0024] Preferably, the smart broadcasting ecological method based on artificial intelligence multi-terminal application disclosed in the embodiment of the present application can be embedded in Figure 5 The audio playback management platform shown in the figure includes a system cloud platform, a solution setting module, an application terminal, a playback terminal and a user terminal.

[0025] In one implementation, users can create phrase audio files through the audio AI synthesis technology provided by the platform, and then undergo AI + manual review. When they meet the preset requirements, the corresponding phrase audio files are stored in the audio resource library of the system cloud platform; then the background can use an intelligent broadcast ecological method based on artificial intelligence multi-terminal application provided in an embodiment of the present application to analyze each phrase audio file that matches each user; the background staff then sets the playback order of the phrase audio files that match each user; and distributes them to user playback terminals of schools, communities, factories, enterprises and other organizations, such as computers, mobile phones, large screens, etc., through the playback end; when the user logs in through the application end, including computers, mobile phones, speakers, etc., the phrase audio files are played in the set playback order that matches each user, and at the same time, various playback data are displayed to the regulatory layer on a visual large screen.

[0026] In a further implementation method, the audio playback management platform also provides an insert function. When receiving information that needs to be inserted, you can enter the system through a mobile phone or computer to broadcast emergency-related audio files, or you can play it through oral broadcast.

[0027] The audio playback management platform provided in the embodiment of the present application can achieve highly flexible playback through the temporary insertion function. Through an intelligent broadcast ecological method based on artificial intelligence multi-terminal applications provided in the embodiment of the present application, it can achieve the selection of playback information with a high degree of fit with users, and can achieve flexible switching of multi-terminal applications by distributing them to multi-terminal applications, thereby realizing intelligent playback in schools, communities, enterprises, factories and other units in places with a large number of people.

[0028] The audio playback management platform provided in the embodiment of the present application can be widely used in news information, online lectures, education, communities, factories, enterprises and other fields. By providing personalized content recommendations, intelligent interactive experience and accurate data analysis services, it can significantly improve user experience and satisfaction, and enhance user stickiness and loyalty. At the same time, the system is also highly scalable and flexible, and can be customized and optimized and upgraded according to the needs of different industries.

[0029] Since AI audit technology, temporary insertion technology and information distribution technology are relatively mature, the embodiments of the present application are not described in detail. The following mainly explains the implementation process of an intelligent broadcast ecological method based on artificial intelligence multi-terminal application provided by the embodiments of the present application:

[0030] S10: Obtaining a user interest tag matrix of a first online user, wherein the user interest tag matrix has a steady-state recording duration, and the steady-state recording duration is a continuous duration during which the user interest tag matrix is ​​not updated;

[0031] In a preferred embodiment of the present invention, the first online user refers to the user who is currently logged in online; the user interest tag matrix is ​​a multidimensional data structure used to describe user preferences, and each dimension represents the user's interest in a specific topic or content, such as: food, military, Hong Kong style clothing, early childhood education, history, etc. Steady-state recording duration refers to the time since the user interest tag matrix was last updated, which is used to measure the timeliness of the user interest tag.

[0032] Preferably, the process of the user interest tag matrix involves collecting user behavior data, such as browsing history, search history, purchase behavior, etc., and then counting the user's click frequency and the points of interest with longer browsing time as user interest tags. The calculation of the steady-state recording time is achieved by recording the last update time of the user interest tag matrix and comparing it with the current time.

[0033] The construction of the user interest tag matrix is ​​to deeply understand the user's interests and preferences, so as to provide more personalized content recommendations. The monitoring of steady-state recording duration ensures that the recommendation system can respond to changes in user interests in a timely manner, avoiding the impact of outdated interest tags on the relevance and accuracy of recommendations, and ultimately improving user satisfaction and the overall effectiveness of the recommendation system.

[0034] S20: When the steady-state recording duration is greater than or equal to a preset update period, a user portrait of the first online user is obtained, wherein the user portrait includes a basic user portrait, a user behavior portrait of the time zone covered by the recording of the steady-state recording duration, and a user browsing content portrait;

[0035] In a preferred embodiment of the present invention, the preset update cycle refers to the shortest update duration of the user interest tag matrix preset by the user, preferably 3 months. User portrait is a comprehensive user information model, which is constructed by analyzing the user's behavior and preferences. The user basic portrait involves the user's basic information, such as age, gender and occupation. The user behavior portrait focuses on the user's behavior patterns, such as browsing time. The user browsing content portrait focuses on the user's browsing habits for specific content, such as the type and frequency of content browsed.

[0036] When the user's steady-state recording time reaches or exceeds the system's preset update cycle, the system will trigger the update process of the user profile. This includes collecting and analyzing the user's latest data, such as the user's basic information update, the user's behavior pattern changes in the covered time zone, and the user's browsing habits for different content. This data can be obtained through the user's online activity log, interaction record, and content consumption record.

[0037] The step of updating the user profile is crucial to ensure that the recommendation system can provide content that matches the user's current interests and behaviors. By regularly updating the user profile, the system can capture the dynamic changes in user interests, thereby improving the relevance and personalization of recommended content. In addition, updating the user profile helps the system more accurately predict user needs, enhance the user experience, and potentially increase user engagement and satisfaction with recommended content.

[0038] S30: performing adjacent interest retrieval according to the user basic profile, the user behavior profile and the user browsing content profile to obtain an adjacent interest tag set;

[0039] In a preferred embodiment of the present invention, adjacent interest retrieval is a data statistical algorithm customized by the embodiment of the present invention. It introduces the user basic profile, the user behavior profile and the user browsing content profile as the screening benchmark, screens similar user samples, and then extracts the interest points of similar user samples to obtain an adjacent interest tag set.

[0040] Preferably, the implementation of adjacent interest retrieval generally involves the following steps, which will be described in detail later, and are briefly described here:

[0041] Build an index: First, build an index based on user basic portraits (such as age, gender, occupation), user behavior portraits (such as browsing time) and user browsing content portraits (such as content type, browsing frequency). Similarity calculation: Then, identify adjacent users by calculating the similarity between user portraits to obtain adjacent user interests. Preferably, this can be achieved through similarity measurement methods such as cosine similarity and Jaccard index. Tag set generation: Finally, based on the similarity calculation results, extract tags that are similar or related to the user's current interest tags to form a set of adjacent interest tags.

[0042] Generally speaking, the interests of adjacent users with similar behaviors, similar browsing content, and similar basic information may also interest the first online user. Compared with the content recommended solely based on the user's own behavioral data, the content will be richer and more personalized than traditional enrichment methods, which is conducive to improving user experience.

[0043] S40: updating the user interest tag matrix according to the adjacent interest tag set to obtain a user interest tag update matrix, and resetting the steady-state recording time to 0 to start timing;

[0044] In a preferred embodiment of the present invention, the user interest tag update matrix refers to a new matrix updated based on the original user interest tag matrix according to the adjacent interest tag set: that is, the interest tags in the adjacent interest tag set that do not appear in the original user interest tag matrix are added to the original user interest tag matrix to obtain a richer user interest tag update matrix. At the same time, the steady-state recording duration is reset to 0 and a new timing cycle is started to track the timing of the next update.

[0045] S50: Matching the playback information according to the user interest tag update matrix, and pushing the matching playback information to multi-terminal applications for playback.

[0046] In a preferred embodiment of the present invention, playback information matching refers to finding matching content based on the user's interest tag matrix. Preferably, when the current playback information is uploaded, various tags will be configured based on its content, and the tags are consistent with the user's interest tags. The playback information with the user's interest tags in the user's interest tag update matrix is ​​used as the matching playback information. Multi-terminal applications refer to information push terminals, knowledge question and answer interactive terminals, audio phrase playback, PC management terminals, mobile operation terminals, etc., exemplarily PC terminals, mobile phone terminals, display terminals, hardware playback terminals, audio, etc., and managers can customize the configuration. When the matching playback information is matched, the matching playback information is distributed to the multi-terminal application logged in by the first online user, preferably pushed to the mobile phone terminal, and the push method can be in the form of audio and video messages, text message lights, to achieve the technical effect of improving the intelligence of playback information push, and achieving balanced richness and personalization.

[0047] Further, when the steady-state recording duration is greater than or equal to a preset update period, a user portrait of the first online user is obtained, and step S20 includes the following steps:

[0048] S21: The basic user profile includes age information, user gender and user occupation;

[0049] S22: The user behavior profile includes user browsing time information;

[0050] S23: The user browsing content portrait includes the user browsing content type and user browsing frequency information;

[0051] S24: wherein the user browsing time information, the user browsing content type and the user browsing frequency information correspond one to one.

[0052] In a preferred embodiment of the present invention, the basic user portrait only needs to include age information, user gender and user occupation. Other basic user information such as birthplace, education, nationality, etc. do not need to be considered. This is because if too many tags are added to the basic portrait, it will limit the collected adjacent users, causing the adjacent users to be too similar to the first online user, which will cause the adjacent interest tags that may be collected to be highly overlapped with the original interest tags. Therefore, the basic user portrait only needs to include age information, user gender and user occupation. User behavior portraits mainly refer to the multiple durations of users browsing various types of content in the time zone covered by the record. User browsing content portraits refer to the types of content browsed by users in the time zone covered by the record, as well as the browsing frequency of each type. User browsing duration information, user browsing content type and user browsing frequency information are used as important reference data for indexing adjacent users in the next step.

[0053] By collecting a small amount of basic user information, it is possible to prevent users from deviating too much from the first online user. By using the user browsing time information, the user browsing content type and the user browsing frequency information as the main screening conditions, adjacent users with the same behavior portrait can be selected, thereby selecting the target sample users as adjacent users, and considering their interest tags as adjacent interest tags to enrich the interest tag matrix of the first online user, which is conducive to improving the richness of the recommended playback information of the first online user, while increasing the probability of matching with the first online user, which is conducive to improving the user browsing experience.

[0054] Further, adjacent interest retrieval is performed based on the user basic profile, the user behavior profile and the user browsing content profile to obtain an adjacent interest tag set. Step S30 includes the following steps:

[0055] S31: collecting adjacent users according to the user basic portrait, the user behavior portrait and the user browsing content portrait, and obtaining a first-level adjacent user basic portrait, a first-level adjacent user behavior portrait and a first-level adjacent user browsing content portrait;

[0056] S32: collecting adjacent users according to the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait, and obtaining the second-level adjacent user basic portrait, the second-level adjacent user behavior portrait and the second-level adjacent user browsing content portrait;

[0057] S33: until adjacent users are collected according to the basic portraits of K-level adjacent users, the behavioral portraits of K-level adjacent users, and the browsing content portraits of K-level adjacent users, and the basic portraits of K+1-level adjacent users, the behavioral portraits of K+1-level adjacent users, and the browsing content portraits of K+1-level adjacent users are obtained, 2≤K+1≤5;

[0058] S34: Perform adjacent interest frequency analysis based on the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, up to the K+1 level adjacent user behavior portrait and the K+1 level adjacent user browsing content portrait to obtain the adjacent interest tag set.

[0059] In a preferred embodiment of the present invention, adjacent interest retrieval refers to a customized user portrait-based data mining technology, which is used to discover other interest points that are similar or related to the user's current interest. Adjacent user collection refers to finding other users with similar characteristics to the target user through the user's portrait information. First-level adjacent users are user groups that are directly similar to the target user. Second-level adjacent users refer to user groups that are similar to first-level adjacent users and have a distant direct relationship with the target user. K-level adjacent users: user groups obtained by expanding K-layer similar user relationships. Adjacent interest tag set: a tag set that may be of interest to the target user, obtained based on adjacent user portrait analysis.

[0060] The detailed process is as follows: First, the user's basic information (age, gender, occupation), behavioral data (browsing time) and content preferences (browsing content type and frequency) are used to determine the first-level adjacent user group that is most similar to the user. Then, the same analysis is performed on the portraits of the first-level adjacent users to find the second-level adjacent users who are similar to them. This process can be extended to K-level adjacent users until enough data is obtained or the preset upper limit of the adjacent level (2≤K+1≤5) is reached. Finally, interest frequency analysis is performed on the portraits of adjacent users at all levels to calculate which interest tags appear most frequently among these users, and these tags are used as the adjacent interest tag set.

[0061] Through the above steps, a comprehensive set of adjacent interest tags can be constructed. This method not only considers the user's direct interests, but also extends to the interests of other user groups similar to the user, thereby providing a richer and more diverse set of interest tags. This helps the recommendation system discover new content that the user may not have directly expressed but is potentially interested in, thereby improving the novelty of the recommendation and user satisfaction. In addition, through multi-level adjacent user collection, the system can capture the diversity and complexity of user interests in a more detailed manner, further improving the performance of the recommendation system and user experience.

[0062] Preferably, the specific steps of determining adjacent users are as follows, and the determination process of the first-level adjacent users is used as an example for illustration, and the determination methods of other adjacent users are the same:

[0063] Further, adjacent users are collected according to the user basic portrait, the user behavior portrait and the user browsing content portrait to obtain the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait. S31 includes the steps of:

[0064] S311: constructing a user basic profile adjacency condition according to the user basic profile;

[0065] Further, based on the user basic portrait, a user basic portrait adjacency condition is constructed, and step S311 includes the following steps:

[0066] S3111: The user basic portrait includes type basic portrait attributes and quantitative basic portrait attributes;

[0067] S3112: When any one of the attributes of the type basic portrait is different, the user basic portrait adjacency condition is not satisfied;

[0068] S3113: traverse the quantitative basic image attributes and configure the attribute deviation threshold;

[0069] S3113: When any quantized basic portrait attribute deviation is greater than or equal to the attribute deviation threshold, the user basic portrait adjacency condition is not met.

[0070] S312: constructing a user behavior profile adjacency condition according to the user behavior profile;

[0071] S313: constructing a user browsing content portrait adjacency condition according to the user browsing content portrait;

[0072] S314: When the first user satisfies the user basic profile adjacency condition, the user behavior profile adjacency condition, and the user browsing content profile adjacency condition at the same time, adding the first-level adjacent user;

[0073] S315: When the number of first-level adjacent users meets N users, obtain the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait, where N≥5.

[0074] In a preferred embodiment of the present invention, the configuration rules of the user basic portrait adjacency conditions, the user behavior portrait adjacency conditions, and the user browsing content portrait adjacency conditions are the same, and the configuration process of the user basic portrait adjacency conditions is used as an example to illustrate:

[0075] The user basic portrait includes type basic portrait attributes and quantitative basic portrait attributes. Type basic portrait attributes refer to labels that cannot be quantified using numbers, such as gender and occupation, and quantitative basic portrait attributes refer to basic portrait attributes that can be quantified, such as age.

[0076] For the type basic portrait attributes, if any one of them is different, the user basic portrait adjacency condition is not met; for the quantitative basic portrait attributes, the quantitative basic portrait attributes are traversed through the management end, and the attribute deviation threshold is configured. The attribute deviation threshold corresponds to the quantitative basic portrait attribute one by one. When the deviation of any quantitative basic portrait attribute of the sample user is greater than or equal to the attribute deviation threshold, the user basic portrait adjacency condition is not met. When all the type basic portrait attributes are the same, and when the deviation of each quantitative basic portrait attribute is less than the attribute deviation threshold, the user basic portrait adjacency condition is met.

[0077] Similarly, the user behavior portrait adjacency conditions and the user browsing content portrait adjacency conditions are constructed. When the first user satisfies the user basic portrait adjacency conditions, the user behavior portrait adjacency conditions, and the user browsing content portrait adjacency conditions at the same time, the first-level adjacent user is added. Furthermore, when the first-level adjacent user satisfies the number of N users, the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait, and the first-level adjacent user browsing content portrait are obtained, where N≥5. A minimum of 5 is required, and in fact, the maximum should not exceed 7 to avoid obtaining too many interest tags, which will drown out the interest tags of the first online user itself, that is, 7≥N≥5.

[0078] Further, according to the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, until the K+1-level adjacent user behavior portrait and the K+1-level adjacent user browsing content portrait, the adjacent interest frequency analysis is performed to obtain the adjacent interest tag set, and step S34 includes the steps of:

[0079] S341: extracting the first browsing content type, the first browsing duration sum and the first browsing frequency sum, up to the Mth browsing content type, the Mth browsing duration sum and the Mth browsing frequency sum, according to the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, up to the K+1-level adjacent user behavior portrait and the K+1-level adjacent user browsing content portrait;

[0080] S342: performing interest analysis on the first browsing content type according to the sum of the first browsing duration and the sum of the first browsing frequency to obtain a first interest index;

[0081] S343: until an interest analysis is performed on the Mth browsing content type according to the Mth sum of browsing duration and the Mth sum of browsing frequency to obtain an Mth interest index;

[0082] S344: Extract browsing content type portrait tags whose first interest index is greater than or equal to the interest index threshold value from the first interest index to the Mth interest index, and set them as the adjacent interest tag set.

[0083] In a preferred embodiment of the present invention, the browsing content types according to the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, and up to the K+1-level adjacent user behavior portrait and the K+1-level adjacent user browsing content portrait are sorted to obtain the first browsing content type, the second browsing type, and up to the Mth browsing content type. Further, the sum of the browsing time of the first browsing content type is counted to obtain the first browsing time sum, and the sum of the browsing time of the second browsing content type is counted to obtain the second browsing time sum, until the sum of the browsing time of the Mth browsing content type is counted to obtain the Mth browsing time sum. Further, the sum of the frequency of the first browsing content type is counted to obtain the first browsing frequency sum, and until the sum of the browsing frequency of the Mth browsing content type is counted to obtain the Mth browsing frequency sum. The first interest index refers to the interest level of the first browsing content type calculated according to the first browsing time sum and the first browsing frequency sum according to the set rules. The larger the value, the greater the interest level. Until the Mth browsing content type is analyzed for interest according to the sum of the Mth browsing time and the sum of the Mth browsing frequency, the Mth interest index is obtained. The first interest index is extracted until the browsing content type portrait tags in the Mth interest index that are greater than or equal to the interest index threshold are set as the adjacent interest tag set. The interest index threshold refers to the sorting threshold preset by the user. Each browsing content type has a corresponding interest tag, which is set as the adjacent interest tag set. Preferably, M is an integer, M≥1, and M represents the total number of browsing content types.

[0084] By selecting interest tags with higher interest indexes of neighboring users, it is possible to enrich the interest tags while maximizing the interest tags that are consistent with the first online user, avoiding useless pushes that affect the user's browsing experience.

[0085] Further, interest analysis is performed on the first browsing content type according to the first browsing duration sum and the first browsing frequency sum to obtain a first interest index. Step S342 includes the following steps:

[0086] S3421: The sum of the first browsing durations has a first weight, and the sum of the first browsing frequencies has a second weight;

[0087] S3422: Perform weighted addition of the first normalized parameter of the sum of the first browsing durations and the second normalized parameter of the sum of the first browsing frequencies using the first weight and the second weight to obtain the first interest index.

[0088] In a preferred embodiment of the present invention, the preferred process of interest analysis is as follows:

[0089] The management terminal configures a first weight for the browsing duration and a second weight for the browsing frequency in advance, and the sum of the second weight and the first weight is equal to 1. Further, the first normalized parameter of the sum of the first browsing duration and the second normalized parameter of the sum of the first browsing frequency are weightedly added by the first weight and the second weight to obtain the first interest index.

[0090] An intelligent broadcasting ecological method based on artificial intelligence multi-terminal application provided by an embodiment of the present invention has at least the following technical effects:

[0091] By obtaining the user interest tag matrix of the first online user and recording its steady-state duration, that is, the continuous duration that the user interest tag matrix has not been updated. When the steady-state recording duration reaches or exceeds the preset update cycle, the user portrait of the user is obtained, and the user portrait includes the user basic portrait, the user behavior portrait and the user browsing content portrait. Adjacent interest retrieval is performed based on the user portrait to obtain a set of adjacent interest tags. The user interest tag matrix is ​​updated using the adjacent interest tag set to form a user interest tag update matrix, and the steady-state recording duration is reset. The technical solution of matching playback information according to the updated user interest tag matrix and pushing the matched playback information to multi-terminal applications for playback uses the user basic portrait, the user behavior portrait and the user browsing content portrait as the index basis of adjacent users, collects the interest tags of other users with similar user basic portraits, behaviors and browsing content, and uses them as indicators to enrich the user interest tag matrix. Compared with the traditional push method, it not only increases diversity, but also increases the probability of matching with users, achieving the technical effect of improving user experience.

[0092] Embodiment 2:

[0093] like Figure 2 As shown, based on the same inventive concept as the smart broadcast ecological method based on artificial intelligence multi-terminal application provided in Embodiment 1, the embodiment of the present invention further provides a smart broadcast ecological system based on artificial intelligence multi-terminal application, including:

[0094] An interest tag matrix retrieval module, used to obtain a user interest tag matrix of a first online user, wherein the user interest tag matrix has a steady-state recording duration, and the steady-state recording duration is a continuous duration during which the user interest tag matrix is ​​not updated;

[0095] A user portrait retrieval module, configured to obtain a user portrait of the first online user when the steady-state recording duration is greater than or equal to a preset update period, wherein the user portrait includes a basic user portrait, a user behavior portrait of the time zone covered by the recording of the steady-state recording duration, and a user browsing content portrait;

[0096] An adjacent interest retrieval module is used to perform adjacent interest retrieval based on the user basic profile, the user behavior profile and the user browsing content profile to obtain an adjacent interest tag set;

[0097] An interest tag updating module, used to update the user interest tag matrix according to the adjacent interest tag set, obtain the user interest tag update matrix, and reset the steady-state recording time to 0 to start timing;

[0098] The smart broadcast execution module is used to update the matrix according to the user interest tags to match the playback information, and push the matching playback information to multi-terminal applications for playback.

[0099] Furthermore, the user portrait retrieval module executes the following steps:

[0100] The user basic profile includes age information, user gender and user occupation;

[0101] The user behavior profile includes user browsing time information;

[0102] The user browsing content portrait includes the user browsing content type and user browsing frequency information;

[0103] The user browsing time information, the user browsing content type and the user browsing frequency information correspond to each other.

[0104] Furthermore, the adjacent interest retrieval module performs the following steps:

[0105] Collecting adjacent users according to the user basic portrait, the user behavior portrait and the user browsing content portrait to obtain a first-level adjacent user basic portrait, a first-level adjacent user behavior portrait and a first-level adjacent user browsing content portrait;

[0106] Collecting adjacent users according to the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait, and obtaining the second-level adjacent user basic portrait, the second-level adjacent user behavior portrait and the second-level adjacent user browsing content portrait;

[0107] Until adjacent users are collected based on the basic portraits of K-level adjacent users, the behavioral portraits of K-level adjacent users, and the browsing content portraits of K-level adjacent users, and the basic portraits of K+1-level adjacent users, the behavioral portraits of K+1-level adjacent users, and the browsing content portraits of K+1-level adjacent users are obtained, 2≤K+1≤5;

[0108] The adjacent interest frequency analysis is performed based on the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, up to the K+1 level adjacent user behavior portrait and the K+1 level adjacent user browsing content portrait to obtain the adjacent interest tag set.

[0109] Furthermore, the adjacent interest retrieval module executes the steps including:

[0110] According to the user basic profile, construct a user basic profile adjacency condition;

[0111] According to the user behavior profile, construct a user behavior profile adjacency condition;

[0112] According to the user browsing content portrait, constructing a user browsing content portrait adjacency condition;

[0113] When the first user simultaneously satisfies the user basic profile adjacency condition, the user behavior profile adjacency condition, and the user browsing content profile adjacency condition, a first-level adjacent user is added;

[0114] When the number of first-level adjacent users meets N users, the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait are obtained, wherein N≥5.

[0115] Furthermore, the adjacent interest retrieval module executes the steps including:

[0116] The user basic portrait includes type basic portrait attributes and quantitative basic portrait attributes;

[0117] When any one of the attributes of the type basic portrait is different, the user basic portrait adjacency condition is not met;

[0118] Traversing the quantitative basic image attributes and configuring attribute deviation thresholds;

[0119] When any quantized basic profile attribute deviation is greater than or equal to the attribute deviation threshold, the user basic profile adjacency condition is not met.

[0120] Furthermore, the adjacent interest retrieval module executes the steps including:

[0121] Extract the first browsing content type, the first browsing duration sum, and the first browsing frequency sum, up to the Mth browsing content type, the Mth browsing duration sum, and the Mth browsing frequency sum, according to the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, up to the K+1-level adjacent user behavior portrait and the K+1-level adjacent user browsing content portrait;

[0122] performing interest analysis on the first browsing content type according to the sum of the first browsing duration and the sum of the first browsing frequency to obtain a first interest index;

[0123] until an interest analysis is performed on the Mth browsing content type according to the sum of the Mth browsing duration and the sum of the Mth browsing frequency to obtain an Mth interest index;

[0124] Extract browsing content type portrait tags that are greater than or equal to the interest index threshold from the first interest index to the Mth interest index, and set them as the adjacent interest tag set.

[0125] Furthermore, the adjacent interest retrieval module executes the steps including:

[0126] The sum of the first browsing durations has a first weight, and the sum of the first browsing frequencies has a second weight;

[0127] The first interest index is obtained by performing weighted addition on a first normalized parameter obtained by summing the first browsing durations and a second normalized parameter obtained by summing the first browsing frequencies using the first weight and the second weight.

[0128] Embodiment three:

[0129] See also Figure 3 , Figure 3 Schematic diagram of an electronic device provided by an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device 500, including a memory 510, a processor 520, and a computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 511, the following steps are implemented:

[0130] Obtaining a user interest tag matrix of a first online user, wherein the user interest tag matrix has a steady-state recording duration, and the steady-state recording duration is a continuous duration during which the user interest tag matrix is ​​not updated;

[0131] When the steady-state recording duration is greater than or equal to a preset update period, a user portrait of the first online user is obtained, wherein the user portrait includes a basic user portrait, a user behavior portrait of the time zone covered by the recording of the steady-state recording duration, and a user browsing content portrait;

[0132] Perform adjacent interest retrieval based on the user basic profile, the user behavior profile, and the user browsing content profile to obtain an adjacent interest tag set;

[0133] According to the adjacent interest tag set, the user interest tag matrix is ​​updated to obtain a user interest tag update matrix, and at the same time, the steady-state recording time is reset to 0 to start timing;

[0134] Matching the playback information according to the user interest tag update matrix, and pushing the matching playback information to multi-terminal applications for playback;

[0135] Among them, performing adjacent interest retrieval according to the user basic profile, the user behavior profile and the user browsing content profile to obtain an adjacent interest tag set includes:

[0136] Collecting adjacent users according to the user basic portrait, the user behavior portrait and the user browsing content portrait to obtain a first-level adjacent user basic portrait, a first-level adjacent user behavior portrait and a first-level adjacent user browsing content portrait;

[0137] Collecting adjacent users according to the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait, and obtaining the second-level adjacent user basic portrait, the second-level adjacent user behavior portrait and the second-level adjacent user browsing content portrait;

[0138] Until adjacent users are collected based on the basic portraits of K-level adjacent users, the behavioral portraits of K-level adjacent users, and the browsing content portraits of K-level adjacent users, and the basic portraits of K+1-level adjacent users, the behavioral portraits of K+1-level adjacent users, and the browsing content portraits of K+1-level adjacent users are obtained, 2≤K+1≤5;

[0139] The adjacent interest frequency analysis is performed based on the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, up to the K+1 level adjacent user behavior portrait and the K+1 level adjacent user browsing content portrait to obtain the adjacent interest tag set.

[0140] Embodiment 4:

[0141] See also Figure 4 , Figure 4 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. Figure 4As shown, this embodiment provides a computer-readable storage medium 600, on which a computer program 611 is stored. When the computer program 611 is executed by a processor, the following steps are implemented:

[0142] Obtaining a user interest tag matrix of a first online user, wherein the user interest tag matrix has a steady-state recording duration, and the steady-state recording duration is a continuous duration during which the user interest tag matrix is ​​not updated;

[0143] When the steady-state recording duration is greater than or equal to a preset update period, a user portrait of the first online user is obtained, wherein the user portrait includes a basic user portrait, a user behavior portrait of the time zone covered by the recording of the steady-state recording duration, and a user browsing content portrait;

[0144] Perform adjacent interest retrieval based on the user basic profile, the user behavior profile, and the user browsing content profile to obtain an adjacent interest tag set;

[0145] According to the adjacent interest tag set, the user interest tag matrix is ​​updated to obtain a user interest tag update matrix, and at the same time, the steady-state recording time is reset to 0 to start timing;

[0146] Matching the playback information according to the user interest tag update matrix, and pushing the matching playback information to multi-terminal applications for playback;

[0147] Among them, performing adjacent interest retrieval according to the user basic profile, the user behavior profile and the user browsing content profile to obtain an adjacent interest tag set includes:

[0148] Collecting adjacent users according to the user basic portrait, the user behavior portrait and the user browsing content portrait to obtain a first-level adjacent user basic portrait, a first-level adjacent user behavior portrait and a first-level adjacent user browsing content portrait;

[0149] Collecting adjacent users according to the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait, and obtaining the second-level adjacent user basic portrait, the second-level adjacent user behavior portrait and the second-level adjacent user browsing content portrait;

[0150] Until adjacent users are collected based on the basic portraits of K-level adjacent users, the behavioral portraits of K-level adjacent users, and the browsing content portraits of K-level adjacent users, and the basic portraits of K+1-level adjacent users, the behavioral portraits of K+1-level adjacent users, and the browsing content portraits of K+1-level adjacent users are obtained, 2≤K+1≤5;

[0151] The adjacent interest frequency analysis is performed based on the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, up to the K+1 level adjacent user behavior portrait and the K+1 level adjacent user browsing content portrait to obtain the adjacent interest tag set.

[0152] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0153] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0154] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0155] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0156] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1The steps for the functions specified in one or more boxes.

[0157] Although preferred embodiments of the present invention have been described, additional changes and modifications may occur to these embodiments once those skilled in the art understand the basic inventive concepts.

[0158] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the present invention and its equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. An intelligent broadcast ecosystem based on artificial intelligence multi-terminal applications, characterized in that: include: An interest tag matrix retrieval module, used to obtain a user interest tag matrix of a first online user, wherein the user interest tag matrix has a steady-state recording duration, and the steady-state recording duration is a continuous duration during which the user interest tag matrix is ​​not updated; A user portrait retrieval module, configured to obtain a user portrait of the first online user when the steady-state recording duration is greater than or equal to a preset update period, wherein the user portrait includes a basic user portrait, a user behavior portrait of the time zone covered by the recording of the steady-state recording duration, and a user browsing content portrait; An adjacent interest retrieval module is used to perform adjacent interest retrieval based on the user basic profile, the user behavior profile and the user browsing content profile to obtain an adjacent interest tag set; An interest tag updating module, used to update the user interest tag matrix according to the adjacent interest tag set, obtain the user interest tag update matrix, and reset the steady-state recording time to 0 to start timing; A smart broadcast execution module is used to update the matrix according to the user interest tags to match the playback information, and push the matching playback information to multi-terminal applications for playback; The adjacent interest retrieval module executes the following steps: Collecting adjacent users according to the user basic portrait, the user behavior portrait and the user browsing content portrait to obtain a first-level adjacent user basic portrait, a first-level adjacent user behavior portrait and a first-level adjacent user browsing content portrait; Collecting adjacent users according to the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait, and obtaining the second-level adjacent user basic portrait, the second-level adjacent user behavior portrait and the second-level adjacent user browsing content portrait; Until adjacent users are collected based on the basic portraits of K-level adjacent users, the behavioral portraits of K-level adjacent users, and the browsing content portraits of K-level adjacent users, and the basic portraits of K+1-level adjacent users, the behavioral portraits of K+1-level adjacent users, and the browsing content portraits of K+1-level adjacent users are obtained, 2≤K+1≤5; Performing adjacent interest frequency analysis based on the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, up to the K+1-level adjacent user behavior portrait and the K+1-level adjacent user browsing content portrait, to obtain the adjacent interest tag set; The adjacent interest retrieval module executes the following steps: According to the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, until the K+1-level adjacent user behavior portrait and the K+1-level adjacent user browsing content portrait, extract the first browsing content type, the first browsing duration sum, and the first browsing frequency sum, until the M-th browsing content type, the M-th browsing duration sum, and the M-th browsing frequency sum, where M is an integer, M≥1, and M represents the total number of browsing content types; performing interest analysis on the first browsing content type according to the sum of the first browsing duration and the sum of the first browsing frequency to obtain a first interest index; until an interest analysis is performed on the Mth browsing content type according to the sum of the Mth browsing duration and the sum of the Mth browsing frequency to obtain an Mth interest index; Extracting browsing content type portrait tags whose first interest index is greater than or equal to the interest index threshold from the first interest index to the Mth interest index, and setting them as the adjacent interest tag set; The adjacent interest retrieval module executes the following steps: The sum of the first browsing durations has a first weight, and the sum of the first browsing frequencies has a second weight; The first interest index is obtained by performing weighted addition on a first normalized parameter obtained by summing the first browsing durations and a second normalized parameter obtained by summing the first browsing frequencies using the first weight and the second weight.

2. The system according to claim 1, characterized in that The user portrait retrieval module executes the steps including: The user basic profile includes age information, user gender and user occupation; The user behavior profile includes user browsing time information; The user browsing content portrait includes the user browsing content type and user browsing frequency information; The user browsing time information, the user browsing content type and the user browsing frequency information correspond to each other.

3. The system according to claim 1, characterized in that The adjacent interest retrieval module performs the following steps: According to the user basic profile, construct a user basic profile adjacency condition; According to the user behavior profile, construct a user behavior profile adjacency condition; According to the user browsing content portrait, constructing a user browsing content portrait adjacency condition; When the first user simultaneously satisfies the user basic profile adjacency condition, the user behavior profile adjacency condition, and the user browsing content profile adjacency condition, a first-level adjacent user is added; When the number of first-level adjacent users meets N users, the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait are obtained, wherein N≥5.

4. The system according to claim 3, characterized in that The adjacent interest retrieval module performs the following steps: The user basic portrait includes type basic portrait attributes and quantitative basic portrait attributes, wherein the type basic portrait attributes refer to labels that cannot be quantified using numbers, and the quantitative basic portrait attributes refer to basic portrait attributes that can be quantified; When any one of the attributes of the type basic portrait is different, the user basic portrait adjacency condition is not met; Traversing the quantitative basic image attributes and configuring attribute deviation thresholds; When any quantized basic profile attribute deviation is greater than or equal to the attribute deviation threshold, the user basic profile adjacency condition is not met.

5. A smart broadcasting ecological method based on artificial intelligence multi-terminal application, characterized in that: include: Obtaining a user interest tag matrix of a first online user, wherein the user interest tag matrix has a steady-state recording duration, and the steady-state recording duration is a continuous duration during which the user interest tag matrix is ​​not updated; When the steady-state recording duration is greater than or equal to a preset update period, a user portrait of the first online user is obtained, wherein the user portrait includes a basic user portrait, a user behavior portrait of the time zone covered by the recording of the steady-state recording duration, and a user browsing content portrait; Perform adjacent interest retrieval based on the user basic profile, the user behavior profile, and the user browsing content profile to obtain an adjacent interest tag set; According to the adjacent interest tag set, the user interest tag matrix is ​​updated to obtain a user interest tag update matrix, and at the same time, the steady-state recording time is reset to 0 to start timing; Matching the playback information according to the user interest tag update matrix, and pushing the matching playback information to multi-terminal applications for playback; Among them, performing adjacent interest retrieval according to the user basic profile, the user behavior profile and the user browsing content profile to obtain an adjacent interest tag set includes: Collecting adjacent users according to the user basic portrait, the user behavior portrait and the user browsing content portrait to obtain a first-level adjacent user basic portrait, a first-level adjacent user behavior portrait and a first-level adjacent user browsing content portrait; Collecting adjacent users according to the first-level adjacent user basic portrait, the first-level adjacent user behavior portrait and the first-level adjacent user browsing content portrait, and obtaining the second-level adjacent user basic portrait, the second-level adjacent user behavior portrait and the second-level adjacent user browsing content portrait; Until adjacent users are collected based on the basic portraits of K-level adjacent users, the behavioral portraits of K-level adjacent users, and the browsing content portraits of K-level adjacent users, and the basic portraits of K+1-level adjacent users, the behavioral portraits of K+1-level adjacent users, and the browsing content portraits of K+1-level adjacent users are obtained, 2≤K+1≤5; The adjacent interest tag set is obtained by performing adjacent interest frequency analysis according to the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, and up to the K+1-level adjacent user behavior portrait and the K+1-level adjacent user browsing content portrait, including: According to the first-level adjacent user behavior portrait, the first-level adjacent user browsing content portrait, the second-level adjacent user behavior portrait, the second-level adjacent user browsing content portrait, until the K+1-level adjacent user behavior portrait and the K+1-level adjacent user browsing content portrait, extract the first browsing content type, the first browsing duration sum, and the first browsing frequency sum, until the M-th browsing content type, the M-th browsing duration sum, and the M-th browsing frequency sum, where M is an integer, M≥1, and M represents the total number of browsing content types; performing interest analysis on the first browsing content type according to the sum of the first browsing duration and the sum of the first browsing frequency to obtain a first interest index; until an interest analysis is performed on the Mth browsing content type according to the sum of the Mth browsing duration and the sum of the Mth browsing frequency to obtain an Mth interest index; Extracting browsing content type portrait tags whose first interest index is greater than or equal to the interest index threshold from the first interest index to the Mth interest index, and setting them as the adjacent interest tag set; The step of performing interest analysis on the first browsing content type according to the sum of the first browsing duration and the sum of the first browsing frequency to obtain a first interest index includes: The sum of the first browsing durations has a first weight, and the sum of the first browsing frequencies has a second weight; The first interest index is obtained by performing weighted addition on a first normalized parameter obtained by summing the first browsing durations and a second normalized parameter obtained by summing the first browsing frequencies using the first weight and the second weight.

6. An electronic device, characterized in that: include: Memory for storing computer software programs; A processor is used to read and execute the computer software program, thereby implementing the intelligent broadcast ecological method based on artificial intelligence multi-terminal application as described in claim 5.

7. A non-transitory computer-readable storage medium, characterized in that: The storage medium stores a computer software program, which, when executed by the processor, implements the intelligent broadcast ecological method based on artificial intelligence multi-terminal application as described in claim 5.

Images