Play information recommendation method and device, intelligent equipment and storage medium

Abstract

The invention provides a broadcaster information recommendation method and device, intelligent equipment and a storage medium, and belongs to the technical field of information recommendation, and the method comprises the steps: determining a first broadcaster recommendation type based on user behavior data; the user behavior data comprises the broadcasting behavior data. And determining a target frequency based on the global podcaster data and the podcaster behavior data, and determining a second podcaster recommendation type based on the target frequency. And determining a podcaster recommendation list based on the first podcaster recommendation type and the second podcaster recommendation type. According to the broadcast information recommendation method and device, the intelligent equipment and the storage medium provided by the invention, the problem that a traditional recommendation mode is difficult to accurately match personalized demands can be solved.

Description

Technical Field

[0001] The present disclosure belongs to the technical field of information recommendation, and more specifically, to a podcast information recommendation method, apparatus, smart device and storage medium. Background Art

[0002] With the rapid development of Internet technology and the widespread popularity of digital media, podcasts, as a new form of audio content dissemination, have played an increasingly important role in people's daily lives. Podcasts are rich and diverse, covering many fields, with low production barriers. Creators can produce and publish programs using simple equipment and software, attracting a large number of listeners. Moreover, podcasts are not restricted by time and region, and listeners can flexibly choose how to listen.

[0003] The popularity of mobile devices provides a convenient terminal for podcasts, allowing people to listen to them in fragmented time in a variety of scenarios, while the high-speed and stable network infrastructure ensures the smooth playback of podcast programs. These factors together have promoted the widespread use of podcasts.

[0004] However, the number of programs on podcast platforms is growing exponentially due to the large number of creators and rich content. Users have difficulty choosing from a vast number of programs. Due to the large individual differences among users, traditional recommendation methods are difficult to accurately match personalized needs. Summary of the invention

[0005] The purpose of the present disclosure is to provide a podcast information recommendation method, apparatus, intelligent device and storage medium to solve the problem that traditional recommendation methods are difficult to accurately match personalized needs.

[0006] A first aspect of an embodiment of the present disclosure provides a podcast information recommendation method, comprising: Determining a first podcast recommendation type based on user behavior data; the user behavior data comprising podcast behavior data; Determine a target frequency based on the global podcast data and the podcast behavior data, and determine a second podcast recommendation type based on the target frequency; Determine a podcast recommendation list based on the first podcast recommendation type and the second podcast recommendation type.

[0007] A second aspect of an embodiment of the present disclosure provides a podcast information recommendation device, including: A first calculation module, configured to determine a first podcast recommendation type based on user behavior data; the user behavior data comprising podcast behavior data; A second calculation module, configured to determine a target frequency based on the global podcast data and the podcast behavior data, and determine a second podcast recommendation type based on the target frequency; A third calculation module is used to determine a podcast recommendation list based on the first podcast recommendation type and the second podcast recommendation type.

[0008] According to a third aspect of an embodiment of the present disclosure, a smart device is provided, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the above-mentioned podcast information recommendation method when executing the computer program.

[0009] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned podcast information recommendation method are implemented.

[0010] The podcast information recommendation method, apparatus, smart device, and storage medium provided by the embodiments of the present disclosure have the following beneficial effects: The disclosed embodiment determines the first podcast recommendation type based on user behavior data, which can accurately capture the user's direct interest preferences. Through in-depth analysis of podcast behavior data, the user's existing interest areas can be quickly located, and highly relevant program types can be provided for recommendation, which greatly improves the accuracy of the recommendation and makes the recommended content more in line with the user's actual preferences.

[0011] The disclosed embodiment combines global podcast data and podcast behavior data to determine the target frequency to determine the second podcast recommendation type, thereby achieving an organic combination of personalization and popular trends. By referring to popular trends in global data, it can explore users' potential interests and avoid recommending content that is too limited. It effectively solves the problem that traditional recommendations are based only on users' known preferences and ignore potential needs, making recommendations more comprehensive and personalized, broadening users' horizons, and enriching the listening experience.

[0012] The disclosed embodiment combines the first and second podcast recommendation types to determine a podcast recommendation list, which can provide a diverse and accurately sorted program selection, overcome the shortcomings of traditional recommendation methods in accurately matching personalized needs, and enhance user satisfaction and loyalty to the podcast platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 A flowchart of a podcast information recommendation method provided by an embodiment of the present disclosure; Figure 2 A structural block diagram of a podcast information recommendation device provided by an embodiment of the present disclosure; Figure 3A schematic block diagram of a smart device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0015] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present disclosure. However, it should be clear to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present disclosure with unnecessary details.

[0016] In order to make the purpose, technical solutions and advantages of the present disclosure more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0017] Please refer to Figure 1 , Figure 1 This is a flowchart of a podcast information recommendation method provided in an embodiment of the present disclosure. The method may include S101 to S103.

[0018] S101: Determine a first podcast recommendation type based on user behavior data. The user behavior data includes podcast behavior data.

[0019] In this embodiment, user behavior data refers to the record information generated by the user performing various operations on the podcast platform or related applications. User behavior data may include podcast behavior data and associated behavior data. Among them, podcast behavior data is the operation behavior information directly related to the podcast program. Podcast behavior data may include the playback record of the podcast program, pause and resume operations, fast forward / rewind operations, search records, collection and subscription behaviors, comment and like operations, sharing behaviors, and browsing tracks within the application. User behavior data can directly reflect the user's preference for different podcast content and behavior patterns.

[0020] The first podcast recommendation type is a type of podcast program that is initially screened out based on the user's individual type characteristics and behavior patterns and is suitable for the user. For example, if the user type is a technology enthusiast and often listens to podcast programs related to artificial intelligence, then the first podcast recommendation type may include podcast programs related to artificial intelligence technology, technology industry trends, etc. This is a direct recommendation direction determined based on the user's existing clear behavior and type.

[0021] Exemplarily, user operations are fully tracked in the podcast application to record all user behavior data. The collected data is cleaned to remove invalid data, and then key features in the podcast behavior data are extracted, such as calculating the proportion of different types of program playback time to the total playback time, the frequency of different operations, etc. Based on the feature analysis results, the first podcast recommendation type is determined through a machine learning algorithm or a rule-based system (such as setting a threshold to determine the user's preference for a certain type of program). For example, if the playback time ratio of a certain type of program exceeds a preset time threshold, the type of program to which this type belongs is determined as the first recommended type. Alternatively, the type distribution data of the podcast programs that the user has listened to in the past fixed period of time is calculated from the time dimension, and the various podcast program types are sorted based on the user's listening time, thereby obtaining the user's preference for different types of podcasts.

[0022] S102: Determine a target frequency based on the global podcast data and the podcast behavior data, and determine a second podcast recommendation type based on the target frequency.

[0023] In this embodiment, the target frequency is determined based on the global podcast data and the podcast behavior data, including: A first frequency is determined based on the podcast behavior data.

[0024] The first frequency is updated based on the global podcast data to obtain a target frequency.

[0025] In this embodiment, determining the second podcast recommendation type based on the target frequency includes: Determine target time points based on target frequency and podcast behavior data.

[0026] A second podcast recommendation type is determined based on the target time node.

[0027] In this embodiment, global podcast data refers to the comprehensive information data of all podcast programs on the entire podcast platform within a fixed time period. Global podcast data may include program attribute data, program dissemination data, program content feature data, program hot list, program update time, region, etc. These data reflect the overall situation and trend of podcast programs on the platform.

[0028] The second podcast recommendation type is based on the first podcast recommendation type, further exploring the potential interests of users and adapting to the overall trend of the platform to determine the podcast program type. It not only combines the user's existing behavioral performance, but also combines the hot trends reflected by the global podcast data, program popularity changes and other factors, aiming to provide users with a wider range of recommended content that is more in line with their potential needs, enrich the diversity of recommendations, and avoid recommendations being too limited to users' known preferences. The screening indicators of the second podcast recommendation type can include hot trend parameters, user potential interest parameters, time and geographical factor parameters, etc.

[0029] The first frequency is the frequency of recommending the second podcast recommended type program to the user. The first frequency is a frequency indicator related to the user's listening habits calculated based on the podcast behavior data. The first frequency may include the playback frequency, operation frequency, playback duration frequency in a specific time period, etc. of various types of programs. The frequency of changes in user preferences can be analyzed through the user's behavior data over a period of time, and then the appropriate first frequency can be obtained. The second podcast recommended type program can be recommended to the user based on the first frequency.

[0030] The target frequency is a frequency indicator that is adjusted and optimized based on the first frequency and combined with global podcast data. It can better reflect the relative situation of user behavior in the entire platform podcast ecosystem.

[0031] The target time node is a time point or time period with specific significance determined based on the target frequency and podcast behavior data. It may include the time point when the user's listening behavior changes significantly and the time point related to the global popular program trends, which is used to further determine the second podcast recommendation type.

[0032] Exemplarily, frequency-related information of various behaviors is extracted from the podcast behavior data to calculate the first frequency, and to gain a preliminary understanding of the user's personal listening habits and preferred frequency characteristics. Global podcast data is introduced to compare and correlate the user's first frequency with the overall program situation of the platform, and the first frequency is updated to obtain the target frequency, making the frequency index more global and representative. The target time node is jointly determined based on the target frequency and podcast behavior data, and the behavior patterns and global program trends around the time node are analyzed to determine the second podcast recommendation type, thereby exploring the user's potential interests and program types that conform to the platform's popular trends.

[0033] For example, the corresponding frequency calculation method is set according to different behavior types. For example, when calculating the playing frequency of a specific type of program, the playing frequency of the program of the type is obtained by counting the number of times a user plays a certain type of program in a certain period of time and dividing it by the total number of times the program is played.

[0034] For operation frequency, the number of operations per unit time is calculated. At the same time, the playback duration frequency is calculated, such as the proportion of program playback duration in different duration intervals in a specific time period to the total playback duration.

[0035] Based on the program dissemination data in the global podcast data, the average playback frequency, popularity index, etc. of various programs on the platform are calculated.

[0036] The user's first frequency is compared with the global average frequency, and the first frequency is updated by weighted calculation or other methods. For example, if the user's playback frequency of a certain type of program is higher than the global average frequency by a certain percentage, the weight of this type of program in the target frequency is appropriately increased, and vice versa. The first frequency can also be adjusted in combination with global factors such as program update frequency and regional popular trends to obtain a target frequency that better reflects the user's behavior in the global situation.

[0037] Analyze the change trend of the target frequency over time, and combine the timestamp information in the podcast behavior data to find the time point or time period when the frequency changes significantly, and determine it as the target time node. For example, the starting time point when the frequency of a certain type of program suddenly increases significantly, or the period of time after the user first listens to the relevant program when a certain type of popular program on the platform becomes popular.

[0038] The second podcast recommendation type is determined based on the podcast behavior data and global podcast data around the target time node. For example, if the popularity of a program related to an emerging technology topic on the platform increases after the target time node, the podcast type related to the emerging technology topic can be determined as the second recommendation type. At the same time, the recommendation type can be further refined based on the user's behavioral changes before and after the target time node. For example, based on the user's attention to programs related to a specific technology field during this time period, a more specific second recommendation type can be determined as an artificial intelligence-related podcast program, including sub-type programs such as technological progress, industry applications, and expert interviews.

[0039] S103: Determine a podcast recommendation list based on the first podcast recommendation type and the second podcast recommendation type.

[0040] In this embodiment, the podcast recommendation list is a sorted list of podcast programs that is finally presented to the user. These programs are selected from program resources related to the first podcast recommendation type and the second podcast recommendation type to help users quickly find podcast programs that suit their interests for listening.

[0041] In this embodiment, weights can be set for the first podcast recommendation type and the second podcast recommendation type respectively, and the weights can be determined based on factors such as the frequency and intensity of the user's historical operations on different types of programs and the stability of the popularity of various types of programs in the global data. For example, if the user listens to the first recommended type of program for a long time and with a high frequency, and the second recommended type of program is of potential interest but has less related behavior, then the weight of the first recommended type may be higher.

[0042] The relevance of each podcast program to the first podcast recommendation type and the second podcast recommendation type can be calculated by the matching degree of program theme keywords, the matching degree of host style and user preference, the fit between program content field and recommended type, etc. In addition, it is also necessary to combine the program audio quality, content quality, user evaluation and other aspects.

[0043] Exemplarily, based on the first podcast recommendation type and the second podcast recommendation type tags, program records belonging to these two recommendation types are queried and screened out in the podcast program database to obtain a preliminary set of recommended program resources.

[0044] For each screened program, its recommendation type weight parameter is calculated. For example, the proportion of the number of operations of the first recommendation type program by the user in the past period of time to the total number of operations is counted as the first recommendation type weight, and the second recommendation type weight is calculated similarly.

[0045] Calculate program relevance parameters and analyze the relevance between program content features and recommended types. For example, calculate the matching degree between program theme keywords and recommended type keywords, and use calculation methods such as cosine similarity algorithm to obtain a relevance value between 0 and 1.

[0046] Evaluate program quality parameters and perform quantitative calculations based on program quality-related data. For example, for audio quality, you can set certain scoring standards based on audio format, bit rate, etc. For content quality, you can analyze the evaluation keywords of the content in user ratings and comments and set corresponding scoring rules; for user reviews, directly use indicators such as the average program rating or the rate of favorable comments as some program quality parameters.

[0047] According to the set weight, calculate the comprehensive score of each program. For example, the comprehensive score of the program = the weight of the first recommendation type × 0.4 + the relevance of the program × 0.3 + the quality parameter of the program × 0.3. Use the sorting algorithm to sort the programs in the recommended program resource set in descending order according to the comprehensive score to obtain a podcast recommendation list.

[0048] The generated podcast recommendation list is presented to the user and can be displayed through the recommendation page of the podcast application. The displayed content includes the program name, cover image, introduction, host information, etc., to facilitate users to choose to listen.

[0049] During the interaction between users and the recommendation list, user feedback data is collected and used to optimize and adjust the subsequent recommendation algorithm. For example, if a user quickly skips a recommended program after clicking play, it indicates that the program does not match the user's actual interests. Based on this feedback, the relevant parameter calculation and recommendation type determination methods can be adjusted to improve the accuracy of recommendations and user satisfaction.

[0050] It can be concluded from the above that the disclosed embodiment determines the first podcast recommendation type based on user behavior data, which can accurately capture the user's direct interest preferences. Through in-depth analysis of podcast behavior data, the user's existing interest areas can be quickly located, and highly relevant program types can be provided for recommendation, which greatly improves the accuracy of the recommendation and makes the recommended content more in line with the user's actual preferences.

[0051] The disclosed embodiment combines global podcast data and podcast behavior data to determine the target frequency to determine the second podcast recommendation type, thereby achieving an organic combination of personalization and popular trends. By referring to popular trends in global data, it can explore users' potential interests and avoid recommending content that is too limited. It effectively solves the problem that traditional recommendations are based only on users' known preferences and ignore potential needs, making recommendations more comprehensive and personalized, broadening users' horizons, and enriching the listening experience.

[0052] The disclosed embodiment combines the first and second podcast recommendation types to determine a podcast recommendation list, which can provide a diverse and accurately sorted program selection, overcome the shortcomings of traditional recommendation methods in accurately matching personalized needs, and enhance user satisfaction and loyalty to the podcast platform.

[0053] In one embodiment of the present disclosure, the user behavior data also includes associated behavior data; The determining the first podcast recommendation type based on the user behavior data includes: Determining a first recommendation identifier and the podcast behavior data based on the user type; Determining a second recommendation identifier based on the podcast behavior data and the associated behavior data; The first podcast recommendation type is determined based on the first recommendation identifier and the second recommendation identifier.

[0054] In this embodiment, determining the first recommendation identifier and the podcast behavior data based on the user type includes: Determine behavior analysis rules and data statistical dimensions based on user types.

[0055] Determine podcast behavior data based on behavioral analysis rules and statistical dimensions.

[0056] A first recommendation identifier is determined based on the user type.

[0057] In this embodiment, the user type may be determined based on the user basic information.

[0058] In this embodiment, the associated behavior data refers to the user's behavior records related to podcast interests in addition to podcast listening operations in the podcast platform or related ecosystem. For example, data generated by browsing articles, participating in topic discussions, following specific accounts or community sections in the platform can indirectly reflect the user's potential interest in listening to podcasts. Another example is the user's behavior data in the music listening software.

[0059] The first recommendation mark is a mark for the user type. For example, if the user type is a college student group (aged about 18-22, currently studying). The first recommendation mark may include: "Academic Assistance": including sharing of learning methods related to professional courses and podcasts on subject knowledge, which will help them better master professional knowledge and cope with academic pressure.

[0060] "Interest expansion": College students have more spare time and will be interested in content in various interest areas, such as podcasts on photography techniques, painting art, music appreciation, etc., to expand their knowledge and interests.

[0061] "Campus Life": Podcast content related to campus culture, club activities, student social interaction, etc., can help them better integrate into university life and understand various campus dynamics.

[0062] The second recommendation identifier is a identifier that is further refined based on the podcast behavior data and the associated behavior data and can more accurately reflect the user's potential podcast interest tendency. The second recommendation identifier supplements and refines the first recommendation identifier, making the determined first podcast recommendation type more in line with the user's actual interests.

[0063] Behavior analysis rules are methods and guidelines for analyzing podcast behavior data for different user types. For example, for the "student" type, the focus may be on rules such as Internet hot topics, talk shows, listening after school, and the playback time of podcasts in specific subjects; for the "elderly" type, the focus may be on the analysis of the operating rules of daytime and health care podcasts.

[0064] The data statistical dimension is a perspective that quantifies podcast behavior data from multiple aspects, such as time dimension, content dimension, device dimension, interaction dimension, etc.

[0065] Basic user information refers to the basic attribute information provided by users during registration or use of the platform. Basic user information may include demographic information, geographic location information, contact information, social account information, and descriptive information such as interests, hobbies, and living habits filled out by users. Basic user information is an important basis for preliminary classification of users and understanding of their backgrounds, which helps to provide more targeted services and recommendations for different types of users.

[0066] In this embodiment, user types are divided according to user basic information to determine behavior analysis rules and data statistical dimensions, and then key feature information is extracted from the podcast behavior data, and a first recommendation identifier is assigned based on the user type. The podcast behavior data and associated behavior data are combined to determine the second recommendation identifier through data mining and analysis technology, and finally the first recommendation identifier and the second recommendation identifier are combined to accurately locate the first podcast recommendation type, realizing the gradual derivation and refinement from user basic characteristics to specific podcast recommendation types.

[0067] Exemplarily, basic information from multiple channels such as user registration information and questionnaire filling is collected, and users are divided into different types, such as "office workers", "students", "freelancers", etc., using classification algorithms or preset rule systems. Build exclusive behavior analysis rule libraries and data statistical dimension sets for each type of user, and store them in databases or configuration files for easy access. Collect all user operation data in the podcast application, and use data processing tools to clean, organize and calculate the data according to the rules and dimension requirements, such as calculating the proportion of playback time of a certain type of podcast in a specific time period. Collect related behavior data within the platform, extract key information through text analysis and other technologies, and associate and match it with the podcast program library to determine the second recommendation identifier. After obtaining the first recommendation identifier based on the user type, combine it with the second recommendation identifier, and use weighted calculations or logical judgments to determine the first podcast recommendation type, providing a precise starting point for the subsequent recommendation process.

[0068] This embodiment divides user types based on basic user information and determines exclusive rules and dimensions, which can accurately extract key features of podcast behavior, make the first recommendation identifier fit user characteristics, and preliminarily locate the scope of interest. The introduction of related behavior data can expand reference data, explore potential interests, and determine the second recommendation identifier to refine and supplement the first recommendation identifier. The combination of the two can accurately locate the first podcast recommendation type and improve the accuracy of the recommendation. Analyzing data from multiple dimensions, fully understanding users, and providing targeted services for different types of users can effectively improve user satisfaction, enhance user stickiness, and promote the good development of the podcast platform and user activity.

[0069] In one embodiment of the present disclosure, determining a podcast recommendation list based on a first podcast recommendation type and a second podcast recommendation type includes: A first podcast recommendation resource is determined based on the first podcast recommendation type, and a second podcast recommendation resource is determined based on the second podcast recommendation type.

[0070] The first recommendation index is calculated based on the first podcast recommendation resource and user behavior data.

[0071] A second recommendation index is calculated based on the second podcast recommendation resource and the user behavior data.

[0072] The target ratio is determined based on the first recommendation index and the second recommendation index.

[0073] Determine a podcast recommendation list from the first podcast recommendation resource and the second podcast recommendation resource based on the target ratio.

[0074] In this embodiment, the first podcast recommendation resource is a collection of podcast programs that are highly matched with the first podcast recommendation type. These programs are consistent with the user interest range preliminarily determined based on user behavior data in terms of theme, style, host, etc.

[0075] The second podcast recommendation resource corresponds to a podcast program group of the second podcast recommendation type. The second podcast recommendation resource is potential programs of interest further mined through comprehensive global podcast data and user behavior data, including emerging hot programs or programs related to potential user interests.

[0076] The first recommendation index is a quantitative indicator that measures the degree of fit between the first podcast recommendation resource and the user's actual interests, reflecting the user's possible liking for the programs in this type of resource.

[0077] The second recommendation index is used to evaluate the attractiveness of the second podcast recommendation resource to users, reflecting its relevance to users based on users' potential interests and platform trends.

[0078] The target ratio is determined based on the first recommendation index and the second recommendation index, and the relative proportion of the first podcast recommendation resource and the second podcast recommendation resource in the final podcast recommendation list to balance the program recommendations of direct interest and potential interest.

[0079] Exemplarily, the first podcast recommendation resource and the second podcast recommendation resource are found in the podcast program library by matching type tags or the like. For the first recommendation index, the historical number of times users play the program types in the first podcast recommendation resource, the average playback time, the collection ratio and other data are counted, and the result is obtained by comprehensive weighted calculation. Similarly, the second recommendation index is calculated by analyzing the potential interest-related data related to the second podcast recommendation resource in the user behavior data. After normalizing the first recommendation index and the second recommendation index, the target ratio is calculated, such as the first recommendation index ratio = first recommendation index / (first recommendation index + second recommendation index). Programs are selected from the two types of resources according to the target ratio, and programs with high comprehensive scores (combined with factors such as resource quality and popularity) can be given priority to form a podcast recommendation list to be displayed to users.

[0080] This embodiment determines resources through the first podcast recommendation type and the second podcast recommendation type, which not only pays attention to the direct interests that users have shown, but also takes into account potential interests, can broaden the recommendation field of view, and avoid the monotony of recommended content. By rigorously calculating the first recommendation index and the second recommendation index, and determining the target ratio based on data-driven, the generation of the recommendation list is highly scientific and objective, accurately reflecting the diverse interests and needs of users. Programs with high comprehensive scores are given priority to ensure the quality of the recommendation list, which can effectively save users' screening time, improve the efficiency of users discovering their favorite podcast programs, enhance users' favorability and loyalty to the podcast platform, and promote user retention and expansion of the platform.

[0081] In one embodiment of the present disclosure, determining a podcast recommendation list from a first podcast recommendation resource and a second podcast recommendation resource based on a target ratio includes: A third podcast recommendation resource is determined from the first podcast recommendation resource and the second podcast recommendation resource based on the target ratio. The third podcast recommendation resource includes multiple podcast programs.

[0082] Determine a priority sequence of multiple podcast programs in a third podcast recommendation resource based on the third podcast recommendation resource and user behavior data.

[0083] A podcast recommendation list is determined based on a priority sequence.

[0084] In this embodiment, the third podcast recommendation resource is a transitional podcast program collection that is screened from the first podcast recommendation resource and the second podcast recommendation resource according to a target ratio to generate a final podcast recommendation list, which integrates programs mined based on the user's direct interests and potential interests.

[0085] The priority sequence is the order in which multiple podcast programs in the third podcast recommendation resource are sorted according to the degree of fit with the user's interests and other relevant factors. The programs ranked at the front are more in line with user needs and should be recommended first.

[0086] In this embodiment, some programs are selected from the two types of resources according to the target ratio to form a third podcast recommendation resource. Combined with user behavior data and factors such as the popularity and quality of the program itself, a priority score is calculated for each program in the third podcast recommendation resource, and a priority sequence is determined based on the score. The programs are arranged in order according to the priority sequence to form a podcast recommendation list, and the programs that are most likely to meet the user's interests are placed at the front.

[0087] Exemplarily, programs are selected from the first podcast recommendation resource and the second podcast recommendation resource according to the calculated target ratio. For example, if the target ratio is 3:2 and the total number of recommendations is 10, then 6 programs are selected from the first podcast recommendation resource and 4 programs are selected from the second podcast recommendation resource to form the third podcast recommendation resource.

[0088] For each program in the third podcast recommendation resource, calculate its priority score. For example, give a certain weight based on the similarity between the program and the user's historical playback behavior; give a certain weight based on the popularity of the program in the global podcast data; give a certain weight based on the program update time and program quality score. Calculate the priority score of each program through weighted summation.

[0089] The programs in the third podcast recommendation resource are sorted in descending order according to the priority scores to obtain a priority sequence. Then, programs are selected in sequence according to the sequence to form a podcast recommendation list and presented to the user. For example, the program with the highest priority is ranked first in the recommendation list, and so on, so that the user can quickly find the podcast program that best suits his or her interests.

[0090] This embodiment selects the third podcast recommendation resource through the target ratio, taking into account the user's direct interest and potential interest programs, making the recommended content richer and more diverse. The priority score is calculated based on multiple factors to determine the sequence, which can accurately analyze the program's fit with the user's interests, popularity, quality, etc., and improve the accuracy of the recommendation. The recommendation list is composed according to the priority sequence, and the programs that best meet the user's interests are placed in front, which effectively saves the user's screening time and improves the efficiency of the user in discovering the program they like, thereby enhancing the user's satisfaction and loyalty to the podcast platform.

[0091] Corresponding to the podcast information recommendation method in the above embodiment, Figure 2 This is a structural block diagram of a podcast information recommendation device provided by an embodiment of the present disclosure. For ease of description, only the parts related to the embodiment of the present disclosure are shown. Figure 2 The podcast information recommendation device 20 includes: a first calculation module 21, a second calculation module 22 and a third calculation module 23.

[0092] The first calculation module 21 is used to determine the first podcast recommendation type based on user behavior data. The user behavior data includes podcast behavior data.

[0093] The second calculation module 22 is used to determine the target frequency based on the global podcast data and the podcast behavior data, and determine the second podcast recommendation type based on the target frequency.

[0094] The third calculation module 23 is used to determine a podcast recommendation list based on the first podcast recommendation type and the second podcast recommendation type.

[0095] In one embodiment of the present disclosure, the second calculation module 22 is specifically configured to determine the first frequency based on the podcast behavior data.

[0096] The first frequency is updated based on the global podcast data to obtain a target frequency.

[0097] In one embodiment of the present disclosure, the second calculation module 22 is further configured to determine a target time node based on the target frequency and the podcast behavior data.

[0098] A second podcast recommendation type is determined based on the target time node.

[0099] In an embodiment of the present disclosure, the user behavior data also includes associated behavior data. The first calculation module 21 is specifically configured to determine the first recommendation identifier and the podcast behavior data based on the user type.

[0100] A second recommendation identifier is determined based on the podcast behavior data and the associated behavior data.

[0101] A first podcast recommendation type is determined based on the first recommendation identifier and the second recommendation identifier.

[0102] In an embodiment of the present disclosure, the first calculation module 21 is further configured to determine behavior analysis rules and data statistical dimensions based on user types.

[0103] Determine podcast behavior data based on behavioral analysis rules and statistical dimensions.

[0104] A first recommendation identifier is determined based on the user type.

[0105] In one embodiment of the present disclosure, the third calculation module 23 is specifically used to determine a first podcast recommendation resource based on the first podcast recommendation type, and determine a second podcast recommendation resource based on the second podcast recommendation type.

[0106] The first recommendation index is calculated based on the first podcast recommendation resource and user behavior data.

[0107] A second recommendation index is calculated based on the second podcast recommendation resource and the user behavior data.

[0108] The target ratio is determined based on the first recommendation index and the second recommendation index.

[0109] Determine a podcast recommendation list from the first podcast recommendation resource and the second podcast recommendation resource based on the target ratio.

[0110] In one embodiment of the present disclosure, the third calculation module 23 is further configured to determine a third podcast recommendation resource from the first podcast recommendation resource and the second podcast recommendation resource based on the target ratio. The third podcast recommendation resource includes a plurality of podcast programs.

[0111] Determine a priority sequence of multiple podcast programs in a third podcast recommendation resource based on the third podcast recommendation resource and user behavior data.

[0112] A podcast recommendation list is determined based on a priority sequence.

[0113] See also Figure 3 , Figure 3 This is a schematic block diagram of a smart device provided by an embodiment of the present disclosure. Figure 3 The smart device 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303 and one or more memories 304. The processors 301, input devices 302, output devices 303 and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of each module in the above-mentioned device embodiments, such as Figure 2 The functions of modules 21 to 23 are shown.

[0114] It should be understood that in the embodiment of the present disclosure, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0115] The input device 302 may include a touch panel, a fingerprint collection sensor (for collecting the user's fingerprint information and fingerprint direction information), a microphone, etc., and the output device 303 may include a display (LCD, etc.), a speaker, etc.

[0116] The memory 304 may include a read-only memory and a random access memory, and provide instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store information about the device type.

[0117] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiments of the present disclosure may execute the implementation methods described in the first and second embodiments of the podcast information recommendation method provided in the embodiments of the present disclosure, and may also execute the implementation methods of the smart device 300 described in the embodiments of the present disclosure, which will not be repeated here.

[0118] In another embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by the processor, all or part of the processes in the above-mentioned embodiment method are implemented, and the computer program can also be completed by instructing the relevant hardware through the computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, the steps of each of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0119] The computer-readable storage medium may be an internal storage unit of the smart device of any of the aforementioned embodiments, such as a hard disk or memory of the smart device. The computer-readable storage medium may also be an external storage device of the smart device, such as a plug-in hard disk equipped on the smart device, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the computer-readable storage medium may also include both an internal storage unit of the smart device and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the smart device. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0120] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0121] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the smart devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

[0122] In the several embodiments provided in the present application, it should be understood that the disclosed intelligent devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or it can be an electrical, mechanical or other form of connection.

[0123] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present disclosure.

[0124] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0125] The above are only specific embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present disclosure, and these modifications or replacements should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

Claims

1. A podcast information recommendation method, characterized in that: include: Determine a first podcast recommendation type based on user behavior data; The user behavior data includes podcast behavior data; Determine a target frequency based on the global podcast data and the podcast behavior data, and determine a second podcast recommendation type based on the target frequency; Determine a podcast recommendation list based on the first podcast recommendation type and the second podcast recommendation type.

2. The podcast information recommendation method according to claim 1, characterized in that: The determining the target frequency based on the global podcast data and the podcast behavior data includes: determining a first frequency based on the podcast behavior data; The first frequency is updated based on the global podcast data to obtain the target frequency.

3. The podcast information recommendation method according to claim 1, characterized in that: The determining a second podcast recommendation type based on the target frequency includes: Determine a target time node based on the target frequency and the podcast behavior data; The second podcast recommendation type is determined based on the target time node.

4. The podcast information recommendation method according to claim 1, characterized in that: The user behavior data also includes associated behavior data; The determining the first podcast recommendation type based on the user behavior data includes: Determining a first recommendation identifier and the podcast behavior data based on the user type; Determining a second recommendation identifier based on the podcast behavior data and the associated behavior data; The first podcast recommendation type is determined based on the first recommendation identifier and the second recommendation identifier.

5. The podcast information recommendation method according to claim 4, characterized in that: The determining the first recommendation identifier and the podcast behavior data based on the user type includes: Determine behavior analysis rules and data statistical dimensions based on the user type; Determining the podcast behavior data based on the behavior analysis rules and the statistical dimensions; The first recommendation identifier is determined based on the user type.

6. The podcast information recommendation method according to claim 1, characterized in that: The determining of the podcast recommendation list based on the first podcast recommendation type and the second podcast recommendation type includes: Determine a first podcast recommendation resource based on the first podcast recommendation type, and determine a second podcast recommendation resource based on the second podcast recommendation type; Calculate a first recommendation index based on the first podcast recommendation resource and the user behavior data; Calculate a second recommendation index based on the second podcast recommendation resource and the user behavior data; determining a target ratio based on the first recommendation index and the second recommendation index; The podcast recommendation list is determined from the first podcast recommendation resource and the second podcast recommendation resource based on the target ratio.

7. The podcast information recommendation method according to claim 6, characterized in that: The determining the podcast recommendation list from the first podcast recommendation resource and the second podcast recommendation resource based on the target ratio includes: Determine a third podcast recommendation resource from the first podcast recommendation resource and the second podcast recommendation resource based on the target ratio; the third podcast recommendation resource includes a plurality of podcast programs; Determine a priority sequence of multiple podcast programs in the third podcast recommendation resource based on the third podcast recommendation resource and the user behavior data; The podcast recommendation list is determined based on the priority sequence.

8. A podcast information recommendation device, characterized in that: include: A first calculation module, configured to determine a first podcast recommendation type based on user behavior data; The user behavior data includes podcast behavior data; A second calculation module, configured to determine a target frequency based on the global podcast data and the podcast behavior data, and determine a second podcast recommendation type based on the target frequency; A third calculation module is used to determine a podcast recommendation list based on the first podcast recommendation type and the second podcast recommendation type.

9. An intelligent device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images