Big data-based network television program intelligent recommendation method and system

Abstract

The present application relates to the technical field of data processing, in particular to a network movie program intelligent recommendation method and system based on big data, comprising: obtaining the number of viewers of each network movie program, the total time length of each user watching network movie programs, the number of different network movie programs watched in each viewing time period and the time length of watching each network movie program, obtaining the network movie program watching enthusiasm degree of each user, the popular degree of each network movie program and the updating recommendation degree of each user for each network movie program; according to the updating recommendation degree of each user for each network movie program, providing network movie program recommendation for each user. Thus, the present application determines the recommendation degree of network movie programs for each user by taking into account the user's own interests and the current popular content by analyzing the network movie program watching enthusiasm activity degree of the user and the popular degree of each network movie program, thereby improving the accuracy of network movie program recommendation.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method and system for intelligent recommendation of online film and television programs based on big data. Background Technology

[0002] The online film and television program recommendation field is in a phase of rapid development. As users' demand for personalized content continues to increase, personalized recommendations have become the mainstream trend for online film and television platforms. Major platforms analyze users' viewing history, preferences, ratings, and favorites data, using recommendation algorithms to provide personalized recommendations for online film and television programs. This aims to offer more accurate, personalized, and diversified recommendation services to meet users' growing needs. Collaborative filtering algorithms are commonly used for recommending online film and television programs to meet the need for recommending programs that users may be interested in. However, because algorithms often tend to recommend programs similar to those previously selected by users, ignoring the novelty and diversity of content, the recommended content may become overly personalized. This can limit users' perspectives and viewpoints, causing them to miss content that may be different from their previous interests but equally appealing. To some extent, this restricts the diversity of users' understanding of different types of online film and television programs, resulting in inappropriate recommendations. Summary of the Invention

[0003] This invention provides a method and system for intelligent recommendation of online film and television programs based on big data, in order to solve existing problems.

[0004] The intelligent recommendation method and system for online film and television programs based on big data of the present invention adopts the following technical solution:

[0005] One embodiment of the present invention provides an intelligent recommendation method for online film and television programs based on big data, the method comprising the following steps:

[0006] Within a preset time period, obtain the number of viewers for each online film and television program each day, the total viewing time of each user for online film and television programs each day, the number of different online film and television programs watched by each user in each viewing time period each day, and the viewing time of each online film and television program.

[0007] The user's enthusiasm for watching online movies and TV shows is calculated based on the total daily viewing time of each user, the number of different online movies and TV shows watched during each viewing period each day, and the viewing time of each online movie and TV show.

[0008] The popularity of each online film and television program is determined by the number of viewers it receives each day.

[0009] Based on each user's enthusiasm and activity in watching online movies and TV shows and the popularity of each online movie and TV show, the update recommendation rate for each online movie and TV show for each user is obtained.

[0010] Based on the update recommendation rate of each online film and television program for each user, online film and television program recommendations are provided to each user.

[0011] Furthermore, the process of determining each user's online film and television viewing enthusiasm based on their total daily viewing time, the number of different online film and television programs watched during each viewing session, and the duration of each online film and television program viewed includes the following specific steps:

[0012] The normalized value of the number of different online film and television programs watched by the i-th user during the nth viewing time in the most recent day is denoted as the threshold of the number of viewing time during the nth viewing time in the most recent day for the i-th user.

[0013] The user's enthusiasm for watching online movies and TV shows is calculated based on the total daily viewing time of each user, the threshold number of viewing sessions per user in the most recent day, and the viewing time of each online movie and TV show.

[0014] Furthermore, the specific formula for calculating each user's enthusiasm for watching online video programs, based on each user's total daily viewing time of online video programs, the threshold number of viewing sessions per user in the most recent day, and the viewing time of each online video program, is as follows:

[0015]

[0016] Among them, Q i T represents the activity level of the i-th user in watching online movies and TV shows. i Let N represent the total viewing time of the i-th user on online video programs in the most recent day, where T represents the duration of the day, and N represents the total viewing time of the i-th user. i m represents the number of times the i-th user watched the video during the most recent day. i,n This represents the threshold for the number of times the i-th user views a content during the n-th time period within the most recent day. This represents the average duration of all online video programs watched by the i-th user during the nth viewing period within the most recent day, where max(t) represents the average duration of all online video programs watched by the i-th user. i,n ) represents the maximum duration of all online video programs watched by the i-th user during the nth viewing period in the most recent day, and norm() represents the linear normalization function.

[0017] Furthermore, the specific steps for determining the popularity of each online film and television program based on the number of viewers per day are as follows:

[0018] Within a preset time period, the average number of viewers of the j-th online film and television program across all days is recorded as the average threshold of the j-th online film and television program.

[0019] The popularity of each online film and television program is determined by the number of viewers and the average threshold for each program per day.

[0020] Furthermore, the specific formula for calculating the popularity of each online film and television program based on the number of viewers and the average threshold per day is as follows:

[0021]

[0022] Among them, E j This indicates the popularity of the j-th online film and television program; Represents the average threshold of the j-th online film and television program; K represents the number of days within the preset time period, r j,k+1 Let r represent the number of viewers of the j-th online video program on the (k+1)th day within the preset time period; j,k This represents the number of viewers of the j-th online video program on the k-th day within a preset time period; norm() represents the linear normalization function.

[0023] Furthermore, the specific steps involved in determining the update recommendation level for each online film and television program for each user based on each user's enthusiasm and activity level in watching online film and television programs and the popularity of each online film and television program are as follows:

[0024] Using traditional collaborative filtering algorithms, we obtain the recommendation score for each online movie and TV show for each user;

[0025] The update recommendation rate for each online film and television program is calculated based on the recommendation rate of each user for each program, the user's enthusiasm and activity in watching online film and television programs, and the popularity of each program.

[0026] Furthermore, the specific formula for calculating the update recommendation rate of each online film and television program for each user, based on the recommendation rate of each user for each online film and television program, the viewing enthusiasm and activity level of each user for each online film and television program, and the popularity of each online film and television program, is as follows:

[0027] G i,j =Q i *norm(X i,j )+(1-Q i )*E j

[0028] Among them, G i,jQ represents the update recommendation rate for the j-th online film and television program for the i-th user; i X represents the activity level of the i-th user in watching online movies and TV shows; i,j E represents the recommendation score for the j-th online movie / TV show for the i-th user. j represents the popularity of the j-th online film and television program; norm() represents the linear normalization function.

[0029] Furthermore, the specific steps for providing online film and television program recommendations to each user based on the updated recommendation rate of each online film and television program are as follows:

[0030] A recommendation sequence is obtained based on the update recommendation rate of each online film and television program for each user;

[0031] Based on the obtained recommendation ranking, online film and television programs are recommended to users in sequence.

[0032] Furthermore, the specific steps involved in obtaining a recommendation sequence based on the updated recommendation score for each online film and television program for each user are as follows:

[0033] The recommendation scores for all online movies and TV shows for each user will be sorted from highest to lowest to obtain a recommendation score sequence.

[0034] This invention also proposes an intelligent recommendation system for online film and television programs based on big data, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program stored in the memory to implement the steps of the aforementioned intelligent recommendation method for online film and television programs based on big data.

[0035] The beneficial effects of the technical solution of the present invention are:

[0036] In this embodiment of the invention, within a preset time period, the number of viewers for each online film and television program per day, the total viewing time for each user per day, the number of different online film and television programs watched by each user during each viewing period per day, and the viewing time of each online film and television program are obtained. Based on the total viewing time for each user per day, the number of different online film and television programs watched by each user during each viewing period per day, and the viewing time of each online film and television program, the viewing enthusiasm of each user is obtained. Based on the analysis of the viewing time and quantity of online film and television programs watched by each user per day, the viewing habits and preferences of users are understood, and the viewing enthusiasm of each user is determined. Based on the number of viewers for each online film and television program per day, the popularity of each online film and television program is obtained. Based on the analysis of the number of viewers for each online film and television program, the current popularity of the program is determined, improving the accuracy of online film and television program recommendations for each user. Based on each user's activity level in watching online movies and TV shows and the popularity of each program, an update recommendation level for each user is determined. This analysis of each user's activity level and popularity further improves the accuracy of online movie and TV show recommendations for each user. Based on the update recommendation level for each user, online movie and TV show recommendations are provided to each user. Thus, this invention, by analyzing users' activity level in watching online movies and TV shows and the popularity of each program, determines the recommendation level for each user's online movies and TV shows, taking into account both their personal interests and currently popular content, thereby improving the accuracy of online movie and TV show recommendations. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart illustrating the steps of the intelligent recommendation method for online film and television programs based on big data according to the present invention. Detailed Implementation

[0039] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the intelligent recommendation method and system for online film and television programs based on big data proposed in accordance with the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0041] The following description, in conjunction with the accompanying drawings, details the specific scheme of the intelligent recommendation method and system for online film and television programs based on big data provided by this invention.

[0042] Please see Figure 1 The diagram illustrates a flowchart of a big data-based intelligent recommendation method for online film and television programs according to an embodiment of the present invention. The method includes the following steps:

[0043] Step S001: Within a preset time period, obtain the number of viewers for each online film and television program each day, the total viewing time of each user for online film and television programs each day, the number of different online film and television programs watched by each user in each viewing time period each day, and the viewing time of each online film and television program.

[0044] In the database of an online video streaming website, data is collected on a preset time period, including the number of viewers for each online video program each day, the total viewing time for each user each day, the number of different online video programs viewed by each user during each viewing period each day, and the viewing time for each online video program.

[0045] It should be noted that: the preset time period is the most recent week, each user's daily viewing time is from the start to the end of each viewing session, the viewing time is allocated to each day based on the viewing start time, the number of viewers is the number of different user IDs, and the viewing duration is in hours. This will be used as an example for the following description.

[0046] Step S002: Based on each user's total daily viewing time of online video programs, the number of different online video programs watched by each user during each viewing period each day, and the viewing time of each online video program, obtain each user's enthusiasm for watching online video programs.

[0047] The total time each user spends watching online movies and TV shows each day determines their activity level. Higher activity levels indicate stronger evidence-based analysis of the user's past viewing history and more accurate viewing preferences. However, this accuracy only reflects the user's previous interests and cannot identify content that might be different from their previous interests but still attract them. Therefore, we analyze each user's enthusiasm by examining their switching frequency while watching online movies and TV shows. A higher switching frequency suggests that the user may have less enthusiasm for previously interesting content, indicating a certain level of boredom. Thus, we determine each user's online movie and TV show viewing enthusiasm and activity level based on their activity and enthusiasm.

[0048] The normalized value of the number of different online film and television programs watched by the i-th user during the nth viewing period in the most recent day is denoted as the threshold of the number of viewing periods in the nth viewing period in the most recent day for the i-th user. Here, sigmoid{} is used to normalize the number of different online film and television programs watched by the i-th user during the nth viewing period in the most recent day.

[0049] It should be noted that sigmoid{} represents the normalization function.

[0050] Taking the i-th user as an example, the formula for calculating the i-th user's online movie and TV show viewing enthusiasm and activity level is:

[0051]

[0052] Among them, Q i T represents the activity level of the i-th user in watching online movies and TV shows. i Let N represent the total viewing time of online video programs by the i-th user in the most recent day, where T represents the length of a day, i.e., 24 hours; i m represents the number of times the i-th user watched the video during the most recent day. i,n This represents the threshold for the number of times the i-th user views a content during the n-th time period within the most recent day. This represents the average duration of all online video programs watched by the i-th user during the nth viewing period within the most recent day, where max(t) represents the average duration of all online video programs watched by the i-th user. i,n ) represents the maximum duration of all online video programs watched by the i-th user during the nth viewing period in the most recent day, and norm() represents the linear normalization function.

[0053] It should be noted that: This represents the online video viewing activity of the i-th user, serving as an adjustment coefficient for the overall formula. It compares the total viewing time of the i-th user in the most recent day to the total viewing time of the day. The larger the ratio, the more time the i-th user spends watching online video programs each day, indicating higher activity. Therefore, the i-th user is more affected by overly personalized online video program recommendations, potentially impacting their enthusiasm for watching online video programs. If the i-th user did not watch any online video programs in the most recent day, that day is not analyzed. This represents the average duration of all online video programs watched by the i-th user during the nth viewing period within the most recent day. The average duration does not include the maximum duration of all online video programs watched by the i-th user during the nth viewing period within the most recent day.

[0054] This represents the viewing enthusiasm of the i-th user for online movies and TV shows, serving as the baseline value for the entire formula. Since users may watch online movies and TV shows at different times of the day, browsing and switching between videos to determine if the content is of interest to them, the video with the longest viewing time corresponds to the user's final selected program. During the browsing and switching process, the percentage of viewing time for other online movies and TV shows relative to the final selected program reflects the user's enthusiasm for the system-recommended programs. The fewer online movies and TV shows a user browses (indicating fewer switches), and the smaller the percentage of the average viewing time for other online movies and TV shows relative to the final selected program, the higher the user's enthusiasm for the system-recommended programs.

[0055] User activity level reflects the total time each user spends watching online movies and TV shows each day. The longer the total time, the higher the user's enthusiasm for the online movies and TV shows recommended by the system. Enthusiasm reflects the user's interest in the content recommended by the system. Therefore, by analyzing each user's online movie and TV show viewing activity and enthusiasm, we can determine each user's online movie and TV show viewing enthusiasm and activity level.

[0056] Following the above method, this embodiment can obtain the online video program viewing enthusiasm and activity level of each user.

[0057] Step S003: Calculate the popularity of each online film and television program based on the number of viewers per day.

[0058] For each online film and television program, its content quality and reputation significantly impact its popularity. The constant emergence and iteration of online film and television programs also cause different programs to reach varying levels of popularity at different stages. This is most directly reflected in metrics such as the number of viewers, the level of discussion surrounding the topic, and the surge in viewership. The popularity of an online film and television program determines the current level of public interest in it. Analyzing the popularity of online film and television programs helps in subsequent recommendations to users, especially those who find the system's recommendations uninteresting. Therefore, by analyzing the number of daily viewers for each online film and television program over the past week, we can determine its popularity.

[0059] Within a preset time period, the average number of viewers of the j-th online film and television program across all days is recorded as the average threshold of the j-th online film and television program.

[0060] Taking the j-th online film and television program as an example, the formula for calculating the popularity of the j-th online film and television program is:

[0061]

[0062] Among them, E j This indicates the popularity of the j-th online film and television program; represents the average threshold value of the j-th online film and television program; K represents the number of days within the preset time period, i.e., 7 days; r j,k+1 Let r represent the number of viewers of the j-th online video program on the (k+1)th day within the preset time period; j,k This represents the number of viewers of the j-th online video program on the k-th day within the preset time period;

[0063] norm() represents the linear normalization function.

[0064] It should be noted that: This represents the average threshold of the j-th online film and television program. The popularity of the j-th online film and television program is determined by using the average number of viewers within a preset time period as a benchmark. Since the popularity of online film and television programs generally shows a certain trend, the popularity trend of the j-th online film and television program is determined by comparing the number of viewers of the j-th online film and television program each day within the preset time period with the number of viewers the following day.

[0065] Following the above method, this embodiment can obtain the popularity of each online film and television program.

[0066] Step S004: Based on each user's enthusiasm and activity in watching online movies and TV shows and the popularity of each online movie and TV show, obtain the update recommendation rate for each online movie and TV show for each user.

[0067] For each user, a recommendation score is assigned to each online movie / TV show recommended using traditional collaborative filtering algorithms. However, traditional collaborative filtering may limit the recommendations to only a portion of the user's interests, ignoring other content the user might be interested in. Popular online movies / TV shows represent the overall preferences of the general public. Therefore, by analyzing each user's viewing activity level, we can determine their level of liking for the recommended shows using traditional collaborative filtering. Higher viewing activity indicates a higher level of liking for the recommended shows, thus assigning greater weight to these shows; conversely, lower viewing activity indicates a lower level of liking, thus assigning less weight to the popularity characteristics of the recommended shows.

[0068] Using traditional collaborative filtering algorithms, we obtain the recommendation score for each online movie / TV show for each user.

[0069] It should be noted that traditional collaborative filtering algorithms are well-known technologies, and the specific methods will not be introduced here.

[0070] Taking the recommendation of the j-th online film and television program to the i-th user as an example, the formula for calculating the update recommendation score of the j-th online film and television program to the i-th user is as follows:

[0071] G i,j =Q i *norm(X i,j )+(1-Q i )*E j

[0072] Among them, G i,j Q represents the update recommendation rate of the j-th online film and television program for the i-th user; i X represents the activity level of the i-th user in watching online movies and TV shows; i,j E represents the recommendation score for the j-th online film and television program for the i-th user. j This represents the popularity of the j-th online film and television program; norm() represents the linear normalization function.

[0073] It should be noted that: Q i This represents the activity level of the i-th user in watching online movies and TV shows, expressed through Q. iAs a weight for the overall formula, the greater the activity level of the i-th user in watching online movies and TV shows, the greater the weight given to the online movies and TV shows recommended by the traditional collaborative filtering algorithm; the smaller the activity level of the i-th user in watching online movies and TV shows, the greater the weight given to the popularity characteristics of the current online movies and TV shows, so as to avoid making users feel bored and dull due to overly personalized recommendations.

[0074] Following the above method, this embodiment can obtain the update recommendation rate for each online film and television program for each user.

[0075] Step S005: Based on the update recommendation rate of each online film and television program for each user, provide online film and television program recommendations for each user.

[0076] Through the above process, the update recommendation rate for each online film and television program for each user is obtained. This rate is determined by analyzing both the user's personal preferences and current popular trends. In this way, the recommendation of online film and television programs for users is no longer affected by excessive personalization, which improves the user's viewing comfort and the adaptability of the system recommendations.

[0077] The recommendation scores for all online movies and TV shows for each user will be sorted from highest to lowest to obtain a recommendation score sequence.

[0078] Online film and television programs are recommended to each user in sequence based on the recommendation level.

[0079] This invention is now complete.

[0080] In summary, in this embodiment of the invention, within a preset time period, the following are obtained: the number of viewers for each online film and television program per day; the total viewing time for each user per day; the number of different online film and television programs watched by each user during each viewing period per day; and the viewing time for each online film and television program. Based on the total viewing time for each user per day, the number of different online film and television programs watched by each user during each viewing period per day, and the viewing time for each online film and television program, the user's enthusiasm for watching online film and television programs is obtained. Based on the number of viewers for each online film and television program per day, the popularity of each online film and television program is obtained. Based on the user's enthusiasm for watching online film and television programs and the popularity of each online film and television program, the update recommendation rate for each online film and television program for each user is obtained. Based on the update recommendation rate for each online film and television program for each user, online film and television program recommendations are provided to each user. Thus, by analyzing users' enthusiasm and activity in watching online film and television programs and the popularity of each online film and television program, this invention determines the recommendation level of online film and television programs for each user, taking into account both their own interests and currently popular content, thereby improving the accuracy of online film and television program recommendations.

[0081] The present invention also provides an intelligent recommendation system for online film and television programs based on big data, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program stored in the memory to implement the steps of the aforementioned intelligent recommendation method for online film and television programs based on big data.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent recommendation of online film and television programs based on big data, characterized in that, The method includes the following steps: Within a preset time period, obtain the number of viewers for each online film and television program each day, the total viewing time of each user for online film and television programs each day, the number of different online film and television programs watched by each user in each viewing time period each day, and the viewing time of each online film and television program. Based on each user's total daily viewing time of online video programs, the number of different online video programs watched by each user during each viewing session per day, and the viewing time of each online video program, the user's online video program viewing enthusiasm and activity level is obtained, including: The user's last day The normalized value of the number of different online video programs watched within the viewing period is denoted as the i-th. The user's last day The threshold for the number of viewing sessions is used to calculate each user's online video viewing activity level. This is based on the total daily viewing time of each user, the threshold for the number of viewing sessions per user in the most recent day, and the viewing time of each online video program. The specific calculation formula is as follows: in, Indicates the first The level of enthusiasm and activity of individual users in watching online movies and TV shows. Indicates the first Total time spent watching online movies and TV shows by each user in the past day Indicates the length of a day. Indicates the first The number of times a user watched content during the most recent day. Indicates the first The user's last day The threshold for the number of viewing sessions. Indicates the first The user's last day The average duration of all online video programs watched within a single viewing period. Indicates the first The user's last day The maximum duration of watching all online video programs within a single viewing period. Represents a linear normalization function; The popularity of each online film and television program is determined by the number of viewers it receives each day. Based on each user's engagement with online movies and TV shows and the popularity of each program, the update recommendation score for each user is calculated. This includes: using a traditional collaborative filtering algorithm to obtain the recommendation score for each user; and using the recommendation score, engagement with online movies and TV shows, and popularity of each program to calculate the update recommendation score for each user. The specific calculation formula is as follows: in, Indicates the first The user performed the first The update recommendation rate of online film and television programs; Indicates the first The level of enthusiasm and activity of individual users in watching online movies and TV shows; Indicates the first The user performed the first Recommendation rate of online film and television programs; Indicates the first The popularity of online film and television programs; Represents a linear normalization function; Based on the update recommendation rate of each online film and television program for each user, online film and television program recommendations are provided to each user.

2. The intelligent recommendation method for online film and television programs based on big data according to claim 1, characterized in that, The process of determining the popularity of each online video program based on the number of viewers per day includes the following specific steps: Within the preset time period, the first The average number of viewers for a given online video program over all days is denoted as the [number]. The average threshold of online film and television programs; The popularity of each online film and television program is determined by the number of viewers and the average threshold for each program per day.

3. The intelligent recommendation method for online film and television programs based on big data according to claim 2, characterized in that, The specific formula for calculating the popularity of each online film and television program based on the number of viewers and the average threshold per day is as follows: in, Indicates the first The popularity of online film and television programs; Indicates the first The average threshold of online film and television programs; This indicates the number of days within a preset time period. Indicates the first The first online film and television program within the preset time period The number of viewers per day; Indicates the first The first online film and television program within the preset time period The number of viewers per day; This represents the linear normalization function.

4. The intelligent recommendation method for online film and television programs based on big data according to claim 1, characterized in that, The specific steps involved in providing online film and television program recommendations to each user based on the updated recommendation rate for each online film and television program are as follows: A recommendation sequence is obtained based on the update recommendation rate of each online film and television program for each user; Based on the obtained recommendation ranking, online film and television programs are recommended to users in sequence.

5. The intelligent recommendation method for online film and television programs based on big data according to claim 4, characterized in that, The process of obtaining a recommendation sequence based on the updated recommendation score for each online film and television program for each user includes the following specific steps: The recommendation scores for all online movies and TV shows for each user will be sorted from highest to lowest to obtain a recommendation score sequence.

6. A big data-based intelligent recommendation system for online film and television programs, comprising a memory, a processor, and a computer program stored on the memory and capable of running on the processor, characterized in that: When the computer program is executed by the processor, it implements the steps of the intelligent recommendation method for online film and television programs based on big data as described in any one of claims 1-5.

Images