A video recommendation method, apparatus, device and medium

Abstract

This invention discloses a video recommendation method, apparatus, device, and medium. The method includes: acquiring target user features and video features; wherein the target user features include target user category tags, and the video features include video tags; inputting the target user features and the video features into a video recommendation model to obtain an initial video recommendation list; and refining the initial video recommendation list based on the target user category tags and the video tags to obtain a target video recommendation list. This technical solution, by refining the initial video recommendation list based on target user category tag and video tag data, can recommend content suitable for the healthy growth of target users.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a video recommendation method, apparatus, device, and medium. Background Technology

[0002] As families with young children continue to grow, nearly half of the streaming traffic on television and large screens is children's content. Children are the main audience in front of the television, and children's video content is an important way for them to acquire information, develop emotional intelligence, and cultivate abilities. While nearly half of the streaming traffic on television and large screens is children's content, this is problematic because children have not yet developed good discernment and self-control.

[0003] The animation market is currently chaotic, and there are no clear evaluation standards for animation in China. Children aged 0-12 are in a developmental stage, and if their viewing behavior is not properly prevented, they may become addicted to unhealthy content while watching animation. Summary of the Invention

[0004] This invention provides a video recommendation method, apparatus, device, and medium that refines an initial video recommendation list based on target user category tags and video tag data, and can recommend content suitable for the healthy growth of target users.

[0005] According to one aspect of the present invention, a video recommendation method is provided, comprising:

[0006] Obtain target user features and video features; wherein, the target user features include target user category tags, and the video features include video tags;

[0007] Input the target user features and the video features into the video recommendation model to obtain an initial video recommendation list;

[0008] The initial video recommendation list is revised based on the target user category tags and the video tags to obtain the target video recommendation list.

[0009] Optionally, the target user category label is represented by the predicted viewing time, which is obtained by the target user's historical viewing time and a set psychological optimization algorithm; the target category label includes a first type of target user, a second type of target user, a third type of target user, and a fourth type of target user, and the predicted viewing time corresponding to the first type of target user, the second type of target user, the third type of target user, and the fourth type of target user gradually decreases.

[0010] Optionally, the video tags include video category tags and video sentiment tags; the initial video recommendation list is modified based on the target user category tags and the video tags, including:

[0011] The initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users are revised based on the video type tags.

[0012] The initial video recommendation list for the fourth type of target users is revised based on the video emotion tags.

[0013] Optionally, the initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users are modified based on the video type tags, including:

[0014] Each video type is determined based on the historical viewing records of the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users.

[0015] Determine whether the proportion of videos of the first preset type in the video types exceeds the first preset threshold;

[0016] If the proportion of videos of the first set type exceeds the first set threshold, then at least a portion of the videos of the first set type in the initial video list will be replaced with videos of the second set type.

[0017] Optionally, the initial video recommendation list for the fourth type of target users is modified based on the video emotion tags, including:

[0018] Based on the historical viewing records of the fourth type of target users, extract videos with positive and negative emotions;

[0019] Determine whether the proportion of the negative emotion videos exceeds a second preset threshold;

[0020] If the proportion of negative emotion videos exceeds the second set threshold, then at least a portion of the negative emotion videos in the initial video recommendation list will be replaced with positive emotion videos.

[0021] Optionally, the target user feature may further include a target user identifier; the video feature may further include a video identifier and basic video information.

[0022] Optionally, acquiring video features includes:

[0023] Emotional identification is performed on each video based on its content to obtain the video emotion tags;

[0024] Text extraction is performed on the video content to obtain the basic information of the video.

[0025] According to another aspect of the present invention, a video recommendation device is provided, comprising:

[0026] The feature acquisition module is used to acquire target user features and video features; wherein, the target user features include target user category tags, and the video features include video tags;

[0027] The initial list acquisition module is used to input the target user features and the video features into the video recommendation model to obtain an initial video recommendation list;

[0028] The target list acquisition module is used to modify the initial video recommendation list based on the target user category tags and the video tags to obtain the target video recommendation list.

[0029] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0030] At least one processor; and

[0031] A memory communicatively connected to the at least one processor; wherein,

[0032] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the video recommendation method according to any embodiment of the present invention.

[0033] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the video recommendation method according to any embodiment of the present invention.

[0034] The technical solution of this invention involves acquiring target user features and video features; wherein the target user features include target user category tags, and the video features include video tags; inputting the target user features and the video features into a video recommendation model to obtain an initial video recommendation list; and refining the initial video recommendation list based on the target user category tags and the video tags to obtain a target video recommendation list. This technical solution, by refining the initial video recommendation list based on target user category tag and video tag data, can recommend content suitable for the healthy growth of target users.

[0035] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0037] Figure 1 This is a flowchart of a video recommendation method provided in Embodiment 1 of the present invention;

[0038] Figure 2 This is a flowchart of a video recommendation method provided according to Embodiment 2 of the present invention;

[0039] Figure 3 This is a schematic diagram of the structure of a video recommendation device according to Embodiment 3 of the present invention;

[0040] Figure 4 This is a schematic diagram of the structure of an electronic device provided according to Embodiment 4 of the present invention. Detailed Implementation

[0041] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0042] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0043] Example 1

[0044] Figure 1This is a flowchart of a video recommendation method according to Embodiment 1 of the present invention. This embodiment is applicable to the situation of recommending videos to children and adolescents. The method can be executed by a video recommendation device, which can be implemented in hardware and / or software, and can be configured in an electronic device with data processing capabilities. Figure 1 As shown, the method includes:

[0045] The target users of this embodiment are children aged 0 to 12. This embodiment analyzes children's psychological characteristics and behavioral data, and uses machine learning algorithms to train a video recommendation model based on this data to achieve more accurate personalized recommendations. This can help parents solve the psychological problems caused by children watching a large number of videos, and the recommendation results give parents more peace of mind. This embodiment can be applied to children's programs on large-screen TVs.

[0046] S110. Obtain target user characteristics and video characteristics.

[0047] The target user characteristics can be understood as the users for whom a recommended video list is generated. Target user characteristics include target user category tags, and video characteristics include video tags. Target users can be children of a specified age, for example, children aged 0 to 12. Target user characteristics can represent the characteristic information of the target user, and may include target user category tags, target user identifiers, and other target user characteristics. Specifically, target user category tags can be understood as tag information that classifies target users into different categories based on their viewing behavior and classification algorithms; target user identifiers can be used to identify the user's identity information, for example, the target user ID information; other target user characteristic information may include the user's region information, etc.

[0048] Video features are used to characterize video information and can include video tags, video identifiers, and basic video information. Specifically, video tags can include video emotion tags and video type tags; the emotion tag can be obtained through video content recognition. The video identifier identifies the video's source and can be a video ID. Basic video information can include the video title, director, and main actors, as well as other information such as the video's place of origin, year, and language.

[0049] In this embodiment, the target users for whom videos need to be recommended and a large amount of video content containing video feature information can be obtained.

[0050] In this embodiment, optionally, obtaining video features includes: performing emotion recognition on each video based on the video content to obtain video emotion tags; and extracting text from the video content to obtain basic video information.

[0051] Here, video content can be understood as the specific content of each video. Video emotion tags can be obtained by identifying the emotional meaning contained in the video content. For example, video emotion tags can be positive emotions such as joy or happiness, and negative emotions such as sadness or anger.

[0052] In this embodiment, emotion recognition can be performed on each video based on its content. By identifying the meaning of the emotion, a corresponding emotion tag can be obtained for each video. Furthermore, in this embodiment, video creators can be required to associate their uploaded video content with corresponding video emotion tags. In this embodiment, video emotion tags can be obtained by directly identifying the video emotions associated by the video creator.

[0053] Text extraction involves recognizing and extracting the text content from the video. In this embodiment, the video content is input into a trained text convolutional network for text extraction to obtain the corresponding basic video information. This configuration allows for the determination of sentiment tags and basic video information for each video based on its content, facilitating the recommendation of content that meets the psychological needs of the target user.

[0054] S120. Input the target user features and video features into the video recommendation model to obtain the initial video recommendation list.

[0055] The video recommendation model can be a pre-trained neural network model. In this embodiment, the video recommendation model can be trained extensively using a dataset based on the user's historical viewing records and user data. During training, the model is continuously optimized by collecting data from various users until it meets the user's video recommendation needs. The structure of the video recommendation model can consist of an embedding layer, two fully connected layers, and a similarity matrix. The initial video recommendation list can be the initial list content output by the video recommendation model. The number of items in the initial list is fixed and can be set according to the user's actual needs. For example, the initial video recommendation list can include recommendation information for 12 videos. In this embodiment, target user features and video features can be input into the video recommendation model to output the corresponding initial video recommendation list content.

[0056] S130. Based on the target user category tags and video tags, the initial video recommendation list is revised to obtain the target video recommendation list.

[0057] The "correction" can be understood as replacing and adjusting the videos in the initial video recommendation list. The target video recommendation list can be the video recommendation list obtained after correcting the initial video recommendation list. In this embodiment, the videos in the initial video recommendation list can be replaced and adjusted according to the target user category tags, video emotion tags, and video type tags to obtain the target video recommendation list. This embodiment can adjust the initial video list for different categories of users based on video emotion characteristics and type characteristics, which can fully consider children's psychological development and achieve healthy guidance for children's mental development.

[0058] The technical solution of this invention involves acquiring target user features and video features; wherein the target user features include target user category tags, and the video features include video tags; inputting the target user features and video features into a video recommendation model to obtain an initial video recommendation list; and refining the initial video recommendation list based on the target user category tags and video tags to obtain a target video recommendation list. This technical solution, by refining the initial video recommendation list based on target user category tag and video tag data, can recommend content suitable for the healthy growth of target users.

[0059] Example 2

[0060] Figure 2 This is a flowchart of a video recommendation method according to Embodiment 2 of the present invention. This embodiment is an optimization based on the above embodiment. Specifically, the optimization includes: video tags including video category tags and video sentiment tags; revising the initial video recommendation list based on the target user category tags and video tags, including: revising the initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users based on video type tags; and revising the initial video recommendation list for the fourth type of target users based on video sentiment tags.

[0061] like Figure 2 As shown, the method includes:

[0062] S210. Obtain target user characteristics and video characteristics.

[0063] The target user features include target user category tags, and the video features include video tags. Video tags include video category tags and video sentiment tags.

[0064] In this embodiment, optionally, the target user category label is represented by the predicted viewing time, which is obtained by the target user's historical viewing time and a set psychological optimization algorithm; the target category label includes a first type of target user, a second type of target user, a third type of target user, and a fourth type of target user, and the predicted viewing time corresponding to the first type of target user, the second type of target user, the third type of target user, and the fourth type of target user gradually decreases.

[0065] The psychological optimization algorithm can be set as the Student Psychology Based Optimization Algorithm (SPBO). In this embodiment, target users can be classified according to their historical viewing time, and a large amount of algorithm training is performed for each category of target users to determine the target user category corresponding to the predicted viewing time. In this embodiment, target users are divided into four categories based on predicted viewing time: Category 1, Category 2, Category 3, and Category 4, with the predicted viewing time for these four categories decreasing progressively. It is understood that Category 1 target users have the longest predicted viewing time, Category 2 target users have a shorter predicted viewing time than Category 1, Category 3 target users have a shorter predicted viewing time than Category 2, and Category 4 target users have a shorter predicted viewing time than Category 3.

[0066] For example, the SPBO (Special Programming Optimization) algorithm is set up to assume that children's behavior depends on their psychology. A type of video is categorized into four groups of children based on viewing time: Group 1, Group 2, Group 3, and Group 4. A separate big data algorithm is trained for each group. Specifically, N represents the total number of users on the children's channel platform, and D represents the number of categories (dimensions). and Let represent the viewing time of the first type of child and the viewing time of child i (at the t-th viewing time), and record the viewing time of the first type of child and child i in the variables respectively. and

[0067] Where j = 1, ..., D; i = 1, ..., N. For the (t+1)th viewing, the viewing time for each of the four categories of children can be updated as follows:

[0068] (1) Children in Category I:

[0069] Among them, child i is a child randomly selected from a children's online user group in a certain province; Let represent the viewing duration (t-th viewing) of video j under category i for children; rand is a random number in [0,1]; k is a randomly selected 1 or 2.

[0070] (2) Children in the second category:

[0071]

[0072]

[0073] Among them, X mean,j Let represent the average number of videos across all platforms for the j-th video. The second type of child selects either equation (ii) or (iii) as follows: randomly select two numbers r1 and r2 from (0,1). If r2 < r1, then choose equation (ii); otherwise, choose equation (iii).

[0074] (3) Third category of children:

[0075]

[0076] X mean,j This indicates that the average value was taken, and there was no significant bias in children's preferences across all categories.

[0077] (4) Children in the fourth category:

[0078]

[0079] Among them, X max,j X min,j Let represent the maximum and minimum durations of the j-th video, respectively. The above formula indicates that the maximum and minimum durations are random numbers between the maximum and minimum durations.

[0080] The iterative formulas for the four stages above show that children in the first three categories are influenced by the viewing habits of others (i.e., the intelligent recommendation lists from big data algorithms). The first category of children receives the most recommendations from the video classification algorithm from X people. best,j Impact; Children in the second category may receive X best,j The impact may also be affected by general video X. mean，j The third type of child is only affected by X mean,j The fourth category of children is not affected by any recommendations; their range is simply randomly selected between the maximum and minimum values.

[0081] In this embodiment, by setting up a category tag for target users based on historical viewing time and a set psychological algorithm, different videos are recommended for different categories of target users. This makes it easier to provide content recommendations that are more in line with the interests, preferences and needs of children, thereby improving user satisfaction and user experience, and providing high-quality services for children.

[0082] S220. Input the target user features and video features into the video recommendation model to obtain the initial video recommendation list.

[0083] S230. Based on video type tags, revise the initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users.

[0084] The video type tag can be the video type associated by the video creator when uploading. For example, the video type can be comedy, action, educational, or heartwarming. This embodiment can modify the initial video lists for the first, second, third, and fourth target users based on the video type tags associated by the video creator when uploading. For example, for children who prefer war content, their psychological development should be fully considered from a scientific perspective. Therefore, the war content type in the initial video recommendation list can be replaced with teamwork or heartwarming animation products, which can reduce the possibility of their preference for violence and achieve healthy guidance for children's mental development.

[0085] In this embodiment, optionally, the initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users are modified based on video type tags, including: determining each video type based on the historical viewing records of the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users; determining whether the proportion of videos of the first set type in the video types exceeds a first set threshold; if the proportion of videos of the first set type exceeds the first set threshold, then at least some of the videos of the first set type in the initial video list are replaced with videos of the second set type.

[0086] In this context, video type can be understood as the type characteristic corresponding to the video content. In this embodiment, each video type can be determined based on the historical viewing records of different target users. The first set type can be any video type, such as war or science / education. The proportion of videos of the first set type can be understood as the percentage of videos of the first set type in the historical viewing records of each video type. The first set threshold can be a pre-set threshold, which can be set according to actual needs; for example, the first set threshold can be set to 90%. The second set type can be any video type, such as war or science / education. The first set type and the second set type are either different from each other or opposite types. In this embodiment, the number of at least some of the first set types can be determined based on the proportion of videos of the first set type; if the proportion of videos of the first set type is large, then the number of videos of at least some of the first set types will be large.

[0087] In this embodiment, it can be determined whether the proportion of videos of a first predetermined type in each video category exceeds a first predetermined threshold. If the proportion of videos of the first predetermined type exceeds the first predetermined threshold, at least a portion of the videos of the first predetermined type in the initial video list can be replaced with video content of a second predetermined type, thereby obtaining a revised video recommendation list. Through this setting, the initial recommendation list for different categories of target users can be revised based on the video type, thereby reducing children's preference for a particular type of video and guiding children's viewing behavior in a more scientific way.

[0088] S240. The initial video recommendation list for the fourth type of target users is revised based on video emotion tags.

[0089] In this embodiment, the viewing time for the fourth type of target users is the shortest. The initial video recommendation list for this user group may lack a clear direction. Therefore, the initial video recommendation list for the fourth type of target users can be modified based on video sentiment tags. In this embodiment, other categories of target users can also be modified based on video sentiment tags as needed.

[0090] In this embodiment, optionally, the initial video recommendation list for the fourth type of target users is modified based on video emotion tags, including: extracting positive emotion videos and negative emotion videos based on the historical viewing records of the fourth type of target users; determining whether the proportion of negative emotion videos exceeds a second preset threshold; if the proportion of negative emotion videos exceeds the second preset threshold, then replacing at least some of the negative emotion videos in the initial video recommendation list with positive emotion videos.

[0091] In this context, positive emotion videos can be understood as videos containing positive or optimistic emotions. Negative emotion videos can be understood as videos containing negative or pessimistic emotions. The proportion of negative emotion videos can be understood as the percentage of negative emotion videos in the historical viewing records. The second set threshold can be a pre-set threshold, which can be set according to actual needs; for example, the second set threshold can be 60%. The number of at least some negative emotion videos in this embodiment can be determined according to the second set threshold. If the value of the second set threshold is larger, the number of at least some negative emotion videos will be greater, which can also be determined according to actual needs.

[0092] Furthermore, in this embodiment, different click behaviors of child users will generate positive and negative feedback in real time, which can affect the video recommendation results and achieve more personalized intelligent recommendations. These different click behaviors can include actions such as liking or saving.

[0093] In this embodiment, positive and negative emotion videos can be extracted based on the historical viewing records of the fourth type of target users. When the proportion of negative emotion videos exceeds a second preset threshold, at least a portion of the negative emotion videos in the initial video recommendation category can be replaced with positive emotion videos. This embodiment, through this setting, can adjust the user's mindset by replacing negative emotion videos with positive emotion videos when the proportion of a certain emotion, especially negative emotion, in the target user's preferences shows an increasing trend. This helps alleviate the user's persistent negative "self-stimulation" and achieves healthy guidance for children's mental development.

[0094] The technical solution of this embodiment modifies the recommendation list based on video type tags or emotion tags for different categories of child users. This can prevent young children from being exposed to inappropriate animated videos and avoid negative impacts on their value formation. It can also help school-aged children filter out children's animations to match their mental development and guide children's viewing behavior in a more scientific way. This can play a positive role in children's value establishment and personality development, and better serve the children's user group on large screens.

[0095] S250, Obtain the target video recommendation list.

[0096] The technical solution of this invention involves acquiring target user features and video features. The target user features include target user category tags, and the video features include video tags. These features are then input into a video recommendation model to obtain an initial video recommendation list. The initial video recommendation lists for the first, second, third, and fourth types of target users are then revised based on video type tags. Finally, the initial video recommendation list for the fourth type of target user is revised based on video emotion tags to obtain the target video recommendation list. This technical solution, by revising the initial video recommendation list based on target user category tag and video tag data, can recommend content suitable for the healthy growth of target users.

[0097] Example 3

[0098] Figure 3 This is a schematic diagram of a video recommendation device according to Embodiment 3 of the present invention.

[0099] like Figure 3 As shown, the device includes:

[0100] The feature acquisition module 310 is used to acquire target user features and video features.

[0101] Among them, target user features include target user category tags, and video features include video tags.

[0102] The initial list acquisition module 320 is used to input the target user features and video features into the video recommendation model to obtain the initial video recommendation list.

[0103] The target list acquisition module 330 is used to refine the initial video recommendation list based on the target user category tags and video tags to obtain the target video recommendation list.

[0104] Optionally, the target user category label is represented by the predicted viewing time, which is obtained by the target user's historical viewing time and a set psychological optimization algorithm. The target category label includes the first type of target user, the second type of target user, the third type of target user, and the fourth type of target user, with the predicted viewing time corresponding to the first type of target user, the second type of target user, the third type of target user, and the fourth type of target user gradually decreasing.

[0105] Optionally, video tags include video category tags and video mood tags.

[0106] The target list includes module 330, which includes:

[0107] The first correction unit is used to correct the initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users based on video type tags.

[0108] The second correction unit is used to correct the initial video recommendation list for the fourth type of target users based on video sentiment tags.

[0109] Optionally, the first correction unit is specifically used for:

[0110] The video types are determined based on the historical viewing records of the first, second, third, and fourth target users;

[0111] Determine whether the proportion of videos of the first set type in the video types exceeds the first set threshold;

[0112] If the proportion of videos of the first set type exceeds the first set threshold, then at least a portion of the videos of the first set type in the initial video list will be replaced with videos of the second set type.

[0113] Optionally, the second correction unit is specifically used for:

[0114] Extract positive and negative emotion videos based on the historical viewing records of the fourth type of target users;

[0115] Determine whether videos with negative emotions exceed a second set threshold;

[0116] If the number of negative emotion videos exceeds the second set threshold, at least some of the negative emotion videos in the initial video recommendation list will be replaced with positive emotion videos.

[0117] Optionally, target user features may also include target user identifiers; video features may also include video identifiers and basic video information.

[0118] Optionally, the feature acquisition module 310 is specifically used for:

[0119] Emotional identification is performed on each video based on its content to obtain video emotion tags;

[0120] Text extraction is performed on the video content to obtain basic video information.

[0121] The video recommendation device provided in this embodiment of the invention can execute a video recommendation method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method execution.

[0122] Example 4

[0123] Figure 4 This is a schematic diagram of an electronic device according to Embodiment 4 of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0124] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0125] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0126] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as video recommendation methods.

[0127] In some embodiments, the video recommendation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the video recommendation method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the video recommendation method by any other suitable means (e.g., by means of firmware).

[0128] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0129] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0130] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0131] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0132] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0133] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0134] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0135] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A video recommendation method, characterized in that, include: Obtain target user features and video features; wherein, the target user features include target user category tags, and the video features include video tags; Input the target user features and the video features into the video recommendation model to obtain an initial video recommendation list; The initial video recommendation list is modified based on the target user category tags and the video tags to obtain the target video recommendation list; The target user category label is represented by the predicted viewing time, which is obtained by the target user's historical viewing time and a set psychological optimization algorithm; the target user category label includes a first type of target user, a second type of target user, a third type of target user, and a fourth type of target user, and the predicted viewing time corresponding to the first type of target user, the second type of target user, the third type of target user, and the fourth type of target user gradually decreases; The video tags include video category tags and video sentiment tags; the initial video recommendation list is revised based on the target user category tags and the video tags, including: The initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users are revised based on the video category tags. The initial video recommendation list for the fourth type of target users is revised based on the video emotion tags.

2. The method according to claim 1, characterized in that, The initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users are revised based on the video category tags, including: Each video type is determined based on the historical viewing records of the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users. Determine whether the proportion of videos of the first preset type in the video category exceeds the first preset threshold; If the proportion of videos of the first set type exceeds the first set threshold, then at least a portion of the videos of the first set type in the initial video list will be replaced with videos of the second set type.

3. The method according to claim 1, characterized in that, The initial video recommendation list for the fourth type of target users is revised based on the video emotion tags, including: Based on the historical viewing records of the fourth type of target users, extract videos with positive and negative emotions; Determine whether the proportion of the negative emotion videos exceeds a second preset threshold; If the proportion of negative emotion videos exceeds the second set threshold, then at least a portion of the negative emotion videos in the initial video recommendation list will be replaced with positive emotion videos.

4. The method according to claim 1, characterized in that, The target user features also include a target user identifier; the video features also include a video identifier and basic video information.

5. The method according to claim 4, characterized in that, The acquisition of video features includes: Emotional identification is performed on each video based on its content to obtain video emotion tags; Text extraction is performed on the video content to obtain the basic information of the video.

6. A video recommendation device, characterized in that, include: The feature acquisition module is used to acquire target user features and video features; wherein, the target user features include target user category tags, and the video features include video tags; The initial list acquisition module is used to input the target user features and the video features into the video recommendation model to obtain an initial video recommendation list; The target list acquisition module is used to modify the initial video recommendation list based on the target user category tags and the video tags to obtain the target video recommendation list; The target user category label is represented by the predicted viewing time, which is obtained by the target user's historical viewing time and a set psychological optimization algorithm; the target user category label includes a first type of target user, a second type of target user, a third type of target user, and a fourth type of target user, and the predicted viewing time corresponding to the first type of target user, the second type of target user, the third type of target user, and the fourth type of target user gradually decreases; The target list acquisition module includes: The first correction unit is used to correct the initial video recommendation lists for the first type of target users, the second type of target users, the third type of target users, and the fourth type of target users based on video category tags; The second correction unit is used to correct the initial video recommendation list for the fourth type of target users based on video sentiment tags.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the video recommendation method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the video recommendation method according to any one of claims 1-5.

Images