Recommended playlist sorting method and device, storage medium and electronic equipment

Abstract

The invention relates to a sorting method and device of recommended playlists, a storage medium and electronic equipment. The method comprises the steps of obtaining a target to-be-recommended playlist set; determining a first to-be-recommended playlist in the target to-be-recommended playlist set as a first recommended playlist of the recommended playlist set after the first sorting; performing duplicate removal and sorting on the to-be-recommended playlist set used in the Nth sorting to obtain the to-be-recommended playlist set after the Nth duplicate removal and sorting; determining the first to-be-recommended broadcast list in the to-be-recommended broadcast list set after the Nth duplicate removal and sorting as the Nth recommended broadcast list in the recommended broadcast list set after the Nth sorting; and updating N = N + 1, and executing N + 1th sorting until N is equal to the target threshold value, thereby obtaining a recommended broadcast list set after the last sorting. The technical problem that the diversity of recommendation results cannot be ensured while the recommendation efficiency is ensured by an existing playlist sorting strategy is solved.

Description

Technical Field

[0001] This application relates to the field of intelligent analysis technology, and in particular, to a method, device, storage medium, and electronic device for sorting recommended playlists. Background Art

[0002] In the video playlist recommendation scenario, about 3 playlists are displayed per screen, and each playlist contains about 6 videos. Playlist recommendation not only needs to consider the user's preference for videos, but also takes into account the user's preference for playlists. Since there may be a large number of duplicate videos between playlists of the same theme, traditional video recommendation sorting is likely to cause the exposure of duplicate content, affecting the user experience. Therefore, in playlist recommendation, effective deduplication and sorting strategies need to be adopted to improve diversity and relevance. The existing sorting and diversity control solutions include three steps: First, sort the videos within the playlist first, and then sort the playlists; Second, the deduplication operation is carried out in sequence according to the playlist sorting, that is, each subsequent playlist needs to be deduplicated with the 6 videos of the previously selected playlists; Third, re-score and sort after deduplication to ensure that there are no duplicates among the 6 videos of each playlist. However, the deduplication order in the second step has a great impact on the final result. The playlists deduplicated first have a smaller overlap with the previous playlists and retain more content, while the subsequent playlists need to be deduplicated with more previous playlists, which may result in a lower score. Therefore, the deduplication order has a great impact on the final recommendation result. The playlists deduplicated first retain more content, and the subsequent playlists may lose more content, resulting in a lower score; In addition, a balance needs to be found between diversity and efficiency, and the combined strategy of deduplication and sorting needs to be optimized. Summary of the Invention

[0003] This application provides a method, device, storage medium, and electronic device for sorting recommended playlists to solve the technical problem that the existing playlist sorting strategy cannot ensure the diversity of the recommendation result while ensuring the recommendation efficiency.

[0004] In a first aspect, the present application provides a method for sorting recommended playlists, including: obtaining a set of target playlists to be recommended, where the set of target playlists to be recommended contains M playlists to be recommended, M is a positive integer, and each playlist to be recommended includes at least one object; obtaining the set of target playlists to be recommended used for the first sorting, determining the first playlist to be recommended in the set of target playlists to be recommended as the first recommended playlist in the recommended playlist set after the first sorting, and determining all playlists to be recommended in the set of target playlists to be recommended except the first playlist to be recommended as the set of playlists to be recommended after the first sorting; obtaining the recommended playlist set and the set of playlists to be recommended used for the Nth sorting, and de-duplicating and sorting the set of playlists to be recommended used for the Nth sorting according to the recommended playlist set used for the Nth sorting, to obtain the set of playlists to be recommended after the Nth de-duplication and sorting, where N is an integer greater than 1 and less than M, the recommended playlist set used for the Nth sorting is the recommended playlist set after the (N - 1)th sorting, and the set of playlists to be recommended used for the Nth sorting is the set of playlists to be recommended after the (N - 1)th sorting; determining the first playlist to be recommended in the set of playlists to be recommended after the Nth de-duplication and sorting as the Nth recommended playlist in the recommended playlist set after the Nth sorting, and determining all playlists to be recommended in the set of playlists to be recommended after the Nth de-duplication and sorting except the first playlist to be recommended as the set of playlists to be recommended after the Nth sorting; updating N = N + 1, and performing the (N + 1)th sorting until N is equal to the target threshold, to obtain the recommended playlist set after the last sorting.

[0005] Second aspect, the present application provides a sorting device for recommended playlists, including: an acquisition module, configured to acquire a set of target playlists to be recommended, where the set of target playlists to be recommended contains M playlists to be recommended, M is a positive integer, and each of the playlists to be recommended includes at least one object; a first sorting module, configured to acquire the set of target playlists to be recommended used for the first sorting, determine the first playlist to be recommended in the set of target playlists to be recommended as the first recommended playlist in the recommended playlist set after the first sorting, and determine all playlists to be recommended in the set of target playlists to be recommended except the first playlist to be recommended as the set of playlists to be recommended after the first sorting; a second sorting module, configured to acquire the recommended playlist set and the playlist set to be recommended used for the Nth sorting, and perform deduplication and sorting on the playlist set to be recommended used for the Nth sorting according to the recommended playlist set used for the Nth sorting, to obtain the playlist set to be recommended after the Nth deduplication and sorting, where N is an integer greater than 1 and less than M, the recommended playlist set used for the Nth sorting is the recommended playlist set after the (N - 1)th sorting, and the playlist set to be recommended used for the Nth sorting is the playlist set to be recommended after the (N - 1)th sorting; a third sorting module, configured to determine the first playlist to be recommended in the playlist set to be recommended after the Nth deduplication and sorting as the Nth recommended playlist in the recommended playlist set after the Nth sorting, and determine all playlists to be recommended in the playlist set to be recommended after the Nth deduplication and sorting except the first playlist to be recommended as the playlist set to be recommended after the Nth sorting; a fourth sorting module, configured to update N = N + 1, and perform the (N + 1)th sorting until N is equal to the target threshold, to obtain the recommended playlist set after the last sorting.

[0006] As an optional example, the acquisition module includes: a scoring sub-module, configured to input the M playlists to be recommended into a sorting model, so that the sorting model scores each of the playlists to be recommended and scores each object of each of the playlists to be recommended, to obtain the recommended score of each of the playlists to be recommended and the recommended score of each object of each of the playlists to be recommended; a first sorting sub-module, configured to sort the M playlists to be recommended in descending order according to the recommended score of each of the playlists to be recommended, and sort all objects of each of the playlists to be recommended in descending order according to the recommended score of each object of each of the playlists to be recommended, to obtain the set of target playlists to be recommended.

[0007] As an alternative example, the second sorting module described above includes: a first acquisition sub-module, configured to acquire a deduplication rule, where the deduplication rule includes complete non-duplication and partial duplication; a first deduplication sub-module, configured to, when the deduplication rule is complete non-duplication, deduplicate the set of playlists to be recommended used in the Nth sorting based on the set of recommended playlists used in the Nth sorting, to obtain the set of playlists to be recommended after the Nth deduplication, where each playlist to be recommended in the set of playlists to be recommended after the Nth deduplication does not contain all objects in the set of recommended playlists used in the Nth sorting; a second sorting sub-module, configured to input the set of playlists to be recommended after the Nth deduplication into the sorting model, so that the sorting model sorts the set of playlists to be recommended after the Nth deduplication, to obtain the set of playlists to be recommended after the Nth deduplication and sorting.

[0008] As an alternative example, the second sorting module further includes: a second acquisition sub-module, configured to, after acquiring the deduplication rule, when the deduplication rule is partial duplication, acquire the optimal deduplicated playlist for each playlist to be recommended in the set of playlists to be recommended used in the Nth sorting, to obtain the set of playlists to be recommended after the Nth deduplication; a third sorting sub-module, configured to input the set of playlists to be recommended after the Nth deduplication into the sorting model, so that the sorting model sorts the set of playlists to be recommended after the Nth deduplication, to obtain the set of playlists to be recommended after the Nth deduplication and sorting.

[0009] As an alternative example, the second acquisition sub-module includes: a processing unit, configured to use each playlist to be recommended in the set of playlists to be recommended used in the Nth sorting as a first playlist to be recommended, and perform the following operations on the first playlist to be recommended: acquire P duplicate objects in the first playlist to be recommended and the set of recommended playlists used in the Nth sorting, where P is a positive integer; delete Q target objects in the first playlist to be recommended to obtain a second playlist to be recommended, and calculate the optimization score of the second playlist to be recommended, where Q is a positive integer less than or equal to P, and the target object is the object with the lowest recommendation score among the duplicate objects; determine the second playlist to be recommended with the highest optimization score as the optimal deduplicated playlist of the first playlist to be recommended.

[0010] As an alternative example, the processing unit is further configured to calculate the optimization score of the second playlist to be recommended through the following formula: S = x * f(d) - y * d; where S is the optimization score of the second playlist to be recommended, x is a first preset value, d is the duplication degree between the second playlist to be recommended and the set of recommended playlists used in the Nth sorting, and y is a second preset value.

[0011] As an optional example, the above processing unit is further configured to, before calculating the optimization score of the second to-be-recommended playlist, obtain all duplicate objects of the second to-be-recommended playlist and the set of recommended playlists used in the Nth sorting; obtain the object duplication degree of each of the above duplicate objects and the set of recommended playlists used in the Nth sorting; calculate the sum of all the above object duplication degrees to obtain the duplication degree of the second to-be-recommended playlist and the set of recommended playlists used in the Nth sorting.

[0012] In a third aspect, the present application provides a storage medium storing a computer program, wherein the computer program, when run by a processor, executes the above sorting method of the recommended playlist.

[0013] In a fourth aspect, the present application further provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the above sorting method of the recommended playlist through the computer program.

[0014] The above technical solutions provided by the embodiments of the present application have the following advantages compared with the prior art:

[0015] This application adopts a method of obtaining a target set of playlists to be recommended. Among them, the above-mentioned target set of playlists to be recommended contains M playlists to be recommended, where M is a positive integer, and each of the above-mentioned playlists to be recommended includes at least one object; obtain the above-mentioned target set of playlists to be recommended used for the first sorting, determine the first playlist to be recommended in the above-mentioned target set of playlists to be recommended as the first recommended playlist in the recommended playlist set after the first sorting, and determine all playlists to be recommended in the above-mentioned target set of playlists to be recommended except the first playlist to be recommended as the set of playlists to be recommended after the first sorting; obtain the recommended playlist set and the playlist set to be recommended used for the Nth sorting, and perform deduplication and sorting on the above-mentioned playlist set to be recommended used for the Nth sorting according to the above-mentioned recommended playlist set used for the Nth sorting, to obtain the playlist set to be recommended after the Nth deduplication and sorting, where N is an integer greater than 1 and less than M, the above-mentioned recommended playlist set used for the Nth sorting is the recommended playlist set after the (N - 1)th sorting, and the above-mentioned playlist set to be recommended used for the Nth sorting is the playlist set to be recommended after the (N - 1)th sorting; determine the first playlist to be recommended in the above-mentioned playlist set to be recommended after the Nth deduplication and sorting as the Nth recommended playlist in the recommended playlist set after the Nth sorting, and determine all playlists to be recommended in the above-mentioned playlist set to be recommended after the Nth deduplication and sorting except the first playlist to be recommended as the playlist set to be recommended after the Nth sorting; update N = N + 1, and perform the (N + 1)th sorting until N is equal to the target threshold, to obtain the recommended playlist set after the last sorting. Since in the above method, the playlist recommendation is optimized through multiple rounds of iterative sorting and deduplication. First, the first playlist is selected from the target set of playlists to be recommended and sorted, and the remaining playlists are used as the playlist set to be recommended. Next, in each round of sorting, through deduplication and re - sorting operations, it is ensured that the objects of each playlist have less repetition and are sorted according to user preferences. During each round of deduplication, the current playlist set to be recommended is deduplicated with the content of the selected playlists, and re - sorted according to the deduplicated set, gradually optimizing the recommendation result until the predetermined number of sorting rounds is reached, to obtain the final recommendation result, thereby achieving the purpose of improving the quality of the recommendation result, avoiding premature removal of high - quality playlists, providing a more accurate and diverse recommendation experience, and further solving the technical problem that the existing playlist sorting strategy cannot ensure the diversity of the recommendation result while ensuring the recommendation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings here are incorporated into the specification and form a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0017] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] One or more embodiments are exemplarily illustrated by the pictures in the corresponding drawings. These exemplary illustrations do not limit the embodiments. Elements with the same reference numerals in the drawings are represented as similar elements. Unless otherwise stated, the drawings in the figures do not constitute a proportional limitation.

[0019] Figure 1 is a flowchart of an optional method for sorting recommended playlists according to an embodiment of the present application;

[0020] Figure 2 is an overall implementation schematic diagram of an optional method for sorting recommended playlists according to an embodiment of the present application;

[0021] Figure 3 is a schematic structural diagram of an optional device for sorting recommended playlists according to an embodiment of the present application;

[0022] Figure 4 is a schematic diagram of an optional electronic device according to an embodiment of the present application. Detailed implementation manners

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of them. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0024] The following disclosure provides many different embodiments or examples for implementing different structures of the present application. To simplify the disclosure of the present application, the components and settings of specific examples are described below. Of course, they are only examples and are not intended to limit the present application. In addition, the present application may repeat reference numerals and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0025] According to the first aspect of the embodiments of the present application, an optional method for sorting recommended playlists is provided. Optionally, as Figure 1 shown, the above method includes:

[0026] S102. Obtain a set of target playlists to be recommended, where the set of target playlists to be recommended contains M playlists to be recommended, M is a positive integer, and each playlist to be recommended includes at least one object.

[0027] S104. Obtain the set of target playlists to be recommended used for the first sorting, determine the first playlist to be recommended in the set of target playlists to be recommended as the first recommended playlist in the recommended playlist set after the first sorting, and determine all playlists to be recommended in the set of target playlists to be recommended except the first playlist to be recommended as the set of playlists to be recommended after the first sorting.

[0028] S106. Obtain the set of recommended playlists and the set of playlists to be recommended used for the Nth sorting, and perform deduplication and sorting on the set of playlists to be recommended used for the Nth sorting according to the set of recommended playlists used for the Nth sorting, to obtain the set of playlists to be recommended after the Nth deduplication and sorting, where N is an integer greater than 1 and less than M, the set of recommended playlists used for the Nth sorting is the set of recommended playlists after the (N - 1)th sorting, and the set of playlists to be recommended used for the Nth sorting is the set of playlists to be recommended after the (N - 1)th sorting.

[0029] S108. Determine the first playlist to be recommended in the set of playlists to be recommended after the Nth deduplication and sorting as the Nth recommended playlist in the set of recommended playlists after the Nth sorting, and determine all playlists to be recommended in the set of playlists to be recommended after the Nth deduplication and sorting except the first playlist to be recommended as the set of playlists to be recommended after the Nth sorting.

[0030] S110. Update N = N + 1, and perform the (N + 1)th sorting until N is equal to the target threshold, to obtain the set of recommended playlists after the last sorting.

[0031] Optionally, in this embodiment, a recommendation system for cyclic sorting and deduplication is involved, which is used to generate a finally recommended playlist set, and gradually removes duplicate content and optimizes the recommendation order through multiple rounds of iteration to ensure that the finally recommended playlist is both diverse and in line with user preferences. Specifically, first, a set of target playlists to be recommended containing M playlists to be recommended needs to be obtained, where M is a positive integer, and each playlist can contain at least one video object as the content to be recommended.

[0032] During the first sorting, select the first playlist from the set of target playlists to be recommended and use it as the first recommended playlist in the recommended playlist set after the first sorting. Then, the remaining playlists in the set of target playlists to be recommended are determined as the set of playlists to be recommended after the first sorting. The set of target playlists to be recommended has been sorted according to a certain standard (which can be video content, playlist type, etc.), and the position of each playlist is fixed.

[0033] When entering the Nth sorting (N is greater than 1 and less than M), the set of playlists to be recommended in the Nth sorting is deduplicated and re-sorted according to the set of recommended playlists after the N-1th sorting. The specific steps are: except for some and / or all objects in the set of playlists to be recommended that are repeated with the selected playlists. At this time, the previously selected playlists will affect the sorting of subsequent playlists to prevent repeated exposure. In the set of playlists to be recommended after deduplication, the playlists are rearranged according to the sorting rules. After sorting, the first playlist in the set of recommended playlists after the Nth sorting will be determined as the Nth recommended playlist and removed from the set of playlists to be recommended after deduplication.

[0034] After each sorting, N increases by 1 and enters the next round of sorting. This process is repeated until N is equal to the target threshold, that is, the predetermined number of sorting is reached and the target threshold is less than or equal to M. After all iterations and deduplication, the final recommended playlist set will try to avoid duplicate videos and be sorted by user preference priority.

[0035] Optionally, in this embodiment, if Figure 2 As shown in the schematic diagram of the arrangement implementation, first, the target set of playlists to be recommended is obtained. In the first sorting, the first optimal playlist in the target set of playlists to be recommended is used as the first recommended playlist (playlist 1), and this playlist is deleted from the target set of playlists to be recommended. In the second sorting, the updated target set of playlists to be recommended is deduplicated and sorted, the optimal playlist is taken as the second recommended playlist (playlist 2), and this playlist is deleted from the updated target set of playlists to be recommended. Similarly, in the Nth sorting, the target set of playlists to be recommended updated in the last sorting is deduplicated and sorted, the optimal playlist is taken as the Nth recommended playlist (playlist N), and finally N recommended playlists are obtained.

[0036] Optionally, in this embodiment, each round of sorting performs a deduplication operation on the set of recommended playlists to ensure that there are as few duplicate objects as possible between the videos and playlists in each playlist, thereby avoiding redundant information in the user experience. Through each round of sorting and deduplication operations, the diversity of the recommendation results is gradually ensured, and users are prevented from seeing playlists with the same theme and similar objects multiple times, thereby improving the user experience. Each round of sorting is optimized based on user preferences to ensure that the final recommended playlist is not only highly relevant, but also meets diversity requirements and enhances the personalized recommendation effect. Since the order of deduplication operations is performed step by step, more playlist options can be retained in the early stage to avoid losing potential high-quality recommended content due to premature deduplication. The iterative process of multiple rounds of deduplication and sorting helps to better balance diversity and relevance in playlist recommendations, and provide a more accurate and diversified user experience.

[0037] As an optional example, obtaining a target set of playlists to be recommended includes:

[0038] Input M playlists to be recommended into a sorting model, so that the sorting model scores each playlist to be recommended and scores each object in each playlist to be recommended, obtaining the recommendation score of each playlist to be recommended and the recommendation score of each object in each playlist to be recommended;

[0039] Sort the M playlists to be recommended in descending order according to the recommendation scores of each playlist to be recommended, and sort all the objects in each playlist to be recommended in descending order according to the recommendation scores of each object in each playlist to be recommended, obtaining a set of target playlists to be recommended.

[0040] Optionally, in this embodiment, first, input the M playlists to be recommended in the set of target playlists to be recommended into a sorting model. The sorting model is a trained sorting model. The sorting model will score each playlist to be recommended according to specific user preferences and content features, generating the recommendation score of each playlist. At the same time, the sorting model will score each object (video or other content) in each playlist and generate the recommendation score of each object. Sort the M playlists to be recommended in descending order according to the recommendation scores of each playlist, ensuring that playlists with higher scores are displayed first. For each playlist, sort the objects in each playlist in descending order according to the recommendation scores of the respective objects included therein, so as to ensure that the content within each playlist is also displayed according to user preferences. Finally, according to the above sorting results, obtain a set of target playlists to be recommended, where the playlists and their objects have been prioritized according to the user's preferences.

[0041] Optionally, in this embodiment, through the scoring and sorting model, the user's interests and preferences can be more accurately reflected, ensuring that the recommended playlists and videos highly match the user's needs. Not only sort according to the overall recommendation scores of the playlists, but also sort each object within the playlists, thereby achieving more refined personalized recommendations and enhancing the user experience. Through reasonable sorting and optimization, the finally recommended playlists and content are more in line with the user's interests, improving the user's satisfaction and stickiness to the platform.

[0042] As an optional example, de-duplicate and sort the set of playlists to be recommended used in the Nth sorting according to the set of playlists to be recommended used in the Nth sorting, and the set of playlists to be recommended after de-duplication and sorting in the Nth time includes:

[0043] Obtain a de-duplication rule, where the de-duplication rule includes complete non-duplication and partial duplication;

[0044] In the case where the deduplication rule is complete non - repetition, the set of candidate playlists to be recommended used in the N - th sorting is deduplicated according to the set of recommended playlists used in the N - th sorting, and the set of candidate playlists to be recommended after the N - th deduplication is obtained. Among them, each candidate playlist in the set of candidate playlists to be recommended after the N - th deduplication does not contain all the objects in the set of recommended playlists used in the N - th sorting;

[0045] The set of candidate playlists to be recommended after the N - th deduplication is input into the sorting model so that the sorting model sorts the set of candidate playlists to be recommended after the N - th deduplication, and the set of candidate playlists to be recommended after the N - th deduplication and sorting is obtained.

[0046] Optionally, in this embodiment, the deduplication rules are divided into two types: complete non - repetition and partial repetition. Complete non - repetition requires that all objects of each playlist in the set of candidate playlists to be recommended are completely different from all objects in the set of recommended playlists, that is, there are no duplicate objects. Partial repetition allows duplicates, but a certain deduplication strategy still needs to be maintained between the set of recommended playlists and the set of candidate playlists to be recommended. In the case of complete non - repetition, the following steps are required for deduplication:

[0047] First, according to the set of recommended playlists used in the N - th sorting, a deduplication operation is performed on the set of candidate playlists to be recommended used in the N - th sorting. Specifically, it is necessary to ensure that each playlist in the set of candidate playlists to be recommended does not contain any content that is the same as any object in the set of recommended playlists. That is to say, each playlist in the set of candidate playlists to be recommended should be completely different from the set of recommended playlists. After deduplication, the set of candidate playlists to be recommended after the N - th deduplication is obtained, and all objects in each playlist will not be repeated with any object in the already selected set of recommended playlists. The set of candidate playlists to be recommended after the N - th deduplication is input into the sorting model, and the sorting model will re - score and sort these deduplicated playlists according to the user's interests and preferences. The sorting model generates a set of candidate playlists to be recommended after the N - th deduplication and sorting based on the characteristics and recommendation scores of each playlist. This set has been optimized according to the deduplication rules and conforms to the user's preferences.

[0048] Optionally, in this embodiment, through the deduplication rule of complete non - repetition, it can be ensured that the content displayed in each recommended playlist will not be repeated with the content of the previous playlists, avoiding redundant exposure and improving the user experience. The deduplicated set of playlists is re - sorted and can be accurately recommended according to the user's interests and preferences, ensuring that the displayed content is closer to the user's needs. Through the deduplication operation, it is possible to avoid displaying too much content with the same theme or video type, thereby increasing the diversity of the recommended content and maintaining the user's freshness and exploration interest. Two deduplication methods, complete non - repetition and partial repetition, are provided, which can be flexibly adjusted according to different scenarios and requirements, ensuring both the diversity of the content and avoiding the problem of too few recommended contents caused by excessive deduplication.

[0049] As an optional example, after obtaining the deduplication rule, the above method further includes:

[0050] In the case where the deduplication rule is partial duplication, obtain the optimal deduplicated playlist for each recommended playlist in the set of recommended playlists to be used for the Nth sorting, and obtain the set of recommended playlists after the Nth deduplication;

[0051] Input the set of recommended playlists after the Nth deduplication into the sorting model, so that the sorting model sorts the set of recommended playlists after the Nth deduplication to obtain the set of recommended playlists after the Nth deduplication and sorting.

[0052] Optionally, in this embodiment, under the deduplication rule of partial duplication, it is not required to completely avoid duplication, but it is allowed that some objects overlap between the recommended playlists. However, in order to maximize diversity, an optimal deduplication selection will be made for each recommended playlist. That is, for each recommended playlist, an optimal playlist (with the least duplication and most in line with user interests) will be selected to be added to the final recommended set. To achieve this goal, the content overlap degree between each recommended playlist and the set of already selected recommended playlists will be evaluated, and the optimal deduplicated playlist will be selected according to the overlap degree and the weight of user preferences. The selection of these optimal deduplicated playlists ensures that within a certain duplication range, the diversity and relevance of the recommended content are the strongest. After the selection of the optimal deduplicated playlist, the set of recommended playlists after the Nth deduplication is obtained. In this set, the content included in each recommended playlist overlaps with the content of the playlists that have been recommended before, but the overlap degree is controlled within a certain range to ensure the diversity and freshness of the recommended set. Input the set of recommended playlists after the Nth deduplication into the sorting model, and the sorting model will re-rate and sort these playlists according to the user's preferences, playlist content, and other features. The sorting model will sort the deduplicated playlist set according to the recommended score of each playlist to obtain the set of recommended playlists after the Nth deduplication and sorting. This step ensures that each playlist is optimally sorted according to the user's needs, and finally generates the recommended result.

[0053] Optionally, in this embodiment, through the partial duplication deduplication strategy, the occurrence of duplicate content can be reduced, but the content that the user is interested in is retained without completely eliminating duplication, thereby enhancing the diversity of recommendations. Allowing partial duplication but controlling the duplication degree through the selection of the optimal playlist ensures that the recommended content not only conforms to the user's preferences but also has a certain degree of diversity, avoiding the content from being too single or identical. The partial duplication deduplication provides flexible control over the overlap degree, and the allowed duplication degree can be adjusted according to different recommendation scenarios, which not only ensures the relevance of the recommended content but also avoids information overload or excessive similarity of content. After the optimization of the deduplication and sorting model, the finally recommended playlists are more in line with the user's interests, and effectively avoid excessive duplication, improving the user experience.

[0054] As an optional example, obtaining the deduplicated optimal playlist for each recommended playlist in the set of recommended playlists used for the Nth sorting includes:

[0055] Taking each recommended playlist in the set of recommended playlists used for the Nth sorting as the first recommended playlist, and performing the following operations on the first recommended playlist:

[0056] Obtaining P duplicate objects in the first recommended playlist that are in the set of recommended playlists used for the Nth sorting, where P is a positive integer;

[0057] Deleting Q target objects in the first recommended playlist to obtain a second recommended playlist, and calculating the optimization score of the second recommended playlist, where Q is a positive integer less than or equal to P, and the target object is the object with the lowest recommendation score among the duplicate objects;

[0058] Determining the deduplicated optimal playlist of the first recommended playlist as the second recommended playlist with the highest optimization score.

[0059] Optionally, in this embodiment, each recommended playlist (denoted as the first recommended playlist) in the set of recommended playlists used for the Nth sorting is processed, and the content to be deduplicated is selected by comparing with the set of recommended playlists. For each recommended playlist, check its duplication degree with the objects in the set of recommended playlists used for the Nth sorting, and find P duplicate objects in the set of recommended playlists, where P is a positive integer representing the number of duplicate objects. Next, Q target objects (where Q ≤ P) will be selected from the first recommended playlist. These target objects are the objects with the lowest recommendation scores among the duplicate objects. That is to say, the duplicate objects will be sorted according to the recommendation scores of each object, and those duplicate objects with lower recommendation scores will be selected for deletion. After deleting the duplicate objects, a second recommended playlist is obtained, which has higher content diversity than the original first recommended playlist because it has removed some duplicate videos or objects. For example, when Q is equal to 1, one object with the lowest recommendation score is deleted to obtain a second recommended playlist. When Q is equal to 2, two objects with the lowest recommendation scores are deleted to obtain a second recommended playlist, and so on until Q is equal to P, obtaining multiple second recommended playlists. For all the second recommended playlists after deleting the duplicate objects, their optimization scores will be recalculated. The optimization score takes into account factors such as user preferences and the diversity of playlist content. The goal is to make the second recommended playlist maintain a high recommendation value while meeting the deduplication requirements. By comparing the optimization scores of all the optimized second recommended playlists, the second recommended playlist with the highest optimization score is selected as the deduplicated optimal playlist of the first recommended playlist. Thus, the deduplicated optimal playlist for each recommended playlist in the set of recommended playlists used for the Nth sorting can be obtained.

[0060] Optionally, in this embodiment, through the deduplication operation, duplicate low-scoring content is removed, ensuring that the content in the recommended playlist is more diverse while still meeting the user's interests. By selecting some duplicate objects for deletion, the intensity of deduplication can be flexibly adjusted, taking into account both the diversity of content and the relevance of recommendations. The calculation of the optimized score ensures that each deduplicated playlist can appear in the recommendation list with higher quality and in a manner that meets the user's interests.

[0061] As an optional example, calculating the optimized score of the second playlist to be recommended includes:

[0062] The optimized score of the second playlist to be recommended is calculated by the following formula:

[0063] S = x * f(d) - y * d;

[0064] Where S is the optimized score of the second playlist to be recommended, x is the first preset value, d is the duplication degree of the second playlist to be recommended with the set of recommended playlists used for the Nth sorting, and y is the second preset value.

[0065] Optionally, in this embodiment, the purpose of calculating the optimized score of the second playlist to be recommended is to quantitatively evaluate the deduplicated playlist, so as to determine its priority in the recommendation system.

[0066] Specifically, the following formula is used to calculate the optimized score:

[0067] S = x * f(d) - y * d;

[0068] Let \(S\) be the optimization score of the second playlist to be recommended, which reflects the recommendation value of the playlist after deduplication. The higher the score, the more in line with the user's interests the playlist is, and a balance is achieved between diversity and relevance. Let \(x\) be the first preset value, which is a weight coefficient used to adjust the influence of \(f(d)\). Here, \(f(d)\) represents the efficiency value, and \(x\) represents the adjustment coefficient of the efficiency value, which can control the diversity of recommendations. \(f(d)\) is the duplication function, that is, the efficiency value, used to calculate the diversity of the playlist. Let \(d\) be the duplication degree of the second playlist to be recommended and other playlists in the set of recommended playlists used for the \(N\)th sorting. The higher the duplication degree, the larger the value of \(d\). The function \(f(d)\) returns a corresponding value according to the size of \(d\), usually decreasing, indicating that the higher the duplication degree, the lower the diversity of the playlist. Let \(y\) be the second preset value, which is another weight coefficient used to adjust the influence of the duplication degree \(d\), representing the cost of deduplication. A high duplication degree may lead to a decline in the user experience. Therefore, the larger \(y\) is, the greater the cost of deduplication, that is, when \(d\) is higher, the optimization score \(S\) will be lower. Here, \(d\) represents the duplication degree of the second playlist to be recommended and other playlists in the set of recommended playlists used for the \(N\)th sorting. The higher the duplication degree, the more similar or identical content there is between the playlists, which may lead to a decline in the user experience.

[0069] Optionally, in this embodiment, by calculating the optimization score, a balance can be achieved between the diversity of deduplication and the trade-off of duplication degree, so that the recommendation results are neither redundant nor can they meet the relevance of user preferences. The \(x\) and \(y\) coefficients used in the formula can be adjusted according to different scenarios, enabling the system to flexibly adjust the recommendation strategy according to different needs and optimize the user experience. By considering the influence of the duplication degree and the content quality after deduplication, the most suitable playlist can be accurately selected to ensure that the recommended content is both diverse and relevant, improving user satisfaction and engagement.

[0070] As an optional example, before calculating the optimization score of the second playlist to be recommended, the above method further includes:

[0071] Obtain all duplicate objects of the second playlist to be recommended and the set of recommended playlists used for the \(N\)th sorting;

[0072] Obtain the object duplication degree of each duplicate object and the set of recommended playlists used for the \(N\)th sorting;

[0073] Calculate the sum of all object duplication degrees to obtain the duplication degree of the second playlist to be recommended and the set of recommended playlists used for the \(N\)th sorting.

[0074] Optionally, in this embodiment, all duplicate objects of the second playlist to be recommended and the set of recommended playlists used in the Nth sorting are obtained. A duplicate object refers to the same video or content that appears in the second playlist to be recommended and the set of recommended playlists used in the Nth sorting. To find duplicate objects, matching is performed based on the unique identifier of the video or other features (such as title, ID, etc.). For example, assume the second playlist to be recommended has the following videos: [Video A, Video B, Video C, Video D], and the set of recommended playlists used in the Nth sorting contains [Video A, Video D, Video E, Video D, Video F]. Then Video A and Video D are duplicate objects. For each duplicate object, calculate its duplication degree with other objects in the set of recommended playlists used in the Nth sorting. The duplication degree of duplicate object Video A is 1, and the duplication degree of duplicate object Video D is 2. Calculate the sum of the duplication degrees of all objects, and the duplication degree of the second playlist to be recommended and the set of recommended playlists used in the Nth sorting is 3. This total duplication degree reflects the similarity and overlap between the second playlist to be recommended and the set of recommended playlists used in the Nth sorting. The higher the duplication degree, the more similar the content in the second playlist to be recommended is to the content in the set of recommended playlists, which may reduce diversity and affect the user's perception.

[0075] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to this application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0076] According to another aspect of the embodiments of the present application, there is also provided a sorting device for recommended playlists, as Figure 3 shown, including:

[0077] An obtaining module 302, configured to obtain a set of target playlists to be recommended, where the set of target playlists to be recommended contains M playlists to be recommended, M is a positive integer, and each playlist to be recommended includes at least one object;

[0078] A first sorting module 304, configured to obtain the set of target playlists to be recommended used in the first sorting, determine the first playlist to be recommended in the set of target playlists to be recommended as the first recommended playlist in the set of recommended playlists after the first sorting, and determine all playlists to be recommended in the set of target playlists to be recommended except the first playlist to be recommended as the set of playlists to be recommended after the first sorting;

[0079] The second sorting module 306 is configured to obtain the recommended playlist set and the to-be-recommended playlist set used for the Nth sorting, and de-duplicate and sort the to-be-recommended playlist set used for the Nth sorting according to the recommended playlist set used for the Nth sorting, so as to obtain the to-be-recommended playlist set after the Nth de-duplication and sorting, where N is an integer greater than 1 and less than M, the recommended playlist set used for the Nth sorting is the recommended playlist set after the (N - 1)th sorting, and the to-be-recommended playlist set used for the Nth sorting is the to-be-recommended playlist set after the (N - 1)th sorting;

[0080] The third sorting module 308 is configured to determine the first to-be-recommended playlist in the to-be-recommended playlist set after the Nth de-duplication and sorting as the Nth recommended playlist in the recommended playlist set after the Nth sorting, and determine all the to-be-recommended playlists except the first to-be-recommended playlist in the to-be-recommended playlist set after the Nth de-duplication and sorting as the to-be-recommended playlist set after the Nth sorting;

[0081] The fourth sorting module 310 is configured to update N = N + 1, and perform the (N + 1)th sorting until N is equal to the target threshold, so as to obtain the recommended playlist set after the last sorting.

[0082] It should be noted that the obtaining module 302 in this embodiment may be used to execute step S102 in the embodiment of the present application, the first calculation module 304 in this embodiment may be used to execute step S104 in the embodiment of the present application, the second calculation module 306 in this embodiment may be used to execute step S106 in the embodiment of the present application, the determining module 308 in this embodiment may be used to execute step S108 in the embodiment of the present application, and the determining module 310 in this embodiment may be used to execute step S110 in the embodiment of the present application.

[0083] As an optional example, the obtaining module includes:

[0084] A scoring sub-module, configured to input M to-be-recommended playlists into a sorting model, so that the sorting model scores each to-be-recommended playlist and each object of each to-be-recommended playlist, to obtain the recommendation score of each to-be-recommended playlist and the recommendation score of each object of each to-be-recommended playlist;

[0085] A first sorting sub-module, configured to sort the M to-be-recommended playlists in descending order according to the recommendation score of each to-be-recommended playlist, and sort all the objects of each to-be-recommended playlist in descending order according to the recommendation score of each object of each to-be-recommended playlist, so as to obtain a target to-be-recommended playlist set.

[0086] As an optional example, the second sorting module includes:

[0087] The first acquisition sub-module is used to acquire the deduplication rules, where the deduplication rules include completely non-duplicate and partially duplicate;

[0088] The first deduplication sub-module is used to deduplicate the set of to-be-recommended playlists used in the Nth sorting according to the set of recommended playlists used in the Nth sorting when the deduplication rule is completely non-duplicate, so as to obtain the set of to-be-recommended playlists after the Nth deduplication, where each to-be-recommended playlist in the set of to-be-recommended playlists after the Nth deduplication does not contain all objects in the set of recommended playlists used in the Nth sorting;

[0089] The second sorting sub-module is used to input the set of to-be-recommended playlists after the Nth deduplication into the sorting model, so that the sorting model sorts the set of to-be-recommended playlists after the Nth deduplication to obtain the set of to-be-recommended playlists after the Nth deduplication and sorting.

[0090] As an optional example, the second sorting module further includes:

[0091] The second acquisition sub-module is used to, after acquiring the deduplication rules, when the deduplication rule is partially duplicate, acquire the optimal deduplicated playlists of each to-be-recommended playlist in the set of to-be-recommended playlists used in the Nth sorting, so as to obtain the set of to-be-recommended playlists after the Nth deduplication;

[0092] The third sorting sub-module is used to input the set of to-be-recommended playlists after the Nth deduplication into the sorting model, so that the sorting model sorts the set of to-be-recommended playlists after the Nth deduplication to obtain the set of to-be-recommended playlists after the Nth deduplication and sorting.

[0093] As an optional example, the second acquisition sub-module includes:

[0094] The processing unit is used to take each to-be-recommended playlist in the set of to-be-recommended playlists used in the Nth sorting as the first to-be-recommended playlist, and perform the following operations on the first to-be-recommended playlist:

[0095] Acquire P duplicate objects in the first to-be-recommended playlist and the set of recommended playlists used in the Nth sorting, where P is a positive integer;

[0096] Delete Q target objects in the first to-be-recommended playlist to obtain the second to-be-recommended playlist, and calculate the optimization score of the second to-be-recommended playlist, where Q is a positive integer less than or equal to P, and the target object is the object with the lowest recommendation score among the duplicate objects;

[0097] Determine the optimal deduplicated playlist of the first to-be-recommended playlist as the second to-be-recommended playlist with the highest optimization score.

[0098] As an optional example, the processing unit is further used to calculate the optimization score of the second to-be-recommended playlist through the following formula:

[0099] S = x * f(d) - y * d;

[0100] Wherein, S is the optimization score of the second to-be-recommended playlist, x is the first preset value, d is the repetition degree between the second to-be-recommended playlist and the recommended playlist set used in the Nth sorting, and y is the second preset value.

[0101] As an optional example, the processing unit is further configured to obtain all duplicate objects between the second to-be-recommended playlist and the recommended playlist set used in the Nth sorting before calculating the optimization score of the second to-be-recommended playlist;

[0102] Obtain the object repetition degree between each duplicate object and the recommended playlist set used in the Nth sorting;

[0103] Calculate the sum of all object repetition degrees to obtain the repetition degree between the second to-be-recommended playlist and the recommended playlist set used in the Nth sorting.

[0104] For other examples of this embodiment, please refer to the above examples and will not be elaborated here.

[0105] Figure 4 is a schematic diagram of an optional electronic device according to an embodiment of the present application, as Figure 4 shown, including a processor 402, a communication interface 404, a memory 406, and a communication bus 408. Among them, the processor 402, the communication interface 404, and the memory 406 complete mutual communication through the communication bus 408, where

[0106] The memory 406 is used to store computer programs;

[0107] The processor 402, when executing the computer program stored on the memory 406, implements the following steps:

[0108] Obtain a target to-be-recommended playlist set, where the target to-be-recommended playlist set contains M to-be-recommended playlists, M is a positive integer, and each to-be-recommended playlist includes at least one object;

[0109] Obtain the target to-be-recommended playlist set used in the first sorting, determine the first to-be-recommended playlist in the target to-be-recommended playlist set as the first recommended playlist in the recommended playlist set after the first sorting, and determine all the to-be-recommended playlists in the target to-be-recommended playlist set except the first to-be-recommended playlist as the to-be-recommended playlist set after the first sorting;

[0110] Obtain the recommended playlist set and the playlist set to be recommended used for the Nth sorting, and deduplicate and sort the playlist set to be recommended used for the Nth sorting according to the recommended playlist set used for the Nth sorting, to obtain the playlist set to be recommended after the Nth deduplication and sorting, where N is an integer greater than 1 and less than M, the recommended playlist set used for the Nth sorting is the recommended playlist set after the (N - 1)th sorting, and the playlist set to be recommended used for the Nth sorting is the playlist set to be recommended after the (N - 1)th sorting;

[0111] Determine the first playlist to be recommended in the playlist set to be recommended after the Nth deduplication and sorting as the Nth recommended playlist in the recommended playlist set after the Nth sorting, and determine all the playlists to be recommended in the playlist set to be recommended after the Nth deduplication and sorting except the first playlist to be recommended as the playlist set to be recommended after the Nth sorting;

[0112] Update N = N + 1, and perform the (N + 1)th sorting until N is equal to the target threshold to obtain the recommended playlist set after the last sorting.

[0113] Optionally, in this embodiment, the above communication bus may be a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. This communication bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 4 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus. The communication interface is used for communication between the above electronic device and other devices.

[0114] The memory may include a RAM, and may also include a non-volatile memory, for example, at least one disk memory. Optionally, the memory may also be at least one storage device located far from the aforementioned processor.

[0115] As an example, the above memory 406 may but is not limited to include the acquisition module 302, the first sorting module 304, the second sorting module 306, the third sorting module 308, and the fourth sorting module 310 in the sorting device of the above recommended playlist. In addition, it may also include but is not limited to other module units in the sorting device of the above recommended playlist, which will not be elaborated in this example.

[0116] The above-mentioned processor can be a general-purpose processor, including but not limited to: CPU (Central Processing Unit, central processing unit), NP (Network Processor, network processor), etc.; it can also be a DSP (Digital Signal Processing, digital signal processor), ASIC (Application Specific Integrated Circuit, application-specific integrated circuit), FPGA (Field-Programmable Gate Array, field-programmable gate array), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0117] Optionally, the specific examples in this embodiment can refer to the examples described in the above-mentioned embodiment, and will not be elaborated herein.

[0118] Those of ordinary skill in the art can understand that Figure 4 The structure shown is only schematic. The device for implementing the sorting method of the above-mentioned recommended playlist can be a terminal device, which can be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a handheld computer, and a mobile Internet device (Mobile Internet Devices, MID), a PAD, and other terminal devices. Figure 4 It does not limit the structure of the above-mentioned electronic device. For example, the electronic device may further include more or fewer components (such as a network interface, a display device, etc.) than those shown in Figure 4 or have a different configuration from that shown in Figure 4 Those shown.

[0119] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above-mentioned embodiments can be completed by instructing the relevant hardware of the terminal device through a program, and the program can be stored in a computer-readable storage medium. The storage medium can include: a flash drive, a ROM, a RAM, a magnetic disk, an optical disk, etc.

[0120] According to another aspect of the embodiments of the present application, there is also provided a computer-readable storage medium, in which a computer program is stored, and when the computer program is run by a processor, it executes the steps in the sorting method of the above-mentioned recommended playlist.

[0121] Optionally, in this embodiment, those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by a program instructing the relevant hardware of the terminal device. The program can be stored in a computer-readable storage medium, and the storage medium can include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, an optical disk, etc.

[0122] The serial numbers of the embodiments of the present application above are only for description and do not represent the superiority or inferiority of the embodiments.

[0123] If the integrated unit in the above embodiments is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in the above computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing one or more computer devices (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application.

[0124] In the above embodiments of the present application, the descriptions of the various embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0125] In the several embodiments provided by the present application, it should be understood that the disclosed client can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some interfaces, and the indirect coupling or communication connection of the units or modules can be in an electrical or other form.

[0126] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0127] In addition, in each embodiment of the present application, each functional unit may be integrated into one processing unit, may exist separately as individual physical units, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware or in the form of software functional units.

[0128] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. A method for sorting a recommended playlist, characterized in that: include: Acquire a target set of playlists to be recommended, wherein the target set of playlists to be recommended includes M playlists to be recommended, M is a positive integer, and each of the playlists to be recommended includes at least one object; Obtain the target to-be-recommended playlist set used for the first sorting, determine the first to-be-recommended playlist in the target to-be-recommended playlist set as the first recommended playlist in the recommended playlist set after the first sorting, and determine all to-be-recommended playlists in the target to-be-recommended playlist set except the first to-be-recommended playlist in the target to-be-recommended playlist set as the to-be-recommended playlist set after the first sorting; Acquire the recommended playlist set and the to-be-recommended playlist set used for the Nth sorting, and perform deduplication and sorting on the to-be-recommended playlist set used for the Nth sorting according to the recommended playlist set used for the Nth sorting, to obtain the to-be-recommended playlist set after the Nth deduplication and sorting, wherein N is an integer greater than 1 and less than M, the recommended playlist set used for the Nth sorting is the recommended playlist set after the N-1th sorting, and the to-be-recommended playlist set used for the Nth sorting is the to-be-recommended playlist set after the N-1th sorting; Determine the first to-be-recommended playlist in the to-be-recommended playlist set after the N-th deduplication and sorting as the N-th recommended playlist in the recommended playlist set after the N-th sorting, and determine all to-be-recommended playlists except the first to-be-recommended playlist in the to-be-recommended playlist set after the N-th deduplication and sorting as the to-be-recommended playlist set after the N-th sorting; Update N=N+1, and perform the N+1th sorting until N is equal to the target threshold, to obtain the recommended playlist set after the last sorting.

2. The method according to claim 1, characterized in that The acquisition of the target set of playlists to be recommended includes: Input the M playlists to be recommended into the sorting model, so that the sorting model scores each of the playlists to be recommended and scores each object of each of the playlists to be recommended, and obtains a recommendation score for each of the playlists to be recommended and a recommendation score for each object of each of the playlists to be recommended; According to the recommendation score of each of the playlists to be recommended, the M playlists to be recommended are sorted in descending order, and according to the recommendation score of each object of each of the playlists to be recommended, all objects of each of the playlists to be recommended are sorted in descending order to obtain the target set of playlists to be recommended.

3. The method according to claim 1, characterized in that Deduplication and sorting of the to-be-recommended playlist set used in the Nth sorting according to the recommended playlist set used in the Nth sorting to obtain the to-be-recommended playlist set after deduplication and sorting for the Nth time comprises: Obtaining a deduplication rule, wherein the deduplication rule includes completely non-repetitive and partially repetitive; In the case where the deduplication rule is completely non-repetitive, deduplication is performed on the set of playlists to be recommended used in the Nth sorting according to the set of recommended playlists used in the Nth sorting to obtain the set of playlists to be recommended after the Nth deduplication, wherein each of the playlists to be recommended in the set of playlists to be recommended after the Nth deduplication does not contain all the objects in the set of recommended playlists used in the Nth sorting; The N-th deduplicated set of playlists to be recommended is input into the sorting model, so that the sorting model sorts the N-th deduplicated set of playlists to be recommended to obtain the N-th deduplicated and sorted set of playlists to be recommended.

4. The method according to claim 3, characterized in that After obtaining the deduplication rule, the method further includes: In the case where the deduplication rule is partial duplication, obtaining the best deduplication playlist for each to-be-recommended playlist in the to-be-recommended playlist set used for the Nth sorting, and obtaining the to-be-recommended playlist set after the Nth deduplication; The N-th deduplicated set of playlists to be recommended is input into the sorting model, so that the sorting model sorts the N-th deduplicated set of playlists to be recommended to obtain the N-th deduplicated and sorted set of playlists to be recommended.

5. The method according to claim 4, characterized in that The step of obtaining the optimal playlist without duplication for each to-be-recommended playlist in the to-be-recommended playlist set used for the Nth sorting includes: Each to-be-recommended playlist in the to-be-recommended playlist set used in the Nth sorting is used as a first to-be-recommended playlist, and the following operations are performed on the first to-be-recommended playlist: Obtain P duplicate objects of the first to-be-recommended playlist set and the recommended playlist set used for the Nth sorting, where P is a positive integer; Delete Q target objects in the first playlist to be recommended to obtain a second playlist to be recommended, and calculate the optimization score of the second playlist to be recommended, wherein Q is a positive integer less than or equal to P, and the target object is the object with the lowest recommendation score among the repeated objects; The second to-be-recommended playlist with the highest optimization score is determined as the optimal playlist after deduplication of the first to-be-recommended playlist.

6. The method according to claim 5, characterized in that The calculating the optimization score of the second to-be-recommended playlist includes: The optimization score of the second playlist to be recommended is calculated by the following formula: S = x*f(d)-y*d; Among them, S is the optimization score of the second playlist to be recommended, x is a first preset value, d is the repetition degree between the second playlist to be recommended and the recommended playlist set used for the Nth sorting, and y is a second preset value.

7. The method according to claim 5, characterized in that Before calculating the optimization score of the second to-be-recommended playlist, the method further includes: Acquire all duplicate objects of the second playlist to be recommended and the recommended playlist set used for the Nth sorting; Obtaining the object repetition degree of each of the repeated objects and the recommended playlist set used for the Nth sorting; The sum of the repetitions of all the objects is calculated to obtain the repetition of the second playlist to be recommended and the recommended playlist set used for the Nth sorting.

8. A device for sorting recommended playlists, characterized in that: include: An acquisition module, used for acquiring a target set of playlists to be recommended, wherein the target set of playlists to be recommended includes M playlists to be recommended, M is a positive integer, and each of the playlists to be recommended includes at least one object; A first sorting module is used to obtain the target set of playlists to be recommended used for the first sorting, determine the first playlist to be recommended in the target set of playlists to be recommended as the first recommended playlist in the recommended playlist set after the first sorting, and determine all the playlists to be recommended in the target set of playlists to be recommended except the first playlist to be recommended as the set of playlists to be recommended after the first sorting; A second sorting module is used to obtain a recommended playlist set and a to-be-recommended playlist set used for the Nth sorting, and to deduplicate and sort the to-be-recommended playlist set used for the Nth sorting according to the recommended playlist set used for the Nth sorting, to obtain the to-be-recommended playlist set after the Nth deduplication and sorting, wherein N is an integer greater than 1 and less than M, the recommended playlist set used for the Nth sorting is the recommended playlist set after the N-1th sorting, and the to-be-recommended playlist set used for the Nth sorting is the to-be-recommended playlist set after the N-1th sorting; A third sorting module is used to determine the first to-be-recommended playlist in the to-be-recommended playlist set after the N-th deduplication and sorting as the N-th recommended playlist in the recommended playlist set after the N-th sorting, and to determine all to-be-recommended playlists except the first to-be-recommended playlist in the to-be-recommended playlist set after the N-th deduplication and sorting as the to-be-recommended playlist set after the N-th sorting; The fourth sorting module is used to update N=N+1 and perform the N+1th sorting until N is equal to the target threshold, thereby obtaining the recommended playlist set after the last sorting.

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

10. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to execute the method according to any one of claims 1 to 7 through the computer program.

Images