Movie and video sheet recommendation method and device, equipment and medium

Abstract

The embodiment of the invention discloses a film and video sheet recommendation method and device, equipment and a medium. The method comprises the following steps: acquiring a mutual enhancement graph model and a to-be-recommended film and video sheet list; wherein the mutual enhancement graph model is composed of a plurality of nodes and a connecting edge between any two nodes; the nodes are used for representing movie and video film lists and users, and a connecting edge between any two nodes is determined based on a historical movie and video film list association relationship, a user association relationship and an association relationship between the users and the film lists; inputting the to-be-recommended film and video sheet list into the mutual enhancement graph model, and determining node features of each to-be-recommended film and video sheet in the mutual enhancement graph model based on a random sampling algorithm; and determining a target recommended film and video film list according to the node features of the to-be-recommended film and video film list. According to the technical scheme, the user interaction data can be mutually enhanced through the mutual enhancement graph model, so that the sheet recommendation result is determined, and more diversified and richer recommendation effects are achieved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method, device, equipment and medium for recommending a film and television playlist. Background Art

[0002] Personalized recommendation algorithms play a vital role in many online services, such as e-commerce, content sharing platforms, and social media. In film and television software, personalized recommendation algorithms can recommend relevant film and television works to users based on their interests and viewing history, thereby improving user experience and platform stickiness. With the continuous increase in film and television content, the amount of information users face has also increased dramatically. Personalized recommendation algorithms are particularly important in helping users find films of interest.

[0003] Currently popular recommendation technologies, such as collaborative filtering recommendation technology and content similarity-based algorithms, can perform intelligent recommendations to a certain extent. However, they still face problems such as data sparsity and cold start problems in actual applications. Moreover, recommendations based purely on content similarity can easily lead to users watching a small number of similar film and television lists, resulting in a lack of diversity in the recommendation results and poor results. Summary of the invention

[0004] The present invention provides a method, device, equipment and medium for recommending film and television playlists, which can mutually enhance user interaction data through a mutual enhancement graph model, thereby determining the playlist recommendation results and achieving more diverse and richer recommendation effects.

[0005] According to one aspect of the present invention, a method for recommending a film and television playlist is provided, comprising:

[0006] Obtaining a mutual enhancement graph model and a list of film and television playlists to be recommended; wherein the mutual enhancement graph model is composed of a plurality of nodes and edges between any two nodes; the nodes are used to represent film and television playlists and users, and the edges between any two nodes are determined based on historical film and television playlist associations, user associations, and user-playlist associations;

[0007] Inputting the list of film and television playlists to be recommended into the mutual enhancement graph model, and determining the node features of each film and television playlist to be recommended in the mutual enhancement graph model based on a random sampling algorithm;

[0008] A target recommended film and television list is determined according to the node features of the film and television list to be recommended.

[0009] Optionally, before obtaining the mutual enhancement graph model and the list of movies and TV shows to be recommended, the following steps are also included:

[0010] Get historical user information and historical film and television list information of a set duration;

[0011] Determine the user association relationship, the film and television list association relationship, and the association relationship between the user and the film list according to the historical user information and the historical film and television list information;

[0012] A mutual enhancement graph model is constructed based on the user association relationship, the film and television list association relationship, and the association relationship between the user and the list.

[0013] Optionally, determining the user association relationship, the film and television list association relationship, and the association relationship between the user and the film list according to the historical user information and the historical film and television list information includes:

[0014] Determine historical user similarity characteristics, movie list similarity characteristics, and historical user viewing behavior characteristics;

[0015] The user association relationship, the film list association relationship and the association relationship between the user and the film list are determined according to the historical user similarity characteristics, the film list similarity characteristics and the historical user viewing behavior characteristics.

[0016] Optionally, constructing a mutual enhancement graph model according to the user association relationship, the film and television playlist association relationship, and the association relationship between the user and the playlist includes:

[0017] Representing the user and the film and television playlist as a user node and a playlist node respectively;

[0018] Representing the user association relationship, the film and television playlist association relationship, and the association relationship between the user and the playlist as edges of the mutual enhancement graph model;

[0019] The mutual enhancement graph model is constructed according to the user nodes, the sheet nodes and the edges.

[0020] Optionally, the node feature is represented by a node weight;

[0021] Inputting the list of films and TV series to be recommended into the mutual enhancement graph model, and determining node features of the films and TV series to be recommended in the mutual enhancement graph model based on a random sampling algorithm, including:

[0022] Starting from each film and television list in the list of film and television lists to be recommended, a random walk is performed to obtain a transfer path;

[0023] The node weight of the to-be-recommended film and television playlist is determined according to the transfer path.

[0024] Optionally, starting from each film and television list in the list of film and television lists to be recommended, a random walk is performed to obtain a transfer path, including:

[0025] Determining the transition probability between nodes of the mutual reinforcement graph model;

[0026] Starting from each film and television list in the list of film and television lists to be recommended, transferring along the node according to the transfer probability, and performing random judgment after the node transfer to obtain random data;

[0027] If the random data meets the end condition, the random walk ends and the transfer path is recorded.

[0028] Optionally, determining a target recommended film and television playlist according to the node features of the to-be-recommended film and television playlist includes:

[0029] Arranging the to-be-recommended film and television list in a set order according to the node weights of the to-be-recommended film and television list to obtain an arrangement result;

[0030] A set number of to-be-recommended film and television playlists in the arrangement result are determined as target recommended film and television playlists.

[0031] According to another aspect of the present invention, there is provided a device for recommending a movie list, comprising:

[0032] A data acquisition module, used to acquire a mutual enhancement graph model and a list of film and television playlists to be recommended; wherein the mutual enhancement graph model is composed of a plurality of nodes and edges between any two nodes; the nodes are used to represent film and television playlists and users, and the edges between any two nodes are determined based on historical film and television playlist associations, user associations, and user-playlist associations;

[0033] A node feature determination module, used for inputting the list of film and television plays to be recommended into the mutual enhancement graph model, and determining the node features of each film and television playlist to be recommended in the mutual enhancement graph model based on a random sampling algorithm;

[0034] The target recommended film and television playlist determination module is used to determine the target recommended film and television playlist according to the node characteristics of the film and television playlist to be recommended.

[0035] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0036] at least one processor; and

[0037] a memory communicatively connected to the at least one processor; wherein,

[0038] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the film and television list recommendation method described in any embodiment of the present invention.

[0039] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for recommending a film and television list described in any embodiment of the present invention when executed.

[0040] The technical solution of the embodiment of the present invention is to obtain a mutually enhanced graph model and a list of film and television lists to be recommended; wherein the mutually enhanced graph model is composed of multiple nodes and edges between any two nodes; the nodes are used to represent film and television lists and users, and the edges between any two nodes are determined based on the historical film and television list association relationship, the user association relationship, and the user-list association relationship; the list of film and television lists to be recommended is input into the mutually enhanced graph model, and the node features of each film and television list to be recommended in the mutually enhanced graph model are determined based on a random sampling algorithm; the target recommended film and television list is determined according to the node features of the film and television list to be recommended. This technical solution can mutually enhance user interaction data through a mutually enhanced graph model, thereby determining the list recommendation results and achieving more diverse and richer recommendation effects.

[0041] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 is a flow chart of a method for recommending a film and television playlist according to the first embodiment of the present invention;

[0044] Figure 2 is a flow chart of a method for recommending a movie list according to Embodiment 2 of the present invention;

[0045] Figure 3 is an example diagram of a mutual enhancement graph model provided according to Embodiment 2 of the present invention;

[0046] Figure 4 is a structural schematic diagram of a device for recommending movie and TV playlists according to Embodiment 3 of the present invention;

[0047] Figure 5 It is a schematic diagram of the structure of an electronic device provided according to Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0048] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0049] It should be noted that the terms "first", "second" and "target" in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. The acquisition, storage, use, processing, etc. of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0050] Embodiment 1

[0051] Figure 1 This is a flow chart of a film and television list recommendation method provided according to the first embodiment of the present invention. This embodiment is applicable to the case of intelligently recommending film and television lists. The method can be executed by a film and television list recommendation device. The film and television list recommendation device can be implemented in the form of hardware and / or software. The film and television list recommendation device can be configured in an electronic device with data processing capabilities. Figure 1 As shown, the method includes:

[0052] The technical solution in this embodiment can be applied to the film and television recommendation function in any film and television software. By building a mutually enhanced graph model through the association between users, users and film lists, and film lists, the recommendation algorithm combining collaborative filtering and content similarity is combined, so that the recommendation results can take into account both the viewing history of users with similar behaviors and the content similarity of the film itself for comprehensive recommendation, achieving more diverse and richer recommendation effects.

[0053] S110: Obtain a mutual enhancement graph model and a list of films and TV shows to be recommended.

[0054] Among them, the mutual enhancement graph model is composed of multiple nodes and the edges between any two nodes; the nodes are used to represent the film and television list and the user, and the edges between any two nodes are determined based on the historical film and television list association relationship, the user association relationship, and the association relationship between the user and the list. The historical film and television list association relationship can be understood as the association relationship between each list and the list in the film and television list recommendation using the multimodal information of the list. Among them, the multimodal information of the list can include text information, picture information, and video information of the list. The user association relationship can be understood as the association relationship between users. In this embodiment, the association relationship between each user can be determined by performing similarity features on each user. The association relationship between the user and the list can be understood as the interactive relationship between the user and the list. In this embodiment, the interactive relationship between the user and the list can be determined based on the user's historical viewing behavior. The list of film and television lists to be recommended can include a set number of film and television lists to be recommended. In this embodiment, a set number of film and television lists to be recommended can be obtained from the film and television list data resource library as the list of film and television lists to be recommended. The list of film and television lists to be recommended can include film and television lists of various film and television types and styles.

[0055] In this embodiment, the mutually enhanced graph model can be a model constructed based on the full amount of data in the film and television list recommendation source of the set duration and the relationship between each user. Among them, the set duration can be a period of one month or two months, etc., which can be set according to actual needs. In this embodiment, each node in the mutually enhanced graph model can be represented as a film and television list and a user, and the edges between each node are determined based on the historical film and television list relationship and the user relationship. The mutually enhanced graph model is formed by multiple nodes and the edges between each node. In this embodiment, a mutually enhanced graph model determined based on the historical film and television list relationship and the user relationship and a list of film and television lists to be recommended containing various types and styles can be obtained.

[0056] S120: Input the list of film and television programs to be recommended into the mutual enhancement graph model, and determine the node features of each film and television program to be recommended in the mutual enhancement graph model based on a random sampling algorithm.

[0057] Among them, the random sampling algorithm is an algorithm that extracts the list of movies and TV shows to be recommended from the mutual enhancement graph model based on randomness for analysis or inference. Exemplarily, the random sampling algorithm in this embodiment can be a Monte Carlo sampling algorithm. The node feature can be understood as the weight of the node, which can reflect the importance of the node.

[0058] In this embodiment, the Monte Carlo sampling method is used to evaluate the weights of each film and television list to be recommended based on the list of film and television lists to be recommended; the number of operations of the random sampling algorithm can be determined according to actual needs; the number of operations of the sampling method can be determined according to the actual computing efficiency requirements. In scenarios requiring high time efficiency, fewer operations can be performed, while scenarios requiring more accurate recommendations can be performed more times to obtain more accurate recommendation results. At the same time, the number of operations can also be determined based on the system load. In periods of heavy load, fewer operations can be performed to ensure response timeliness, while in periods of light load, more operations can be performed to improve the accuracy of the results. The technical solution of this embodiment can adapt to recommendation scenarios with different loads, different computing power, and different precision requirements, and can adjust the algorithm capabilities more conveniently.

[0059] In this embodiment, the list of film and television lists to be recommended can be input into the mutual enhancement graph model, and random walks can be performed from each film and television list in the list of film and television lists to be recommended to record the transfer path; the node weight of each film and television list node to be recommended in the mutual enhancement graph model is determined according to the transfer path. In addition, the source to be recommended in this embodiment can be a list of film and television lists to be recommended, or it can be a user to be recommended; it can be understood that in this embodiment, the corresponding film and television list can be recommended based on the specified user as input; it can also be based on multiple recommended film and television list data source input to recommend film and television lists that meet the user's preferences.

[0060] In this embodiment, optionally, node features are represented by node weights; the list of film and television lists to be recommended is input into the mutual reinforcement graph model, and the node features of the film and television lists to be recommended in the mutual reinforcement graph model are determined based on a random sampling algorithm, including: starting from each film and television list in the list of film and television lists to be recommended, performing a random walk to obtain a transfer path; and determining the node weight of the film and television list to be recommended based on the transfer path.

[0061] The transfer path may be a path obtained by performing a random walk starting from each movie list. In this embodiment, a random walk may be performed based on a transfer probability matrix determined by weights of edges between nodes in a mutually enhanced graph model.

[0062] In this embodiment, the list of film and television lists to be recommended can be input into the mutual reinforcement graph model as a starting point for recommendation, and then each film and television list in the list of film and television lists to be recommended can be started from, and a random walk can be performed to record the transfer path. In this embodiment, the specific method of determining the node weight of the film and television list to be recommended according to the transfer path can be based on the recorded transfer path, and the node weight of the film and television list node can be increased by the number of times the film and television list node is passed through to determine the final node weight of each film and television list to be recommended. It can be understood that the more times a certain film and television list node is passed through, the greater the weight of the film and television list.

[0063] Through such a setting in this embodiment, the transfer path can be determined according to the random walk path of the random sampling algorithm, and the node weight of each film and television list node can be determined based on the transfer path, so as to determine the film and television list that meets the user's needs and preferences.

[0064] In this embodiment, optionally, a random walk is performed starting from each film and television list in the list of film and television lists to be recommended to obtain a transfer path, including: determining the transfer probability between each node of the mutually reinforcing graph model; starting from each film and television list in the list of film and television lists to be recommended, transferring along the node according to the transfer probability, and performing random judgment after the node transfer to obtain random data; if the random data meets the end condition, the random walk is terminated and the transfer path is recorded.

[0065] Among them, the transition probability can be understood as the estimated relationship between the nodes of the mutually enhanced graph model. The end condition can be determined based on the set probability value. The set probability value can be represented by d. For example, d can be set to 0.7 or 0.8; it can be set according to actual needs, and this embodiment is not limited to this. The end condition in this embodiment can set the data to meet the probability value of 1-d. In this embodiment, if the random data meets the probability value of 1-d, the random walk can be ended. Random data can be a random value obtained by random judgment.

[0066] Specifically, in the embodiment, the weights of the film and television list association relationship, the user association relationship, and the association relationship between the user and the film list can be parameterized based on the mutual reinforcement graph model, and the probability matrix of the transition between each node is determined according to the weights of the film and television list association relationship, the user association relationship, and the association relationship between the user and the film list. In this embodiment, the similarity can be used to measure the film and television list association relationship, the user association relationship, and the association relationship between the user and the film list to obtain the similarity between the nodes and convert it into a transition probability.

[0067] In this embodiment, when random walking is performed from each movie list in the list of movies to be recommended, it is necessary to transfer between each node according to the transition probability matrix of the mutually enhanced graph model. In this embodiment, when transferring along the node according to the transition probability, when random walking, it is necessary to transfer along each node according to the transition probability matrix of the mutually enhanced graph model, and continue to walk with the probability of the set probability value d. After each node transfer, random judgment is performed to obtain random data. If the random data reaches the probability value of 1-d, the random walk is ended and the transfer path is recorded.

[0068] Through such a setting in this embodiment, a random walk can be performed in the mutual reinforcement graph model according to the random sampling method based on the transition probability to obtain a transfer path, thereby determining the node weight of each film and television list, and obtaining a more accurate recommendation result.

[0069] S130: Determine a target recommended film and television list according to node features of the film and television list to be recommended.

[0070] Among them, the node feature can be understood as the node weight of the film and television list to be recommended. The target recommended film and television list can be understood as the film and television list finally recommended to the user. In this embodiment, the number of target recommended film and television lists can be set based on the actual needs of the user. For example, if a user needs 6 recommended film and television lists, 6 target recommended film and television lists can be determined.

[0071] In this embodiment, the node weight of each film and television list to be recommended is different according to the path of random walk, and the node weight of each film and television list to be recommended is also different. Based on the node weights of each film and television list to be recommended, they can be sorted in order of size, and then starting from the film and television list with the highest node weight, the film and television list to be recommended that meets the user's required quantity can be selected as the target recommended film and television list.

[0072] In this embodiment, optionally, a target recommended film and television list is determined based on the node characteristics of the film and television list to be recommended, including: arranging the film and television list to be recommended in a set order according to the node weights of the film and television list to be recommended to obtain an arrangement result; and determining a set number of film and television lists to be recommended in the arrangement result as the target recommended film and television list.

[0073] Among them, the setting order can be understood as an arrangement order from large to small, or an arrangement order from small to large. In this embodiment, the setting order can be set according to actual needs. The arrangement result can be a list of movies to be recommended obtained by arranging the node weights of the list of movies to be recommended according to the set order. The set number can be understood as the number of target recommended movie lists required by the user. The set number in this embodiment can be set according to the actual needs of the user. Exemplarily, the set number can be 6.

[0074] In this embodiment, the movie and TV playlists to be recommended may be arranged in descending order according to their node weights, and a set number of movie and TV playlists to be recommended may be selected as target recommended movie and TV playlists according to the arrangement order.

[0075] Through such a setting in this embodiment, the target recommended film and television list can be determined according to the node weights of each film and television list to be recommended in the mutual reinforcement graph model, so that the recommendation results are more accurate and meet the needs and preferences of each user.

[0076] The technical solution of the embodiment of the present invention is to obtain a mutually enhanced graph model and a list of film and television lists to be recommended; wherein the mutually enhanced graph model is composed of multiple nodes and edges between any two nodes; the nodes are used to represent film and television lists and users, and the edges between any two nodes are determined based on the historical film and television list association relationship, the user association relationship, and the user-list association relationship; the list of film and television lists to be recommended is input into the mutually enhanced graph model, and the node features of each film and television list to be recommended in the mutually enhanced graph model are determined based on a random sampling algorithm; the target recommended film and television list is determined according to the node features of the film and television list to be recommended. This technical solution can mutually enhance user interaction data through a mutually enhanced graph model, thereby determining the list recommendation results and achieving more diverse and richer recommendation effects.

[0077] Embodiment 2

[0078] Figure 2 This is a flowchart of a film and television playlist recommendation method provided according to the second embodiment of the present invention. This embodiment is optimized based on the above embodiment. The specific optimization is: before obtaining the mutual enhancement graph model and the list of film and television playlists to be recommended, it also includes: obtaining historical user information and historical film and television playlist information of a set time; determining user association relationships, film and television playlist association relationships, and the relationship between users and playlists based on historical user information and historical film and television playlist information; constructing a mutual enhancement graph model based on user association relationships, film and television playlist association relationships, and the relationship between users and playlists. Figure 2 As shown, the method includes:

[0079] S210, obtaining historical user information and historical film and television list information of a set duration.

[0080] Among them, the set duration can be understood as a set duration. The set duration in this embodiment can be a duration such as one month or two months, which can be set according to actual needs. Historical user information may include user feature attributes and / or user viewing habit information. User feature attributes may include attribute information such as the user's authorized location information, age, or set tags, and historical film and television list information can be understood as multimodal information of historical film and television list, which may include but is not limited to text information of the film and television list (such as field information such as text title or text description), picture information (information such as the header image of the list), and video information (such as promotional videos or video clips, etc.).

[0081] In this embodiment, historical user personal information and / or user's viewing habit information within a set time period and multi-modal information of historical film and television playlists can be obtained in the film and television application software.

[0082] S220. Determine user associations, film and television list associations, and associations between users and lists based on historical user information and historical film and television list information.

[0083] Among them, the user association relationship is understood as the association between users. In this embodiment, the user association relationship can be obtained based on the user's characteristic attributes and the user similarity characteristics determined by the similarity algorithm; it can also be established by dividing users into different groups by clustering, and this embodiment does not limit this. The film and television list association relationship can be understood as the association between film and television lists. In this embodiment, it can be constructed by the similarity of the content of the film and television list. The specific construction method is to use a pre-trained multimodal large model to construct the list features, and then calculate the similarity between the lists to obtain the film and television clip association relationship. The association relationship between users and lists can be understood as the interactive relationship between users and film and television lists. In this embodiment, the interactive relationship between users and film and television lists can be constructed through the user's viewing behavior; if the user watches a certain film and television list for a longer time or a longer time ratio, it can be determined as the interactive relationship between the user and the film and television list. Among them, the specified time can be set according to actual needs, such as the specified time can be one hour or two hours.

[0084] In addition, in determining the association relationship of film and television lists, this embodiment can construct the characteristics of the list through different modal information; specifically, a pre-trained multimodal large model with modality alignment can be used, such as ImageBind, which can project information features of different modalities into the same feature space, and then the multimodal information of the list can be used to construct similarity relationships between the lists, that is, to obtain the association relationship of the film and television lists.

[0085] In this embodiment, the association between users, the association between film and television lists, and the association between users and lists can be determined based on historical user personal information and / or user viewing habit information within a set time period and multimodal information of historical film and television lists.

[0086] In this embodiment, optionally, user association relationships, film and television list association relationships, and association relationships between users and film lists are determined based on historical user information and historical film and television list information, including: determining historical user similarity characteristics, film list similarity characteristics, and historical user viewing behavior characteristics; determining user association relationships, film and television list association relationships, and association relationships between users and film lists based on historical user similarity characteristics, film list similarity characteristics, and historical user viewing behavior characteristics.

[0087] In this embodiment, the similarity features between historical users and the similarity features between film and television lists can be determined based on the content similarity algorithm, or the similarity features can be extracted by a feature extraction model, which is not limited in this embodiment. The historical user viewing behavior features can be extracted based on the user's historical viewing behavior.

[0088] In this embodiment, a pre-trained multimodal large model can be used to extract similarity features to obtain similarity features of historical users, similarity features of film lists, and historical user viewing behavior features, and then build user association relationships based on the obtained user similarity features, build association relationships of film and television film lists based on film list similarity features, and build association relationships between users and film lists based on historical user viewing behavior features.

[0089] Through such a setting in this embodiment, the association relationship between the user and the film and television list can be established through the user's viewing behavior, the association relationship within the film and television list can be established through multimodal feature similarity, and the association relationship within the user can be established through user feature attributes. The three parts of information enhance each other to achieve more diverse and richer recommendation effects.

[0090] S230: construct a mutual enhancement graph model based on user association relationships, film and television playlist association relationships, and association relationships between users and playlists.

[0091] In this embodiment, users and movie lists can be represented as nodes in the mutual enhancement graph model respectively. The connection relationship between the nodes is determined according to the user association relationship, the movie list association relationship, and the association relationship between the user and the movie list, that is, the edge of the mutual enhancement graph model, so as to construct a mutual enhancement graph model.

[0092] In this embodiment, optionally, a mutually reinforcing graph model is constructed based on user association relationships, film and television list association relationships, and the association relationships between users and film lists, including: representing users and film and television list nodes as user nodes and film list nodes, respectively; representing user association relationships, film and television list association relationships, and the association relationships between users and film lists as edges of the mutually reinforcing graph model, respectively; and constructing a mutually reinforcing graph model based on user nodes, film and television list nodes, and edges.

[0093] In this embodiment, users can be represented as user nodes, and movie lists can be represented as movie list nodes; user associations, movie list associations, and the associations between users and movie lists can be represented as edges between user nodes and movie list nodes in a mutually reinforcing graph model, respectively, so that a mutually reinforcing graph model can be constructed based on user nodes, movie list nodes, and edges between nodes.

[0094] An example diagram of the mutual enhancement graph model in this embodiment is shown in FIG. Figure 3As shown. The user relationship graph part is used to reflect the relationship between users, and its circle shape can be represented as a user node; the movie list relationship graph part reflects the relationship between movie lists, and its circle shape can be represented as a movie list node. In the mutual enhancement graph model, the relationship between users is obtained by dividing users into different groups through the similarity of population characteristics or by clustering, which is used as the edge between each user node; the edge between movie lists is obtained by calculating the relationship between movie lists through the similarity between the movie list contents, which is used as the edge between each movie list node; the interactive relationship between users and movie lists is constructed through the user's viewing behavior, which is used as the edge between users and movie lists, and the mutual enhancement graph model is constructed through each edge and each node.

[0095] In this embodiment, a mutually enhanced graph model is obtained through such a setting, which has good scalability. By adding more ways to determine the association relationship between movie lists, users, and users, recommendation algorithms with different effects can be implemented, the similarity robustness can be increased, and the adaptation range can be wider.

[0096] S240: Obtain a mutual enhancement graph model and a list of movies and TV shows to be recommended.

[0097] Among them, the mutual enhancement graph model is determined based on the association relationship between historical film and television lists and the user association relationship. The mutual enhancement graph model is composed of multiple nodes and edges between any two nodes.

[0098] S250: Input the list of film and television programs to be recommended into the mutual enhancement graph model, and determine the node features of each film and television program to be recommended in the mutual enhancement graph model based on a random sampling algorithm.

[0099] S260: Determine a target recommended film and television list according to node features of the film and television list to be recommended.

[0100] The technical solution of the embodiment of the present invention is to obtain historical user information and historical film and television list information of a set time; determine the user association relationship, the film and television list association relationship and the association relationship between the user and the list according to the historical user information and the historical film and television list information; construct a mutual enhancement graph model according to the user association relationship, the film and television list association relationship and the association relationship between the user and the list; obtain the mutual enhancement graph model and the list of film and television lists to be recommended; wherein the mutual enhancement graph model is composed of multiple nodes and the edges between any two nodes; the nodes are used to represent the film and television list and the user, and the edges between any two nodes are determined based on the historical film and television list association relationship, the user association relationship and the association relationship between the user and the list; input the list of film and television lists to be recommended into the mutual enhancement graph model, determine the node features of each film and television list to be recommended in the mutual enhancement graph model based on a random sampling algorithm; determine the target recommended film and television list according to the node features of the film and television list to be recommended. This technical solution can mutually enhance the user interaction data through the mutual enhancement graph model, thereby determining the list recommendation results, and achieving more diverse and richer recommendation effects.

[0101] Embodiment 3

[0102] Figure 4 1 is a schematic diagram of the structure of a device for recommending movie lists according to Embodiment 3 of the present invention. Figure 4 As shown, the device comprises:

[0103] The data acquisition module 410 is used to acquire a mutual enhancement graph model and a list of film and television playlists to be recommended; wherein the mutual enhancement graph model is composed of a plurality of nodes and edges between any two nodes; the nodes are used to represent film and television playlists and users, and the edges between any two nodes are determined based on the historical film and television playlist association relationship, the user association relationship, and the association relationship between the user and the playlist;

[0104] A node feature determination module 420 is used to input the list of movies and TV series to be recommended into the mutual enhancement graph model, and determine the node features of each movie and TV series to be recommended in the mutual enhancement graph model based on a random sampling algorithm;

[0105] The target recommended film and television playlist determination module 430 is used to determine the target recommended film and television playlist according to the node characteristics of the film and television playlist to be recommended.

[0106] Optionally, the device further comprises:

[0107] A historical information acquisition module, used to acquire historical user information and historical film and television list information of a set duration before acquiring the mutual enhancement graph model and the list of films and television lists to be recommended;

[0108] An association determination module, used to determine user associations, film and television list associations, and associations between users and lists based on historical user information and historical film and television list information;

[0109] The mutual enhancement graph model construction module is used to construct a mutual enhancement graph model based on user association relationships, film and television list association relationships, and the association relationship between users and list relationships.

[0110] Optionally, an association relationship determination module is specifically used to determine historical user similarity characteristics, film list similarity characteristics, and historical user viewing behavior characteristics; based on historical user similarity characteristics, film list similarity characteristics, and historical user viewing behavior characteristics, user association relationships, film and television film list association relationships, and association relationships between users and film lists are determined.

[0111] Optionally, a mutual enhancement graph model construction module is specifically used to represent users and film and television lists as user nodes and film and television lists nodes, respectively; represent user associations, film and television list associations, and the associations between users and film lists as edges of the mutual enhancement graph model, respectively; and construct a mutual enhancement graph model based on user nodes, film and television list nodes and edges.

[0112] Optionally, node features are represented by node weights;

[0113] The node feature determination module 420 includes:

[0114] A transfer path acquisition unit, used to perform random walks from each film and television list in the list of films and television lists to be recommended to obtain a transfer path;

[0115] The node weight determination unit is used to determine the node weight of the film and television list to be recommended according to the transfer path.

[0116] Optionally, a transfer path acquisition unit is specifically used to determine the transfer probability between each node of the mutually reinforced graph model; starting from each film and television list in the list of film and television lists to be recommended, transfer along the node according to the transfer probability, and perform random judgment after the node transfer to obtain random data; if the random data meets the end condition, the random walk is ended and the transfer path is recorded.

[0117] Optionally, the target recommended film and television list determination module 430 is specifically used to determine the target recommended film and television list based on the node characteristics of the film and television list to be recommended, including: arranging the film and television list to be recommended in a set order according to the node weights of the film and television list to be recommended to obtain an arrangement result; and determining a set number of film and television lists to be recommended in the arrangement result as the target recommended film and television list.

[0118] A device for recommending a film and television list provided in an embodiment of the present invention can execute a method for recommending a film and television list provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0119] Embodiment 4

[0120] Figure 5 1 is a schematic diagram of the structure of an electronic device provided 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 processing, cellular phones, smart phones, 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 examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0121] like Figure 5 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0122] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0123] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a film and television playlist recommendation method.

[0124] In some embodiments, the film and television list recommendation method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the film and television list recommendation method described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to execute the film and television list recommendation method in any other appropriate manner (for example, by means of firmware).

[0125] Various implementations 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 chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0126] Computer programs for implementing 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, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0127] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0128] To provide interaction with a user, the systems and techniques described herein may 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 a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0129] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0130] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0131] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0132] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for recommending a movie list. It is characterized in that include: Obtaining a mutual enhancement graph model and a list of film and television playlists to be recommended; wherein the mutual enhancement graph model is composed of a plurality of nodes and edges between any two nodes; the nodes are used to represent film and television playlists and users, and the edges between any two nodes are determined based on historical film and television playlist associations, user associations, and user-playlist associations; Inputting the list of film and television playlists to be recommended into the mutual enhancement graph model, and determining the node features of each film and television playlist to be recommended in the mutual enhancement graph model based on a random sampling algorithm; A target recommended film and television list is determined according to the node features of the film and television list to be recommended.

2. The method according to claim 1, It is characterized in that Before obtaining the mutual enhancement graph model and the list of movies and TV shows to be recommended, the following steps are also included: Get historical user information and historical film and television list information of a set duration; Determine the user association relationship, the film and television list association relationship, and the association relationship between the user and the film list according to the historical user information and the historical film and television list information; A mutual enhancement graph model is constructed based on the user association relationship, the film and television list association relationship, and the association relationship between the user and the list.

3. The method according to claim 2, It is characterized in that Determining the user association relationship, the film and television list association relationship, and the association relationship between the user and the film list according to the historical user information and the historical film and television list information includes: Determine historical user similarity characteristics, movie list similarity characteristics, and historical user viewing behavior characteristics; The user association relationship, the film list association relationship and the association relationship between the user and the film list are determined according to the historical user similarity characteristics, the film list similarity characteristics and the historical user viewing behavior characteristics.

4. The method according to claim 2, It is characterized in that Constructing a mutual enhancement graph model according to the user association relationship, the film and television playlist association relationship, and the association relationship between the user and the playlist includes: Representing the user and the film and television playlist as a user node and a playlist node respectively; Representing the user association relationship, the film and television playlist association relationship, and the association relationship between the user and the playlist as edges of the mutual enhancement graph model; The mutual enhancement graph model is constructed according to the user nodes, the sheet nodes and the edges.

5. The method according to claim 1, It is characterized in that The node feature is represented by a node weight; Inputting the list of films and TV series to be recommended into the mutual enhancement graph model, and determining node features of the films and TV series to be recommended in the mutual enhancement graph model based on a random sampling algorithm, including: Starting from each film and television list in the list of film and television lists to be recommended, a random walk is performed to obtain a transfer path; The node weight of the to-be-recommended film and television playlist is determined according to the transfer path.

6. The method according to claim 5, It is characterized in that Starting from each film and television list in the list of film and television lists to be recommended, a random walk is performed to obtain a transfer path, including: Determining the transition probability between nodes of the mutual reinforcement graph model; Starting from each film and television list in the list of film and television lists to be recommended, transferring along the node according to the transfer probability, and performing random judgment after the node transfer to obtain random data; If the random data meets the end condition, the random walk ends and the transfer path is recorded.

7. The method according to claim 5, It is characterized in that Determining a target recommended film and television playlist according to the node features of the to-be-recommended film and television playlist includes: Arranging the to-be-recommended film and television list in a set order according to the node weights of the to-be-recommended film and television list to obtain an arrangement result; A set number of to-be-recommended film and television playlists in the arrangement result are determined as target recommended film and television playlists.

8. A device for recommending movie and TV series, It is characterized in that include: A data acquisition module, used to acquire a mutual enhancement graph model and a list of film and television playlists to be recommended; wherein the mutual enhancement graph model is composed of a plurality of nodes and edges between any two nodes; the nodes are used to represent film and television playlists and users, and the edges between any two nodes are determined based on historical film and television playlist associations, user associations, and user-playlist associations; A node feature determination module, used for inputting the list of film and television plays to be recommended into the mutual enhancement graph model, and determining the node features of each film and television playlist to be recommended in the mutual enhancement graph model based on a random sampling algorithm; The target recommended film and television playlist determination module is used to determine the target recommended film and television playlist according to the node characteristics of the film and television playlist to be recommended.

9. An electronic device, It is characterized in that The electronic device comprises: 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, and the computer program is executed by the at least one processor so that the at least one processor can execute the film and television list recommendation method described in any one of claims 1-7.

10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the film and television list recommendation method described in any one of claims 1-7 when executed.

Images