Multimedia resource recommendation method and system, electronic equipment and storage medium

Abstract

The invention relates to a multimedia resource recommendation method and system, electronic equipment and a storage medium. The method can comprise the steps of obtaining respective object attribute information and object historical operation information of a plurality of objects; the object historical operation information is used for representing operation information generated by the multiple objects in the multimedia resource browsing process; performing data cleaning and feature extraction processing on the historical operation information of the object to obtain key historical operation information of the object; according to the object attribute information and the object key historical operation information, performing clustering processing on the plurality of objects to obtain a plurality of object clustering groups; and screening out first multimedia resources corresponding to the plurality of object clustering groups from the plurality of to-be-recommended multimedia resources, and recommending the first multimedia resources corresponding to the object clustering groups to the objects in the object clustering groups. According to the technical scheme provided by the invention, the recommendation accuracy of the multimedia resources can be improved, and the retention rate of the multimedia resource playing platform objects can also be improved.

Description

Technical Field

[0001] The present application relates to the field of Internet application technology, and in particular to a multimedia resource recommendation method, system, electronic device and storage medium. Background Art

[0002] In the current era of rapid development of the Internet and mobile applications, users are faced with a vast amount of multimedia resources, such as short dramas and short videos, which provide users with a rich variety of choices. However, with the explosive growth of the amount of content, the recommendation system faces challenges and it is difficult to accurately recommend multimedia resources that each user is really interested in from the resource library. This not only reduces the user's browsing experience, but may also lead to gradual loss of users, affecting the user retention rate of the platform. Summary of the invention

[0003] The present application provides a multimedia resource recommendation method, system, electronic device and storage medium to at least solve the problems in the related art such as how to reduce the frequency of occurrence of short sentences that users do not care about and improve the user retention rate. The technical solution of the present application is as follows:

[0004] According to a first aspect of an embodiment of the present application, a multimedia resource recommendation method is provided, comprising:

[0005] Acquire object attribute information and object historical operation information of each of the plurality of objects; the object historical operation information is used to represent the operation information generated by each of the plurality of objects in the process of browsing multimedia resources;

[0006] Performing data cleaning and feature extraction processing on the historical operation information of the object to obtain key historical operation information of the object;

[0007] Clustering the multiple objects according to the object attribute information and the object key historical operation information to obtain multiple object cluster groups;

[0008] The first multimedia resources corresponding to each of the plurality of object clusters are screened out from the plurality of multimedia resources to be recommended, and the first multimedia resources corresponding to each object cluster are recommended to the objects in each object cluster.

[0009] According to a second aspect of an embodiment of the present application, a multimedia resource recommendation system is provided, including:

[0010] An information acquisition module, used to acquire object attribute information and object historical operation information of each of the multiple objects; the object historical operation information is used to represent the operation information generated by each of the multiple objects in the process of browsing multimedia resources;

[0011] A feature extraction module is used to perform data cleaning and feature extraction processing on the object historical operation information to obtain key historical operation information of the object;

[0012] A clustering processing module, used for performing clustering processing on the multiple objects according to the object attribute information and the key historical operation information of the objects to obtain multiple object cluster groups;

[0013] The multimedia resource recommendation module is used to select the first multimedia resources corresponding to each of the plurality of object clusters from the plurality of multimedia resources to be recommended, and recommend the first multimedia resources corresponding to each object cluster to the objects in each object cluster.

[0014] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement a method as described in any one of the above-mentioned first aspects.

[0015] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute any method described in the first aspect of the embodiment of the present application.

[0016] According to a fifth aspect of an embodiment of the present application, a computer program product is provided, comprising computer instructions, which, when executed by a processor, enable a computer to execute any one of the methods described in the first aspect of the embodiment of the present application.

[0017] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application.

[0018] The technical solution provided by the embodiments of the present application brings at least the following beneficial effects:

[0019] By obtaining the object attribute information and object historical operation information of multiple objects, data cleaning and feature extraction processing are performed on the object historical operation information to obtain the key historical operation information of the object, which can remove noise and redundant information and improve the accuracy of multimedia resource recommendation;

[0020] Based on the object attribute information and the object key historical operation information, multiple objects are clustered to obtain multiple object cluster groups, and then the first multimedia resources corresponding to the multiple object cluster groups are screened out from the multiple multimedia resources to be recommended, and the first multimedia resources corresponding to each object cluster group are recommended to the objects in each object cluster group. This can reduce the recommendation of multimedia resources that users are not interested in, achieve accurate clustering of multiple users and accurate recommendation of multimedia resources to users of each object cluster group, thereby improving the utilization rate of multimedia resources and the user retention rate of the playback platform where the multimedia resources are located.

[0021] Other features and aspects of the present application will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions and advantages of the embodiments of this specification or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 A multimedia resource recommendation method according to an exemplary embodiment is shown in FIG. Figure 1 .

[0024] Figure 2 A multimedia resource recommendation method according to an exemplary embodiment is shown in FIG. Figure 2 .

[0025] Figure 3 The present invention is a device block diagram of a multimedia resource recommendation system according to an exemplary embodiment.

[0026] Figure 4 A frame of an electronic device for multimedia resource recommendation according to an exemplary embodiment is shown Figure 1 .

[0027] Figure 5 A frame of an electronic device for multimedia resource recommendation according to an exemplary embodiment is shown Figure 2 . DETAILED DESCRIPTION

[0028] In order to enable ordinary persons in the art to better understand the technical solutions of the present application, the technical solutions in the embodiments of the present application specification will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application specification. Obviously, the described embodiments are only part of the embodiments of the specification, not all of the embodiments. Based on the embodiments in the present specification, all other embodiments obtained by ordinary persons in the art without creative work are within the scope of protection of the present invention.

[0029] It should be noted that the terms "first", "second", etc. 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 server 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.

[0030] Various exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0031] The word "exemplary" used exclusively herein means "serving as an example, embodiment or illustrative". Any embodiment described herein as "exemplary" is not necessarily to be construed as being superior or better than other embodiments. The term "and / or" herein is merely a description of an association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent the following three situations: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "at least one" herein represents any combination of at least two of any one or more of a plurality of types. For example, including at least one of A, B, and C may represent any one or more elements selected from the set consisting of A, B, and C.

[0032] Unless otherwise specified, the directions in this document should be understood as follows: the direction closer to the user is the front, and the direction away from the user is the rear.

[0033] In addition, in order to better illustrate the present application, numerous specific details are provided in the following specific embodiments. It should be understood by those skilled in the art that the present application can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present application.

[0034] It should be noted that the following figure shows a possible order of steps, which is not actually limited to this order. Some steps can be executed in parallel without relying on each other. The user information (including but not limited to user device information, user personal information, user behavior information, etc.) and data (including but not limited to data for display, training data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0035] Figure 1 FIG. 1 is a flowchart of a multimedia resource recommendation method according to an exemplary embodiment. Figure 2 As shown, the following steps may be included.

[0036] In step S101, object attribute information and object historical operation information of each of a plurality of objects are obtained.

[0037] In the embodiments of this specification, the object may refer to a user corresponding to the multimedia resource platform.

[0038] The object attribute information may refer to user characteristic information. For example, in a multimedia resource recommendation system, the object attribute information may include the user's age, gender, occupation, geographic location, etc., which is not limited in this application.

[0039] The object historical operation information can be used to represent the operation information generated by multiple objects in the process of browsing multimedia resources. The object historical operation information can include the type, duration, likes, comments, sharing and other behavior data of the user browsing multimedia resources, which is not limited in this application.

[0040] Multimedia resources may refer to resources played in multimedia form. Multimedia resources may include short plays, short videos, long videos, etc., which are not limited in this application.

[0041] like Figure 2As shown, the object attribute information and object historical operation information of all objects are collected from the multimedia resource platform. Exemplarily, during the user's registration on the multimedia resource platform, multiple object attribute information can be set to be filled in, such as age, gender, geographic location, occupation, etc., and the multimedia resource recommendation system will obtain the object attribute information corresponding to the user. After the user registers the object attribute information, the multimedia resource platform can be browsed directly. Furthermore, during the browsing process, the user will generate various operation information, such as clicks, viewing time, likes, frequency, attention, sharing, etc. The multimedia resource recommendation system will collect and update the user's object historical operation information based on the user's real-time browsing situation.

[0042] In step S103, data cleaning and feature extraction are performed on the object historical operation information to obtain key historical operation information of the object.

[0043] In the embodiments of the present specification, the object key historical operation information may refer to key information used to describe user behavior characteristics obtained after data cleaning and feature extraction.

[0044] In one possible implementation, historical operation information of objects whose multimedia browsing volume is less than a second preset browsing volume is determined as noise data information; historical operation information of objects that contain abnormal data information and noise data information is eliminated to obtain historical operation information of target objects; and feature extraction is performed on the historical operation information of target objects to obtain key historical operation information of objects corresponding to multiple objects.

[0045] In the embodiments of this specification, multimedia browsing volume may refer to the number of multimedia resources browsed by a user within a certain period of time. The second preset browsing volume may refer to the preset number of user browsing times, which is used to determine whether the object historical operation information is noise data information. Exemplarily, within the preset time period, the number of user browsing on the multimedia platform is 30, and the number of multimedia browsing of the user is 30 at this time. This application does not limit the number of multimedia browsing and the corresponding time period.

[0046] Noise data information may refer to information that interferes with data analysis or processing results. Abnormal data information may refer to abnormal information existing in a multimedia resource recommendation system. For example, garbled characters generated during the transmission of object historical operation information or the system cannot recognize user behavior information.

[0047] After the object history operation information is obtained, the object history operation information is immediately cleaned. Exemplarily, a second preset browsing volume corresponding to a certain time period is preset, for example, 50. When the corresponding multimedia browsing volume of a user in each time period is 30, the multimedia browsing volume 30 of the user in each time period is less than the second preset browsing volume 50, which indicates that the user rarely browses multimedia resources. Therefore, the object history operation information of the user cannot be clustered at the same level with the object history operation information whose multimedia browsing volume in each time period is greater than 1000, which can improve the accuracy of user clustering. Therefore, the object history operation information whose multimedia browsing volume corresponding to each time period is less than the second preset browsing volume is determined as noise data information.

[0048] Furthermore, the garbled information received in the system, the garbled information cannot correctly read the IP address, or the system cannot recognize the user's incompatible information of the user's behavior information, etc., which is not limited by this application. The above-mentioned information is determined as abnormal data information. Then, the object historical operation information with abnormal data information and noise data information is eliminated to obtain the target object historical operation information.

[0049] After data cleaning, feature extraction is performed on the historical operation information of the target object to extract key historical operation information of multiple objects.

[0050] Among them, the browsing frequency VF of a certain multimedia resource can be extracted from the key historical operation information of the object, and the total number of views and the total number of viewers of all short dramas under a certain playback type within a preset time period can be obtained; then, according to the total number of playbacks of all multimedia resources divided by the total number of viewers and then divided by the number of minutes in the preset time period, the average number of views per user per minute within a preset time period can be determined, that is, the browsing frequency VF. Exemplarily, the preset time period is from 7:00 on March 1, 2024 to 8:00 on March 1, 2024, and the playback type is animation type. Within the preset time period, the total number of views of all short dramas of the animation type is 500,000 times, and the total number of viewers is 10,000 times. Then, it can be determined that the browsing frequency VF of the animation type within the preset time period is 500,000 divided by the total number of viewers 10,000, and then divided by 60. The larger the browsing frequency VF value, the faster the multimedia resource switching, that is, the content of this type of multimedia resource is not attractive enough to the audience, and the content of this type needs to be improved.

[0051] The total browsing time VT is extracted from the key historical operation information of the object, that is, the sum of the browsing time of all short dramas under a certain playback type by all users within the preset time period. The single browsing rate V-PR is extracted from the key historical operation information of the object, that is, the total number of playbacks of a single multimedia resource under a certain playback type within the preset time period is obtained. For example, in the preset time period from 7:00 on March 1, 2024 to 8:00 on March 1, 2024, the total number of playbacks of a single multimedia resource under the costume drama playback type is 100,000 times. The single browsing rate V-PR can locate the quality of each multimedia resource. When the single browsing rate V-PR is high, it can be concluded that the multimedia resource is high-quality content, and then the multimedia resource will be recommended to other users under the corresponding playback type.

[0052] The multimedia payment rate V-RR can also be extracted from the key historical operation information of the object, that is, the number of payments for a single episode in a single multimedia resource of a certain type divided by the total number of payments for a single multimedia in a preset time period. For example, in the preset time period from 7:00 on March 1, 2024 to 8:00 on March 1, 2024, the total number of payments for a single episode in a single multimedia resource of the ancient costume drama playback type is divided by the total number of payments for a single multimedia resource to obtain the payment rate V-RR of the single multimedia resource. The multimedia payment number V-RP can also be extracted from the key historical operation information of the object, that is, the total number of payments for a single multimedia resource of a certain playback type in a preset time period. This application does not limit the content of feature extraction, which can be determined according to actual conditions.

[0053] In step S105, a plurality of objects are clustered according to the object attribute information and the object key historical operation information to obtain a plurality of object cluster groups.

[0054] In the embodiment of this specification, the object cluster may refer to a user group obtained by clustering all users on the multimedia platform.

[0055] In one possible implementation, object weight information corresponding to each of the multiple objects is determined from the object's key historical operation information and object attribute information; vector information of each of the multiple objects is determined based on the object weight information; and adjacent information of each of the multiple objects is clustered based on a density-based spatial clustering algorithm to obtain multiple object cluster groups.

[0056] In the embodiment of this specification, the object weight information can be used to quantify the index of each object in the clustering process. Exemplarily, the object weight information may include the playback type (a): the weight value of romance drama = 1.5, the weight value of costume drama = 2, the weight value of science fiction drama = 0.8, etc. The present application does not limit the number of playback types, which can be determined according to the actual situation. The object weight information also includes the user's daily online time (b): 2 hours to 3 hours weight value = 1.2, less than two hours weight value = 1.8, more than 3 hours weight value = 0.7, the present application does not limit the specific duration of the online time. The object weight information may also include age (c): 19 to 30 years old weight value = 1.2, 30 to 45 years old weight value = 2, 45 years old and above weight value = 1, the present application does not limit age. The present application may also include object weight information related to other object required information and object key historical operation information, and the present application does not limit the object weight information.

[0057] The vector information of each of the multiple objects may refer to the vector obtained by performing mathematical operations on the object weight information. The density-based spatial clustering algorithm (Density-Based Spatial Clustering of Applications with Noise, referred to as DBSCAN) can divide areas with sufficiently high density into clusters, and can find clusters of arbitrary shapes in a spatial database with noise. The density-based spatial clustering algorithm does not need to know the number of clusters to be formed in advance, and can divide data points into core points, boundary points, and noise points. A core point is a point with at least MinPts points (including itself) in its ε-neighborhood; a boundary point is a point that is located in the ε-neighborhood of a core point but is not a core point itself; a noise point is a point that is neither a core point nor a boundary point. The present application uses a density-based spatial clustering algorithm without the need to determine the data of the specified cluster in advance, thereby enhancing the flexibility and adaptability of clustering and subsequent recommendations.

[0058] First, the system can initially classify users based on the object attribute information of multiple users, such as the gender, age, IP address, occupation, etc. of the object. This can reduce the use of density-based spatial clustering algorithms for complex clustering processing and avoid, for example, recommending short plays such as rugby and boxing to female sales staff. This application does not limit this.

[0059] Furthermore, according to the key historical operation information and object attribute information of the object, the object weight information corresponding to each of the multiple objects is determined, namely, the playback type (a), the user's daily online time (b), age (c), etc. After determining the object weight information of each user, the corresponding vector information can be determined according to the object weight information. The specific solution formula for the vector information is as follows:

[0060] Among them, p i is the vector information of the i-th object, and the vector information of different objects is different. i is the playback type weight information corresponding to the i-th object, b i is the weight information of the daily online time of the user corresponding to the i-th object, c i is the age weight information corresponding to the i-th object.

[0061] For example, the weight value corresponding to a user's playback type (a) is 2, which means that the user focuses on costume dramas; the weight value corresponding to the user's daily online time (b) is 1.2, which means that the user browses multimedia resources between 2 and 3 hours a day. The weight value corresponding to age (c) is 1, which means that the user is over 45 years old. It can be concluded that the vector information of the user is:

[0062] After determining the vector information corresponding to all objects, the corresponding neighborhood radius ε in the density-based spatial clustering algorithm and the minimum number of points, that is, the minimum number of data points required in the neighborhood, including itself, are determined according to the multimedia resource recommendation system to determine whether it is a core point.

[0063] The density-based spatial clustering algorithm performs clustering processing on the vector information of multiple objects to obtain multiple object cluster groups.

[0064] Specifically, the formula of the density-based spatial clustering algorithm is:

[0065] Among them, N ε (p) is the point p i ε neighborhood of q i is the data point p i Points in the neighborhood of , q i The calculation method of p i The calculation method is the same, and both use the above vector information solution formula for calculation. i In p i In the neighborhood of point, we get object p i and object q i Belong to the same group. Yes i to q i The distance between them, in the vector information p i and q i After the calculation is completed, you can directly Simplified to This application does not limit the neighborhood radius ε.

[0066] According to the above clustering process, all users are traversed and p i and q i All objects whose distance is less than the distance of the ε domain are clustered into the same group, and then the object clustering group can be obtained. i Points can be used to obtain multiple object clusters.

[0067] In another possible implementation, a target object other than the multiple object clusters is obtained from the multiple objects, and a second multimedia resource whose object browsing volume exceeds a first preset browsing volume is obtained; and the second multimedia resource is recommended to the target object.

[0068] After traversing all objects, select objects whose distance between p and q is greater than the ε domain, and identify such objects as target objects. Target objects represent that such objects belong to outlier users, for example, the browsing time is very short, and the types of multimedia resources browsed are many, etc. This application does not limit this. Further, the target objects are classified into target object cluster groups, and the users of the target object cluster groups are recommended as a backup, for example, a second multimedia resource whose object browsing volume exceeds the first preset browsing volume is obtained, that is, a currently popular multimedia resource is obtained, and the second multimedia resource is recommended to the target object cluster group. This application does not limit the backup recommendation method.

[0069] In a possible implementation, object operation information with abnormal data information and noise data information is determined as abnormal historical operation information, and the object corresponding to the abnormal historical operation information is also classified into the target object cluster group. Similar to the above-mentioned bottom-up recommendation method, the second multimedia resource is recommended to all users in the target object cluster group.

[0070] In step S107, first multimedia resources corresponding to each of a plurality of object clusters are screened out from a plurality of multimedia resources to be recommended, and the first multimedia resources corresponding to each object cluster are recommended to the objects in each object cluster.

[0071] In the embodiment of the specification, the multimedia resource to be recommended may refer to a set of multimedia content that can be recommended to the user in the multimedia resource recommendation system. The first multimedia resource may refer to a multimedia resource selected from a plurality of multimedia resources to be recommended and corresponding to the object cluster group.

[0072] In a possible implementation, multiple multimedia resources to be recommended and multimedia feature information to be recommended corresponding to each of the multiple multimedia resources to be recommended are obtained; based on the multimedia feature information to be recommended, feature vector information to be recommended corresponding to each of the multiple multimedia resources to be recommended is determined; based on the cosine similarity formula, preset vector information corresponding to each of the multiple object clustering groups and angle information of the feature vector information to be recommended are determined; target angle information that meets the preset angle conditions corresponding to each of the object clustering groups is filtered out from the angle information, and the multimedia resource to be recommended corresponding to the target angle information is confirmed as the first multimedia resource.

[0073] In the embodiments of the present specification, the feature information of the multimedia resources to be recommended may refer to information describing the features of each multimedia resource to be recommended. The feature vector information to be recommended may refer to the feature information of the multimedia resources to be recommended converted into a vector in a multi-dimensional space.

[0074] The multimedia resource recommendation system can obtain the multimedia feature information corresponding to the multimedia resource to be recommended from the key historical operation information of the object, for example, the browsing frequency VF of the multimedia resource to be recommended, the total browsing time VT of the multimedia resource to be recommended, the single browsing rate V-PR of the multimedia resource to be recommended, the payment rate V-RR of the multimedia resource to be recommended, the number of payment times V-RP of the multimedia resource to be recommended, etc., which are not limited in this application.

[0075] The multiple feature necklace information to be recommended are converted into the corresponding feature vector information to be recommended. For example, the browsing frequency VF of drama A is 0.6, the total browsing time VT of drama A is 0.8, the single browsing rate V-PR of drama A is 0.7, the payment rate V-RR of drama A is 0.5, and the number of payment times V-RP of drama A is 0.4, thereby determining the feature vector information to be recommended of drama A = [VF A , VT A , V-PR A , V-RR A , V-RP A ]=[0.6,0.8,0.7,0.5,0.4]. Each object cluster group has its own corresponding preset vector information, for example, the preset vector information of object cluster group B=[VF B , VT B , V-PR B , V-RR B , V-RP B ]=[0.8, 0.7, 0.6, 0.8, 0.5].

[0076] According to the cosine similarity formula, the similarity between two vectors is evaluated by calculating the cosine value of the angle between them in multidimensional space. Specifically, the cosine similarity formula is

[0077]

[0078] Among them, A i is the feature vector information to be recommended for the i-th feature vector, B i is the preset vector information of the i-th object cluster group. The feature vector information to be recommended of A and the preset vector information of B are input into the cosine similarity formula to calculate

[0079]

[0080] Among them, Cosθ is the angle information, VF B is the browsing frequency corresponding to the cluster of object B, VT B is the total browsing time corresponding to the B object cluster, V-PR B is the single browsing rate corresponding to the B object cluster, V-RR B is the payment rate corresponding to the B object cluster, V-RP B The number of payments corresponding to the B object cluster.

[0081] According to the above calculation, the angle information Cosθ between A and B can be obtained, and then analyzed and processed according to the preset angle condition set by the B object cluster group. When the angle information Cosθ meets the preset angle condition, that is, the angle information is greater than the threshold in the preset angle condition, the multimedia resource to be recommended corresponding to the A drama is confirmed as the first multimedia resource, and the first multimedia resource is recommended to the objects in the B object cluster group.

[0082] In another possible implementation, drama A whose angle information Cosθ does not satisfy the preset angle condition set by object cluster group B is not recommended to objects in object cluster group B, and cosine similarity calculation is performed between drama A and preset vector information of other object cluster groups to re-determine the corresponding angle information, which can improve the accuracy of multimedia resource recommendations.

[0083] Figure 3 is a device block diagram of a multimedia resource recommendation system according to an exemplary embodiment. Figure 3 , the device may include:

[0084] The information acquisition module 301 is used to acquire object attribute information and object historical operation information of each of the multiple objects; the object historical operation information is used to represent the operation information generated by each of the multiple objects in the process of browsing multimedia resources;

[0085] A feature extraction module 303 is used to perform data cleaning and feature extraction processing on the object historical operation information to obtain key historical operation information of the object;

[0086] In a possible implementation, the feature extraction module 303 includes:

[0087] A noise data information determination module, configured to determine historical operation information of an object whose multimedia browsing amount is less than a second preset browsing amount as noise data information;

[0088] A removal module, used for removing the object historical operation information containing abnormal data information and the noise data information, to obtain the target object historical operation information;

[0089] The object key historical operation information confirmation module is used to perform feature extraction processing on the target object historical operation information to obtain the object key historical operation information corresponding to the multiple objects.

[0090] In another possible implementation, the feature extraction module 303 further includes:

[0091] An abnormal historical operation information confirmation module is used to determine the object historical operation information with abnormal data information and noise data information as abnormal historical operation information;

[0092] A target object cluster group classification module is used to classify objects corresponding to abnormal historical operation information into target object cluster groups;

[0093] A second multimedia resource acquisition module, used to acquire a second multimedia resource whose object browsing amount exceeds a first preset browsing amount;

[0094] The second multimedia resource recommendation module is used to recommend the second multimedia resource to the objects in the target object cluster group.

[0095] A clustering processing module 305 is used to perform clustering processing on the multiple objects according to the object attribute information and the key historical operation information of the objects to obtain multiple object cluster groups;

[0096] In a possible implementation, the clustering processing module 305 includes:

[0097] An object weight information determination module is used to determine object weight information corresponding to each of a plurality of objects from object key historical operation information and object attribute information;

[0098] A vector information determination module, used to determine the vector information of each of the multiple objects according to the object weight information;

[0099] The object clustering module is used for density-based spatial clustering algorithm to perform clustering processing on the vector information of multiple objects to obtain multiple object cluster groups.

[0100] In another possible implementation, the clustering processing module 305 further includes:

[0101] A target object acquisition module, used to acquire a target object other than a plurality of object clusters from a plurality of objects, and to acquire a second multimedia resource whose object browsing volume exceeds a first preset browsing volume;

[0102] The target object recommendation module is used to recommend the second multimedia resource to the target object.

[0103] The multimedia resource recommendation module 307 is used to select the first multimedia resources corresponding to each of the plurality of object clusters from the plurality of multimedia resources to be recommended, and recommend the first multimedia resources corresponding to each object cluster to the objects in each object cluster.

[0104] In a possible implementation, the multimedia resource recommendation module 307 includes:

[0105] A module for acquiring characteristic information of multimedia to be recommended, used for acquiring a plurality of multimedia resources to be recommended and characteristic information of multimedia to be recommended corresponding to each of the plurality of multimedia resources to be recommended;

[0106] A module for determining feature vector information to be recommended, used for determining feature vector information to be recommended corresponding to each of a plurality of multimedia resources to be recommended according to the feature information of the multimedia resources to be recommended;

[0107] An angle information determination module, used to determine angle information of preset vector information and feature vector information to be recommended corresponding to each of a plurality of object clusters based on a cosine similarity formula;

[0108] The first multimedia resource determination module is used to filter out target angle information that meets preset angle conditions corresponding to multiple object clusters from the angle information, and confirm the multimedia resource to be recommended corresponding to the target angle information as the first multimedia resource.

[0109] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0110] Figure 4 is a block diagram of an electronic device for multimedia resource recommendation according to an exemplary embodiment. The electronic device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 4As shown. The electronic device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for realizing multimedia resources is implemented. The display screen of the electronic device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the electronic device can be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the electronic device, or an external keyboard, touchpad or mouse, etc.

[0111] Those skilled in the art will understand that Figure 4 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0112] Figure 5 is a block diagram of an electronic device for multimedia resource recommendation according to an exemplary embodiment. The electronic device may be a server, and its internal structure diagram may be as shown in FIG. Figure 5 As shown. The electronic device includes a processor, a memory and a network interface connected through a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for recommending multimedia resources is implemented.

[0113] Those skilled in the art will understand that Figure 5 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0114] In an exemplary embodiment, an electronic device is also provided, including: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement the multimedia resource recommendation method in the embodiment of the present application.

[0115] In an exemplary embodiment, a computer-readable storage medium is also provided, and when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device can perform the multimedia resource recommendation method in the embodiment of the present application. The computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0116] In an exemplary embodiment, a computer program product including instructions is also provided. When the computer program product is executed on a computer, the computer executes the multimedia resource recommendation method in the embodiment of the present application.

[0117] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing related hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0118] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0119] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A multimedia resource recommendation method, characterized in that: include: Acquire object attribute information and object historical operation information of each of the plurality of objects; the object historical operation information is used to represent the operation information generated by each of the plurality of objects in the process of browsing multimedia resources; Performing data cleaning and feature extraction processing on the historical operation information of the object to obtain key historical operation information of the object; Clustering the multiple objects according to the object attribute information and the object key historical operation information to obtain multiple object cluster groups; The first multimedia resources corresponding to each of the plurality of object clusters are screened out from the plurality of multimedia resources to be recommended, and the first multimedia resources corresponding to each object cluster are recommended to the objects in each object cluster.

2. The multimedia resource recommendation method according to claim 1, characterized in that: The clustering of the multiple objects according to the object attribute information and the object key historical operation information to obtain multiple object cluster groups includes: Determining object weight information corresponding to each of the plurality of objects from the object key historical operation information and the object attribute information; Determining vector information of each of the plurality of objects according to the object weight information; A density-based spatial clustering algorithm is used to perform clustering processing on the vector information of each of the plurality of objects to obtain the plurality of object cluster groups.

3. The multimedia resource recommendation method according to claim 1, characterized in that: The step of selecting the first multimedia resources corresponding to each of the plurality of object clusters from the plurality of multimedia resources to be recommended comprises: Acquire the plurality of multimedia resources to be recommended and the multimedia feature information to be recommended corresponding to each of the plurality of multimedia resources to be recommended; Determining, according to the multimedia feature information to be recommended, feature vector information to be recommended corresponding to each of the plurality of multimedia resources to be recommended; Based on the cosine similarity formula, determining the angle information between the preset vector information corresponding to each of the plurality of object clusters and the feature vector information to be recommended; Target angle information satisfying preset angle conditions corresponding to each of the plurality of object clusters is screened out from the angle information, and the multimedia resource to be recommended corresponding to the target angle information is confirmed as the first multimedia resource.

4. The multimedia resource recommendation method according to claim 1 or 2, characterized in that: The method further comprises: Acquire a target object other than the plurality of object clusters from the plurality of objects, and acquire a second multimedia resource whose object browsing volume exceeds a first preset browsing volume; The second multimedia resource is recommended to the target object.

5. The multimedia resource recommendation method according to claim 1, characterized in that: The data cleaning and feature extraction processing of the object historical operation information to obtain the object key historical operation information includes: Determine the object historical operation information whose multimedia browsing amount is less than a second preset browsing amount as noise data information; Eliminating the object historical operation information that contains abnormal data information and the noise data information to obtain the target object historical operation information; Feature extraction processing is performed on the target object historical operation information to obtain the object key historical operation information corresponding to the multiple objects.

6. The multimedia resource recommendation method according to claim 5, characterized in that: The method further comprises: determining the object historical operation information having the abnormal data information and the noise data information as abnormal historical operation information; Classifying the objects corresponding to the abnormal historical operation information into target object cluster groups; Acquire a second multimedia resource whose object browsing amount exceeds a first preset browsing amount; The second multimedia resource is recommended to the objects in the target object cluster.

7. A multimedia resource recommendation system, characterized in that: include: An information acquisition module, used to acquire object attribute information and object historical operation information of each of the multiple objects; the object historical operation information is used to represent the operation information generated by each of the multiple objects in the process of browsing multimedia resources; A feature extraction module is used to perform data cleaning and feature extraction processing on the object historical operation information to obtain key historical operation information of the object; A clustering processing module, used for performing clustering processing on the multiple objects according to the object attribute information and the key historical operation information of the objects to obtain multiple object cluster groups; The multimedia resource recommendation module is used to select the first multimedia resources corresponding to each of the plurality of object clusters from the plurality of multimedia resources to be recommended, and recommend the first multimedia resources corresponding to each object cluster to the objects in each object cluster.

8. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the multimedia resource recommendation method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the multimedia resource recommendation method as claimed in any one of claims 1 to 6.

10. A computer program product comprising computer instructions, characterized in that When the computer instructions are executed by a processor, the multimedia resource recommendation method according to any one of claims 1 to 6 is implemented.

Images