A multimedia resource recommendation method, device and storage medium

Abstract

The present disclosure discloses a multimedia resource recommendation method, device and storage medium, which relates to the field of mobile Internet technology, in order to at least solve the problem of low accuracy of recommending multimedia resources in related technologies. In this method, multimedia resources are selected through multiple dimensions according to the feature information of candidate multimedia resources, a set of multimedia resources under different dimensions is obtained, and the multimedia resources selected in each dimension are comprehensively sorted to obtain the final screening result. In this way, by selecting multimedia resources in multiple dimensions, it can be ensured that high-quality candidate multimedia resources are not filtered out, thereby improving the accuracy of recommending multimedia resources.

Description

Technical Field

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

[0002] With the development of the times, electronic devices have become a part of life. In order to meet the needs of users, various application software are usually downloaded and installed in electronic devices. Users can watch videos, listen to music, etc. through application software. In order to recommend multimedia resources such as videos, music or news that users are interested in, application software uses recommendation algorithms to select multimedia resources that users are interested in from the resource library.

[0003] However, the existing recommendation algorithm obtains a set of candidate multimedia resources, and then filters the candidate multimedia resources according to a unified rule to determine the recommendation result, such as filtering according to the number of times the candidate multimedia resources are displayed. However, due to the use of rigid rules to filter candidate multimedia resources, some high-quality candidate multimedia resources will be mistakenly killed, and experiments have found that the proportion of false positives is high, so the accuracy of recommending multimedia resources by the existing recommendation method is low. Summary of the invention

[0004] The embodiments of the present disclosure provide a multimedia resource recommendation method, device and storage medium to improve the accuracy of recommending multimedia resources.

[0005] According to a first aspect of an embodiment of the present disclosure, a multimedia resource recommendation method is provided, comprising:

[0006] Acquire a candidate multimedia resource set; wherein the candidate multimedia resource set includes at least two candidate multimedia resources;

[0007] According to the feature information of the candidate multimedia resources, the candidate multimedia resource set is screened in at least two dimensions to obtain a set of multimedia resources to be recommended corresponding to each dimension; wherein the feature information is used to characterize the state information and attribute information of the candidate multimedia resources;

[0008] The weights of the multimedia resources in each set of multimedia resources to be recommended are modified by using a preset weight factor to obtain screening parameters of each multimedia resource;

[0009] The multimedia resources are filtered according to the filtering parameters.

[0010] In a possible implementation, the at least two dimensions include the historical display times of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained in the following manner:

[0011] Acquire the historical display times of the candidate multimedia resource from the feature information of the candidate multimedia resource;

[0012] Classifying the candidate multimedia resources according to a preset number of historical displays and a number of historical displays of the candidate multimedia resources;

[0013] Performing Bayesian statistical calculation on the historical display times of the candidate multimedia resources in each category to determine a first threshold for each category;

[0014] A set of candidate multimedia resources in each category whose historical display times are greater than the first threshold is used as a first set of multimedia resources to be recommended.

[0015] In a possible implementation, the at least two dimensions include characteristic parameters of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained in the following manner:

[0016] Determine the characteristic parameters of the candidate multimedia resources by extracting the characteristic information of the candidate multimedia resources;

[0017] A set of candidate multimedia resources whose characteristic parameters are greater than a second threshold is used as a second set of multimedia resources to be recommended.

[0018] In a possible implementation, the at least two dimensions include the enhancement parameters of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained in the following manner:

[0019] Acquire the life cycle state of the candidate multimedia resource from the feature information of the candidate multimedia resource;

[0020] Inputting the life cycle state of the candidate multimedia resource into a life cycle reinforcement learning neural network model to obtain reinforcement parameters of the candidate multimedia resource;

[0021] A set of candidate multimedia resources whose enhanced parameters are greater than a third threshold is used as a third set of multimedia resources to be recommended.

[0022] In a possible implementation, if the same multimedia resources exist in each set of multimedia resources to be recommended, the weights of the multimedia resources in each set of multimedia resources to be recommended are modified by a preset weight factor to obtain screening parameters of each multimedia resource, including:

[0023] The weights of the multimedia resources in the to-be-recommended multimedia resource set corresponding to the screening conditions of each dimension are modified by a preset weight factor to obtain a modified parameter of each multimedia resource;

[0024] The correction parameters of the same multimedia resources in each set of multimedia resources to be recommended are summed to obtain the screening parameters of the multimedia resources.

[0025] According to a second aspect of an embodiment of the present disclosure, a multimedia resource recommendation device is provided, including:

[0026] A first acquisition unit is configured to execute acquisition of a candidate multimedia resource set; wherein the candidate multimedia resource set includes at least two candidate multimedia resources;

[0027] A first screening unit is configured to screen the candidate multimedia resource set in at least two dimensions according to the feature information of the candidate multimedia resource, to obtain a set of multimedia resources to be recommended corresponding to each dimension; wherein the feature information is used to characterize the state information and attribute information of the candidate multimedia resource;

[0028] A correction unit is configured to correct the weights of the multimedia resources in each set of multimedia resources to be recommended by using a preset weight factor to obtain a screening parameter of each multimedia resource;

[0029] The second screening unit is configured to screen multimedia resources according to the screening parameters.

[0030] In a possible implementation, at least two dimensions include the historical display times of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained by the following means:

[0031] An acquisition unit, configured to acquire a historical display count of the candidate multimedia resource from the feature information of the candidate multimedia resource;

[0032] A classification unit, configured to classify the candidate multimedia resources according to a preset historical display number and a historical display number of the candidate multimedia resources;

[0033] A calculation unit configured to perform Bayesian statistical calculation on the historical display times of the candidate multimedia resources in each category to determine a first threshold value for each category;

[0034] The first determining unit is configured to execute the process of taking a set of candidate multimedia resources in each category whose historical display times are greater than the first threshold as a first set of multimedia resources to be recommended.

[0035] In a possible implementation manner, the at least two dimensions include characteristic parameters of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained by the following means:

[0036] A feature extraction unit, configured to perform feature extraction on feature information of the candidate multimedia resource to determine feature parameters of the candidate multimedia resource;

[0037] The second determining unit is configured to take a set of candidate multimedia resources whose characteristic parameters are greater than a second threshold as a second set of multimedia resources to be recommended.

[0038] In a possible implementation manner, the at least two dimensions include the enhancement parameters of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained by the following means:

[0039] A second acquisition unit is configured to acquire the life cycle state of the candidate multimedia resource from the feature information of the candidate multimedia resource;

[0040] An input unit, configured to input the life cycle state of the candidate multimedia resource into a life cycle reinforcement learning neural network model to obtain reinforcement parameters of the candidate multimedia resource;

[0041] The third determining unit is configured to take a set of candidate multimedia resources whose enhanced parameters are greater than a third threshold as a third set of multimedia resources to be recommended.

[0042] In a possible implementation, if the same multimedia resources exist in each set of multimedia resources to be recommended, the correction unit includes:

[0043] The multimedia resource weighting subunit is configured to modify the weights of the multimedia resources in the to-be-recommended multimedia resource set corresponding to the screening conditions of each dimension by using a preset weight factor to obtain modification parameters of each multimedia resource;

[0044] The summing subunit is configured to sum the correction parameters of the same multimedia resources in each set of multimedia resources to be recommended, so as to obtain the screening parameters of the multimedia resources.

[0045] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0046] processor;

[0047] a memory for storing instructions executable by the processor;

[0048] Wherein, the processor is configured to execute the instructions to implement a multimedia resource recommendation method;

[0049] According to a fourth aspect of an embodiment of the present disclosure, a storage medium is provided. When instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute a multimedia resource recommendation method.

[0050] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, comprising at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the multimedia resource recommendation method provided by an embodiment of the present disclosure.

[0051] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:

[0052] According to the characteristic information of the candidate multimedia resources, multimedia resources are selected through multiple dimensions to obtain a set of multimedia resources under different dimensions, and the multimedia resources selected in each dimension are comprehensively sorted to obtain the final screening result. In this way, by selecting multimedia resources through multiple dimensions, it can be ensured that high-quality candidate multimedia resources are not filtered out, thereby improving the accuracy of recommending multimedia resources.

[0053] Other features and advantages of the present disclosure will be described in the following description, and partly become apparent from the description, or be understood by practicing the present disclosure. The purpose and other advantages of the present disclosure can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The drawings described herein are used to provide a further understanding of the present disclosure and constitute a part of the present disclosure. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure. In the drawings:

[0055] Figure 1 A flowchart of a recommended method in the prior art;

[0056] Figure 2 A flowchart of a multimedia resource recommendation method in an embodiment of the present disclosure;

[0057] Figure 3 Schematic diagram of the first screening method in the embodiment of the present disclosure;

[0058] Figure 4 Schematic diagram of the second screening method in the embodiment of the present disclosure;

[0059] Figure 5 Schematic diagram of the third screening method in the embodiment of the present disclosure;

[0060] Figure 6 A schematic diagram of the structure of a multimedia resource recommendation device according to an embodiment of the present disclosure;

[0061] Figure 7 It is a schematic diagram of the terminal device structure in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0062] In order to improve the accuracy of recommending multimedia resources to users, a multimedia resource recommendation method, device and storage medium are provided in the embodiments of the present disclosure. In order to enable ordinary people in the art to better understand the technical solution of the present disclosure, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.

[0063] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure 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 interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0064] The technical solution provided by the embodiments of the present disclosure is described below in conjunction with the accompanying drawings.

[0065] In the information age, both information consumers and information producers are facing great challenges. As an information consumer, it is difficult to find the information of interest from a large amount of information; as an information producer, it is also difficult to make your information stand out and attract the attention of the majority of users. Recommendation system is the main tool to solve this contradiction.

[0066] The recommendation system can recommend multimedia information that information consumers like to information consumers, and it can also recommend high-quality multimedia resources produced by information producers to information consumers. Taking short videos as an example, in a mature video community, producers will continuously upload their own videos. The recommendation system needs to filter out some low-quality works that do not conform to the mainstream of the community under the complex situation of information overload, and reasonably select high-quality works as the candidate set of the recommendation model, so as to push the works that users like to users, while satisfying the personalized interests and diverse experiences of users.

[0067] An industrial recommendation system mainly includes two stages: recall and sorting. There is a screening process between these two stages, which screens the multimedia resources in the recall source and sorts the screened multimedia resources.

[0068] The existing recommendation algorithm obtains a set of candidate multimedia resources from the recall source, then filters the candidate multimedia resources according to a unified rule, determines the recommendation results, and recommends to users after sorting the recommendation results. Figure 1 As shown, it is a recommendation method in the prior art. After receiving the recommendation request, multimedia resources are screened from the i2i recall source, the u2i recall source and the FM recall source according to a unified rule, and the screened multimedia resources are sorted to obtain the recommendation result. However, due to the use of strong rules to filter candidate multimedia resources, some high-quality candidate multimedia resources will be mistakenly killed, and experiments have found that the proportion of mistaken killing is high, so the accuracy of the existing recommendation method in recommending multimedia resources is low.

[0069] In view of this, in order to solve the above problems, the present disclosure provides a multimedia resource recommendation method, which selects multimedia resources in multiple dimensions to obtain multimedia resource sets under different dimensions, and comprehensively sorts the multimedia resources selected in each dimension to obtain the final screening result. By selecting multimedia resources in multiple dimensions, it can be ensured that high-quality candidate multimedia resources are not filtered out, thereby improving the accuracy of recommending multimedia resources.

[0070] For ease of understanding, the technical solution provided by the present disclosure is further described below in conjunction with the accompanying drawings.

[0071] Figure 2 is a flowchart of a multimedia resource recommendation method according to an exemplary embodiment. Figure 2 As shown, the following steps are included.

[0072] In step S21, a candidate multimedia resource set is obtained; wherein the candidate multimedia resource set includes at least two candidate multimedia resources.

[0073] Among them, multimedia resources include music, video, news and other multimedia information.

[0074] The candidate multimedia resource set is a multimedia resource set obtained by roughly sorting the multimedia resources in the recall source.

[0075] In step S22, the candidate multimedia resource set is screened in at least two dimensions according to the feature information of the candidate multimedia resources to obtain a set of multimedia resources to be recommended corresponding to each dimension; wherein the feature information is used to characterize the state information and attribute information of the candidate multimedia resources.

[0076] In the disclosed embodiment, after obtaining the candidate multimedia resource set, the candidate multimedia resource set is screened in multiple dimensions according to the feature information of the candidate multimedia resources. Among them, the feature information of the candidate multimedia resources includes attribute features and statistical features; the attribute features can extract categories of different dimensions of multimedia resources (taking video as an example, categories of different dimensions include various portrait features of the video. For example, content features include scenery, people, animals, etc. For example, behavior features include dancing, driving, sports, etc.), producer attributes (producer attributes refer to the portrait features of the video author, including the gender, age, and region of the video author), etc. Statistical features can extract features such as its historical display times, clicks, reposts, likes, and playback completion.

[0077] In the embodiments of the present disclosure, the above dimensions include any one of the following:

[0078] First, the historical display times of the candidate multimedia resources;

[0079] In the embodiment of the present disclosure, when the candidate multimedia resources are screened based on the historical display times of the candidate multimedia resources as a dimension, the candidate multimedia resources can be classified according to the historical display times of the candidate multimedia resources, and the candidate multimedia resources that meet the conditions in each category are obtained, which can be specifically implemented as steps A1-A4:

[0080] Step A1: Acquire the historical display times of the candidate multimedia resource from the feature information of the candidate multimedia resource.

[0081] Step A2: classify the candidate multimedia resources according to a preset historical display number and the historical display number of the candidate multimedia resources.

[0082] The preset number of historical impressions is at least one. The competition tracks are divided according to the preset number of historical impressions, wherein the preset number of historical impressions is determined based on a statistical learning method.

[0083] For example, if there are two preset historical display times, 100 and 1000, then three competition tracks are divided according to the preset historical display times, and the range of each track is 0-100, 100-1000, and 1000-infinity. The candidate multimedia resources are divided into corresponding tracks according to their own historical display times.

[0084] Step A3: Perform Bayesian statistical calculation on the historical display times of the candidate multimedia resources in each category to determine the first threshold of each category.

[0085] Each track has a first threshold.

[0086] Step A4: taking a set of candidate multimedia resources in each category whose historical display times are greater than the first threshold as a first set of multimedia resources to be recommended.

[0087] In the disclosed embodiment, candidate multimedia resources are screened by comparing the first threshold with the size of the historical display times of the candidate multimedia resources. For example: it is divided into three tracks, and the range of each track is 0-100, 100-1000, and 1000-infinity, and the first threshold of each track is 80, 860, and 5000, respectively. When screening, multimedia resources with a historical display time greater than 80 are selected from the first track; multimedia resources with a historical display time greater than 860 are selected from the second track; multimedia resources with a historical display time greater than 5000 are selected from the third track, and the multimedia resources screened out from the three tracks are used as the first set of multimedia resources to be recommended.

[0088] like Figure 3 As shown in FIG. 1 , it is a flowchart of the first screening method, wherein the candidate multimedia resources are obtained by roughly sorting the multimedia resources in various recall sources.

[0089] By developing several tracks, using the Bayesian statistical model to calculate the first threshold in each track, and screening out the multimedia resources to be recommended based on the first threshold of each track, the quality of the multimedia resources to be recommended can be improved by screening the candidate multimedia resources with different display times with different standards, and the accuracy of the recommended multimedia resources can be further improved.

[0090] Second, characteristic parameters of candidate multimedia resources;

[0091] The feature parameters are obtained by extracting feature information of candidate multimedia resources.

[0092] In the embodiment of the present disclosure, when the candidate multimedia resources are screened by the characteristic parameters of the candidate multimedia resources, the multimedia resources to be recommended are screened by the magnitude relationship between the characteristic parameters of the candidate multimedia resources and the second threshold value, which can be specifically implemented as steps B1-B2:

[0093] Step B1: Determine feature parameters of the candidate multimedia resources by extracting feature information of the candidate multimedia resources.

[0094] By extracting the feature information of the candidate multimedia resource, a feature vector representing the feature information of the candidate multimedia resource is obtained, and the feature vector is scored by a scoring formula trained by a supervised learning method to obtain the feature parameters of the candidate multimedia resource.

[0095] As platform traffic grows and user interests continue to evolve and be mined, the fluctuations in historical statistical indicators of multimedia resources will be more obvious; at the same time, at certain stages of the product cycle, it is also expected to introduce high-quality content with specific characteristics to adjust the sensory tone of multimedia resources. By using the previously collected multimedia resource quality tag data and a portion of manually annotated high-quality data, a supervised learning algorithm is used to train a scoring formula, and the result of the formula can be used as a feature parameter of the multimedia resource. The learning algorithm can use rulefit (rule fitting) with strong explanatory power.

[0096] Step B2: taking a set of candidate multimedia resources whose characteristic parameters are greater than a second threshold as a second set of multimedia resources to be recommended.

[0097] like Figure 4 As shown, it is a flow chart of the second screening method, wherein the feature parameters are obtained by extracting the features of the candidate multimedia resources, and the feature parameters are compared with the second threshold value to determine the second set of multimedia resources to be recommended.

[0098] In this way, the characteristic parameters of candidate multimedia resources are used as an important basis for screening multimedia resources, which improves the generalization performance of the original algorithm and ensures the long-term stability of the effect, thereby improving the quality of the multimedia resources to be recommended and further improving the accuracy of recommending multimedia resources.

[0099] Third, the enhancement parameters of the candidate multimedia resources;

[0100] The enhancement parameter is obtained according to the life cycle status of the candidate multimedia resource.

[0101] In the embodiment of the present disclosure, when the candidate multimedia resources are screened by the enhancement parameters of the candidate multimedia resources, the multimedia resources to be recommended are screened by the magnitude relationship between the enhancement parameters of the candidate multimedia resources and the third threshold value, which can be specifically implemented as steps C1-C3:

[0102] Step C1: Acquire the life cycle status of the candidate multimedia resource from the feature information of the candidate multimedia resource.

[0103] Like anything else, a multimedia resource also goes through stages such as gestation, birth, growth, maturity, and decline, which is generally called the life cycle of multimedia resources. The characteristic information of the candidate multimedia resources can be used to determine which stage of the life cycle the candidate multimedia resources are in.

[0104] Step C2: inputting the life cycle state of the candidate multimedia resource into the life cycle reinforcement learning neural network model to obtain the reinforcement parameters of the candidate multimedia resource.

[0105] Step C3: taking a set of candidate multimedia resources whose enhanced parameters are greater than a third threshold as a third set of multimedia resources to be recommended.

[0106] like Figure 5 As shown, it is a flow chart of the third screening method. In which, the life cycle state of the multimedia resources is determined, and the life cycle state is analyzed to obtain the enhancement parameter, and finally the enhancement parameter is compared with the third threshold to determine the third set of multimedia resources to be recommended.

[0107] In step S23, the weights of the multimedia resources in each set of multimedia resources to be recommended are modified by a preset weight factor to obtain screening parameters of each multimedia resource.

[0108] In step S24, multimedia resources are screened according to the screening parameters.

[0109] In the embodiment of the present disclosure, after obtaining the first set of multimedia resources to be recommended, the second set of multimedia resources to be recommended, and the third set of multimedia resources to be recommended, the three sets need to be screened again.

[0110] The preset weight may be pre-set or determined based on experience. The weight of the multimedia resource is considered to be 1 before being modified.

[0111] In an embodiment of the present disclosure, a preset weight factor is multiplied by the weight of a multimedia resource to obtain a screening parameter for the multimedia resource. For example, if there are three multimedia resources, namely a1, a2 and a3, and their corresponding weight factors are w1, w2 and w3, then after correction, the screening parameters of the three multimedia resources are a1w1, a2w2 and a3w3. And the multimedia resources are screened according to the obtained screening parameters. Multimedia resources can be selected according to the numerical value of the screening parameter, for example, multimedia resources with a higher order are selected in order from large to small. Multimedia resources can also be selected from multimedia resources whose screening parameters exceed a threshold, which is not limited by the present disclosure.

[0112] It should be noted that when filtering, multimedia resources under all dimensions are filtered.

[0113] In this way, by screening again, the quality of multimedia resources can be further improved, thereby improving the accuracy of recommending multimedia resources.

[0114] However, in the embodiment of the present disclosure, among the multimedia resource sets to be recommended determined by the above three dimensions, there may be a multimedia resource that exists in multiple multimedia resource sets to be recommended. In this case, it is necessary to perform weighted summation on the multimedia resource to obtain the final screening parameter, which can be specifically implemented as follows:

[0115] The weights of the multimedia resources in the multimedia resource set to be recommended corresponding to the screening conditions of each dimension are corrected with a preset weight factor to obtain the correction parameters of each multimedia resource; the correction parameters of the same multimedia resources in each multimedia resource set to be recommended are summed to obtain the screening parameters of the multimedia resources.

[0116] For example: if there is a multimedia resource that appears in the set of multimedia resources to be recommended in three dimensions, the correction parameters under each dimension are first determined, and the correction parameters under each dimension are summed to obtain the screening parameters. For example: a multimedia resource a1 has a preset weight of w1 in the first dimension, a preset weight of w2 in the second dimension, and a preset weight of w3 in the third dimension, then the screening parameter of the multimedia resource is (w1+w2+w3)a1.

[0117] In this way, by performing weighted sum calculation on the same multimedia resources in different sets, high-quality multimedia resources can be determined more accurately, thereby improving the accuracy of recommending multimedia resources.

[0118] By selecting multimedia resources in multiple dimensions, a set of multimedia resources under different dimensions is obtained, and the multimedia resources selected in each dimension are comprehensively sorted to obtain the final screening result. In this way, by selecting multimedia resources in multiple dimensions, it can be ensured that high-quality candidate multimedia resources are not filtered out, thereby improving the overall quality of multimedia resources and further improving the accuracy of recommending multimedia resources.

[0119] Based on the same inventive concept, the present disclosure also provides a multimedia resource recommendation device. Figure 6 FIG. 1 is a schematic diagram of a multimedia resource recommendation device provided by the present disclosure. The device includes:

[0120] The first acquisition unit 601 is configured to execute acquisition of a candidate multimedia resource set; wherein the candidate multimedia resource set includes at least two candidate multimedia resources;

[0121] The first screening unit 602 is configured to screen the candidate multimedia resource set in at least two dimensions according to the feature information of the candidate multimedia resource, and obtain a set of multimedia resources to be recommended corresponding to each dimension; wherein the feature information is used to represent the state information and attribute information of the candidate multimedia resource;

[0122] The correction unit 603 is configured to correct the weight of each multimedia resource in the set of multimedia resources to be recommended by using a preset weight factor to obtain a screening parameter of each multimedia resource;

[0123] The second screening unit 604 is configured to screen multimedia resources according to the screening parameters.

[0124] In a possible implementation, at least two dimensions include the historical display times of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained by the following means:

[0125] An acquisition unit, configured to acquire a historical display count of the candidate multimedia resource from the feature information of the candidate multimedia resource;

[0126] A classification unit, configured to classify the candidate multimedia resources according to a preset historical display number and a historical display number of the candidate multimedia resources;

[0127] A calculation unit configured to perform Bayesian statistical calculation on the historical display times of the candidate multimedia resources in each category to determine a first threshold value for each category;

[0128] The first determining unit is configured to execute the process of taking a set of candidate multimedia resources in each category whose historical display times are greater than the first threshold as a first set of multimedia resources to be recommended.

[0129] In a possible implementation manner, the at least two dimensions include characteristic parameters of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained by the following means:

[0130] A feature extraction unit, configured to perform feature extraction on feature information of the candidate multimedia resource to determine feature parameters of the candidate multimedia resource;

[0131] The second determining unit is configured to take a set of candidate multimedia resources whose characteristic parameters are greater than a second threshold as a second set of multimedia resources to be recommended.

[0132] In a possible implementation manner, the at least two dimensions include the enhancement parameters of the candidate multimedia resources, and the set of multimedia resources to be recommended corresponding to the dimension is obtained by the following means:

[0133] A second acquisition unit is configured to acquire the life cycle state of the candidate multimedia resource from the feature information of the candidate multimedia resource;

[0134] An input unit, configured to input the life cycle state of the candidate multimedia resource into a life cycle reinforcement learning neural network model to obtain reinforcement parameters of the candidate multimedia resource;

[0135] The third determining unit is configured to take a set of candidate multimedia resources whose enhanced parameters are greater than a third threshold as a third set of multimedia resources to be recommended.

[0136] In a possible implementation, if the same multimedia resources exist in each set of multimedia resources to be recommended, the correction unit 603 includes:

[0137] The multimedia resource weighting subunit is configured to modify the weights of the multimedia resources in the to-be-recommended multimedia resource set corresponding to the screening conditions of each dimension by using a preset weight factor to obtain modification parameters of each multimedia resource;

[0138] The summing subunit is configured to sum the correction parameters of the same multimedia resources in each set of multimedia resources to be recommended, so as to obtain the screening parameters of the multimedia resources.

[0139] like Figure 7 As shown, based on the same technical concept, the embodiment of the present disclosure further provides an electronic device 70 which may include a memory 701 and a processor 702 .

[0140] The memory 701 is used to store computer programs executed by the processor 702. The memory 701 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the task management device, etc. The processor 702 may be a central processing unit (CPU), or a digital processing unit, etc. The specific connection medium between the memory 701 and the processor 702 is not limited in the embodiments of the present disclosure. The embodiments of the present disclosure are Figure 7 In the embodiment, the memory 701 and the processor 702 are connected via a bus 703. The bus 703 is connected to the processor 702 via a bus 703. Figure 7 The connection between other components is shown by bold lines, which is only for schematic illustration and is not intended to be limiting. The bus 703 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0141] The memory 701 may be a volatile memory, such as a random-access memory (RAM); the memory 701 may also be a non-volatile memory, such as a read-only memory, a flash memory, a hard disk drive (HDD) or a solid-state drive (SSD), or the memory 701 may be any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 701 may be a combination of the above memories.

[0142] The processor 702 is used to execute the following when calling the computer program stored in the memory 701: Figure 2 The method performed by the device in the embodiment shown in FIG.

[0143] In some possible implementations, various aspects of the method provided by the present disclosure may also be implemented in the form of a program product, which includes a program code. When the program product is run on a computer device, the program code is used to enable the computer device to execute the steps of the method according to various exemplary embodiments of the present disclosure described above in this specification. For example, the computer device may execute the following steps: Figure 2 to Figure 6 The method performed by the device in the embodiment shown in FIG.

[0144] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable 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 above.

[0145] Although the preferred embodiments of the present disclosure have been described, those skilled in the art may make additional changes and modifications to these embodiments once they are aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present disclosure. Those skilled in the art will readily come to think of other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The description and embodiments are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0146] It should be understood that the present disclosure is not limited to the exact 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 disclosure is limited only by the appended claims.

Claims

1. A multimedia resource recommendation method, characterized in that: The method comprises: Acquire a candidate multimedia resource set; wherein the candidate multimedia resource set includes at least three candidate multimedia resources; According to the feature information of the candidate multimedia resources, the candidate multimedia resource set is screened in at least three dimensions to obtain a set of multimedia resources to be recommended corresponding to each dimension; wherein the feature information is used to characterize the state information and attribute information of the candidate multimedia resources; The weights of the multimedia resources in each set of multimedia resources to be recommended are modified by using a preset weight factor to obtain screening parameters of each multimedia resource; Filter multimedia resources according to the filtering parameters; The at least three dimensions include a dimension of the number of historical displays of the candidate multimedia resource, a dimension of characteristic parameters of the candidate multimedia resource, and a dimension of enhanced parameters of the candidate multimedia resource; The feature parameters in the feature parameter dimension of the candidate multimedia resource are obtained by extracting the feature information of the candidate multimedia resource to obtain a feature vector for representing the feature information, and scoring the feature vector through a scoring formula trained by a supervised learning method; If the same multimedia resources exist in each set of multimedia resources to be recommended, the weights of the multimedia resources in each set of multimedia resources to be recommended are modified by a preset weight factor to obtain screening parameters of each multimedia resource, including: The weights of the multimedia resources in the to-be-recommended multimedia resource set corresponding to the screening conditions of each dimension are modified by a preset weight factor to obtain a modified parameter of each multimedia resource; The correction parameters of the same multimedia resources in each set of multimedia resources to be recommended are summed to obtain the screening parameters of the multimedia resources.

2. The multimedia resource recommendation method according to claim 1, characterized in that: The set of multimedia resources to be recommended corresponding to the dimension of the historical display times of the candidate multimedia resources is obtained in the following manner: Acquire the historical display times of the candidate multimedia resource from the feature information of the candidate multimedia resource; Classifying the candidate multimedia resources according to a preset number of historical displays and the number of historical displays of the candidate multimedia resources; Performing Bayesian statistical calculation on the historical display times of the candidate multimedia resources in each category to determine a first threshold for each category; A set of candidate multimedia resources in each category whose historical display times are greater than the first threshold is used as a first set of multimedia resources to be recommended.

3. The multimedia resource recommendation method according to claim 1, characterized in that: The set of multimedia resources to be recommended corresponding to the characteristic parameter dimensions of the candidate multimedia resources is obtained in the following manner: The set of candidate multimedia resources whose characteristic parameters are greater than a second threshold is taken as a second set of multimedia resources to be recommended.

4. The multimedia resource recommendation method according to claim 1, characterized in that: The set of multimedia resources to be recommended corresponding to the enhanced parameter dimension of the candidate multimedia resources is obtained in the following manner: Acquire the life cycle state of the candidate multimedia resource from the feature information of the candidate multimedia resource; Inputting the life cycle state of the candidate multimedia resource into a life cycle reinforcement learning neural network model to obtain reinforcement parameters of the candidate multimedia resource; A set of candidate multimedia resources whose enhanced parameters are greater than a third threshold is used as a third set of multimedia resources to be recommended.

5. A multimedia resource recommendation device, characterized in that: include: A first acquisition unit is configured to execute acquisition of a candidate multimedia resource set; wherein the candidate multimedia resource set includes at least three candidate multimedia resources; A first screening unit is configured to screen the candidate multimedia resource set in at least three dimensions according to the feature information of the candidate multimedia resource, and obtain a set of multimedia resources to be recommended corresponding to each dimension; wherein the feature information is used to represent the state information and attribute information of the candidate multimedia resource; A correction unit is configured to correct the weights of the multimedia resources in each set of multimedia resources to be recommended by using a preset weight factor to obtain a screening parameter of each multimedia resource; A second screening unit is configured to screen multimedia resources according to the screening parameters; The at least three dimensions include a dimension of the number of historical displays of the candidate multimedia resource, a dimension of characteristic parameters of the candidate multimedia resource, and a dimension of enhanced parameters of the candidate multimedia resource; The feature parameters in the feature parameter dimension of the candidate multimedia resource are obtained by extracting the feature information of the candidate multimedia resource to obtain a feature vector for representing the feature information, and scoring the feature vector through a scoring formula trained by a supervised learning method; If the same multimedia resources exist in each of the multimedia resource sets to be recommended, the correction unit includes: The multimedia resource weighting subunit is configured to modify the weights of the multimedia resources in the to-be-recommended multimedia resource set corresponding to the screening conditions of each dimension by using a preset weight factor to obtain modification parameters of each multimedia resource; The summing subunit is configured to sum the correction parameters of the same multimedia resources in each set of multimedia resources to be recommended, so as to obtain the screening parameters of the multimedia resources.

6. The device according to claim 5, characterized in that The set of multimedia resources to be recommended corresponding to the dimension of the historical display times of the candidate multimedia resources is obtained by the following means: An acquisition unit, configured to acquire a historical display count of the candidate multimedia resource from the feature information of the candidate multimedia resource; A classification unit, configured to classify the candidate multimedia resources according to a preset historical display number and a historical display number of the candidate multimedia resources; A calculation unit configured to perform Bayesian statistical calculation on the historical display times of the candidate multimedia resources in each category to determine a first threshold value for each category; The first determining unit is configured to execute the process of taking a set of candidate multimedia resources in each category whose historical display times are greater than the first threshold as a first set of multimedia resources to be recommended.

7. The device according to claim 5, characterized in that The set of multimedia resources to be recommended corresponding to the characteristic parameter dimensions of the candidate multimedia resources is obtained by the following means: The second determining unit is configured to take a set of candidate multimedia resources whose characteristic parameters are greater than a second threshold as a second set of multimedia resources to be recommended.

8. The device according to claim 5, characterized in that The set of multimedia resources to be recommended corresponding to the enhanced parameter dimension of the candidate multimedia resources is obtained by the following means: A second acquisition unit is configured to acquire the life cycle state of the candidate multimedia resource from the feature information of the candidate multimedia resource; An input unit, configured to input the life cycle state of the candidate multimedia resource into a life cycle reinforcement learning neural network model to obtain reinforcement parameters of the candidate multimedia resource; The third determining unit is configured to take a set of candidate multimedia resources whose enhanced parameters are greater than a third threshold as a third set of multimedia resources to be recommended.

9. 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 as claimed in any one of claims 1 to 4.

10. A storage medium, characterized in that: When the instructions in the 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 4.

Images