Multimedia resource tag construction method, device, electronic device and storage medium

Abstract

The present disclosure relates to a method, device, electronic device and storage medium for labeling multimedia resources. The method includes: acquiring multimedia resources, and performing content identification on the multimedia resources to obtain multimedia content information. Labeling the multimedia content information based on sub-tag information corresponding to each of a plurality of main tag information to obtain a plurality of labeling results. The plurality of main tag information are independent of each other. Based on the tag attribute information of the multimedia resources, multimedia resources are recommended. The method can obtain a plurality of labeling results at one time, thereby improving the efficiency of labeling. The method determines a matching sub-tag information as the target sub-tag information from each main tag information, labels the multimedia resources, and the plurality of labeling results obtained are independent of each other, which can avoid repeated construction and improve the accuracy of labeling.

Description

Technical Field

[0001] The present disclosure relates to the field of big data technology, and in particular to a method, device, electronic device and storage medium for constructing labels for multimedia resources. Background Art

[0002] With the continuous development of commercial advertising business, advertising information is becoming more and more abundant. Advertising tags can mark advertising information, so that the advertising system can identify the attributes of advertising information through advertising tags, obtain appropriate advertising information from a large amount of advertising information, and accurately deliver appropriate advertising information to users.

[0003] In the related art, the sources of advertising labels are very rich and the number of advertising labels is large, which leads to low efficiency in labeling advertising information and the situation of repeated construction of advertising labels, resulting in low accuracy of label construction. Summary of the invention

[0004] The present disclosure provides a method, device, electronic device and storage medium for labeling multimedia resources, so as to at least solve the problems of low labeling efficiency and low labeling accuracy in the related art. The technical solution of the present disclosure is as follows:

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

[0006] Obtain content information of multimedia resources to be tagged;

[0007] Based on the sub-tag information corresponding to each of the preset multiple main tag information and the content information, the multimedia resource is tagged to obtain multiple tag building results, each tag building result corresponds to a target sub-tag, the target sub-tag is the sub-tag information corresponding to the multimedia resource in each main tag information of the multiple main tag information, the main tag information is the category tag information corresponding to the sub-tag information of the same category, and the sub-tag information is at least one tag information under the category corresponding to each main tag information.

[0008] As an optional embodiment, the labeling of the multimedia resource is performed based on the sub-label information corresponding to each of the preset multiple main label information and the content information, and the multiple labeling results obtained include:

[0009] Matching each sub-tag information corresponding to each main tag information in the plurality of main tag information with the content information;

[0010] Using sub-tag information matching the content information as the target sub-tag;

[0011] A tag construction result between the target sub-tag corresponding to each main tag information and the multimedia resource is established.

[0012] As an optional embodiment, the method further includes:

[0013] Acquire a plurality of tag information to be processed, wherein the tag information to be processed is unclassified tag information obtained based on a plurality of tag sources;

[0014] Performing similarity matching on any two of the plurality of tag information to be processed to obtain a plurality of similarity matching results;

[0015] Classifying the tag information to be processed according to the multiple similarity matching results to obtain multiple target tag information groups, each target tag information group corresponding to a category information;

[0016] Using the category information corresponding to each target label information group as the main label information;

[0017] The tag information to be processed in each target tag information group is used as the sub-tag information corresponding to the main tag information.

[0018] As an optional embodiment, after performing similarity matching on any two of the plurality of tag information to be processed to obtain a plurality of similarity matching results, the method further includes:

[0019] Classifying the tag information to be processed according to the multiple similarity matching results to obtain multiple first tag information groups, wherein the multiple first tag information groups include at least two tag information groups corresponding to the same category information and tag information groups corresponding to different category information respectively;

[0020] Merging at least two label information groups corresponding to the same category of information to obtain at least one second label information group;

[0021] Deduplication processing is performed on the to-be-processed label information in each second label information group;

[0022] The label information groups corresponding to the different category information and the second label information group after deduplication processing are used as the target label information group.

[0023] As an optional embodiment, the method further includes:

[0024] Based on the multiple tag construction results, tag attribute information of the multimedia resource is generated.

[0025] As an optional embodiment, the step of obtaining content information of a multimedia resource to be tagged includes:

[0026] Based on a preset content feature extraction model, content features are extracted from the multimedia resources to obtain the content information.

[0027] As an optional embodiment, the method further includes:

[0028] Get the user tag information corresponding to the target object;

[0029] Acquire a plurality of to-be-recommended construction results corresponding to the to-be-recommended multimedia resources, wherein the plurality of to-be-recommended construction results are a plurality of tag construction results corresponding to the to-be-recommended multimedia resources;

[0030] Comparing the user tag information with the target sub-tag corresponding to each to-be-recommended build result;

[0031] In a case where the user tag information matches the target sub-tag corresponding to the to-be-recommended construction result, the to-be-recommended multimedia resource is recommended to the target object.

[0032] According to a second aspect of an embodiment of the present disclosure, a device for constructing a label of a multimedia resource is provided, the device comprising:

[0033] A multimedia content identification module is configured to obtain content information of a multimedia resource to be tagged;

[0034] The label construction module is configured to execute label construction for the multimedia resource based on the sub-label information corresponding to each of the preset multiple main label information and the content information, and obtain multiple label construction results, each label construction result corresponds to a target sub-label, and the target sub-label is the sub-label information corresponding to the multimedia resource in each main label information of the multiple main label information. The main label information is the category label information corresponding to the sub-label information of the same category, and the sub-label information is at least one label information under the category corresponding to each main label information.

[0035] As an optional embodiment, the label construction module includes:

[0036] a sub-tag matching unit, configured to match each sub-tag information corresponding to each main tag information in the plurality of main tag information with the content information;

[0037] a target sub-tag determining unit, configured to execute sub-tag information matching the content information as the target sub-tag;

[0038] The tag building result acquisition unit is configured to execute the establishment of a tag building result between the target sub-tag corresponding to each main tag information and the multimedia resource.

[0039] As an optional embodiment, the device further includes:

[0040] A tag information acquisition module to be processed is configured to execute acquisition of a plurality of tag information to be processed, wherein the tag information to be processed is unclassified tag information obtained based on a plurality of tag sources;

[0041] A similarity matching result acquisition module is configured to perform similarity matching on any two of the plurality of tag information to be processed to obtain a plurality of similarity matching results;

[0042] A first classification module is configured to classify the tag information to be processed according to the multiple similarity matching results to obtain multiple target tag information groups, each target tag information group corresponding to a category information;

[0043] A main label information determination module is configured to use the category information corresponding to each target label information group as the main label information;

[0044] The sub-tag information determining module is configured to use the tag information to be processed in each target tag information group as the sub-tag information corresponding to the main tag information.

[0045] As an optional embodiment, the device further includes:

[0046] A second classification module is configured to classify the tag information to be processed according to the multiple similarity matching results to obtain multiple first tag information groups, wherein the multiple first tag information groups include at least two tag information groups corresponding to the same category information and tag information groups corresponding to different category information respectively;

[0047] A same category merging module, configured to merge at least two label information groups corresponding to the same category information to obtain at least one second label information group;

[0048] a deduplication module, configured to perform deduplication processing on the to-be-processed label information in each second label information group;

[0049] The target label information group determination module is configured to execute the label information groups corresponding to the different category information and the second label information group after deduplication processing as the target label information group.

[0050] As an optional embodiment, the device further includes:

[0051] The tag attribute information generating module is configured to generate tag attribute information of the multimedia resource based on the multiple tag building results.

[0052] As an optional embodiment, the multimedia content recognition module includes:

[0053] The feature extraction unit is configured to perform content feature extraction on the multimedia resource based on a preset content feature extraction model to obtain the content information.

[0054] As an optional embodiment, the device further includes:

[0055] A user tag information acquisition module is configured to execute acquisition of user tag information corresponding to a target object;

[0056] A tag attribute information acquisition module is configured to execute acquisition of a plurality of to-be-recommended construction results corresponding to the to-be-recommended multimedia resource, wherein the plurality of to-be-recommended construction results are a plurality of tag construction results corresponding to the to-be-recommended multimedia resource;

[0057] A comparison module is configured to compare the user tag information with a target sub-tag corresponding to each to-be-recommended build result;

[0058] The recommendation module is configured to recommend the multimedia resource to be recommended to the target object when the user tag information matches the target sub-tag corresponding to the construction result to be recommended.

[0059] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0060] processor;

[0061] a memory for storing instructions executable by the processor;

[0062] The processor is configured to execute the instructions to implement the method for constructing a label for multimedia resources as described above.

[0063] According to a fourth aspect of an embodiment of the present disclosure, 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 can execute the method for constructing tags for multimedia resources as described above.

[0064] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the computer program implements the above-mentioned method for constructing labels for multimedia resources.

[0065] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:

[0066] Acquire multimedia resources, and perform content recognition on the multimedia resources to obtain multimedia content information. Label the multimedia content information based on the sub-label information corresponding to each of the multiple main label information to obtain multiple label construction results. The multiple main label information are independent of each other. Recommend multimedia resources based on the label attribute information of the multimedia resources. This method can obtain multiple label construction results at one time, thereby improving the efficiency of label construction. This method determines a matching sub-label information as the target sub-label information from each main label information, and labels the multimedia resources. The multiple label construction results obtained are independent of each other, which can avoid repeated construction and improve the accuracy of label construction.

[0067] 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 disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.

[0069] Figure 1 The diagram is a schematic diagram of a scenario of a method for constructing a label for multimedia resources according to an exemplary embodiment.

[0070] Figure 2 The figure is a flow chart of a method for constructing a label of a multimedia resource according to an exemplary embodiment.

[0071] Figure 3 The present invention is a flowchart of obtaining main tag information and sub-tag information in a tag construction method for multimedia resources according to an exemplary embodiment.

[0072] Figure 4 The diagram is a schematic diagram of main tag information and sub-tag information in a tag construction method for multimedia resources according to an exemplary embodiment.

[0073] Figure 5 The present invention is a flowchart of performing deduplication processing in a method for constructing a tag for multimedia resources according to an exemplary embodiment.

[0074] Figure 6 The present invention is a flowchart of label building in a label building method for multimedia resources according to an exemplary embodiment.

[0075] Figure 7 It is a schematic diagram of a matching result between sub-tag information and multimedia content information when constructing a tag in a method for constructing a tag of a multimedia resource according to an exemplary embodiment.

[0076] Figure 8 The present invention is a flowchart of recommending multimedia resources in a method for constructing multimedia resource tags according to an exemplary embodiment.

[0077] Fig. 9 The present invention is a block diagram of a device for constructing a label for multimedia resources according to an exemplary embodiment.

[0078] Fig.10 The invention is a block diagram of a server-side electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0079] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.

[0080] 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.

[0081] Figure 1 The present invention is a schematic diagram of an application scenario of a method for labeling multimedia resources according to an exemplary embodiment, and the application scenario includes a server 110 and a client 120. The server 110 obtains multimedia resources, and performs content recognition on the multimedia resources to obtain multimedia content information. The server 110 labels the multimedia content information based on the sub-label information corresponding to each of the multiple main label information, and obtains multiple label construction results. The server generates label attribute information of the multimedia resources based on the multiple label construction results. The server 110 recommends multimedia resources to the client 120 based on the label attribute information of the multimedia resources.

[0082] In the disclosed embodiment, the server 110 may include an independently operated server, a distributed server, or a server cluster composed of multiple servers. The server 110 may include a network communication unit, a processor, a memory, etc. Specifically, the server 110 may construct a tag for the multimedia content information based on the sub-tag information corresponding to each of the multiple main tag information, and generate tag attribute information of the multimedia resource.

[0083] In the disclosed embodiment, the client 120 includes physical devices such as smart phones, desktop computers, tablet computers, laptops, digital assistants, smart wearable devices, etc., and may also include software running in the physical devices, such as applications, etc. The operating system running on the physical device in the embodiment of the present application may include but is not limited to Android system, IOS system, Linux, Unix, Windows, etc. The multimedia resources sent by the server 110 are received based on API (Application Programming Interface).

[0084] Figure 2 is a flowchart of a method for constructing a label of a multimedia resource according to an exemplary embodiment. Figure 1 As shown, the method is used in a server and includes the following steps.

[0085] S210. Obtain content information of multimedia resources to be tagged;

[0086] As an optional embodiment, the multimedia resources may include advertising information in different forms, such as image advertisements, video advertisements, text advertisements, etc.

[0087] As an optional embodiment, the server can obtain a data stream related to multimedia resources, the data stream related to multimedia resources includes at least one multimedia resource, the server can obtain multimedia identification information, determine the multimedia resources corresponding to the multimedia identification information in the data stream through the multimedia identification information, and construct labels for the multimedia resources corresponding to the multimedia identification information.

[0088] As an optional embodiment, based on a preset content feature extraction model, content features are extracted from multimedia resources to obtain multimedia content information. The content feature extraction model may include an image feature extraction model, a voice feature extraction model, and a text feature extraction model.

[0089] As an optional embodiment, according to the type of multimedia resources, features of the multimedia resources may be extracted by using a content feature extraction model corresponding to the type of multimedia resources.

[0090] When the multimedia resource is video information or image information, image features can be extracted from key frames and image information in the video information based on an image recognition model. When the multimedia resource is video information, voice features can also be extracted from the voice information in the audio track of the video information. The server can also obtain information associated with the multimedia resource, such as the publisher, introduction, theme, etc. of the multimedia resource from the multimedia data stream, and perform feature extraction on the multimedia-related information. For example, when the publisher is a certain game software, it can be determined that the multimedia resource is a multimedia resource related to the game industry. The feature extraction results of the multimedia resource and the feature extraction results of the multimedia-related information can both be used as multimedia content information.

[0091] By performing content recognition on multimedia resources, multimedia content information is obtained, so that in subsequent steps, a target sub-tag corresponding to the multimedia content information is determined, and a corresponding relationship between the multimedia resources and the target sub-tag is established, thereby improving the accuracy of tag construction.

[0092] As an alternative embodiment, see Figure 3 , the method further comprises:

[0093] S310. Acquire multiple tag information to be processed, where the tag information to be processed is unclassified tag information obtained based on multiple tag sources;

[0094] S320. Perform similarity matching on any two of the plurality of tag information to be processed to obtain a plurality of similarity matching results;

[0095] S330. According to the multiple similarity matching results, the tag information to be processed is classified to obtain multiple target tag information groups, each target tag information group corresponds to a category information;

[0096] S340. Using the category information corresponding to each target label information group as the main label information;

[0097] S350. Use the tag information to be processed in each target tag information group as the sub-tag information corresponding to the main tag information.

[0098] As an optional embodiment, the server can obtain multiple tag information to be processed through different tag sources, and the tag sources may include publishers, topics, products, publishing locations, publishing requirements, publishing time, etc. For example, the tag information to be processed is extracted through the relevant information of the publisher, the tag information to be processed is extracted through the relevant information of the topic, and the tag information to be processed is extracted through the relevant information of the product.

[0099] As an optional embodiment, multiple tag information to be processed are classified, and the similarity between any two tag information to be processed among the multiple tag information to be processed is calculated to obtain multiple similarity matching results. A similarity matching result with a similarity greater than a preset threshold is obtained. According to the similarity matching results with a similarity greater than a preset threshold, the tag information to be processed is classified to obtain multiple target tag information groups, and each tag information group to be processed corresponds to the same category information. The same category information corresponding to each target tag information group is used as the main tag information, and the tag information to be processed in each tag information group to be processed is used as the sub-tag information corresponding to the main tag information.

[0100] See also Figure 4 ,like Figure 4 As shown, each main tag information may include multiple sub-tag information. When classifying the tag information to be processed, the tag information to be processed in multiple target tag information groups may be deduplicated, so that after the tags to be processed are classified, the preset tag information corresponding to each main tag information is different from the preset tag information corresponding to other main tag information.

[0101] As an optional embodiment, the tag information to be processed includes tag information to be processed a, tag information to be processed a1, tag information to be processed b1, tag information to be processed b2, and tag information to be processed c1. The similarity between tag information to be processed a and tag information to be processed a1 is greater than a preset threshold, the similarity between tag information to be processed b1 and tag information to be processed b2 is greater than a preset threshold, and the similarity between tag information to be processed c1 and any of tag information to be processed a, tag information to be processed a1, tag information to be processed b1, and tag information b2 is less than a preset threshold. Then the server can classify tag information to be processed a and tag information to be processed a1 into one category, and determine the corresponding category information as main tag information A. Classify tag information to be processed b1 and tag information to be processed b2 into one category, and determine the corresponding category information as main tag information B. Classify tag information to be processed c1 into one category, and determine the corresponding category information as main tag information C.

[0102] As an optional embodiment, when classifying the tag information to be processed, the tag information to be processed may be classified by manual classification, or the tag information to be processed may be classified by a pre-trained classification model.

[0103] Based on the similarity matching results, the tag information to be processed is classified to obtain multiple target tag information groups, thereby improving the efficiency of tag construction.

[0104] As an alternative embodiment, see Figure 5, after performing similarity matching on any two of the plurality of tag information to be processed and obtaining a plurality of similarity matching results, the method further comprises:

[0105] S510. Classify the tag information to be processed according to the multiple similarity matching results to obtain multiple first tag information groups, wherein the multiple first tag information groups include at least two tag information groups corresponding to the same category information and tag information groups corresponding to different category information respectively;

[0106] S520. Merge at least two label information groups corresponding to the same category of information to obtain at least one second label information group;

[0107] S530. De-duplicate the tag information to be processed in each second tag information group;

[0108] S540. Take the label information groups corresponding to the different category information and the second label information group after deduplication processing as the target label information group.

[0109] As an optional embodiment, the label information to be processed is classified according to multiple similarity matching results to obtain multiple first label information groups. Among these multiple first label information groups, there may be first label information groups with repeated categories, that is, two or more first label information groups correspond to the same category. Therefore, these multiple first label information groups can be divided into at least two label information groups corresponding to the same category information, and label information groups corresponding to different category information, wherein the label information groups corresponding to different category information are first label information groups without repeated categories, and each first label information group corresponds to a category.

[0110] As an optional embodiment, at least two tag information groups corresponding to the same category information are merged to obtain at least one second tag information group. For example, if tag information group A and tag information group B in the first tag information group both correspond to a game category, tag information group A and tag information group B are merged into a second tag information group C, and the second tag information group C corresponds to the game category. After obtaining the second tag information group, the tag information to be processed in the second tag information group is deduplicated, and repeated tags in the tag information to be processed are deleted. For example, if tag information group A and tag information group B in the first tag information group both correspond to a game category, and there is a card game tag in tag information group A, and there is also a card game tag in tag information group B, then after merging tag information group A and tag information group B to obtain the second tag information group C, the tag information to be processed in the second tag information group C is deduplicated, and one of the two card game tags is deleted.

[0111] The label information groups corresponding to the different category information in the first label information group and the second label information group after deduplication processing are both label information groups corresponding to one category, and each label information group is independent of each other, there is no repeated category and no repeated label information, then the label information groups corresponding to the different category information in the first label information group and the second label information group after deduplication processing can be used as the target label information group.

[0112] At least two tag information groups corresponding to the same category of information will be merged and deduplicated, which can avoid repeated grouping of the tag information to be processed while classifying, making the target tag information groups independent of each other, avoiding duplication of tag construction results of multimedia resources, and improving the accuracy of tag construction.

[0113] S220. Based on the sub-tag information and content information corresponding to each of the preset multiple main tag information, the multimedia resource is tagged to obtain multiple tag tagged results, each tag tagged result corresponds to a target sub-tag, the target sub-tag is the sub-tag information corresponding to the multimedia resource in each of the multiple main tag information, the main tag information is the category tag information corresponding to the sub-tag information of the same category, and the sub-tag information is at least one tag information under the category corresponding to each main tag information;

[0114] As an optional embodiment, the preset tag information is tag information that carries main tag information. According to the multimedia content, the server can obtain the target sub-tag corresponding to the multimedia content from the preset tag information. When multiple target sub-tags are obtained, the server can simultaneously construct multiple target sub-tags with the multimedia resources when constructing tags for multimedia resources, and obtain the tag construction results of each target sub-tag. Each main tag information is independent of each other, that is, each main tag information does not overlap. Therefore, the target sub-tag belongs to different main tag information, and the tag construction results obtained are independent of each other based on the target sub-tags corresponding to the different main tag information.

[0115] As an alternative embodiment, see Figure 6 , based on the sub-tag information and content information corresponding to each of the preset multiple main tag information, tag construction is performed to obtain multiple tag construction results including:

[0116] S610. Match each sub-tag information corresponding to each main tag information in the plurality of main tag information with the content information;

[0117] S620. Using the sub-tag information matching the content information as the target sub-tag;

[0118] S630. Establish a tag construction result between the target sub-tag corresponding to each main tag information and the multimedia resource.

[0119] As an alternative embodiment, see Figure 7 , match the sub-tag information corresponding to each main tag information with the multimedia content information. Since there can be multiple sub-tag information under different main tag information, multiple matching results can be obtained, and a target sub-tag is determined from these multiple matching results. For example, the industry main tag information can include the education industry, the game industry, the publishing industry, etc. When the multimedia content information is the advertising information of a certain online course software, in the matching results corresponding to each sub-tag information of the industry main tag information, it can be determined that the education industry is the sub-tag information that matches the multimedia content information, and the education industry is used as the target sub-tag.

[0120] Among the multiple target subtags that match the multimedia content, different target subtags belong to different main tag information. Each target subtag is matched with the multimedia resource to obtain the corresponding relationship between each target subtag and the multimedia resource, and the multimedia resource can be labeled to obtain the label construction result corresponding to each target subtag. Since different target subtags belong to different main tag information, each label construction result is independent of each other, and there will be no repeated label construction results. For example, assuming that the target subtag a belongs to the main tag information A1, and the target subtag a also belongs to the main tag information A2, and the target subtag a matches the multimedia content information, then when the label is constructed, the target subtag a will be obtained from the main tag information A1, and the target subtag a will also be obtained from the main tag information A2. When the corresponding relationship between the target subtag a and the multimedia resource is established, two label construction results corresponding to the target subtag a will be generated, resulting in repeated construction. When the target subtag a only belongs to the main tag information A1 and does not belong to other main tag information, the target subtag a can only be obtained from the main tag information A1. Therefore, when establishing the correspondence between the target subtag a and the multimedia resource, only one tag construction result corresponding to the target subtag a can be obtained.

[0121] As an optional embodiment, there may be no target sub-tag matching the multimedia content information in the sub-tag information under the main tag information.

[0122] By building labels based on the matched target sub-labels, multiple label building results can be obtained at one time, thereby improving the efficiency of label building.

[0123] As an optional embodiment, the method further includes:

[0124] Based on multiple tag building results, tag attribute information of multimedia resources is generated.

[0125] As an optional embodiment, the multimedia resource may be provided with initial attribute information, and corresponding attribute content may be added to the initial attribute information by assignment. Multiple tag construction results may be assigned to the initial attribute information to obtain tag attribute information. Multiple tag construction results correspond to one tag attribute information.

[0126] As an alternative embodiment, see Figure 8 , the method further comprises:

[0127] S810. Obtain user tag information corresponding to the target object;

[0128] S820. Obtain multiple to-be-recommended construction results corresponding to the to-be-recommended multimedia resources, where the multiple to-be-recommended construction results are multiple tag construction results corresponding to the to-be-recommended multimedia resources;

[0129] S830. Compare the user tag information and the target sub-tag corresponding to each recommended build result;

[0130] S840. When the user tag information matches the target sub-tag corresponding to the to-be-recommended construction result, recommend the to-be-recommended multimedia resource to the target object.

[0131] As an optional embodiment, when recommending multimedia resources to a client, the server obtains the user tag information corresponding to the user of the client, and obtains multiple recommended construction results of the multimedia resources to be recommended. If the multimedia resources to be recommended generate tag output information for multiple recommended construction results, the tag attribute information corresponding to the multimedia resources to be recommended can also be obtained. The user tag information is compared with the target sub-tag corresponding to each recommended construction result. When the target sub-tag corresponding to each recommended construction result matches the user tag information, the multimedia resources to be recommended are sent to the client. The client receives the multimedia resources to be recommended and displays them.

[0132] Recommending multimedia resources based on tag attribute information can improve the accuracy of multimedia resource recommendations.

[0133] As an optional embodiment, the method can be used to tag advertising information. The server obtains the advertising information and extracts content features of the advertising information based on the content feature extraction model to obtain advertising content information. When the advertising information is a video, the content feature extraction model may include an image feature extraction model and a voice feature extraction model. When the advertising information is text, the content feature extraction model may include a text feature extraction model.

[0134] The server stores pre-classified main tag information and sub-tag information corresponding to each main tag. Each main tag information is independent of each other, that is, the sub-tag information corresponding to each main tag information is different. The server matches the sub-tag information corresponding to each main tag information with the advertising content information to obtain the matching result between each sub-tag and the advertising information. Based on the matching result between each sub-tag and the advertising information, the server can determine the target sub-tag, establish the corresponding relationship between each target sub-tag and the advertising information, and obtain the tag construction result corresponding to each target sub-tag. The server is based on multiple tag construction results. The server recommends advertising information to the user based on the multiple tag construction results of the advertising information. The server can obtain the user tag information of the user, and push the advertising information to the user when the user tag information matches the target sub-tag in the multiple tag construction results.

[0135] The present disclosure proposes a method for constructing labels for multimedia resources, the method comprising: acquiring multimedia resources, and performing content identification on the multimedia resources to obtain multimedia content information. The multimedia content information is labeled based on the sub-label information corresponding to each of the multiple main label information to obtain multiple label construction results. The multiple main label information are independent of each other. The multimedia resources are recommended based on the label attribute information of the multimedia resources. The method can obtain multiple label construction results at one time, thereby improving the efficiency of label construction. The method determines a matching sub-label information as the target sub-label information from each main label information, and constructs labels for the multimedia resources. The multiple label construction results obtained are independent of each other, which can avoid repeated construction and improve the accuracy of label construction.

[0136] Fig. 9 FIG. 1 is a block diagram of a device for constructing a label of a multimedia resource according to an exemplary embodiment. Fig. 9 , the device comprises:

[0137] The multimedia content identification module 910 is configured to obtain content information of the multimedia resource to be tagged;

[0138] The label construction module 920 is configured to perform label construction on the multimedia resources based on the sub-label information and content information corresponding to each of the preset multiple main label information, and obtain multiple label construction results, each of which corresponds to a target sub-label, and the target sub-label is the sub-label information corresponding to the multimedia resource in each main label information of the multiple main label information. The main label information is the category label information corresponding to the sub-label information of the same category, and the sub-label information is at least one label information under the category corresponding to each main label information.

[0139] As an optional embodiment, the label construction module 920 includes:

[0140] The sub-tag matching unit is configured to match each sub-tag information corresponding to each main tag information in the plurality of main tag information with the content information;

[0141] a target sub-tag determining unit, configured to execute sub-tag information matching the content information as a target sub-tag;

[0142] The tag building result acquisition unit is configured to execute the tag building result between the target sub-tag corresponding to each main tag information and the multimedia resource.

[0143] As an optional embodiment, the device further includes:

[0144] A module for acquiring tag information to be processed is configured to execute acquisition of a plurality of tag information to be processed, where the tag information to be processed is unclassified tag information obtained based on a plurality of tag sources;

[0145] A similarity matching result acquisition module is configured to perform similarity matching on any two of the plurality of tag information to be processed to obtain a plurality of similarity matching results;

[0146] A first classification module is configured to classify the tag information to be processed according to the multiple similarity matching results to obtain multiple target tag information groups, each target tag information group corresponding to a category information;

[0147] A main label information determination module is configured to execute the category information corresponding to each target label information group as the main label information;

[0148] The sub-tag information determination module is configured to use the tag information to be processed in each target tag information group as the sub-tag information corresponding to the main tag information.

[0149] As an optional embodiment, the device further includes:

[0150] A second classification module is configured to classify the tag information to be processed according to the multiple similarity matching results to obtain multiple first tag information groups, wherein the multiple first tag information groups include at least two tag information groups corresponding to the same category information and tag information groups corresponding to different category information respectively;

[0151] A same category merging module is configured to merge at least two label information groups corresponding to the same category information to obtain at least one second label information group;

[0152] a deduplication module, configured to perform deduplication processing on the to-be-processed label information in each second label information group;

[0153] The target label information group determination module is configured to execute label information groups corresponding to different category information and the second label information group after deduplication processing as the target label information group.

[0154] As an optional embodiment, the device further includes:

[0155] The tag attribute information generating module is configured to generate tag attribute information of multimedia resources based on multiple tag building results.

[0156] As an optional embodiment, the multimedia content recognition module includes:

[0157] The feature extraction unit is configured to perform content feature extraction on multimedia resources based on a preset content feature extraction model to obtain content information.

[0158] As an optional embodiment, the device further includes:

[0159] A user tag information acquisition module is configured to execute acquisition of user tag information corresponding to a target object;

[0160] The tag attribute information acquisition module is configured to execute acquisition of a plurality of to-be-recommended construction results corresponding to the to-be-recommended multimedia resources, where the plurality of to-be-recommended construction results are a plurality of tag construction results corresponding to the to-be-recommended multimedia resources;

[0161] A comparison module is configured to compare the user tag information with the target sub-tag corresponding to each to-be-recommended build result;

[0162] The recommendation module is configured to recommend the multimedia resource to be recommended to the target object when the user tag information matches the target sub-tag corresponding to the construction result to be recommended.

[0163] 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.

[0164] Fig.10 is a block diagram of an electronic device for label construction of multimedia resources according to an exemplary embodiment. The electronic device may be a terminal, and its internal structure diagram may be as shown in FIG. Fig.10As 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 constructing a label for 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.

[0165] Those skilled in the art will understand that Fig.10 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present disclosure, and does not constitute a limitation on the electronic device to which the scheme of the present disclosure is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0166] In an exemplary embodiment, a computer-readable storage medium including instructions is also provided, such as a memory 1004 including instructions, and the above instructions can be executed by the processor 1020 of the electronic device 1000 to complete the above method. Alternatively, 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.

[0167] In an exemplary embodiment, a computer program product is also provided, including a computer program, and when the computer program is executed by a processor, the above-mentioned method for constructing a tag for multimedia resources is implemented.

[0168] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0169] 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 method for constructing a label of a multimedia resource, characterized in that: The method comprises: Acquire content information of multimedia resources to be tagged; the multimedia resources include advertising information in different forms; Based on the sub-tag information corresponding to each of the preset multiple main tag information and the content information, the multimedia resource is labeled to obtain multiple tag construction results, each tag construction result corresponds to a target sub-tag, the target sub-tag is the sub-tag information corresponding to the multimedia resource in each of the multiple main tag information, the main tag information is the category tag information corresponding to the sub-tag information of the same category, and the sub-tag information is at least one tag information under the category corresponding to each main tag information; The method further comprises: Acquire a plurality of tag information to be processed, wherein the tag information to be processed is unclassified tag information obtained based on a plurality of tag sources; Performing similarity matching on any two of the plurality of tag information to be processed to obtain a plurality of similarity matching results; Classifying the tag information to be processed according to the multiple similarity matching results to obtain multiple target tag information groups, each target tag information group corresponding to a category information; Using the category information corresponding to each target label information group as the main label information; Using the to-be-processed label information in each target label information group as the sub-label information corresponding to the main label information; After performing similarity matching on any two of the plurality of tag information to be processed to obtain a plurality of similarity matching results, the method further includes: Classify the tag information to be processed according to the multiple similarity matching results to obtain multiple first tag information groups, wherein the multiple first tag information groups include at least two tag information groups corresponding to the same category information and tag information groups corresponding to different category information respectively; Merging at least two label information groups corresponding to the same category of information to obtain at least one second label information group; Deduplication processing is performed on the to-be-processed label information in each second label information group; The label information groups corresponding to the different category information and the second label information group after deduplication processing are used as the target label information group.

2. The method for constructing a label for multimedia resources according to claim 1, characterized in that: The step of constructing labels for the multimedia resources based on the sub-label information corresponding to each of the preset plurality of main label information and the content information to obtain a plurality of label construction results includes: Matching each sub-tag information corresponding to each main tag information in the plurality of main tag information with the content information; Using sub-tag information matching the content information as the target sub-tag; A tag construction result between the target sub-tag corresponding to each main tag information and the multimedia resource is established.

3. The method for constructing a label of multimedia resources according to claim 1, characterized in that: The method further comprises: Based on the multiple tag construction results, tag attribute information of the multimedia resource is generated.

4. The method for constructing a label of multimedia resources according to claim 1, characterized in that: The step of obtaining content information of multimedia resources to be tagged includes: Based on a preset content feature extraction model, content features are extracted from the multimedia resources to obtain the content information.

5. The method for constructing a label of a multimedia resource according to claim 1, characterized in that: The method further comprises: Get the user tag information corresponding to the target object; Acquire a plurality of to-be-recommended construction results corresponding to the to-be-recommended multimedia resources, wherein the plurality of to-be-recommended construction results are a plurality of tag construction results corresponding to the to-be-recommended multimedia resources; Comparing the user tag information with the target sub-tag corresponding to each to-be-recommended build result; In a case where the user tag information matches the target sub-tag corresponding to the to-be-recommended construction result, the to-be-recommended multimedia resource is recommended to the target object.

6. A device for constructing a label of a multimedia resource, characterized in that: The device comprises: A multimedia content identification module is configured to obtain content information of a multimedia resource to be tagged; A label construction module is configured to perform label construction on the multimedia resource based on the sub-label information corresponding to each of the preset multiple main label information and the content information, and obtain multiple label construction results, each of which corresponds to a target sub-label, and the target sub-label is the sub-label information corresponding to the multimedia resource in each of the multiple main label information, the main label information is the category label information corresponding to the sub-label information of the same category, and the sub-label information is at least one label information under the category corresponding to each main label information; The device also includes: A tag information acquisition module to be processed is configured to execute acquisition of a plurality of tag information to be processed, wherein the tag information to be processed is unclassified tag information obtained based on a plurality of tag sources; A similarity matching result acquisition module is configured to perform similarity matching on any two of the plurality of tag information to be processed to obtain a plurality of similarity matching results; A first classification module is configured to classify the tag information to be processed according to the multiple similarity matching results to obtain multiple target tag information groups, each target tag information group corresponding to a category information; A main label information determination module is configured to use the category information corresponding to each target label information group as the main label information; A sub-tag information determination module is configured to execute the tag information to be processed in each target tag information group as the sub-tag information corresponding to the main tag information; The device also includes: A second classification module is configured to classify the tag information to be processed according to the multiple similarity matching results to obtain multiple first tag information groups, wherein the multiple first tag information groups include at least two tag information groups corresponding to the same category information and tag information groups corresponding to different category information respectively; A same category merging module is configured to merge at least two label information groups corresponding to the same category information to obtain at least one second label information group; a deduplication module, configured to perform deduplication processing on the to-be-processed label information in each second label information group; The target label information group determination module is configured to execute the label information groups corresponding to the different category information and the second label information group after deduplication processing as the target label information group.

7. The device for constructing a label of multimedia resources according to claim 6, characterized in that: The label building module includes: a sub-tag matching unit, configured to match each sub-tag information corresponding to each main tag information in the plurality of main tag information with the content information; a target sub-tag determining unit, configured to execute sub-tag information matching the content information as the target sub-tag; The tag building result acquisition unit is configured to execute the establishment of a tag building result between the target sub-tag corresponding to each main tag information and the multimedia resource.

8. The device for constructing a label of multimedia resources according to claim 6, characterized in that: The device also includes: The tag attribute information generating module is configured to generate tag attribute information of the multimedia resource based on the multiple tag building results.

9. The device for constructing a label of multimedia resources according to claim 6, characterized in that: The multimedia content recognition module includes: The feature extraction unit is configured to perform content feature extraction on the multimedia resource based on a preset content feature extraction model to obtain the content information.

10. The device for constructing a label of multimedia resources according to claim 6, characterized in that: The device also includes: A user tag information acquisition module is configured to execute acquisition of user tag information corresponding to a target object; A tag attribute information acquisition module is configured to execute acquisition of a plurality of to-be-recommended construction results corresponding to the to-be-recommended multimedia resource, wherein the plurality of to-be-recommended construction results are a plurality of tag construction results corresponding to the to-be-recommended multimedia resource; A comparison module is configured to compare the user tag information with a target sub-tag corresponding to each to-be-recommended build result; The recommendation module is configured to recommend the multimedia resource to be recommended to the target object when the user tag information matches the target sub-tag corresponding to the construction result to be recommended.

11. 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 method for constructing a label for multimedia resources as claimed in any one of claims 1 to 5.

12. 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 method for building a label for multimedia resources as claimed in any one of claims 1 to 5.

13. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for constructing a label for multimedia resources according to any one of claims 1 to 5 is implemented.

Images