Data transmission method and device, computer equipment, storage medium and program product

Abstract

The invention relates to the technical field of computers, and discloses a data transmission method and device, computer equipment, a storage medium and a program product, and a label management method comprises the following steps: obtaining to-be-identified data in a target platform, and determining an entity label based on the to-be-identified data; performing content identification on the entity tag, and determining a management level of the entity tag according to an identification result; and based on a mounting parameter corresponding to the management level, carrying out a mounting operation on the entity tag, the mounting operation being used for indicating to distribute a corresponding content entity for the entity tag in the target platform. According to the method and the device, the entity tags can be automatically mined and supplemented, the mined entity tags are graded, and the mounting operation is executed according to the management result, so that effective management of the entity tags is realized, the application efficiency of the entity tags is improved, and the application scale of the entity tags is favorably expanded.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a data transmission method, apparatus, computer equipment, storage medium, and program product. Background Art

[0002] Video platforms refer to Internet platforms that provide services such as video content playback, sharing, creation, and social interaction. At the same time, video platforms often provide item acquisition functions, where users can display item information of items to be acquired for other users on the video platform to acquire. Considering that the traffic on video platforms is often large, users need to query content entities such as content or item information that they want to obtain through search keywords. In order to improve the accuracy of search results, entity tags can be attached to content entities on the video platform, and when the entity tags match the search keywords, the corresponding content entities can be displayed.

[0003] However, in related tag management solutions, entity tags are usually obtained by users or operators uploading them themselves, and the entity tags are mounted based on subjective judgment of the content entity. The quality of the entity tags cannot be controlled, and there is a lack of effective management of entity tags, which greatly hinders the application efficiency and scale of the tags. Summary of the invention

[0004] In view of this, the present disclosure provides a data transmission method, apparatus, computer device, storage medium and program product to solve the problem that the lack of effective management of entity tags greatly hinders the application efficiency and scale of tags.

[0005] In a first aspect, the present disclosure provides a tag management method, the method comprising:

[0006] Obtain the data to be identified in the target platform, and determine the entity label based on the data to be identified;

[0007] Perform content identification on entity tags and determine the management level of entity tags based on the identification results;

[0008] Based on the mounting parameters corresponding to the management level, a mounting operation is performed on the entity tag, wherein the mounting operation is used to indicate that a corresponding content entity is allocated to the entity tag in the target platform.

[0009] In a second aspect, the present disclosure provides a tag management device, the device comprising:

[0010] A determination module, used to obtain the data to be identified in the target platform and determine the entity label based on the data to be identified;

[0011] An identification module, used to identify the content of the entity tag and determine the management level of the entity tag according to the identification result;

[0012] The mounting module is used to perform a mounting operation on the entity tag based on the mounting parameters corresponding to the management level, wherein the mounting operation is used to indicate that a corresponding content entity is allocated to the entity tag in the target platform.

[0013] In a third aspect, the present disclosure provides a computer device, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, computer instructions are stored in the memory, and the processor executes the label management method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0014] In a fourth aspect, the present disclosure provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the tag management method of the first aspect or any corresponding embodiment thereof.

[0015] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the first aspect or any corresponding embodiment thereof.

[0016] In the disclosed embodiment, first, the data to be identified in the target platform can be obtained, and the entity tag can be determined based on the data to be identified. Next, the entity tag can be subjected to content identification, and the management level of the entity tag can be determined based on the identification result. Then, the entity tag can be subjected to a mounting operation based on the mounting parameters corresponding to the management level, wherein the mounting operation is used to indicate that the corresponding content entity is assigned to the entity tag in the target platform, thereby automatically mining and supplementing the entity tag, and grading the mined entity tags to perform the mounting operation according to the management results, thereby achieving effective management of the entity tag, so as to improve the application efficiency of the entity tag, and help expand the application scale of the entity tag. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the specific embodiments of the present disclosure or the technical solutions in the prior art, the drawings required for use in the specific embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 is a flowchart of a tag management method according to an embodiment of the present disclosure;

[0019] Figure 2 This is a mounting process of the third type of entity tag according to an embodiment of the present disclosure;

[0020] Figure 3 is a schematic diagram of the association relationship between entity tags according to an embodiment of the present disclosure;

[0021] Figure 4 is a schematic diagram of a unified graph label model according to an embodiment of the present disclosure;

[0022] Figure 5 is a schematic diagram of determining the coverage rate corresponding to each management level according to an embodiment of the present disclosure;

[0023] Figure 6 is a schematic diagram of the automatic mounting process according to an embodiment of the present disclosure;

[0024] Figure 7 is a structural block diagram of a label management device according to an embodiment of the present disclosure;

[0025] Figure 8 It is a schematic diagram of the hardware structure of the computer device of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present disclosure.

[0027] In conjunction with the application scenarios on which the execution of the tag management method depends, the application scenarios are described here.

[0028] Video platforms refer to Internet platforms that provide services such as video content playback, sharing, creation, and social interaction. At the same time, video platforms often provide item acquisition functions, where users can display item information of items to be acquired for other users on the video platform to acquire. Considering that the traffic on video platforms is often large, users need to query content entities such as content or item information that they want to obtain through search keywords. In order to improve the accuracy of search results, entity tags can be attached to content entities on the video platform, and when the entity tags match the search keywords, the corresponding content entities can be displayed.

[0029] However, in related tag management solutions, entity tags are usually obtained by users or operators uploading them themselves, and the entity tags are mounted based on subjective judgment of the content entity. The quality of the entity tags cannot be controlled, and there is a lack of effective management of entity tags, which greatly hinders the application efficiency and scale of the tags.

[0030] Based on this, the embodiment of the present disclosure provides a tag management method, which can first obtain the data to be identified in the target platform, and determine the entity tag based on the data to be identified. Next, the entity tag can be subjected to content identification, and the management level of the entity tag can be determined according to the identification result. Then, the entity tag can be subjected to a mounting operation based on the mounting parameters corresponding to the management level, wherein the mounting operation is used to indicate that the corresponding content entity is assigned to the entity tag in the target platform, thereby automatically mining and supplementing the entity tag, and grading the mined entity tags to perform the mounting operation according to the management results, thereby achieving effective management of the entity tag, so as to improve the application efficiency of the entity tag, and help expand the application scale of the entity tag.

[0031] According to an embodiment of the present disclosure, an embodiment of a tag management method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0032] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0033] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can autonomously choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.

[0034] As an optional but non-limiting implementation, in response to receiving an active request from the user, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0035] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet the relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0036] It is understandable that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and relevant provisions.

[0037] According to an embodiment of the present disclosure, an embodiment of a tag management method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0038] In this embodiment, a tag management method is provided, which can be used in the above-mentioned video platforms, such as long video platforms, short video platforms, etc. Figure 1 is a flowchart of a tag management method according to an embodiment of the present disclosure. Figure 1 As shown, the process includes the following steps:

[0039] Step S101, obtaining the data to be identified in the target platform, and determining the entity tag based on the data to be identified.

[0040] In the disclosed embodiment, the target platform may be the above-mentioned video platform, and the data to be identified may include various types of data such as text, voice, and video. Here, the data to be identified may be determined based on content entities such as pictures, texts, and videos uploaded by users on the target platform. Alternatively, a database may be pre-set for the target platform, and the data to be identified may be obtained based on the database. The database is used to store the data to be identified that is mined based on multiple approaches, for example, identification is performed based on publicly available external multimedia data, and the data to be identified that may be related to the content entity is stored in the database.

[0041] After obtaining the data to be identified, keywords related to the content entity in the data to be identified can be mined to obtain entity tags. Here, the entity tag is processed from the original data and can be directly applied to the business to generate business value. The entity tag is a conceptual and logical definition of the object, which is readable and easy to understand. For example, "cute style" and "white moonlight outfit" are abstract descriptions of items and can be used as labels for items. In addition, the entity tag can be a graph tag, which refers to an entity tag processed by a graph algorithm. For example, an entity tag may contain keywords used to describe items displayed on the target platform, such as item name, item function, item brand, etc. For another example, the keyword may be a keyword used to describe the usage scenarios of items displayed on the target platform, such as the Spring Festival, Dragon Boat Festival, etc.

[0042] Step S102: perform content recognition on the entity tag and determine the management level of the entity tag according to the recognition result.

[0043] In the disclosed embodiment, the obtained entity tags may be graded and hierarchically managed according to the management level. For example, the management levels may include L3, L2, and L1, wherein from L3 to L1, the content requirements of the entity tags are gradually precise and the management requirements are gradually refined.

[0044] Based on this, the entity tag can be subjected to content recognition to determine the management level of the entity tag according to the recognition result, wherein the content recognition can be used to identify whether the semantics of the entity tag is accurate. For example, the entity tag directly identified based on the data to be identified is determined as L3, the entity tag with an identification result accuracy of 80% in L3 is determined as L2, and the entity tag with an identification result accuracy of 100% in L2 is determined as L1.

[0045] Step S103: performing a mounting operation on the entity tag based on the mounting parameters corresponding to the management level, wherein the mounting operation is used to indicate that a corresponding content entity is allocated to the entity tag in the target platform.

[0046] In the disclosed embodiment, specific operation strategies can be formulated in advance for entity tags of different management levels. For example, corresponding mounting parameters are set for different management levels, wherein the mounting parameters can be used to indicate the automatic mounting ratio for the entity tag, the mounted content entity type, the mounted content entity field, the mounting scenario, etc.

[0047] For example, if the above management levels include L3, L2, and L1, the management requirements of entity tags are gradually refined from L3 to L1. Here, the mounting scenario set for L3 can be a scenario without mounting accurate recall requirements, the mounting scenario set for L2 can be a scenario with low mounting accurate recall requirements, such as a search relevance scenario, and the mounting scenario set for L3 can be a scenario with high mounting accurate recall requirements, such as a scenario for search screening of content entities.

[0048] It can be seen from the above description that in the embodiment of the present disclosure, first, the data to be identified in the target platform can be obtained, and the entity tag can be determined based on the data to be identified. Next, the entity tag can be subjected to content identification, and the management level of the entity tag can be determined based on the identification result. Then, the entity tag can be mounted based on the mounting parameters corresponding to the management level, wherein the mounting operation is used to indicate the assignment of the corresponding content entity to the entity tag in the target platform, thereby automatically mining and supplementing the entity tag, and grading the mined entity tags to perform the mounting operation according to the management results, thereby achieving effective management of the entity tags, so as to improve the application efficiency of the entity tags, and help expand the application scale of the entity tags.

[0049] In some optional implementations, the above step S101, obtaining the data to be identified in the target platform, includes:

[0050] Step S11, obtaining multimedia content published on the target platform.

[0051] Step S12, performing recognition based on the text in the multimedia content to obtain data to be recognized.

[0052] In the disclosed embodiment, the multimedia content may be audio content, video content, graphic content, etc., which are published by the user to the target platform. Here, the multimedia content published in the target platform may be obtained in real time to perform automatic tag mining to obtain the above-mentioned L3 level entity tags.

[0053] Here, the text in the multimedia content can be extracted to obtain the data to be identified, thereby reducing the difficulty of entity label extraction. For example, if the multimedia content is a video for displaying melons, then the text in the video can be extracted to obtain the data to be identified "Freshly picked XX melon, crisp, sweet, juicy and delicious #thin-skinned melon #direct from the place of origin". For another example, if the multimedia content is a combination of pictures and texts for displaying accessories, then the text in the combination of pictures and texts can be extracted to obtain the data to be identified "The thing that time cannot change is called love #atmosphere beauty needs a set of accessories #beauty with a sense of luxury".

[0054] In the disclosed embodiments, multimedia content published on the target platform can be acquired in real time for automated tag mining, thereby replenishing the number of entity tags, reducing the need for manual tag filling, improving tag generation efficiency, and further improving the application efficiency of entity tags.

[0055] In some optional implementations, the above step S101, determining the entity tag based on the data to be identified, includes:

[0056] Step S21, obtaining a preset tag type.

[0057] Step S22, performing semantic recognition on the data to be recognized, obtaining keywords matching the preset tag type, and determining entity tags based on the keywords.

[0058] In the disclosed embodiment, a language model may be pre-established to perform semantic recognition on the data to be recognized, thereby obtaining keywords that may be related to the item description and scene description of the displayed items in the target platform.

[0059] For example, a graph sorting algorithm can be used to determine keywords by calculating the weight of nodes. This graph sorting algorithm can take into account the semantic association between words and is more effective in extracting keywords with semantic coherence. For example, in a food copy, words such as "ingredients" and "cooking methods" may be identified as key product keywords by the graph sorting algorithm.

[0060] For example, for the above data to be identified, "Freshly picked XX melon, crisp, sweet, juicy and delicious #thin-skin melon #straight from the place of origin", the keywords obtained by semantic recognition can be: XX melon and thin-skin melon and other item words. For the above data to be identified, "The thing that time cannot change is love #atmosphere beauty needs a set of accessories #beauty with a sense of luxury", the keywords obtained by semantic recognition can be: atmosphere, luxury and other item description words.

[0061] Specifically, the above-mentioned preset tag types include: a first type and a second type. The above-mentioned step S22, performing semantic recognition on the data to be recognized, and obtaining keywords matching the preset tag types, specifically includes:

[0062] Keywords whose semantic recognition results in the data to be identified are object description information are determined as keywords matching the first type, and keywords whose semantic recognition results in the data to be identified are phenomenon description information are determined as keywords matching the second type.

[0063] In the disclosed embodiment, the first type may be an atomic tag, which is used to trigger from the perspective of the supply and production of items, and to characterize the concrete description of the items, such as the brand and category (food, fresh produce, electronic products, etc.) of the items. The second type may be a high-order tag, which may be divided into descriptions of various aspects of phenomena such as actual stimulation, symbolic vocabulary, and social stimulation according to the user's item acquisition behavior decision. Here, the actual stimulation may be a trendy label, such as white moonlight. The symbolic vocabulary may be a scene label, such as outdoor fishing. Social stimulation may be a description related to a social phenomenon, such as a commuting artifact.

[0064] Based on this, semantic recognition can be performed on the data to be recognized, so as to identify keywords that match the first type and the second type definitions according to semantic recognition. Specifically, grammatical rules can be established in advance for the first type and the second type to extract words that conform to the grammatical rules as keywords. Then, the parts of speech of the words in the data to be recognized can be annotated, and the dependency relationship between the words in the sentence can be analyzed based on the annotation results to determine the core words and modifiers, so as to generate the above keywords based on the combination of the core words and the modifiers. Here, the core words are usually words that express the main concepts, and the modifiers describe and define the core words.

[0065] In addition, the above-mentioned preset tag type also includes: the third type, then, the above-mentioned step S22, performing semantic recognition on the data to be recognized to obtain keywords matching the preset tag type, also includes the following process:

[0066] Semantic recognition is performed on keywords corresponding to the first type and the second type to obtain a third type that matches at least part of the target entity, wherein the target entity includes at least one of the following: a content publishing object, an item entity, a registered object in the target platform, and a content entity.

[0067] In the embodiments of the present disclosure, different from the non-cross-entity tags of the first and second types mentioned above (i.e., each entity tag corresponds to one type of entity), the third type of entity tag can be a cross-entity tag, i.e., one entity tag of the third type can mark multiple types of entities at the same time.

[0068] For example, for the entity tag "New Chinese Style", the entity tag can be used to mark multiple target entities at the same time, including content publishing objects (users who publish items to be acquired in the target platform), item entities (items to be acquired in the target platform), registration objects in the target platform (users who publish content in the target platform), and content entities (videos and other content published in the target platform).

[0069] Here, if Figure 2 The mounting process of the third type of entity tag "New Chinese Style" mentioned above is shown. Specifically, the "New Chinese Style" tag can be mounted to a content publishing object in which the proportion of New Chinese style items in the published items is relatively high to obtain a first object tag. The "New Chinese Style" tag can be mounted to an item entity with a style of "New Chinese Style" in the target platform to obtain an item tag. The "New Chinese Style" tag can be mounted to a registered object with a public preference for New Chinese style in the target platform to obtain a second object tag. The "New Chinese Style" tag can be mounted to a content entity whose main content published in the target platform is related to New Chinese style to obtain a content tag, wherein the content associated with the "New Chinese Style" item tag can be marked with the same tag, or based on the understanding of the content of the content entity, the content entity with a high concentration of "New Chinese Style" can be marked with the "New Chinese Style" content tag.

[0070] In the disclosed embodiment, the preset tag types can be divided into a first type, a second type, and a third type, and the entity tag is determined based on the keyword whose semantic recognition result in the data to be identified matches the preset tag type, thereby enriching the types of entity tags to obtain entity tags with comprehensive description dimensions, thereby better responding to the mounting needs of various types of entities with increasingly rich content and different characteristics in the target platform.

[0071] In some optional implementations, the above step S22 further includes the following process:

[0072] Step a1: Analyze the phenomenon description information corresponding to the second type of entity tag and the item description information corresponding to the first type of entity tag to obtain the second type of entity tag that is associated with the first type of entity tag.

[0073] Step a2: establishing association information according to the association relationship.

[0074] In the disclosed embodiment, an association relationship between the first type and the second type of entity tags can be established. In addition, an association relationship between the first type and the first type of entity tags can be established, or an association relationship between the second type and the second type of entity tags can be established. Here, the association relationship can include a containment relationship, a similarity relationship, a belonging relationship, etc. It should be understood that the similarity relationship is a bidirectional relationship (A is similar to B), and the containment relationship and the belonging relationship are unidirectional relationships (A contains B, or B belongs to A).

[0075] Specifically, the association relationship between the entity tags of the second type and the first type may be a containment relationship, that is, the second type contains the first type. The association relationship between the entity tags of the first type and the second type may be a belonging relationship, that is, the first type belongs to the second type. The association relationship between the entity tags of the first type and the first type, and the association relationship between the entity tags of the second type and the second type may be a similarity relationship, that is, the first type is similar to the first type, and the second type is similar to the second type.

[0076] When determining the association relationship of entity tags, it can be determined based on the relationship between the semantics of keywords in the entity tags. Here, the belonging relationship between the above-mentioned first type and second type entity tags is analyzed as an example. Specifically, the belonging relationship between the item description information corresponding to the first type of entity tag and the phenomenon description information corresponding to the second type of entity tag can be analyzed. For example, if the first type of entity tag includes a white dress and the second type of entity tag includes a white moonlight outfit, there is an association between the two, and it should be determined that the white dress belongs to the white moonlight outfit.

[0077] Here, if Figure 3 The figure shows the relationship between entity tags, where the first type of tags include: white dress, cute headband, and the second type of entity tags include: white moonlight outfit, retro pure outfit. The relationship between white dress and cute headband is a similarity relationship, the relationship between white moonlight outfit and retro pure outfit is a similarity relationship, the relationship between white dress, cute headband and white moonlight outfit, retro pure outfit is a belonging relationship, and the relationship between white moonlight outfit, retro pure outfit and white dress, cute headband is an inclusion relationship.

[0078] In addition, the present disclosure can manage the acquired entity tags through a unified graph tag model. Figure 4 The figure shows a schematic diagram of a unified graph label model, wherein the unified graph label model may be a PV model, where P is a label item and V is an object label.

[0079] When managing entity tags based on the above-mentioned unified model of graph tags, the root node in the tag tree is the industry attribute, and the branch nodes corresponding to the root node are used to indicate specific item categories under the industry attribute, such as clothing, shoes, bags, digital products, etc., where the level of branch nodes is up to 3 levels. In the tag tree, the tag items are leaf nodes, which are used to indicate the specific subdivision of each item category, such as the leaf nodes style, craftsmanship, and crowd corresponding to clothing, shoes, bags, etc.

[0080] It should be understood that each entity tag can be classified into a leaf node in the tag tree. For example, the entity tags corresponding to the style node may include: cute style, princess dress, etc., the entity tags corresponding to the craft node may include: lace, embroidery, etc., and the entity tags corresponding to the crowd node may include: male, female, etc.

[0081] Here, the above-mentioned unified graph label model has the following characteristics: label items (P) and labels (V) are strictly constrained, and the primary key of PV is globally unique to ensure the uniqueness of the data, so that each label item and label can be accurately identified and distinguished, avoiding data duplication and confusion. In addition, a label can be mounted on multiple entities, and the specific mounting rules are determined by the P rule, V rule and entity type. In addition, association relationships can be established between labels, and these relationships can be unidirectional or bidirectional, so as to facilitate the management of entity labels according to the association relationship.

[0082] In the disclosed embodiment, an association relationship may be established between entity tags of the first type and the second type, and association information may be determined based on the association relationship, thereby establishing a connection between the entity tags to facilitate management of a large number of entity tags in the target platform.

[0083] In some optional implementations, the above step S101, obtaining the data to be identified in the target platform, further includes:

[0084] Step S31, obtaining coverage rates corresponding to various query contents in the target platform, wherein the query contents include at least one keyword for obtaining corresponding content from the target platform, and the coverage rate is used to indicate the hit rate of entity tags in the target platform to the keywords.

[0085] Step S32, determining the data to be identified according to the query content whose coverage is lower than the second threshold.

[0086] In the disclosed embodiment, keywords in the query content input by the user can be identified, and the ratio of the number of keywords hit by the existing entity tags in the query content to the total number of keywords can be determined to obtain the coverage rate.

[0087] For example, if the query content entered by the user is XXX same style red high-end short dress, the keywords may be XXX same style, red, high-end, short, dress, and the entity tags stored in the entity tag library of the target platform include: red, high-end, short, dress, then it is considered that the number of keywords hit by the entity tag in the query content is 4, and the total number of keywords in the query content is 5, then the coverage rate for the query content is 80%.

[0088] Here, a second threshold value can be set in advance, for example, 60%. If the coverage rate corresponding to the query content is 50%, which is lower than the second threshold value, the source data related to the query content can be determined in the target platform, such as text containing keywords in the query content or having an association with the keywords, and the text can be determined as data to be identified.

[0089] In addition, when determining the coverage of the query content, it can also be graded according to the management level to obtain the corresponding coverage of each management level, so as to formulate a corresponding mining strategy for the data to be identified for each management level, so as to improve the coverage of the query content by the entity tags in each management level and supplement the entity tags in a targeted manner.

[0090] For example, if the management levels include the above L1-L3, then the coverage rate corresponding to each management level can be determined separately. Figure 5 The figure shows a schematic diagram for determining the coverage rate corresponding to each management level, wherein, if the query content input by the user is XXX same style red high-end short dress, the keywords may be XX same style, red, high-end, short, dress.

[0091] For L3, the entity tags that hit the query content include red, high-end, short style, dress, short skirt, and dress. For the 15-word query content, the intersection of the entity tags in L3 and the query content is 10 words (the words of the key color are the intersection words), and the coverage rate is 65%. For L2, the entity tags that hit the query content include red, short style, dress, short skirt, and dress. For the 15-word query content, the intersection of the entity tags in L2 and the query content is 8 words, and the coverage rate is 53%. For L1, the entity tags that hit the query content include short skirt and dress. For the 15-word query content, the intersection of the entity tags in L1 and the query content is 4 words, and the coverage rate is 27%.

[0092] Next, the data to be identified corresponding to each management level can be determined based on the missed keywords at each management level to increase the number of entity tags for the missed keyword field at that management level. For example, for the L3 management level, if the missed keyword is XXX same item, then the copy related to XXX same item can be obtained in the target platform to obtain the data to be identified, for example, copy containing AAA same item, BBB same item, etc.

[0093] In the disclosed embodiments, the coverage rate of the entity tag library in the target platform for the query content input by the user can be obtained, and when the coverage rate is low, the data to be identified is formulated according to the query content, so as to carry out targeted mining of the entity tags, so as to comprehensively improve the coverage rate of the entity tag system for the query content, thereby improving the user's satisfaction during use.

[0094] In some optional implementations, the management level includes: a first level, a second level, and a third level. The step S102, determining the management level of the entity tag according to the recognition result, includes:

[0095] Step S1021, based on the recognition result, obtain a semantically identifiable third-level entity tag.

[0096] Step S1022: Determine the entity tag of the second level based on the identification information of the entity tag of the third level.

[0097] Step S1023, obtaining the mounting accuracy of the second-level entity tags, and determining the entity tags whose mounting accuracy exceeds the first threshold as the first level.

[0098] In the disclosed embodiment, the first level is the above-mentioned L1, the second level is the above-mentioned L2, and the third level is the above-mentioned L3. Here, the definition for L3 can be structured classified mining data, that is, through recognition algorithms such as language models, the entity label is directly identified according to the data to be identified. The definition for L2 can be data that has been manually reviewed and has correct semantics, where the entity label of L2 requires complete semantics. The definition for L1 can be highly standardized data with business semantics.

[0099] Here, when mining the entity tags of L2 in the entity tags corresponding to L3, the above step S1022 determines the entity tags of the second level based on the identification information of the entity tags of the third level, and specifically includes the following process:

[0100] Step b1, based on the heat identifier, determine the heat value of the third-level entity tag.

[0101] Step b2: when the heat value meets the heat condition, the third-level entity tag is determined as the second-level entity tag.

[0102] In the disclosed embodiment, on the basis of determining that the semantics of the entity tags in L3 are correct, the determined entity tags may be further screened based on the heat standard to obtain the entity tags of L2.

[0103] Specifically, first, the heat identification of each determined L3 entity tag can be obtained to determine the heat value of the L3 entity tag according to the heat identification. For example, the heat value may include: high heat, medium heat, and low heat. Then, the heat condition can be obtained to determine the L3 entity tag that meets the heat condition as the L2 entity tag. Here, the heat condition is used to filter out entity tags with higher heat. For example, the heat condition can be used to filter out entity tags with heat values ​​above medium heat.

[0104] In the disclosed embodiment, the third-level entity tags can be screened by heat value to obtain the second-level entity tags that meet the heat condition, thereby reducing the number of second entity tags and reducing the pressure on the entity tag management system.

[0105] In some optional implementations, the management level includes a first level, a second level, and a third level, and the mounting parameters include an automatic mounting ratio, wherein the automatic mounting ratio is used to indicate the proportion of automatic mounting operations when mounting the entity tag, the automatic mounting ratio corresponding to the second level is higher than the third level, and the automatic mounting ratio corresponding to the first level is higher than the second level. The above step S103, based on the mounting parameters corresponding to the management level, performs a mounting operation on the entity tag, including:

[0106] Step S1031, obtaining the automatic mounting ratio pre-set for the management level.

[0107] Step S1032: Based on the automatic mounting ratio, mounting operations are performed on the entity tags corresponding to each management level.

[0108] In the embodiment of the present disclosure, since different management levels have different requirements for the mounting accuracy of entity tags, the corresponding automatic mounting ratios are also different, wherein the higher the automatic mounting ratio, the lower the mounting accuracy.

[0109] For example, the application scenario of L3 entity tags is to mine L2 entity tags, so there is no requirement for mounting accuracy, and the automatic mounting ratio can be 100%.

[0110] For example, the application scenario of L2 entity tags is to mine L1 entity tags and user-side mounts with low requirements for accurate recall, such as search relevance. Therefore, if the mount accuracy is required to reach more than 80%, the automatic mount ratio can be 20%-50%.

[0111] For example, the application scenario of L1 entity tag is the scenario of high-accuracy recall on the user side, such as scene screening, or the application scenario related to the object side of the object provision in the target platform. Therefore, if the mounting accuracy is required to reach more than 90% or even 100%, the automatic mounting ratio can be 0%-10%.

[0112] In the disclosed embodiment, corresponding mounting parameters can be set for different management levels to formulate corresponding mounting strategies respectively, so as to meet the accurate recall requirements in the usage scenarios corresponding to different management levels, thereby improving the management efficiency of entity tags while reducing the pressure of entity tag management.

[0113] In some optional embodiments, the above Figure 1 The corresponding embodiments also include:

[0114] When a new entity tag is detected, a target content entity that matches the new entity tag is determined in the content entities of the target platform, and the target content entity is identified based on the new entity tag; or when a new content entity in the target platform is detected, a target entity tag that matches the new content entity is determined in the entity tags, and the new content entity is identified based on the entity tags.

[0115] In the embodiment of the present disclosure, the newly added entity tags and newly added content entities in the target platform can be mounted to achieve automatic mounting. Specifically, the target content entity that matches the newly added entity tag can be determined as a mounting method for finding a word for an entity, and the target entity tag that matches the newly added content entity can be determined as a mounting method for finding a word for an entity.

[0116] like Figure 6 The figure shows a flow chart of automatic mounting, wherein first, the entity tag management system in the embodiment corresponding to the above step S101 can be used to manage the entity tag, and the entity tag management system can also be used to construct the tag features of the entity tag. Then, the content features of the content entity in the target platform can be constructed by the entity feature construction system. Next, the tag features and content features can be matched through the language model to realize the mounting method of word search for entity and entity search for word. Here, the language model can perform feature matching through text matching, image-text matching, image-image matching and multi-mode matching.

[0117] In summary, in the embodiments of the present disclosure, first, the data to be identified in the target platform can be obtained, and the entity tag can be determined based on the data to be identified. Next, the entity tag can be subjected to content identification, and the management level of the entity tag can be determined based on the identification result. Then, the entity tag can be subjected to a mounting operation based on the mounting parameters corresponding to the management level, wherein the mounting operation is used to indicate that the corresponding content entity is assigned to the entity tag in the target platform, thereby automatically mining and supplementing the entity tag, and grading the mined entity tags to perform the mounting operation according to the management results, thereby achieving effective management of the entity tag, so as to improve the application efficiency of the entity tag, and help expand the application scale of the entity tag.

[0118] In this embodiment, a label management device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated here. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0119] This embodiment provides a label management device, such as Figure 7 As shown, including:

[0120] The determination module 701 is used to obtain the data to be identified in the target platform and determine the entity tag based on the data to be identified;

[0121] An identification module 702 is used to identify the content of the entity tag and determine the management level of the entity tag according to the identification result;

[0122] The mounting module 703 is used to perform a mounting operation on the entity tag based on the mounting parameters corresponding to the management level, wherein the mounting operation is used to indicate that a corresponding content entity is allocated to the entity tag in the target platform.

[0123] In some optional implementations, the management levels include: a first level, a second level, and a third level; the identification module 702 is further used to:

[0124] Based on the recognition results, a semantically identifiable third-level entity label is obtained;

[0125] Determine the entity tag of the second level based on the identification information of the entity tag of the third level;

[0126] The mounting accuracy of the second-level entity tags is obtained, and the entity tags whose mounting accuracy exceeds the first threshold are determined as the first level.

[0127] In some optional implementations, the identification information includes a heat level identification; the identification module 702 is further configured to:

[0128] Based on the heat identifier, determine the heat value of the third-level entity tag;

[0129] When the heat value meets the heat condition, the third-level entity tag is determined as the second-level entity tag.

[0130] In some optional implementations, the determination module 701 is further configured to:

[0131] Obtain multimedia content published in the target platform;

[0132] Recognize the text in the multimedia content to obtain the data to be recognized.

[0133] In some optional implementations, the determination module 701 is further configured to:

[0134] Get the preset tag type;

[0135] Perform semantic recognition on the data to be identified, obtain keywords that match the preset tag type, and determine the entity tag based on the keywords.

[0136] In some optional implementations, the preset tag types include: a first type and a second type; the determination module 701 is further configured to:

[0137] Keywords whose semantic recognition results in the data to be identified are object description information are determined as keywords matching the first type, and keywords whose semantic recognition results in the data to be identified are phenomenon description information are determined as keywords matching the second type.

[0138] In some optional implementations, the preset tag type includes: a third type; the determination module 701 is further used to:

[0139] Semantic recognition is performed on keywords corresponding to the first type and the second type to obtain a third type that matches at least part of the target entity, wherein the target entity includes at least one of the following: a content publishing object, an item entity, a registered object in the target platform, and a content entity.

[0140] In some optional implementations, the determination module 701 is further configured to:

[0141] Analyze the phenomenon description information corresponding to the second type of entity tag and the item description information corresponding to the first type of entity tag to obtain a second type of entity tag that is associated with the first type of entity tag;

[0142] Establish associated information based on the associated relationship.

[0143] In some optional implementations, the determination module 701 is further configured to:

[0144] Obtaining coverage rates corresponding to each query content in the target platform, wherein the query content includes at least one keyword for obtaining corresponding content from the target platform, and the coverage rate is used to indicate a hit rate of the entity tag in the target platform to the keyword;

[0145] The data to be identified is determined according to the query content whose coverage is lower than the second threshold.

[0146] In some optional implementations, the management level includes a first level, a second level, and a third level, and the mounting parameter includes an automatic mounting ratio, wherein the automatic mounting ratio is used to indicate the proportion of automatic mounting operations when mounting the entity tag, the automatic mounting ratio corresponding to the second level is higher than the third level, and the automatic mounting ratio corresponding to the first level is higher than the second level; the mounting module 703 is further used to:

[0147] Get the automatic mount ratio pre-set for the management level;

[0148] Based on the automatic mounting ratio, mount operations are performed on the entity tags corresponding to each management level.

[0149] In some optional embodiments, the device is also used for:

[0150] When a new entity tag is detected, a target content entity matching the new entity tag is determined in the content entities of the target platform, and the target content entity is identified based on the new entity tag; or

[0151] When a new content entity in the target platform is detected, a target entity tag matching the new content entity is determined in the entity tags, and the new content entity is identified based on the entity tag.

[0152] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0153] The label management device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0154] The present disclosure also provides a computer device having the above Figure 7 The label management device shown.

[0155] See also Figure 8 , Figure 8is a schematic diagram of a computer device provided by an optional embodiment of the present disclosure, such as Figure 8 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 8 A processor 10 is taken as an example.

[0156] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0157] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0158] The memory 20 may 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; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0159] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0160] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0161] The embodiments of the present disclosure also provide a computer-readable storage medium. The above-mentioned method according to the embodiments of the present disclosure can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium and downloaded through a network, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0162] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0163] Although the embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present disclosure, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A tag management method, characterized in that: The method comprises: Acquire the data to be identified in the target platform, and determine the entity tag based on the data to be identified; Performing content identification on the entity tag, and determining a management level of the entity tag according to the identification result; Based on the mounting parameters corresponding to the management level, a mounting operation is performed on the entity tag, wherein the mounting operation is used to indicate that a corresponding content entity is allocated to the entity tag in the target platform.

2. The method according to claim 1, characterized in that The management levels include: first level, second level and third level; Determining the management level of the entity tag according to the identification result includes: Based on the recognition result, a semantically identifiable third-level entity label is obtained; Determine the entity tag of the second level based on the identification information of the entity tag of the third level; The mounting accuracy of the second-level entity tags is obtained, and the entity tags whose mounting accuracy exceeds the first threshold are determined as the first level.

3. The method according to claim 2, characterized in that The identification information includes a heat identification; The determining the second-level entity tag based on the identification information of the third-level entity tag includes: Based on the heat identifier, determining a heat value of the entity tag of the third level; When the popularity value meets the popularity condition, the third-level entity tag is determined as the second-level entity tag.

4. The method according to claim 1, characterized in that The step of obtaining the data to be identified in the target platform includes: Acquire multimedia content published in the target platform; Recognition is performed based on the text in the multimedia content to obtain the data to be recognized.

5. The method according to claim 1, characterized in that The determining the entity tag based on the to-be-identified data includes: Get the preset tag type; Semantic recognition is performed on the data to be recognized to obtain keywords that match the preset tag type, and the entity tag is determined according to the keywords.

6. The method according to claim 5, characterized in that The preset label types include: a first type and a second type; The performing semantic recognition on the data to be recognized to obtain keywords matching the preset tag type includes: Keywords whose semantic recognition results in the data to be identified are object description information are determined as keywords matching the first type, and keywords whose semantic recognition results in the data to be identified are phenomenon description information are determined as keywords matching the second type.

7. The method according to claim 6, characterized in that The preset tag types include: a third type; The performing semantic recognition on the data to be recognized to obtain keywords matching the preset tag type includes: Perform semantic recognition on keywords corresponding to the first type and the second type to obtain a third type that matches at least part of the target entity, wherein the target entity includes at least one of the following: a content publishing object, an item entity, a registered object in the target platform, and a content entity.

8. The method according to claim 6, characterized in that The method further comprises: Analyze the phenomenon description information corresponding to the second type of entity tag and the item description information corresponding to the first type of entity tag to obtain a second type of entity tag associated with the first type of entity tag; Establish association information according to the association relationship.

9. The method according to claim 1, characterized in that: The step of obtaining the data to be identified in the target platform further includes: Obtaining coverage rates corresponding to each query content in the target platform, wherein the query content includes at least one keyword for obtaining corresponding content from the target platform, and the coverage rate is used to indicate a hit rate of entity tags in the target platform to the keyword; The data to be identified is determined according to the query content whose coverage is lower than a second threshold.

10. The method according to claim 1, characterized in that The management level includes a first level, a second level and a third level, and the mounting parameter includes an automatic mounting ratio, wherein the automatic mounting ratio is used to indicate a proportion of automatic mounting operations when mounting the entity tag, the automatic mounting ratio corresponding to the second level is higher than the third level, and the automatic mounting ratio corresponding to the first level is higher than the second level; The mounting operation on the entity tag based on the mounting parameters corresponding to the management level includes: Get the automatic mount ratio pre-set for the management level; Based on the automatic mounting ratio, mounting operations are performed on the entity tags corresponding to each management level respectively.

11. The method according to claim 1, characterized in that: The method further comprises: When a new entity tag is detected, determining a target content entity matching the new entity tag in the content entities of the target platform, and identifying the target content entity based on the new entity tag; or When a new content entity in the target platform is detected, a target entity tag matching the new content entity is determined in the entity tags, and the new content entity is identified based on the entity tag.

12. A label management device, characterized in that: The device comprises: A determination module, used to obtain the data to be identified in the target platform, and determine the entity tag based on the data to be identified; an identification module, configured to perform content identification on the entity tag and determine a management level of the entity tag according to the identification result; The mounting module is used to perform a mounting operation on the entity tag based on the mounting parameters corresponding to the management level, wherein the mounting operation is used to indicate that a corresponding content entity is allocated to the entity tag in the target platform.

13. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the label management method according to any one of claims 1 to 11 by executing the computer instructions.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the tag management method according to any one of claims 1 to 1.

15. A computer program product, characterized in that The method comprises computer instructions, wherein the computer instructions are used to enable a computer to execute the tag management method according to any one of claims 1 to 11.

Images