Community content production method and device, electronic equipment and storage medium

Abstract

The invention provides a community content production method and device, electronic equipment and a storage medium, and relates to the technical field of computers, in particular to the technical fields of artificial intelligence, intelligent recommendation, big data processing, information flow, community and content platform development and the like. According to the specific implementation scheme, a plurality of content search terms are obtained; the method comprises the following steps: by taking each content search word in a plurality of content search words as a target search word, constructing virtual theme content related to the target search word; generating virtual comment information for the virtual theme content; generating virtual community content based on the virtual theme content and the virtual comment information; wherein the virtual community content is used for pushing the virtual community content to the target user when the search engine receives an actual search term input by the target user; the actual search terms are search terms related to the virtual community content. According to the invention, the content quality, content diversity and content updating timeliness of the community and the content platform can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to the fields of artificial intelligence, intelligent recommendation, big data processing, information flow, and the development of community and content platforms. Specifically, it relates to a community content production method, apparatus, electronic device, and storage medium. Background Technology

[0002] A community and content platform is a platform that provides users with opportunities for communication and sharing. Users can publish topics, participate in comments, and reply to comments on community and content platforms to achieve the purpose of communication and sharing with other users. Summary of the Invention

[0003] This disclosure provides a method, apparatus, electronic device, and storage medium for producing community content.

[0004] According to a first aspect of this disclosure, a method for producing community content is provided, comprising:

[0005] Retrieve multiple content search terms;

[0006] By using each of the multiple content search terms as the target search term, virtual topic content related to the target search term is constructed;

[0007] Generate virtual comment information for virtual topic content;

[0008] Virtual community content is generated based on virtual topic content and virtual comment information. This virtual community content is used by the search engine to push the virtual community content to the target user when it receives the actual search terms entered by the target user. The actual search terms are search terms related to the virtual community content.

[0009] According to a second aspect of this disclosure, a community content production apparatus is provided, comprising:

[0010] The search term acquisition unit is used to acquire multiple content search terms;

[0011] The topic content construction unit is used to construct virtual topic content related to the target search term by using each of multiple content search terms as the target search term;

[0012] The comment information generation unit is used to generate virtual comment information for virtual topic content;

[0013] The community content generation unit is used to generate virtual community content based on virtual topic content and virtual comment information. The virtual community content is used by the search engine to push the virtual community content to the target user when it receives the actual search terms entered by the target user. The actual search terms are search terms related to the virtual community content.

[0014] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0015] At least one processor;

[0016] The memory that is communicatively connected to the at least one processor;

[0017] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method provided in the first aspect of this disclosure.

[0018] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform the method provided according to a first aspect of this disclosure.

[0019] According to a fifth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to a first aspect of this disclosure.

[0020] Adopting this disclosure can improve the quality, diversity, and timeliness of content updates in communities and content platforms.

[0021] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0022] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0023] Figure 1 A flowchart illustrating a community content production method provided in this embodiment of the disclosure;

[0024] Figure 2 This disclosure provides an auxiliary description of a community content production method for embodiments of the present disclosure. Figure 1 ;

[0025] Figure 3 This disclosure provides an auxiliary description of a community content production method for embodiments of the present disclosure. Figure 2 ;

[0026] Figure 4 This disclosure provides an auxiliary description of a community content production method for embodiments of the present disclosure. Figure 3 ;

[0027] Figure 5A diagram illustrating a method for storing virtual community content provided in an embodiment of this disclosure;

[0028] Figure 6 A schematic diagram of a preset structured template provided in an embodiment of this disclosure;

[0029] Figure 7 Another flowchart of the community content production method provided in this disclosure embodiment;

[0030] Figure 8 A schematic diagram of a preset recommendation result template provided in an embodiment of this disclosure;

[0031] Figure 9 This is a schematic diagram illustrating an application scenario of a community content production method provided in an embodiment of this disclosure;

[0032] Figure 10 A schematic structural block diagram of a community content production device provided in this disclosure embodiment;

[0033] Figure 11 This is a schematic structural block diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0034] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0035] As mentioned earlier, community and content platforms (such as "Baidu Tieba") are platforms that provide users with opportunities for communication and sharing. Users can post topics, participate in comments, and reply to comments on these platforms to achieve the purpose of communication and sharing with other users. However, the inventors' research has found that current community and content platforms suffer from problems such as poor content quality, insufficient content diversity, untimely content updates, and a lack of traffic channels. Ultimately, this may lead to a rapid decline in the number of active users (e.g., Daily Active Users (DAU)).

[0036] To address the above problems, this disclosure provides a community content production method that can be applied to electronic devices. These electronic devices can be servers, workbenches, mainframe computers, conventional computers (e.g., desktop computers, laptops, tablets, etc.), smartphones, personal digital processors, or other similar computing devices. The following will be combined with… Figure 1The flowchart shown illustrates a community content production method provided by an embodiment of this disclosure. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described in the flowchart may be performed in a different order.

[0037] Step S101: Obtain multiple content search terms.

[0038] Among them, multiple content search terms can include at least one of historical popular search terms, vertical search terms, and less popular search terms.

[0039] In this embodiment, popular search terms can be search terms included in a preset search term set and whose historical search count on the content providing platform exceeds a first preset threshold; vertical search terms can be search terms corresponding to vertical categories such as games, animation, education, digital products, automobiles, and healthcare; less popular search terms can be search terms included in the preset search term set and whose historical search count on the content providing platform is less than a second preset threshold. The preset search term set can be set according to application requirements, and this embodiment does not limit its implementation. The content providing platform can include multiple data platforms capable of communicating with electronic devices, such as search engines, community and content platforms, video and entertainment platforms, etc.; the first preset threshold can be a large value, and its specific value can also be set according to application requirements, and this embodiment does not limit its implementation. The second preset threshold can be a small value, and its specific value can also be set according to application requirements, and this embodiment does not limit its implementation.

[0040] Step S102: Take each of the multiple content search terms as the target search term and construct virtual topic content related to the target search term.

[0041] Virtual topics can include, but are not limited to, discussion and exchange topics, sharing and display topics, knowledge popularization topics, resource recommendation topics, and emotional expression topics.

[0042] In this embodiment of the disclosure, existing topic content can be obtained from a content providing platform, and virtual topic content related to the target search term can be constructed accordingly. Alternatively, a large model can be used to construct virtual topic content related to the target search term from scratch. This embodiment of the disclosure does not limit the scope of the invention. The large model can be pre-trained, specifically a Large Language Model (LLM). The LLM can be a pre-trained neural network model (e.g., an autoregressive generative model with a Transformer architecture) that possesses general language knowledge, world knowledge, and domain-specific expertise (e.g., expertise in the field of computer technology).

[0043] Step S103: Generate virtual comment information for the virtual topic content.

[0044] The virtual comment information can include multiple available comment messages.

[0045] In this embodiment of the disclosure, existing comment information can be obtained from the content providing platform, and virtual comment information for virtual topic content can be generated accordingly. Alternatively, a large model can be used to generate virtual comment information for virtual topic content from scratch. This embodiment of the disclosure does not limit this approach.

[0046] Step S104: Generate virtual community content based on virtual topic content and virtual comment information.

[0047] In one example, virtual topic content and virtual evaluation information can be combined to generate virtual community content. Specifically, virtual topic content and virtual evaluation information can be combined according to a preset content template to generate virtual community content. The preset content template can be set according to application requirements, and this embodiment of the disclosure does not limit it. When the community and content platform is "Tieba", a piece of virtual community content can be considered as a "post".

[0048] Furthermore, it should be noted that in this embodiment, the generated virtual community content can be published on the community and content platform, and can also be used by the search engine to push the virtual community content to the target user when it receives the actual search terms entered by the target user, so that the virtual community content pushed by the search engine can enter the community and content platform. The target user can be any user of the search engine; the actual search terms are search terms related to the virtual community content.

[0049] The community content production method provided in this disclosure can acquire multiple content search terms, use each of the multiple content search terms as a target search term, construct virtual topic content related to the target search term, generate virtual comment information for the virtual topic content, and then generate virtual community content based on the virtual topic content and virtual comment information. Moreover, the virtual community content can be used by the search engine to push the virtual community content to the target user when it receives the actual search term (i.e., the search term related to the virtual community content) input by the target user. In this process, on the one hand, multiple virtual community content items can be generated, each corresponding to a different content search term, through virtualization. This addresses the issues of poor content quality, insufficient content diversity, and untimely updates on communities and content platforms, thereby improving the quality, diversity, and timeliness of content updates. On the other hand, after generating a virtual community content item, it can be used by search engines to push the virtual community content to target users when they receive actual search terms. This solves the problem of a lack of traffic channels for communities and content platforms, providing users with richer entry point options, simplifying user operations, and further enhancing the user experience.

[0050] Furthermore, as mentioned above, in this embodiment of the disclosure, the multiple content search terms may include at least one of historical popular search terms, vertical search terms, and unpopular search terms; that is, the target search term may be one of popular search terms, vertical search terms, and unpopular search terms.

[0051] In some optional implementations, when the target search term is a popular search term, existing topic content can be obtained from a content provider platform, and virtual topic content related to the target search term can be constructed accordingly. Based on this, in this embodiment of the disclosure, step S102, "constructing virtual topic content related to the target search term," may include:

[0052] Obtain multiple initial topic selections related to the target search term from content provider platforms;

[0053] Based on multiple thematic evaluation dimensions, N high-quality thematic content are selected from multiple initial shortlisted thematic content;

[0054] Based on N high-quality topic content, construct virtual topic content related to the target search term.

[0055] In one example, a first search tool (e.g., Site) can be used to retrieve multiple initial topic contents related to the target search term from a content provider platform.

[0056] For example, a first retrieval tool can be used to obtain multiple preliminary search terms related to the target search term from a content providing platform, serving as the first preliminary selection of topics. Alternatively, the first retrieval tool can be used to obtain multiple preliminary search terms related to the target search term from a content providing platform, and obtain the Natural Language Processing (NLP) score for each of the multiple preliminary search terms. Based on the NLP score of each of the multiple preliminary search terms, multiple first preliminary selection topics can be selected from the multiple preliminary search terms. For example, preliminary search terms with NLP scores greater than a first preset scoring threshold can be selected as the first preliminary selection of topics. For each of the multiple preliminary search terms, the NLP score can be the relevance score between the preliminary search term and the target search term, specifically the relevance score between the title of the preliminary search term and the target search term. The first preset scoring threshold can be set according to application requirements, and this embodiment does not limit this.

[0057] After obtaining multiple initial shortlisted topics related to the target search term from the content provider platform, N high-quality topics can be selected from these initial shortlisted topics based on multiple topic evaluation dimensions. Then, based on these N high-quality topics, virtual topic content related to the target search term can be constructed. For example, using a large model, the N high-quality topics can be integrated into a single overall topic content, which can then be used as the virtual topic content related to the target search term. The multiple topic evaluation dimensions can include the first positive / negative status, the first subject / object relationship, the first relevance, the first text richness, the first number of interactions, and the first timeliness, etc.; N≥1, and N is an integer.

[0058] Furthermore, in this embodiment of the disclosure, "selecting N high-quality topic contents from multiple first preliminary topic contents based on multiple topic evaluation dimensions" may include:

[0059] Based on the first theme evaluation dimension, multiple initial theme contents are filtered to obtain multiple remaining theme contents;

[0060] Based on the second topic evaluation dimension, the quality of multiple remaining topic contents is ranked to obtain the topic quality ranking results for multiple remaining topic contents.

[0061] Based on the topic quality ranking results, select N high-quality topics from the remaining topics.

[0062] The first theme evaluation dimension can have at least one, specifically including the first pornographic / anti-pornographic state and the first subject-object state. Here, the first pornographic / anti-pornographic state can be used to characterize whether the first preliminary theme content involves pornography or anti-pornography; the first subject-object state can be used to characterize whether the first preliminary theme content can be viewed by the public without barriers on the content providing platform. Generally, when the first preliminary theme content is set to "visible to everyone", it can be considered that it can be viewed by the public without barriers on the content providing platform.

[0063] Based on this, in this embodiment of the disclosure, when filtering multiple first preliminary topic contents based on the first topic evaluation dimension to obtain multiple remaining topic contents, the first preliminary topic contents involving pornography or anti-corruption can be deleted, and the first preliminary topic contents that cannot be accessed by the public without barriers on the content providing platform can be deleted, thereby obtaining multiple remaining topic contents.

[0064] After filtering multiple initial topic contents based on the first topic evaluation dimension to obtain multiple remaining topic contents, the remaining topic contents can be ranked by quality based on the second topic evaluation dimension to obtain the topic quality ranking results for the multiple remaining topic contents.

[0065] The second theme evaluation dimension can have at least one, specifically including primary relevance, primary text richness, primary interaction quantity, and primary timeliness. Here, primary relevance can be used to characterize the primary relevance score between the remaining theme content and the target search term, specifically the primary relevance score between the title of the remaining theme content and the target search term; primary text richness can be used to characterize the richness and diversity of the remaining theme content in terms of specific content, content format, information content, etc.; primary interaction quantity can be used to characterize the number of comments, likes, and shares of the remaining theme content on the content providing platform; primary timeliness can be used to characterize the publication time of the remaining theme content on the content providing platform.

[0066] Based on this, in this embodiment of the disclosure, when ranking the quality of multiple remaining topic content based on the second topic evaluation dimension to obtain the topic quality ranking result for multiple remaining topic content, a first relevance score, a first text richness score, a first interaction quantity score, and a first timeliness score can be obtained for each remaining topic content (for example, the later the publication time, the higher the relevance score). The first relevance score, the first text richness score, the first interaction quantity score, and the first timeliness score are weighted and fused to obtain the topic content quality score for the remaining topic content. Then, according to the topic content score of each remaining topic content, the multiple remaining topic content is ranked in quality to obtain the topic quality ranking result for multiple remaining topic content.

[0067] After ranking the remaining topics based on the second topic evaluation dimension, and obtaining the topic quality ranking results for the remaining topics, N high-quality topics can be selected from the remaining topics based on the topic quality ranking results. For example, when N=3, 3 high-quality topics can be selected from the remaining topics.

[0068] In this embodiment of the present disclosure, multiple initial topic content related to the target search term can be obtained from a content providing platform. Based on multiple topic evaluation dimensions, N high-quality topic content items are selected from these initial topic content items. Then, based on these N high-quality topic content items, virtual topic content related to the target search term is constructed. In other words, in this embodiment of the present disclosure, the virtual topic content is not only constructed based on the content providing platform, but also based on N high-quality topic content items that have undergone quality evaluation through multiple topic evaluation dimensions. Furthermore, when selecting N high-quality topic content items from the multiple initial topic content items based on multiple topic evaluation dimensions, the multiple initial topic content items are filtered based on the first topic evaluation dimension to obtain multiple remaining topic content items. Based on the second topic evaluation dimension, the multiple remaining topic content items are ranked by quality to obtain a topic quality ranking result for the multiple remaining topic content items. Finally, based on the topic quality ranking result, N high-quality topic content items are selected from the multiple remaining topic content items. In this way, not only can virtual subject content be given a realistic simulation effect that "looks like it's been built by a real person," thereby enhancing its credibility, but it can also improve the quality of the virtual subject content. Specifically, it can enhance the legality, text richness, and real-time nature of the virtual subject content.

[0069] Furthermore, in this embodiment of the disclosure, when "constructing virtual topic content related to the target search term" in step S102 includes "obtaining multiple first preliminary topic contents related to the target search term from the content providing platform; selecting N high-quality topic contents from the multiple first preliminary topic contents based on multiple topic evaluation dimensions; and constructing virtual topic content related to the target search term based on the N high-quality topic contents," step S103, that is, "generating virtual comment information for the virtual topic content," may include:

[0070] Obtain multiple initial review comments related to the target search term from the content provider platform;

[0071] Based on multiple primary review evaluation dimensions, multiple primary usable review information is selected from multiple primary initial review information;

[0072] Based on multiple available comment information, generate virtual comment information for the virtual topic content.

[0073] In one example, multiple first coarse-search comment information related to the target search term can be obtained from a content providing platform. An NLP score for each of these first coarse-search comment information is then obtained. Based on the NLP score for each of the first coarse-search comment information, multiple first preliminary selection comment information are selected from the multiple first coarse-search comment information. For example, the first coarse-search comment information whose NLP score is greater than a second preset scoring threshold can be selected as the first preliminary selection comment information. For each of the multiple first coarse-search comment information, the NLP score can be the relevance score between the first coarse-search comment information and the target search term. The second preset scoring threshold can be set according to application requirements, and this embodiment does not limit this setting.

[0074] After obtaining multiple initial review information related to the target search term from the content provider platform, multiple usable reviews can be selected from these initial reviews based on multiple initial review evaluation dimensions. Then, virtual review information for the virtual topic content can be generated based on this usable review information. For example, multiple usable reviews can be combined into a single virtual review page for the virtual topic content, generating a review page that includes the virtual review information (i.e., multiple usable reviews). These multiple initial review evaluation dimensions can include secondary anti-yellow status, secondary subject-object status, secondary text richness, and secondary interaction quantity, among others.

[0075] Furthermore, in this embodiment of the disclosure, "selecting multiple first usable comment information from multiple first preliminary comment information based on multiple first comment evaluation dimensions" may include:

[0076] Based on multiple primary review evaluation dimensions, multiple initial review information is filtered to obtain multiple usable primary review information.

[0077] In this embodiment of the disclosure, the second pornographic / anti-pornographic state can be used to characterize whether the first preliminary selection comment information involves pornography or anti-pornography; the second subject-object state can be used to characterize whether the first preliminary selection comment information can be viewed by the public without barriers on the content providing platform. Under normal circumstances, when the first preliminary selection comment information is set to "visible to everyone", it can be considered that it can be viewed by the public without barriers on the content providing platform; the second text richness can be used to characterize the richness and diversity of the first preliminary selection comment information in terms of specific content, content form, information volume, etc.; the second interaction quantity can be used to characterize the number of comments, likes, and reposts of the first preliminary selection comment information on the content providing platform.

[0078] Based on this, in this embodiment of the disclosure, when filtering multiple first preliminary comment information based on multiple first comment evaluation dimensions to obtain multiple first usable comment information, a second text richness score and a second interaction quantity score can be obtained for each first preliminary comment information. First preliminary comment information involving pornography or anti-corruption will be deleted, as will first preliminary comment information that cannot be accessed by the public without barriers on the content providing platform. Then, first preliminary comment information with a second text richness score lower than a third preset score threshold will be deleted. Finally, first preliminary comment information with a second interaction quantity score lower than a fourth preset score threshold will be deleted, thereby obtaining multiple first usable comment information. The third and fourth preset score thresholds can be set according to application requirements, and this embodiment of the disclosure does not impose any restrictions on them.

[0079] Through the above methods, in this embodiment of the disclosure, multiple first preliminary review information related to the target search term can be obtained from the content providing platform. Based on multiple first review evaluation dimensions, multiple first usable review information is selected from the multiple first preliminary review information. Then, based on the multiple first usable review information, virtual review information for the virtual topic content is generated. In other words, in this embodiment of the disclosure, the virtual review information is not only constructed based on the content providing platform, but also generated based on multiple first usable review information that has undergone quality evaluation through multiple first review evaluation dimensions. Moreover, when selecting multiple first usable review information from the multiple first preliminary review information based on multiple first review evaluation dimensions, the multiple first preliminary review information can be filtered based on multiple first review evaluation dimensions to obtain multiple first usable review information. This not only gives the virtual review information a "realistic" simulation effect, thereby improving its credibility, but also improves the content quality of the virtual review information; specifically, it can improve the legality and text richness of the virtual review information.

[0080] In some optional implementations, when the target search term is a vertical category search term, existing topic content can be obtained from the content providing platform, and virtual topic content related to the target search term can be constructed accordingly. Based on this, in this embodiment of the disclosure, step S102, "constructing virtual topic content related to the target search term," may include:

[0081] Obtain multiple preliminary topic selections related to the target search term from content provider platforms;

[0082] Select multiple valid topics from a pool of preliminary second-round topics;

[0083] Based on multiple valid topic content, construct virtual topic content related to the target search term.

[0084] In one example, a second retrieval tool (e.g., Site) can be used to obtain multiple preliminary topic content related to the target search term from a content provider platform. After obtaining these preliminary topic content, multiple valid topic content can be selected from them. Based on these valid topic content, virtual topic content related to the target search term can be constructed. For example, a large model can be used to integrate multiple valid topic content into a single overall topic content, which can then be used as the virtual topic content related to the target search term. In another example, when selecting multiple valid topic content from the multiple preliminary topic content, each preliminary topic content can be parsed to obtain the videos, images, text, etc., included within it. The validity of these videos, images, and text can be verified to ensure they do not have issues such as being unplayable, displaying garbled text, or containing blank spaces. If these issues are resolved, the preliminary topic content is considered valid topic content.

[0085] Furthermore, it should be noted that in this embodiment of the disclosure, after selecting multiple valid topic contents from multiple second preliminary topic contents, data cleaning operations can be performed on the multiple valid topic contents. Specifically, this can include at least one of black word filtering, watermark removal, and ad removal. Moreover, after performing data cleaning operations on the multiple valid topic contents, deduplication operations can be performed on the multiple valid topic contents to obtain multiple valid topic contents after deduplication. Based on this, in this embodiment of the disclosure, when constructing virtual topic contents related to the target search term based on multiple valid topic contents, virtual topic contents related to the target search term can be constructed based on the multiple valid topic contents after deduplication. For example, using a large model, the multiple valid topic contents after deduplication can be integrated into a whole topic content, and this whole topic content can be used as the virtual topic content related to the target search term.

[0086] Through the above methods, in this embodiment of the disclosure, multiple second preliminary topic contents related to the target search term can be obtained from the content providing platform, and multiple valid topic contents can be selected from these multiple second preliminary topic contents. Then, based on these multiple valid topic contents, virtual topic content related to the target search term can be constructed. In other words, in this embodiment of the disclosure, the virtual topic content is not only constructed based on the content providing platform, but also based on multiple valid topic contents that have undergone validity evaluation. This not only gives the virtual topic content a realistic simulation effect, enhancing its credibility, but also improves its content quality. Specifically, it ensures that the virtual topic content does not have validity issues such as being unplayable, having garbled text, or blank spaces.

[0087] Furthermore, it should be noted that in this embodiment of the disclosure, when the target search term is a vertical category search term, the virtual topic content can also be provided by a vertical category content partner. This vertical category content partner may include game partners, novel content providers, cloud storage services, and photo albums, etc. Specifically, after obtaining the target search term and determining that it is a vertical category search term, multiple third-stage preliminary topic content related to the target search term can be obtained from the data platform provided by the vertical category content partner. Multiple valid topic content is then selected from these multiple third-stage preliminary topic content to construct virtual topic content related to the target search term based on these multiple valid topic content. The specific functions and examples of "selecting multiple valid topic content from multiple third-stage preliminary topic content to construct virtual topic content related to the target search term based on these multiple valid topic content" can be found in the descriptions of similar steps described above, and will not be repeated here.

[0088] Furthermore, in this embodiment of the disclosure, when "constructing virtual topic content related to the target search term" in step S102 includes "obtaining multiple second preliminary topic contents related to the target search term from the content providing platform; selecting multiple valid topic contents from the multiple second preliminary topic contents; and constructing virtual topic content related to the target search term based on the multiple valid topic contents," step S103, that is, "generating virtual comment information for the virtual topic content," may include at least one of the following:

[0089] Based on the first comment generation template and multiple first comment generation requirements, generate virtual comment information for the virtual topic content;

[0090] Using a large model, virtual comment information is generated for virtual topic content based on multiple first comment generation requirements;

[0091] Based on the content delivery platform, virtual comment information is generated for virtual topic content.

[0092] The first comment generation template can be set according to application requirements, and this embodiment does not limit it; multiple first comment generation requirements may include, but are not limited to, not involving pornography or anti-pornography, being highly relevant to the virtual topic content, and having high text richness.

[0093] Furthermore, it should be noted that in this embodiment of the disclosure, "generating virtual comment information for virtual topic content based on the content providing platform" may include:

[0094] Obtain multiple preliminary review comments related to the target search term from the content provider platform;

[0095] Based on multiple secondary review evaluation dimensions, multiple secondary usable review information is selected from multiple secondary preliminary review information;

[0096] Based on multiple available secondary comment information, virtual comment information is generated for the virtual topic content.

[0097] In one example, multiple coarse-search comment pieces related to the target search term can be obtained from a content providing platform. An NLP score for each coarse-search comment piece is then obtained. Based on the NLP score for each coarse-search comment piece, multiple preliminary second-selection comment pieces are selected from the coarse-search comment pieces. For example, a coarse-search comment piece with an NLP score greater than a fifth preset scoring threshold can be selected as the preliminary second-selection comment piece. For each coarse-search comment piece, the NLP score can be a relevance score between the coarse-search comment piece and the target search term. The fifth preset scoring threshold can be set according to application requirements, and this embodiment does not impose any limitations on it.

[0098] After obtaining multiple preliminary second-selection comment information related to the target search term from the content provider platform, multiple usable second-selection comment information can be selected from the multiple preliminary second-selection comment information based on multiple second-selection comment evaluation dimensions. Based on these usable second-selection comment information, virtual comment information for the virtual topic content can be generated. For example, multiple usable second-selection comment information can be combined as virtual comment information for the virtual topic content, and a comment information page including the virtual comment information (i.e., including multiple usable second-selection comment information) can be generated. These multiple second-selection comment evaluation dimensions may include third-party objectivity, third-party subject-object interaction, third-party text richness, and third-party interaction quantity, etc.

[0099] Furthermore, in this embodiment of the disclosure, "selecting multiple second available comment information from multiple second preliminary comment information based on multiple second comment evaluation dimensions" may include:

[0100] Based on multiple secondary review evaluation dimensions, multiple secondary preliminary review information is filtered to obtain multiple secondary usable review information.

[0101] In this embodiment of the disclosure, the third yellow / anti-pornographic state can be used to characterize whether the second preliminary comment information involves pornography or anti-pornography; the third subject / object state can be used to characterize whether the second preliminary comment information can be viewed by the public without barriers on the content providing platform. Generally, when the second preliminary comment information is set to "visible to everyone", it can be considered that it can be viewed by the public without barriers on the content providing platform; the third text richness can be used to characterize the richness and diversity of the second preliminary comment information in terms of specific content, content form, information volume, etc.; the third interaction quantity can be used to characterize the number of comments, likes, and reposts of the second preliminary comment information on the content providing platform.

[0102] Based on this, in this embodiment of the disclosure, when filtering multiple preliminary second-selection comment information based on multiple second-review evaluation dimensions to obtain multiple usable second-selection comment information, a third text richness score and a third interaction quantity score can be obtained for each preliminary second-selection comment information. Preliminary second-selection comment information involving pornography or anti-corruption content is deleted, as are those that cannot be accessed freely by the public on the content providing platform. Furthermore, preliminary second-selection comment information with a third text richness score below a sixth preset scoring threshold is deleted. Finally, preliminary second-selection comment information with a third interaction quantity score below a seventh preset scoring threshold is deleted, thereby obtaining multiple usable second-selection comment information. The sixth and seventh preset scoring thresholds can be set according to application requirements, and this embodiment of the disclosure does not impose any restrictions on them.

[0103] Through the above methods, in this embodiment of the disclosure, virtual comment information for virtual topic content can be generated according to a first comment generation template and based on multiple first comment generation requirements. Alternatively, a large model can be used to generate virtual comment information for virtual topic content based on multiple first comment generation requirements. Furthermore, virtual comment information for virtual topic content can be generated based on a content providing platform. In other words, in this embodiment of the disclosure, the methods for generating virtual comment information can be diversified, thereby enhancing the diversity of virtual comment information.

[0104] Furthermore, when generating virtual comment information for virtual topic content based on a content providing platform, multiple preliminary second-selection comment information related to the target search term can be obtained from the content providing platform. Then, based on multiple second-selection comment evaluation dimensions, multiple usable second-selection comment information is selected from these preliminary second-selection comment information. Finally, based on these usable second-selection comment information, virtual comment information for the virtual topic content is generated. In this process, the virtual comment information is not only constructed based on the content providing platform, but also generated based on multiple usable second-selection comment information that has undergone quality evaluation across multiple second-selection comment evaluation dimensions. Moreover, when selecting multiple usable second-selection comment information from the preliminary second-selection comment information based on these dimensions, the preliminary second-selection comment information can be filtered to obtain the usable second-selection comment information. This not only gives the virtual comment information a realistic simulation effect, enhancing its credibility, but also improves its content quality, specifically enhancing its legality and text richness.

[0105] In some optional implementations, when the target search term is a niche search term, a large model can be used to construct virtual topic content related to the target search term from scratch. Based on this, in this embodiment of the disclosure, step S102, "constructing virtual topic content related to the target search term," may include:

[0106] Obtain multiple content generation requirements;

[0107] By leveraging a large model and based on multiple content generation requirements, virtual topic content related to the target search terms is constructed.

[0108] Among these, content generation requirements may include, but are not limited to, not involving pornography or anti-racism, being highly relevant to the target search term, and having high text richness.

[0109] In addition, it should be noted that when using a large model to construct virtual topic content related to the target search term based on multiple content generation requirements, the large model can be invoked outside of peak traffic periods. For example, the large model can be invoked at night from 00:00 to 6:00 to construct virtual topic content related to the target search term based on multiple content generation requirements.

[0110] Through the above methods, this embodiment of the disclosure can obtain multiple content generation requirements and utilize a large model to construct virtual topic content related to the target search term based on these requirements. Since the application method of the large model is simple and does not require the development of a large number of programs, it can reduce manual input costs and save project costs.

[0111] Furthermore, in this embodiment of the disclosure, when "constructing virtual topic content related to the target search term" in step S102 includes "obtaining multiple content generation requirements; using a large model, based on the multiple content generation requirements, constructing virtual topic content related to the target search term", step S103, that is, "generating virtual comment information for the virtual topic content", may include at least one of the following:

[0112] Based on the second comment generation template and multiple second comment generation requirements, generate virtual comment information for the virtual topic content;

[0113] By utilizing a large model and based on multiple secondary comment generation requirements, virtual comment information is generated for virtual topic content.

[0114] The second comment generation template can be set according to application requirements, and this embodiment does not limit it; multiple requirements for generating second comments may include, but are not limited to, not involving pornography or anti-pornography, being highly relevant to the virtual topic content, and having high text richness.

[0115] In one example, according to the second comment generation template, based on multiple second comment generation requirements, virtual comment information for virtual topic content is generated. This can be done by: according to the second comment generation template, based on multiple second comment generation requirements, generating multiple third available comment information, and using the multiple third available comment information together as virtual comment information for virtual topic content, and then generating a comment information page that includes the virtual comment information (i.e., multiple third available comment information).

[0116] In one example, using a large model, based on multiple second comment generation requirements, to generate virtual comment information for virtual topic content, could be: using a large model, based on multiple second comment generation requirements and multiple fourth available comment information, and combining the multiple fourth available comment information as virtual comment information for virtual topic content, and then generating a comment information page that includes virtual comment information (i.e., multiple fourth available comment information).

[0117] Through the above methods, in this embodiment of the disclosure, virtual comment information for virtual topic content can be generated according to the second comment generation template and based on multiple second comment generation requirements. Alternatively, a large model can be used to generate virtual comment information for virtual topic content based on multiple second comment generation requirements. In other words, in this embodiment of the disclosure, the generation methods of virtual comment information can be diversified, thereby improving the diversity of virtual comment information.

[0118] In some optional implementations, after constructing virtual topic content related to the target search term and generating virtual comment information for the virtual topic content through steps S101, S102, and S103, when performing step S104, that is, "generating virtual community content based on virtual topic content and virtual comment information", the following can be done:

[0119] Initial community content is generated based on virtual topic content and virtual comment information;

[0120] Meta tag optimization is performed on community content to obtain intermediate community content;

[0121] Virtual community content is derived from the content of the intermediate community.

[0122] In one example, virtual topic content and virtual evaluation information can be combined to generate initial community content. Specifically, virtual topic content and virtual evaluation information can be combined according to a preset content template to generate initial community content. The preset content template can be set according to application requirements, and this embodiment does not impose any limitations on it.

[0123] After generating initial community content based on virtual topic content and virtual comment information, meta tag optimization can be performed on the community content to obtain intermediate community content. Meta tag optimization includes at least one of title optimization, description optimization, and keyword optimization. That is, in this embodiment of the disclosure, the intermediate community content can be obtained by optimizing the title, description, and keywords of the initial community content (also known as TDK optimization). In one example, TDK assistants and search engine optimization (SEO) tools can be used to optimize the title, description, and keywords of the initial community content to obtain intermediate community content.

[0124] After performing meta tag optimization on community content to obtain intermediate community content, virtual community content can be obtained based on the intermediate community content. For example, the intermediate community content can be processed according to the communication protocol specifications of the search engine, and the intermediate community content can be configured and optimized with the Uniform Resource Locator (URL) to obtain virtual community content.

[0125] With the above settings, in this embodiment of the disclosure, after constructing virtual topic content related to the target search term and generating virtual comment information for the virtual topic content, the virtual community content is not directly generated based on the virtual topic content and virtual comment information. Instead, initial community content is generated based on the virtual topic content and virtual comment information, and meta tag optimization is performed on the community content to obtain intermediate community content. Then, the virtual community content is obtained based on the intermediate community content. This further improves the content quality of the virtual community content; specifically, it enhances the clarity of the titles and the detail of the descriptions. Simultaneously, it improves the accuracy of pushing virtual community content to target users.

[0126] Furthermore, in this embodiment of the disclosure, after constructing the virtual topic content related to the target search term, the following can also be done:

[0127] Identify multiple candidate virtual accounts;

[0128] Based on multiple account selection dimensions, usable virtual accounts are selected from multiple candidate virtual accounts;

[0129] You can bind available virtual accounts to virtual theme content.

[0130] In this context, multiple candidate virtual accounts can be simply understood as multiple candidate virtual aliases; the selection dimensions for multiple accounts can include the total number of interactions, the number of reports received, and whether there is a ban record. Here, the total number of interactions can be used to represent the number of comments, likes, and shares of all historical community content posted by the candidate virtual accounts on community and content platforms.

[0131] In one example, after identifying multiple candidate virtual accounts, a candidate virtual account with an overall interaction count greater than a first preset threshold, a number of reports less than a third preset threshold, and no ban record can be selected as an available virtual account, and then the available virtual account can be bound to the virtual topic content. The first preset threshold and the third preset threshold can be set according to application requirements, and this embodiment does not impose any restrictions on them.

[0132] Through the above methods, in this embodiment of the disclosure, multiple candidate virtual accounts can be identified, and a usable virtual account can be selected from the multiple candidate virtual accounts based on multiple account selection dimensions. Then, the usable virtual account is bound to virtual topic content. In other words, in this embodiment of the disclosure, after constructing virtual topic content related to the target search term, the virtual topic content can be bound to a high-quality, usable virtual account to improve the credibility of the virtual topic content, thereby reducing the difficulty of promoting the virtual topic content, that is, reducing the difficulty of promoting virtual community content generated based on virtual topic content.

[0133] Correspondingly, in this embodiment of the disclosure, after generating virtual comment information (including multiple available comment information, such as multiple first available comment information, multiple second available comment information, multiple third available comment information, or multiple fourth available comment information) for each available comment information, a candidate virtual account can be selected from multiple candidate virtual accounts (for example, randomly selected), and the candidate virtual account can be bound to the available comment information to improve the authenticity of the available comment information.

[0134] Furthermore, in this embodiment of the disclosure, after generating the virtual community content through steps S101, S102, S103, and S104, at least one of the following can also be performed:

[0135] The content of the virtual community is stored in the first content library (also known as the "Daily Library");

[0136] The virtual community content is stored in a designated storage space within a second content repository (also known as the "designated content repository").

[0137] The designated storage space can be the data storage space corresponding to the topic category of the virtual community content. For example, if the community and content platform is "Tieba" and the topic category of the virtual community content is "Celebrity Liu Moumou", the designated storage space can be the data storage space corresponding to "Celebrity Liu Moumou Bar".

[0138] Through the above methods, this embodiment of the disclosure allows for the storage of virtual community content using various storage methods. Specifically, the virtual community content can be stored in a first content library or in a designated storage space within a second content library. This provides diversified content selection methods for the subsequent virtual community content push process. Specifically, target community content can be selected from the first content library to promote a single community, or multiple specific community content items can be selected from the second content library, and community content recommendation results can be generated accordingly to promote multiple community contents at once, thereby meeting different application needs.

[0139] Furthermore, it should be noted that in this embodiment of the disclosure, after storing the virtual community content in the first content library, an overall quality check can be periodically performed on all community content (including virtual community content and real community content) stored in the first content library to filter out low-quality community content in the first content library; similarly, after storing the virtual community content in a designated storage space in the second content library, an overall quality check can also be periodically performed on all community content (including virtual community content and real community content) stored in the designated storage space to filter out low-quality community content in the designated storage space.

[0140] It should also be noted that, in this embodiment of the disclosure, before storing the virtual community content in the first content library, the topic category of the virtual community content can be determined. For example, when the community and content platform is "Tieba," the "bar" corresponding to the virtual community content can be determined. For example, when the topic category of the virtual community content is "celebrity Liu Moumou," the "bar" corresponding to the virtual community content could be "celebrity Liu Moumou bar." Based on this, in this embodiment of the disclosure, "storing the virtual community content in the first content library" can include:

[0141] A comprehensive quality assessment of the content in the virtual community is conducted to obtain the content assessment results;

[0142] If the content detection results indicate that there are no overall quality problems in the virtual community content, the virtual community content is structured to obtain the target community content.

[0143] Store the target community content in the first content library.

[0144] The content detection results can include dead link detection results and quality detection results. Here, dead link detection results are used to characterize whether the virtual community content is invalid and cannot be accessed normally; quality detection results are used to characterize whether the virtual community content has quality problems such as being unplayable, displaying garbled text, or displaying blank pages. Based on this, in this embodiment of the disclosure, it can be determined that the virtual community content does not have overall quality problems when the virtual community content is not invalid and cannot be accessed normally, and does not have quality problems such as being unplayable, displaying garbled text, or displaying blank pages.

[0145] In one example, dead link detection can be performed on virtual community content from multiple dimensions to obtain dead link detection results for the virtual community content. For instance, when the community and content platform is "Tieba," the multiple dead link detection dimensions can include "bar dimension," "account dimension," and "post dimension." That is, when performing dead link detection on virtual community content from multiple dimensions, it can detect whether the "bar" corresponding to the virtual community content has been closed. If it is determined that the "bar" corresponding to the virtual community content has been closed, a dead link detection result is obtained to characterize the virtual community content as invalid community content that cannot be accessed normally. It can also detect whether an available virtual account bound to the virtual community content exists (i.e., whether it has been cancelled or banned), and if so, a dead link detection result is obtained to characterize the virtual community content as invalid community content that cannot be accessed normally. When the available virtual account bound to the content does not exist, a dead link detection result is obtained to indicate that the virtual community is invalid and cannot be accessed normally. It can also detect whether the virtual community content can be viewed by the public without barriers on the community and content platforms (generally, if the virtual community content has not been deleted and is set to "visible to everyone", it can be considered that it can be viewed by the public without barriers on the community and content platforms). When it is determined that the virtual community content cannot be viewed by the public without barriers on the community and content platforms, a dead link detection result is obtained to indicate that the virtual community content is invalid and cannot be accessed normally.

[0146] After performing an overall quality inspection on the virtual community content and obtaining the content inspection results, and confirming that the content inspection results indicate that the virtual community content does not have overall quality problems, the virtual community content can be structured to obtain the target community content. For example, the virtual community content can be structurally adjusted according to a preset structured template to obtain the target community content. The preset structured template can be set according to application requirements, and this embodiment does not impose any limitations on it.

[0147] Through the above methods, in this embodiment of the disclosure, the overall quality of the virtual community content can be detected to obtain content detection results. If the content detection results indicate that the virtual community content does not have overall quality problems, the virtual community content is then structured to obtain target community content, which is then stored in a first content library. This further improves the content quality of the target community content. Specifically, it ensures that the target community content is valid and can be accessed normally, while also ensuring that the target community content does not have quality problems such as being unplayable, displaying garbled text, or being blank.

[0148] Based on the above, in this embodiment of the disclosure, upon receiving the actual search terms (i.e., search terms related to the virtual community content) sent by the target user through the search engine, the target community content can be obtained from the first content library, and the target community content can be pushed to the target user through the search engine (promotion of single community content). This solves the problem of the lack of traffic channels for communities and content platforms, provides users with a richer selection of entry channels for communities and content platforms, simplifies the complexity of user operations, and improves user experience.

[0149] In some alternative implementations, after storing the virtual community content in a designated storage space within a second content library, the community content production method may further include:

[0150] From the specified storage space, determine M1 specific community content items;

[0151] Upon receiving the actual search terms sent by the target user through the search engine, based on M1 specific community content, community content recommendation results are generated and pushed to the target user through the search engine;

[0152] Alternatively, based on M1 specific community content items, generate community content recommendation results, and upon receiving the actual search terms sent by the target user through the search engine, push the community content recommendation results to the target user through the search engine.

[0153] Where M1≥2 and M1 is an integer, for example, M1=3; specific community content can be manually determined by the community and content platform administrators (e.g., project managers) from the specified storage space according to application requirements, or it can be automatically determined from the specified storage space according to other selection strategies (e.g., high-quality selection strategy, high-popularity selection strategy, etc.).

[0154] After determining M1 specific community content items from the designated storage space, community content recommendation results can be generated based on the M1 specific community content items upon receiving the actual search terms sent by the target user through the search engine, and then pushed to the target user through the search engine; alternatively, community content recommendation results can be generated first based on the M1 specific community content items, and then pushed to the target user through the search engine upon receiving the actual search terms sent by the target user through the search engine. In one example, when generating community content recommendation results based on the M1 specific community content items, community content recommendation results can be generated according to a preset recommendation result template, that is, the M1 specific community content items are organized according to the preset recommendation result template to generate community content recommendation results. The preset recommendation result template can be set according to application requirements, and this embodiment does not limit it.

[0155] Through the above methods, in this embodiment of the disclosure, M1 specific community content items can be determined from a designated storage space. Upon receiving the actual search terms sent by the target user through a search engine, community content recommendation results are generated based on the M1 specific community content items. These recommendations are then pushed to the target user via the search engine (i.e., a one-time promotion of multiple community content items). This provides users with a richer selection of entry channels for communities and content platforms, further simplifying user operations and improving user experience.

[0156] Furthermore, in this embodiment of the disclosure, "determining M1 specific community content items from a specified storage space" may include:

[0157] From the specified storage space, identify M2 initial community content entries that have a preset association with the topic category;

[0158] If M2≥M1, select M1 specific community contents from the M2 initial community contents;

[0159] Alternatively, if M2 < M1, select M1-X2 supplementary community content from the specified storage space, so that the M2 initial community content and the M1-X2 supplementary community content are combined as M1 specific community content.

[0160] Where M2≥0 and M2 is an integer; the initial selection of community content can be manually determined by the administrators of the community and content platform (e.g., project managers) from a designated storage space based on application requirements. Therefore, the number of initial selections of community content is uncertain.

[0161] After identifying M2 initial community content items with a preset association with the topic category from the specified storage space, if M2 ≥ M1, M1 specific community content items can be selected from the M2 initial community content items, for example, M1 specific community content items can be randomly selected from the M2 initial community content items; or, if M2 < M1, M1-X2 supplementary community content items can be selected from the specified storage space, so that the M2 initial community content items and the M1-X2 supplementary community content items are combined as the M1 specific community content items. In one example, "selecting M1-X2 supplementary community content items from the specified storage space" can include:

[0162] If there are M3 popular community content items in the specified storage space, filter the M3 popular community content items according to the content filtering rules to obtain M4 first selected community content items;

[0163] If M4≥M1-X2, select M1-X2 supplementary community contents from the M4 first-selected community contents;

[0164] Alternatively, if M4 < M1-X2, according to the fallback selection strategy, select M1-X2-M4 second-selected community contents from the specified storage space, so that M4 first-selected community contents and M1-X2-M4 second-selected community contents are used together as M1-X2 supplementary community contents.

[0165] Where M3≥1 and M3 is an integer, for example, M3=100; 0≤M4≤M3 and M4 is an integer.

[0166] In this embodiment of the disclosure, the content filtering rules may include at least one of the following:

[0167] Filter popular community content that cannot be accessed normally;

[0168] Filter popular community content that contains pornography or other offensive content;

[0169] Filter out popular community content where the number of comments is less than the second preset threshold and the title index of popular community content is less than the third preset threshold;

[0170] Filter popular community content where the time interval between the latest comment's publication time and the current time exceeds a preset time length;

[0171] Filter popular community content with NLP scores below the eighth prediction score threshold.

[0172] The popular community content can be virtual community content with a recent interaction count greater than or equal to a fourth preset threshold. The recent interaction count can be used to represent the number of comments, likes, and reposts of virtual community content on community and content platforms. The second preset threshold, the third preset threshold, and the eighth prediction scoring threshold can be set according to application requirements, and this embodiment does not impose any restrictions on them. The NLP score can be used to represent the relevance score between popular community content and actual search terms, specifically the relevance score between the title of the popular community content and the target search term.

[0173] Furthermore, in this embodiment of the disclosure, when M4≥M1-X2, the quality of the M4 first selected community contents can be ranked based on at least one content evaluation dimension to obtain the content quality ranking result for the M4 first selected community contents. Then, based on the content quality ranking result, M1-X2 supplementary community contents are selected from the M4 first selected community contents.

[0174] In this embodiment of the disclosure, at least one content evaluation dimension may include a second relevance, a fourth text richness, a fourth interaction quantity, and a second timeliness. Specifically, the second relevance can be used to characterize the second relevance score between the first selected community content and the actual search term; the fourth text richness can be used to characterize the richness and diversity of the first selected community content in terms of specific content, content format, and information content; the fourth interaction quantity can be used to characterize the number of comments (i.e., the number of comment information included in the first selected community content), the number of likes, and the number of reposts of the first selected community content; and the second timeliness can be used to characterize the publication time of the first selected community content on the community and content platform.

[0175] Based on this, in this embodiment of the disclosure, when ranking the quality of multiple first-selected community contents based on at least one content evaluation dimension to obtain a content quality ranking result for the multiple first-selected community contents, a second relevance score, a fourth text richness score, a fourth interaction quantity score, and a second timeliness score can be obtained for each first-selected community content (for example, the later the publication time, the higher the second relevance score). The second relevance score, the fourth text richness score, the fourth interaction quantity score, and the second timeliness score are then weighted and fused to obtain a community content quality score for the first-selected community content. Then, according to the community content quality score of each first-selected community content, the multiple first-selected community contents are ranked in quality to obtain a content quality ranking result for the multiple first-selected community contents.

[0176] After ranking the content of the M4 first-selected communities based on at least one content evaluation dimension, and obtaining the content quality ranking results for the M4 first-selected community content, M1-X2 first-selected community content with higher community content quality scores can be selected from the M4 first-selected community content as supplementary community content.

[0177] Furthermore, as described above, in this embodiment of the disclosure, when M4 < M1-X2, a fallback selection strategy can be used to select M1-X2-M4 second-selected community content items from a specified storage space, so that the M4 first-selected community content items and the M1-X2-M4 second-selected community content items are combined as the M1-X2 supplementary community content items. The fallback selection strategy may include:

[0178] Based on specified keywords, select the second choice of community content from the specified storage space;

[0179] Based on the strategy of selecting high-quality community content, select the second-best community content from the specified storage space;

[0180] The virtual community content targeted by the latest comments is selected as the second-choice community content.

[0181] The selection strategies for designated keywords and high-quality community content can be set according to application requirements, and this embodiment does not impose any restrictions on them.

[0182] Furthermore, it should be noted that in this embodiment of the disclosure, if there are no M3 popular community contents in the specified storage space, the fallback selection strategy can be used to select M1-X2 supplementary community contents from the specified storage space.

[0183] Through the above methods, in this embodiment of the disclosure, M2 initial community content items with a preset association with the topic category can be determined from a designated storage space. If M2 ≥ M1, M1 specific community content items are selected from the M2 initial community content items; or, if M2 < M1, M1-X2 supplementary community content items are selected from the designated storage space, so that the M2 initial community content items and the M1-X2 supplementary community content items are combined as the M1 specific community content items. This process does not involve complex data processing procedures, thus improving the speed of determining the M1 specific community content items, thereby improving the efficiency of generating and pushing community content recommendation results.

[0184] The following describes the overall process of a community content production method provided by the embodiments of this disclosure.

[0185] I. Producing virtual community content based on popular search terms

[0186] like Figure 2 As shown:

[0187] First, multiple candidate popular search terms are obtained, and then filtered according to a first search term filtering rule to obtain multiple content search terms. Each of these content search terms is then used as the target search term. The first search term filtering rule can be set according to application requirements, and this embodiment does not impose any limitations on it.

[0188] Subsequently, multiple initial shortlisted topic contents related to the target search term can be obtained from the content provider platform. Based on multiple topic evaluation dimensions, N high-quality topic contents are selected from the multiple initial shortlisted topic contents. Then, based on the N high-quality topic contents, virtual topic contents related to the target search term are constructed. Where N≥1 and N is an integer.

[0189] In one example, a first retrieval tool can be used to obtain multiple coarse search topics related to the target search term from a content providing platform, and obtain the NLP score of each coarse search topic. Then, based on the NLP score of each coarse search topic, multiple first preliminary topic topics can be selected from the multiple coarse search topics. For example, coarse search topics with NLP scores greater than a first preset scoring threshold can be selected as the first preliminary topic topics.

[0190] In another example, after obtaining multiple initial shortlisted topic content related to the target search term from the content provider platform, N high-quality topic content can be selected from these initial shortlisted topic content based on multiple topic evaluation dimensions. Then, based on these N high-quality topic content, virtual topic content related to the target search term can be constructed. For example, using a large model, the N high-quality topic content can be integrated into a single overall topic content, and this overall topic content can be used as the virtual topic content related to the target search term. The multiple topic evaluation dimensions can include the first positive / negative state, the first subjective / objective state, the first relevance, the first text richness, the first number of interactions, and the first timeliness, etc.

[0191] Next, multiple initial review information related to the target search term can be obtained from the content provider platform. Based on multiple evaluation dimensions of the initial reviews, multiple first available reviews can be selected from the multiple initial review information. Then, based on the multiple first available reviews, virtual review information for the virtual topic content can be generated.

[0192] In one example, multiple first coarse search comments related to the target search term can be obtained from the content providing platform, and the NLP score of each first coarse search comment can be obtained. Then, based on the NLP score of each first coarse search comment, multiple first preliminary selection comments can be selected from the multiple first coarse search comments. For example, the first coarse search comments with an NLP score greater than a second preset scoring threshold can be selected as the first preliminary selection comments.

[0193] In another example, after obtaining multiple initial review information related to the target search term from the content provider platform, multiple usable reviews can be selected from the initial review information based on multiple initial review evaluation dimensions. Based on these usable reviews, virtual review information for the virtual topic content can be generated. For example, multiple usable reviews can be combined as virtual review information for the virtual topic content, and a review page including the virtual review information (i.e., including multiple usable reviews) can be generated. The multiple initial review evaluation dimensions can include secondary anti-yellow status, secondary subject-object status, secondary text richness, secondary interaction quantity, etc.

[0194] Finally, the virtual topic content and virtual evaluation information are combined to generate virtual community content.

[0195] In one example, when combining virtual topic content and virtual review information to generate virtual community content, initial community content can be generated based on the virtual topic content and virtual review information. Meta tag optimization operations are then performed on this community content to obtain intermediate community content. This intermediate community content is then processed according to the search engine's communication protocol specifications, and its configuration and URL are optimized to obtain the final virtual community content. Finally, the virtual community content is stored in a first content library (also known as the "Daily library"). The meta tag optimization operation includes at least one of title optimization, description optimization, and keyword optimization. That is, in this embodiment of the disclosure, the intermediate community content can be obtained by performing TDK optimization on the initial community content.

[0196] In another example, after combining virtual topic content and virtual evaluation information to generate virtual community content, the virtual community content can be stored in a designated storage space in a second content library (also known as the "designated content library").

[0197] II. Producing Virtual Community Content Based on Vertical Search Terms

[0198] like Figure 3 As shown:

[0199] First, multiple candidate vertical search terms are obtained (specifically, these can correspond to verticals such as games, animation, education, digital products, automobiles, and healthcare). These candidate vertical search terms are then filtered according to a second search term filtering rule to obtain multiple content search terms. Each of these content search terms is then used as the target search term. The second search term filtering rule can be set according to application requirements, and this embodiment does not impose any limitations on it.

[0200] Subsequently, multiple preliminary topic contents related to the target search term can be obtained from the content provider platform, and multiple effective topic contents can be selected from the multiple preliminary topic contents. Based on the multiple effective topic contents, virtual topic contents related to the target search term can be constructed.

[0201] In one example, a second retrieval tool can be used to obtain multiple preliminary topic content related to the target search term from a content provider platform. After obtaining these preliminary topic content, several valid topic content can be selected from them. Based on these valid topic content, virtual topic content related to the target search term can be constructed. For example, using a large model, multiple valid topic content can be integrated into a single topic content, which can then be used as the virtual topic content related to the target search term. In a specific example, when selecting multiple valid topic content from the multiple preliminary topic content, each preliminary topic content can be parsed to obtain the videos, images, and text included within it. The validity of these videos, images, and text can be verified to eliminate issues such as unplayable content, garbled text, or blank spaces. If these issues are resolved, the preliminary topic content is considered valid topic content.

[0202] Furthermore, it should be noted that in this embodiment of the disclosure, after selecting multiple valid topic contents from multiple second preliminary topic contents, data cleaning operations can be performed on the multiple valid topic contents. Specifically, this can include at least one of black word filtering, watermark removal, and ad removal. Moreover, after performing data cleaning operations on the multiple valid topic contents, deduplication operations can be performed on the multiple valid topic contents to obtain multiple valid topic contents after deduplication. Based on this, in this embodiment of the disclosure, when constructing virtual topic contents related to the target search term based on multiple valid topic contents, virtual topic contents related to the target search term can be constructed based on the multiple valid topic contents after deduplication. For example, using a large model, the multiple valid topic contents after deduplication can be integrated into a whole topic content, and this whole topic content can be used as the virtual topic content related to the target search term.

[0203] It should also be noted that, in this embodiment of the disclosure, when the target search term is a vertical category search term, the virtual topic content can also be provided by a vertical content partner. This vertical content partner may include game partners, novel content providers, cloud storage services, and photo albums, etc. Specifically, after obtaining the target search term and determining that it is a vertical category search term, multiple third-stage preliminary topic content related to the target search term can be obtained from the data platform provided by the vertical content partner. Multiple valid topic content is then selected from these multiple third-stage preliminary topic content to construct virtual topic content related to the target search term based on these multiple valid topic content. The specific functions and examples of "selecting multiple valid topic content from multiple third-stage preliminary topic content to construct virtual topic content related to the target search term based on these multiple valid topic content" can be found in the descriptions of similar steps described above, and will not be repeated here.

[0204] Next, virtual comment information for the virtual topic content can be generated according to the first comment generation template and based on multiple first comment generation requirements. Alternatively, a large model can be used to generate virtual comment information for the virtual topic content based on multiple first comment generation requirements. Virtual comment information for the virtual topic content can also be generated based on a content providing platform. The first comment generation template can be set according to application requirements, and this embodiment does not limit it. Multiple first comment generation requirements may include, but are not limited to, not involving pornography or anti-corruption, being highly relevant to the virtual topic content, and having high text richness.

[0205] Finally, the virtual topic content and virtual evaluation information are combined to generate virtual community content.

[0206] In one example, after combining virtual topic content and virtual review information to generate virtual community content, initial community content can be generated based on the virtual topic content and virtual review information. Meta tag optimization is then performed on this community content to obtain intermediate community content. This intermediate community content is then processed according to the search engine's communication protocol specifications, and its configuration and URL are optimized to obtain the final virtual community content. Finally, the virtual community content is stored in a first content library. The meta tag optimization operation includes at least one of title optimization, description optimization, and keyword optimization. That is, in this embodiment, the intermediate community content can be obtained by performing TDK optimization on the initial community content.

[0207] In another example, after combining virtual topic content and virtual evaluation information to generate virtual community content, the virtual community content can be stored in a designated storage space in a second content library.

[0208] III. Producing Virtual Community Content Based on Uncommon Search Terms

[0209] like Figure 4 As shown:

[0210] First, multiple candidate less popular search terms are obtained, and then filtered according to a third search term filtering rule to obtain multiple content search terms. Each of these content search terms is then used as the target search term. The third search term filtering rule can be set according to application requirements, and this embodiment does not impose any limitations on it.

[0211] After that, multiple content generation requirements can be obtained, and a large model can be used to construct virtual topic content related to the target search term based on these multiple content generation requirements.

[0212] Among these, content generation requirements may include, but are not limited to, not involving pornography or anti-racism, being highly relevant to the target search term, and having high text richness.

[0213] In addition, it should be noted that when using a large model to construct virtual topic content related to the target search term based on multiple content generation requirements, the large model can be invoked outside of peak traffic periods. For example, the large model can be invoked at night from 00:00 to 6:00 to construct virtual topic content related to the target search term based on multiple content generation requirements.

[0214] Next, virtual comment information for the virtual topic content can be generated according to the second comment generation template and based on multiple second comment generation requirements. Alternatively, a large model can be used to generate virtual comment information for the virtual topic content based on multiple second comment generation requirements. The second comment generation template can be set according to application needs, and this embodiment does not limit this. Multiple second comment generation requirements may include, but are not limited to, not involving pornography or anti-corruption, being highly relevant to the virtual topic content, and having high text richness.

[0215] Finally, the virtual topic content and virtual evaluation information are combined to generate virtual community content.

[0216] In one example, after integrating virtual topic content and virtual review information to generate virtual community content, initial community content can be generated based on the virtual topic content and virtual review information. Meta tag optimization is then performed on this community content to obtain intermediate community content. This intermediate community content is then processed according to the search engine's communication protocol specifications, and its configuration and URL are optimized to obtain the final virtual community content. Finally, the virtual community content is stored in a first content library. The meta tag optimization operation includes at least one of title optimization, description optimization, and keyword optimization. That is, in this embodiment, the intermediate community content can be obtained by performing TDK optimization on the initial community content.

[0217] In another example, after combining virtual topic content and virtual evaluation information to generate virtual community content, the virtual community content can be stored in a designated storage space in a second content library.

[0218] Furthermore, please combine Figure 5 In this embodiment of the disclosure, when storing virtual community content in the first content library, it can be done after obtaining the virtual community content:

[0219] A comprehensive quality assessment of the content in the virtual community is conducted to obtain the content assessment results;

[0220] If the content detection results indicate that there are no overall quality problems in the virtual community content, the virtual community content is structured to obtain the target community content.

[0221] Store the target community content in the first content library.

[0222] The content detection results can include dead link detection results and quality detection results. Here, dead link detection results are used to characterize whether the virtual community content is invalid and cannot be accessed normally; quality detection results are used to characterize whether the virtual community content has quality problems such as being unplayable, displaying garbled text, or displaying blank pages. Based on this, in this embodiment of the disclosure, it can be determined that the virtual community content does not have overall quality problems when the virtual community content is not invalid and cannot be accessed normally, and does not have quality problems such as being unplayable, displaying garbled text, or displaying blank pages.

[0223] In one example, dead link detection can be performed on virtual community content from multiple dimensions to obtain dead link detection results for the virtual community content. For instance, when the community and content platform is "Tieba," the multiple dead link detection dimensions can include "bar dimension," "account dimension," and "post dimension." That is, when performing dead link detection on virtual community content from multiple dimensions, it can detect whether the "bar" corresponding to the virtual community content has been closed. If it is determined that the "bar" corresponding to the virtual community content has been closed, a dead link detection result is obtained to characterize the virtual community content as invalid community content that cannot be accessed normally. It can also detect whether an available virtual account bound to the virtual community content exists (i.e., whether it has been cancelled or banned), and if so, a dead link detection result is obtained to characterize the virtual community content as invalid community content that cannot be accessed normally. When the available virtual account bound to the content does not exist, a dead link detection result is obtained to indicate that the virtual community is invalid and cannot be accessed normally. It can also detect whether the virtual community content can be viewed by the public without barriers on the community and content platforms (generally, if the virtual community content has not been deleted and is set to "visible to everyone", it can be considered that it can be viewed by the public without barriers on the community and content platforms). When it is determined that the virtual community content cannot be viewed by the public without barriers on the community and content platforms, a dead link detection result is obtained to indicate that the virtual community content is invalid and cannot be accessed normally.

[0224] After conducting an overall quality inspection of the virtual community content and obtaining the content inspection results, and confirming that the results indicate the virtual community content does not have overall quality issues, the virtual community content can be structured to obtain the target community content. For example, the virtual community content can be structurally adjusted according to a preset structured template to obtain the target community content. The preset structured template can be set according to application requirements. For example, when the community and content platform is "Tieba," the preset structured template can be as follows: Figure 6 As shown, the embodiments disclosed herein are not limited in this respect.

[0225] After structuring the virtual community content to obtain the target community content, a verification tool (e.g., a schema library) can be used to verify the target community content, and the verified target community content can be stored in the first content library.

[0226] Furthermore, please combine Figure 7 In this embodiment of the disclosure, after storing the virtual community content in a designated storage space in the second content library, the community content production method may further include:

[0227] Step S701: From the specified storage space, determine M2 initial community content items that have a preset association with the topic category.

[0228] Where M2≥0 and M2 is an integer; the initial selection of community content can be manually determined by the administrators of the community and content platform (e.g., project managers) from a designated storage space based on application requirements. Therefore, the number of initial selections of community content is uncertain.

[0229] Step S702: Determine whether M2 is greater than or equal to M1.

[0230] Where M1≥2 and M1 is an integer, for example, M1=3.

[0231] If M2≥M1, proceed to step S703; or if M2<M1, proceed to step S704.

[0232] Step S703: Select M1 specific community content from the M2 initially selected community content.

[0233] For example, randomly select M1 specific community content from M2 initial community content.

[0234] Step S704: Select M1-X2 supplementary community content from the specified storage space, so that the M2 initially selected community content and the M1-X2 supplementary community content are combined as M1 specific community content.

[0235] In one example, "selecting M1-X2 pieces of supplementary community content from a specified storage space" could include:

[0236] Step S704-1: Determine whether there are M3 popular community content items in the specified storage space.

[0237] Where M3≥1 and M3 is an integer, for example, M3=100; popular community content can be virtual community content with a recent interaction count greater than or equal to the fourth preset threshold, and the recent interaction count can be used to represent the number of comments, likes and reposts of virtual community content on community and content platforms.

[0238] If there are M3 popular community content items in the specified storage space, proceed to step S704-2; or, if there are no M3 popular community content items in the specified storage space, proceed to step S704-3.

[0239] Step S704-2: Filter the content of M3 popular communities according to the content filtering rules to obtain M4 first-selected community content.

[0240] Where 0 ≤ M4 ≤ M3, and M4 is an integer.

[0241] In this embodiment of the disclosure, the content filtering rules may include at least one of the following:

[0242] Filter popular community content that cannot be accessed normally;

[0243] Filter popular community content that contains pornography or other offensive content;

[0244] Filter out popular community content where the number of comments is less than the second preset threshold and the title index of popular community content is less than the third preset threshold;

[0245] Filter popular community content where the time interval between the latest comment's publication time and the current time exceeds a preset time length;

[0246] Filter popular community content with NLP scores below the eighth prediction score threshold.

[0247] The second preset quantity threshold, the third preset quantity threshold, and the eighth prediction scoring threshold can be set according to application requirements, and this embodiment does not impose any restrictions on them; the NLP score can be used to characterize the relevance scoring result between popular community content and actual search terms, specifically the relevance scoring result between the title of the popular community content and the target search term.

[0248] Step S704-4: Determine whether M4 is greater than or equal to M1-X2.

[0249] If M4≥M1-X2, proceed to step S704-5; or if M4<M1-X2, proceed to step S704-6.

[0250] Step S704-5: Based on at least one content evaluation dimension, sort the quality of the M4 first-selected community content to obtain the content quality ranking result for the M4 first-selected community content. Then, based on the content quality ranking result, select M1-X2 supplementary community content from the M4 first-selected community content.

[0251] In this embodiment of the disclosure, at least one content evaluation dimension may include a second relevance, a fourth text richness, a fourth interaction quantity, and a second timeliness. Specifically, the second relevance can be used to characterize the second relevance score between the first selected community content and the actual search term; the fourth text richness can be used to characterize the richness and diversity of the first selected community content in terms of specific content, content format, and information content; the fourth interaction quantity can be used to characterize the number of comments (i.e., the number of comment information included in the first selected community content), the number of likes, and the number of reposts of the first selected community content; and the second timeliness can be used to characterize the publication time of the first selected community content on the community and content platform.

[0252] Based on this, in this embodiment of the disclosure, when ranking the quality of multiple first-selected community contents based on at least one content evaluation dimension to obtain a content quality ranking result for the multiple first-selected community contents, a second relevance score, a fourth text richness score, a fourth interaction quantity score, and a second timeliness score can be obtained for each first-selected community content (for example, the later the publication time, the higher the second relevance score). The second relevance score, the fourth text richness score, the fourth interaction quantity score, and the second timeliness score are then weighted and fused to obtain a community content quality score for the first-selected community content. Then, according to the community content quality score of each first-selected community content, the multiple first-selected community contents are ranked in quality to obtain a content quality ranking result for the multiple first-selected community contents.

[0253] After ranking the content of the M4 first-selected communities based on at least one content evaluation dimension, and obtaining the content quality ranking results for the M4 first-selected community content, M1-X2 first-selected community content with higher community content quality scores can be selected from the M4 first-selected community content as supplementary community content.

[0254] Step S704-6: According to the fallback selection strategy, select M1-X2-M4 second-selected community contents from the specified storage space, so that the M4 first-selected community contents and the M1-X2-M4 second-selected community contents are used together as the M1-X2 supplementary community contents.

[0255] Among them, the fallback selection strategy may include:

[0256] Based on specified keywords, select the second choice of community content from the specified storage space;

[0257] Based on the strategy of selecting high-quality community content, select the second-best community content from the specified storage space;

[0258] The virtual community content targeted by the latest comments is selected as the second-choice community content.

[0259] The selection strategies for designated keywords and high-quality community content can be set according to application requirements, and this embodiment does not impose any restrictions on them.

[0260] Step S704-3: Directly follow the fallback selection strategy and select M1-X2 supplementary community content items from the specified storage space.

[0261] Step S704-7: Combine the M2 initial community content and the M1-X2 supplementary community content as M1 specific community content.

[0262] Step S705: Generate community content recommendation results based on M1 specific community content items.

[0263] In one example, community content recommendations can be generated based on M1 specific community content items according to a preset recommendation result template. That is, the M1 specific community content items are organized according to the preset recommendation result template to generate community content recommendations. The preset recommendation result template can be set according to application requirements. For example, when the community and content platform is "Tieba," the preset recommendation result template can be as follows: Figure 8 As shown, the embodiments disclosed herein are not limited in this respect.

[0264] For detailed descriptions and examples of the above steps in this embodiment, please refer to the relevant descriptions of the corresponding steps in the embodiment of the community content production method, which will not be repeated here.

[0265] Please see Figure 9 This is a schematic diagram illustrating an application scenario of a community content production method provided in this embodiment of the disclosure.

[0266] The community content production method provided in this disclosure is applied to electronic devices. These electronic devices can be servers, workbenches, mainframe computers, conventional computers, smartphones, personal digital processors, or other similar computing devices.

[0267] Here, electronic devices are used for:

[0268] Retrieve multiple content search terms;

[0269] By using each of the multiple content search terms as the target search term, virtual topic content related to the target search term is constructed;

[0270] Generate virtual comment information for virtual topic content;

[0271] Virtual community content is generated based on virtual topic content and virtual comment information. This virtual community content is used by the search engine to push the virtual community content to the target user when it receives the actual search terms entered by the target user. The actual search terms are search terms related to the virtual community content.

[0272] It should be noted that, in the embodiments disclosed herein, Figure 9 The application scenario diagrams shown are for illustrative purposes only and are not restrictive. Those skilled in the art can use them as a basis for their own interpretation. Figure 9 The examples may be modified in various obvious ways and / or substitutions, and the resulting technical solutions still fall within the scope of the disclosure of the embodiments of this disclosure.

[0273] To better implement the community content production method, this disclosure also provides a community content production apparatus that can be integrated into an electronic device. The electronic device can be a server, workbench, mainframe computer, conventional computer, smartphone, personal digital processor, or other similar computing device. The following will be combined with… Figure 10 The schematic structural block diagram shown illustrates a community content production device 1000 provided in the disclosed embodiment.

[0274] Community content production equipment 1000, including:

[0275] Search term acquisition unit 1001 is used to acquire multiple content search terms;

[0276] The topic content construction unit 1002 is used to construct virtual topic content related to the target search term by using each of the multiple content search terms as the target search term;

[0277] The comment information generation unit 1003 is used to generate virtual comment information for virtual topic content;

[0278] The community content generation unit 1004 is used to generate virtual community content based on virtual topic content and virtual comment information. The virtual community content is used by the search engine to push the virtual community content to the target user when it receives the actual search terms entered by the target user. The actual search terms are search terms related to the virtual community content.

[0279] In some alternative implementations, the target search term falls into one of the categories of popular search terms, vertical search terms, and niche search terms.

[0280] In some optional implementations, the target search term belongs to popular search terms; the topic content construction unit 1002 is used for:

[0281] Obtain multiple initial topic selections related to the target search term from content provider platforms;

[0282] Based on multiple thematic evaluation dimensions, N high-quality thematic content is selected from multiple initial shortlisted thematic content; where N≥1 and N is an integer.

[0283] Based on N high-quality topic content, construct virtual topic content related to the target search term.

[0284] In some optional implementations, the topic content construction unit 1002 is used for:

[0285] Based on the first theme evaluation dimension, multiple initial theme contents are filtered to obtain multiple remaining theme contents;

[0286] Based on the second topic evaluation dimension, the quality of multiple remaining topic contents is ranked to obtain the topic quality ranking results for multiple remaining topic contents.

[0287] Based on the topic quality ranking results, select N high-quality topics from the remaining topics.

[0288] In some optional implementations, the comment information generation unit 1003 is used for:

[0289] Obtain multiple initial review comments related to the target search term from the content provider platform;

[0290] Based on multiple primary review evaluation dimensions, multiple primary usable review information is selected from multiple primary initial review information;

[0291] Based on multiple available comment information, generate virtual comment information for the virtual topic content.

[0292] In some optional implementations, the comment information generation unit 1003 is used for:

[0293] Based on multiple primary review evaluation dimensions, multiple initial review information is filtered to obtain multiple usable primary review information.

[0294] In some optional implementations, the target search term belongs to a vertical category search term; the topic content construction unit 1002 is used for:

[0295] Obtain multiple preliminary topic selections related to the target search term from content provider platforms;

[0296] Select multiple valid topics from a pool of preliminary second-round topics;

[0297] Based on multiple valid topic content, construct virtual topic content related to the target search term.

[0298] In some optional implementations, the comment information generation unit 1003 is used for at least one of the following:

[0299] Based on the first comment generation template and multiple first comment generation requirements, generate virtual comment information for the virtual topic content;

[0300] Using a large model, virtual comment information is generated for virtual topic content based on multiple first comment generation requirements;

[0301] Based on the content delivery platform, virtual comment information is generated for virtual topic content.

[0302] In some optional implementations, the comment information generation unit 1003 is used for:

[0303] Obtain multiple preliminary review comments related to the target search term from the content provider platform;

[0304] Based on multiple secondary review evaluation dimensions, multiple secondary usable review information is selected from multiple secondary preliminary review information;

[0305] Based on multiple available secondary comment information, virtual comment information is generated for the virtual topic content.

[0306] In some optional implementations, the comment information generation unit 1003 is used for:

[0307] Based on multiple secondary review evaluation dimensions, multiple secondary preliminary review information is filtered to obtain multiple secondary usable review information.

[0308] In some optional implementations, the target search term is a less common search term; the topic content construction unit 1002 is used for:

[0309] Obtain multiple content generation requirements;

[0310] By leveraging a large model and based on multiple content generation requirements, virtual topic content related to the target search terms is constructed.

[0311] In some alternative implementations, the comment information generation unit 1003 is used for one of the following:

[0312] Based on the second comment generation template and multiple second comment generation requirements, generate virtual comment information for the virtual topic content;

[0313] By utilizing a large model and based on multiple secondary comment generation requirements, virtual comment information is generated for virtual topic content.

[0314] In some optional implementations, the community content generation unit 1004 is used for:

[0315] Initial community content is generated based on virtual topic content and virtual comment information;

[0316] Meta tag optimization is performed on community content to obtain intermediate community content; the meta tag optimization includes at least one of title optimization, description optimization, and keyword optimization.

[0317] Virtual community content is derived from the content of the intermediate community.

[0318] In some alternative implementations, the community content production device 1000 also includes an account binding unit for:

[0319] Identify multiple candidate virtual accounts;

[0320] Based on multiple account selection dimensions, usable virtual accounts are selected from multiple candidate virtual accounts;

[0321] You can bind available virtual accounts to virtual theme content.

[0322] In some alternative implementations, the community content production apparatus 1000 further includes a community content management unit for at least one of the following:

[0323] The content of the virtual community is stored in the first content library;

[0324] The virtual community content is stored in a designated storage space within the second content library; wherein, the designated storage space is a data storage space corresponding to the topic category of the virtual community content.

[0325] In some alternative implementations, the community content management unit is used for:

[0326] A comprehensive quality assessment of the content in the virtual community is conducted to obtain the content assessment results;

[0327] If the content detection results indicate that there are no overall quality problems in the virtual community content, the virtual community content is structured to obtain the target community content.

[0328] Store the target community content in the first content library.

[0329] In some optional embodiments, the community content production device 1000 further includes a first community content recommendation unit, used for:

[0330] Upon receiving the actual search terms sent by the target user through a search engine, retrieve the target community content from the first content library;

[0331] By using search engines, content from the target community can be pushed to the target users.

[0332] In some optional embodiments, the community content production device 1000 further includes a second community content recommendation unit, used for:

[0333] From the specified storage space, determine M1 specific community content items; where M1≥2 and M1 is an integer;

[0334] Upon receiving the actual search terms sent by the target user through the search engine, community content recommendation results are generated based on M1 specific community content items;

[0335] By using search engines, community content recommendations are pushed to target users.

[0336] In some optional implementations, the second community content recommendation unit is used for:

[0337] From the specified storage space, determine M2 initial community content items that have a preset association with the topic category; where M2≥0 and M2 is an integer.

[0338] If M2≥M1, select M1 specific community contents from the M2 initial community contents;

[0339] Alternatively, if M2 < M1, select M1-X2 supplementary community content from the specified storage space, so that the M2 initial community content and the M1-X2 supplementary community content are combined as M1 specific community content.

[0340] In some optional implementations, the second community content recommendation unit is used for:

[0341] Given that there are M3 popular community content items in the specified storage space, select M4 first-choice community content items from the M3 popular community content items based on multiple content selection dimensions; where M3≥1 and M2 is an integer; 0≤M4≤M3 and M4 is an integer;

[0342] If M4≥M1-X2, select M1-X2 supplementary community contents from the M4 first-selected community contents;

[0343] Alternatively, if M4 < M1-X2, according to the fallback selection strategy, select M1-X2-M4 second-selected community contents from the specified storage space, so that M4 first-selected community contents and M1-X2-M4 second-selected community contents are used together as M1-X2 supplementary community contents.

[0344] In some optional implementations, the second community content recommendation unit is used for:

[0345] Based on content filtering rules, the content of M3 popular communities is filtered to obtain multiple remaining community contents;

[0346] Based on at least one content evaluation dimension, the quality of multiple remaining community content is ranked to obtain the content quality ranking results for multiple remaining community content.

[0347] Based on the content quality ranking results, M4 top-ranked community content items were selected from M3 popular community content items.

[0348] The specific functions and examples of each unit in the community content production device 1000 in this embodiment can be found in the relevant descriptions of the corresponding steps in the community content production method embodiment, and will not be repeated here.

[0349] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0350] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0351] Figure 11 A schematic structural block diagram of an example electronic device 1100 that can be used to implement embodiments of the present disclosure is shown. Electronic device 1100 is intended to represent various forms of digital computers, such as in-vehicle computing devices, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 1100 may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0352] like Figure 11 As shown, the electronic device 1100 includes a computing unit 1101, which can perform various appropriate actions and processes according to a computer program stored in read-only memory (ROM) 1102 or a computer program loaded from storage unit 1108 into random access memory (RAM) 1103. The RAM 1103 may also store various programs and data required for the operation of the electronic device 1100. The computing unit 1101, ROM 1102, and RAM 1103 are interconnected via a bus 1104. An input / output (I / O) interface 1105 is also connected to the bus 1104.

[0353] Multiple components in electronic device 1100 are connected to I / O interface 1105, including: input unit 1106, such as keyboard, mouse, etc.; output unit 1107, such as various types of renderers, speakers, etc.; storage unit 1108, such as disk, optical disk, etc.; and communication unit 1109, such as network card, modem, wireless transceiver, etc. Communication unit 1109 allows electronic device 1100 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0354] The computing unit 1101 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1101 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. The computing unit 1101 performs the various methods and processes described above, such as community content production methods. For example, in some embodiments, the community content production method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 1108. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 1100 via ROM 1102 and / or communication unit 1109. When the computer program is loaded into RAM 1103 and executed by the computing unit 1101, one or more steps of the community content production method described above may be performed. Alternatively, in other embodiments, computing unit 1101 may be configured for community content production methods in any other suitable manner (e.g., by means of firmware).

[0355] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transferring data and instructions to the storage system, the at least one input device, and the at least one output device.

[0356] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data optimization device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0357] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, erasable programmable read-only memory (EPROM) or flash memory, optical fibers, compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0358] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a rendering device for rendering information to the user (e.g., a cathode ray tube (CRT) renderer or a liquid crystal display (LCD); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices are also used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0359] The systems and technologies described herein can be implemented in computing systems that include back-end components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected via digital data communication (e.g., a communication network) of any form or medium. Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0360] A computer system can include client and server components. Clients and servers are generally located far apart and typically interact via a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, a server in a distributed system, or a server incorporating blockchain technology.

[0361] This disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute a community content production method.

[0362] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements a community content production method.

[0363] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein. Furthermore, in this disclosure, relational terms such as "first," "second," and "third" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Additionally, "multiple" in this disclosure can be understood as at least two.

[0364] The foregoing specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for producing community content, comprising: Retrieve multiple content search terms; By using each of the multiple content search terms as a target search term, virtual topic content related to the target search term is constructed; Generate virtual comment information for the virtual topic content; Based on the virtual topic content and the virtual comment information, virtual community content is generated; wherein, the virtual community content is used by the search engine to push the virtual community content to the target user when it receives the actual search terms entered by the target user; the actual search terms are search terms related to the virtual community content.

2. The method according to claim 1, wherein, The target search term belongs to one of the following categories: popular search terms, vertical search terms, and niche search terms.

3. The method according to claim 2, wherein, The target search term belongs to the popular search terms; The construction of virtual topic content related to the target search term includes: From the content providing platform, obtain multiple initial topic contents related to the target search term; Based on multiple thematic evaluation dimensions, N high-quality thematic content is selected from the multiple first-selection thematic content; where N≥1 and N is an integer. Based on the N high-quality topic content, construct virtual topic content related to the target search term.

4. The method according to claim 3, wherein, The selection of N high-quality topics from the multiple initial shortlisted topics, based on multiple topic evaluation dimensions, includes: Based on the first theme evaluation dimension, the multiple first preliminary theme contents are filtered to obtain multiple remaining theme contents; Based on the second topic evaluation dimension, the quality of the multiple remaining topic contents is ranked to obtain the topic quality ranking results for the multiple remaining topic contents. Based on the topic quality ranking results, N high-quality topics are selected from the remaining topics.

5. The method according to claim 3, wherein, The generation of virtual comment information for the virtual topic content includes: Obtain multiple initial review messages related to the target search term from the content providing platform; Based on multiple first-level comment evaluation dimensions, multiple first-level usable comment information is selected from the multiple first-level preliminary comment information; Based on the multiple available first comment information, virtual comment information is generated for the virtual topic content.

6. The method according to claim 5, wherein, The selection of multiple first usable comment information from the multiple first preliminary comment information based on multiple first comment evaluation dimensions includes: Based on the multiple first comment evaluation dimensions, the multiple first preliminary comment information is filtered to obtain the multiple first usable comment information.

7. The method according to claim 2, wherein, The target search term belongs to the vertical category search term; The construction of virtual topic content related to the target search term includes: From the content providing platform, obtain multiple second preliminary topic contents related to the target search term; Select multiple valid topics from the multiple second preliminary topic selection topics; Based on the multiple valid topic contents, construct virtual topic contents related to the target search term.

8. The method according to claim 7, wherein, The generation of virtual comment information for the virtual topic content includes at least one of the following: Based on the first comment generation template and multiple first comment generation requirements, virtual comment information is generated for the virtual topic content. Using a large model, virtual comment information is generated for the virtual topic content based on the multiple first comment generation requirements; Based on the content providing platform, virtual comment information is generated for the virtual topic content.

9. The method according to claim 8, wherein, The generation of virtual comment information for the virtual topic content based on the content providing platform includes: Obtain multiple preliminary review messages related to the target search term from the content providing platform; Based on multiple second review evaluation dimensions, multiple second usable review information is selected from the multiple second preliminary review information; Based on the multiple available second comment information, virtual comment information is generated for the virtual topic content.

10. The method according to claim 9, wherein, The selection of multiple second available comment information from the multiple second preliminary comment information based on multiple second comment evaluation dimensions includes: Based on the multiple second review evaluation dimensions, the multiple second preliminary review information is filtered to obtain the multiple second available review information.

11. The method according to claim 2, wherein, The target search term belongs to the less common search term; The construction of virtual topic content related to the target search term includes: Obtain multiple content generation requirements; Using a large model, virtual topic content related to the target search term is constructed based on the multiple content generation requirements.

12. The method according to claim 11, wherein, The generation of virtual comment information for the virtual topic content includes at least one of the following: Based on the second comment generation template and multiple second comment generation requirements, virtual comment information is generated for the virtual topic content. Using a large model, virtual comment information is generated for the virtual topic content based on the multiple second comment generation requirements.

13. The method according to claim 1, wherein, The process of generating virtual community content based on the virtual topic content and the virtual comment information includes: Based on the virtual topic content and the virtual comment information, initial community content is generated; Meta tag optimization is performed on the community content to obtain intermediate community content; wherein, the meta tag optimization operation includes at least one of title optimization, description optimization and keyword optimization. The virtual community content is obtained based on the content of the intermediate community.

14. The method according to claim 1, further comprising: Identify multiple candidate virtual accounts; Based on multiple account selection dimensions, usable virtual accounts are selected from the multiple candidate virtual accounts; Bind the available virtual account to the virtual theme content.

15. The method according to any one of claims 1 to 14, further comprising at least one of the following: The virtual community content is stored in the first content library; The virtual community content is stored in a designated storage space within the second content library; wherein... The designated storage space is the data storage space corresponding to the topic category of the virtual community content.

16. The method according to claim 15, wherein, The step of storing the virtual community content in the first content library includes: The overall quality of the virtual community content was tested to obtain the content testing results; If the content detection results indicate that the virtual community content does not have overall quality problems, the virtual community content is subjected to structured processing to obtain the target community content. The target community content is stored in the first content library.

17. The method of claim 16, further comprising: Upon receiving the actual search terms sent by the target user through the search engine, the target community content is retrieved from the first content library; The target community content is pushed to the target users through the search engine.

18. The method according to claim 15, wherein, After storing the virtual community content in a designated storage space within a second content library, the method further includes: From the specified storage space, determine M1 specific community content items; where M1≥2 and M1 is an integer; Upon receiving the actual search terms sent by the target user through the search engine, community content recommendation results are generated based on the M1 specific community content items, and the community content recommendation results are pushed to the target user through the search engine; Alternatively, based on the M1 specific community content items, community content recommendation results can be generated, and upon receiving the actual search terms sent by the target user through the search engine, the community content recommendation results can be pushed to the target user through the search engine.

19. The method according to claim 18, wherein, The step of determining M1 specific community content items from the designated storage space includes: From the specified storage space, determine M2 initial community content items that have a preset association with the topic category; wherein, M2≥0 and M2 is an integer; When M2≥M1, select M1 specific community contents from the M2 initially selected community contents; Alternatively, if M2 < M1, select M1-X2 supplementary community content from the specified storage space, and use the M2 initial community content and the M1-X2 supplementary community content together as the M1 specific community content.

20. The method according to claim 19, wherein, The step of selecting M1-X2 pieces of supplementary community content from the designated storage space includes: If there are M3 popular community content items in the specified storage space, the M3 popular community content items are filtered according to the content filtering rules to obtain M4 first selected community content items; wherein, M3≥1 and M3 is an integer; 0≤M4≤M3 and M4 is an integer; When M4≥M1-X2, select M1-X2 supplementary community contents from the M4 first selected community contents; Alternatively, if M4 < M1-X2, according to the fallback selection strategy, select M1-X2-M4 second selected community contents from the specified storage space, so that the M4 first selected community contents and the M1-X2-M4 second selected community contents are used together as the M1-X2 supplementary community contents.

21. A community content production device, comprising: The search term acquisition unit is used to acquire multiple content search terms; The topic content construction unit is used to construct virtual topic content related to the target search term by using each of the plurality of content search terms as the target search term; A comment information generation unit is used to generate virtual comment information for the virtual topic content; The community content generation unit is used to generate virtual community content based on the virtual topic content and the virtual comment information; wherein, the virtual community content is used by the search engine to push the virtual community content to the target user when it receives the actual search terms input by the target user; the actual search terms are search terms related to the virtual community content.

22. An electronic device, comprising: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 20.

23. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1 to 20.

24. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 20.

Images