Content publishing method and apparatus, electronic device, and storage medium

Abstract

Embodiments of the present application disclose a content publishing method and device, electronic equipment and storage medium, which are applied to the technical field of computers. Embodiments of the present application acquire to-be-published content in a first content interaction system, the to-be-published content including main idea information and subject information; according to the main idea information, a reference content set is acquired from a second content interaction system, the reference content set including at least one reference content matched with the main idea information, the reference content including reference subject information; target reference content is determined from the reference content set according to a subject information similarity between the subject information and the reference subject information; and when a target time interval is less than a preset time interval, the to-be-published content is published, the target time interval being a time interval between a to-be-published time of the to-be-published content and a publishing time of the target reference content, thereby improving the quality of published content in a content interaction system.

Description

Technical Field

[0001] This application relates to the field of computer technology, and more specifically to a content publishing method, apparatus, electronic device, and storage medium. Background Technology

[0002] With the rapid development of information technology, browsing daily news and information through content interaction systems has become a routine part of people's lives. However, the content presented on terminals through content interaction systems comes from various sources. For example, some content publishers may publish content from other content interaction systems to the current one. However, when this content published on other content interaction systems is outdated and no longer relevant, publishing it to the current content interaction system will affect the quality of the already published content. Summary of the Invention

[0003] This application provides a content publishing method, apparatus, electronic device, and storage medium, which can improve the quality of published content in a content interaction system.

[0004] This application provides a content publishing method, including:

[0005] Obtain content to be published from the first content interaction system, the content to be published including theme information and main body information;

[0006] Based on the main theme information, a set of reference content is obtained from the second content interaction system. The set of reference content includes at least one reference content that matches the main theme information, and the reference content includes reference subject information.

[0007] Target reference content is determined from the set of reference content based on the similarity between the subject information and the reference subject information;

[0008] When the target time interval is less than the preset time interval, the content to be published is published. The target time interval is the time interval between the time of publication of the content to be published and the publication time of the target reference content.

[0009] Accordingly, embodiments of this application also provide a content publishing device, including:

[0010] The first acquisition unit is used to acquire content to be published in the first content interaction system, the content to be published including theme information and main body information;

[0011] The second acquisition unit is used to acquire a set of reference content from the second content interaction system based on the theme information. The set of reference content includes at least one set of reference content that matches the theme information. The reference content includes reference subject information.

[0012] A determining unit is configured to determine target reference content from the set of reference content based on the similarity between the subject information and the reference subject information;

[0013] The publishing unit is used to publish the content to be published when the target time interval is less than a preset time interval, wherein the target time interval is the time interval between the time to be published of the content to be published and the time to be published of the target reference content.

[0014] In one embodiment, the second acquisition unit includes:

[0015] The retrieval subunit is used to retrieve content in the second content interaction system based on the theme information to obtain an initial reference content set, wherein the initial reference content set includes at least one reference content that matches the theme information.

[0016] A calculation subunit is used to calculate the number of reference contents in the initial reference content set;

[0017] The filtering subunit is used to filter the reference content in the initial reference content set according to the number of reference content, so as to obtain a reference content set.

[0018] In one embodiment, the filtering subunit can be used to:

[0019] When the number of reference contents exceeds a preset threshold, the reference theme information of the reference contents is obtained;

[0020] Calculate the similarity between the reference topic information and the topic information.

[0021] The main idea similarity is compared with the preset main idea similarity to obtain the comparison result;

[0022] The reference content is processed based on the comparison results to obtain a set of reference content.

[0023] In one embodiment, the determining unit includes:

[0024] A calculation subunit is used to calculate the similarity between the subject information and the reference subject information;

[0025] The first determining subunit is used to determine a candidate reference content set from the reference content set based on the subject information similarity, wherein the candidate reference content set includes at least one reference content whose subject information similarity is less than or equal to a preset subject information similarity.

[0026] The second determining subunit is used to determine the target reference content from the set of candidate reference content.

[0027] In one embodiment, the computing subunit can be used to:

[0028] Calculate the text information similarity between the stated text information and the reference text information;

[0029] Calculate the image information similarity between the image information and the reference image information;

[0030] The similarity between the text information and the image information is calculated to obtain the subject information similarity.

[0031] In one embodiment, the computing subunit can be used to:

[0032] Identify the content type of the subject information and the reference content type of the reference subject information;

[0033] Select a similarity calculation method based on the content type and the reference content type;

[0034] Calculate the subject information similarity between the subject information and the reference subject information.

[0035] This application also provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various alternative embodiments described above.

[0036] Accordingly, this application also provides a storage medium storing instructions that, when executed by a processor, implement any of the content publishing methods provided in this application.

[0037] This embodiment first obtains content to be published from a first content interaction system, the content to be published including theme information and subject information; then, based on the theme information, it obtains a set of reference content from a second content interaction system, wherein the set of reference content includes at least one reference content that matches the theme information; next, it determines target reference content from the set of reference content based on the subject information similarity between the subject information and the reference subject information; when the target time interval is less than a preset time interval, it publishes the content to be published, wherein the target time interval is the time interval between the publication time of the content to be published and the publication time of the target reference content; by filtering reference content similar to the content to be published from the second content interaction system, and based on the subject information similarity between the reference subject information and the subject information in the reference content, it can be determined from the second content system whether there is content similar to the content to be published. If there is content similar to the content to be published in the second content system, it can be determined whether the content to be published is time-sensitive based on the time interval between the publication time of the content to be published and the target reference content. By judging the duplication rate and timeliness of content to be published from three dimensions—thematic information, main information, and time—it is possible to effectively prevent the publication of duplicate and outdated content, thereby improving the quality of published content in the content interaction system. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1a This is a schematic diagram illustrating a scenario of the content publishing method provided in an embodiment of this application;

[0040] Figure 1b This is a schematic diagram illustrating another content publishing method provided in an embodiment of this application;

[0041] Figure 2a This is a flowchart illustrating the content publishing method provided in an embodiment of this application;

[0042] Figure 2b This is a schematic diagram of the content to be published provided in the embodiments of this application;

[0043] Figure 2c This is a flowchart illustrating the SimHash algorithm provided in an embodiment of this application;

[0044] Figure 3a This is another flowchart illustrating the content publishing method provided in this application embodiment;

[0045] Figure 3b This is another schematic diagram of content to be published provided in the embodiments of this application;

[0046] Figure 4a This is a schematic diagram of the structure of the content publishing device provided in the embodiments of this application;

[0047] Figure 4b This is another structural schematic diagram of the content publishing device provided in the embodiments of this application;

[0048] Figure 5 This is a schematic diagram of the terminal structure provided in the embodiments of this application. Detailed Implementation

[0049] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0050] This application provides a content publishing method, apparatus, electronic device, and storage medium.

[0051] This application provides a content publishing system, such as... Figure 1a As shown, it includes: the content publishing device provided in the embodiments of this application, such as a content publishing device suitable for a content publishing terminal, a content publishing device suitable for a publishing server, etc.

[0052] The content publishing terminal can be a smartphone, tablet, laptop, personal computer (PC), smart wearable device, etc. The publishing server can be an interconnecting server between multiple heterogeneous systems or a backend server for a content interaction system. It can also be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, and big data and artificial intelligence platforms, etc.

[0053] For example, such as Figure 1aAs shown, the content publishing system may include a first content interaction system, a second content interaction system, and an interconnection server between the first and second content interaction systems, etc., wherein the first and second content interaction systems are heterogeneous systems. Each content interaction system may include a server (such as a system publishing server or management server) and a terminal (such as a content publishing terminal or a content management terminal). The method of this application will be described below using the first content interaction system as an example.

[0054] In one embodiment, such as Figure 1b As shown, when content needs to be published, a terminal in the first content interaction system, such as a content management terminal, obtains the content to be published from the first content interaction system. The content to be published includes theme information and subject information. Based on the theme information, a set of reference content is obtained from the second content interaction system. The set of reference content includes at least one reference content that matches the theme information, and the reference content includes reference subject information. Then, a target reference content is determined from the set of reference content based on the similarity between the subject information and the reference subject information. When the target time interval is less than a preset time interval, the content to be published is published. The target time interval is the time interval between the publication time of the content to be published and the publication time of the target reference content. The content to be published can be published to the content publishing terminal of the first content interaction system, or it can be published to the content publishing terminal of the second content interaction system, which does not include the content to be published.

[0055] The first content interaction system may include computer hardware and software, as well as network systems, for content presentation and interaction. For example, the first content interaction system may include various news and information applications (APPs), various news and information browsing websites, or various news and information mini-programs, etc. Through the first content interaction system, users can not only browse content on the system, but also leave messages to interact with other users, and so on.

[0056] In this context, "APP" can include various software installed on a device to provide users with a rich user experience. For example, an APP can include various types of software installed on a mobile phone, and so on.

[0057] Among them, the content publishing terminals in the first content interaction system may include terminals that have installed various news and information apps of the first content interaction system or terminals that have logged into news and information browsing websites of the first content interaction system, etc.

[0058] The second content interaction system may also include computer hardware and software, as well as network systems, for content presentation and interaction. For example, the second content interaction system may also include various news and information applications and websites, or various news and information mini-programs, etc.

[0059] The content publishing terminal in the second content interaction system may also include terminals that have installed various news and information apps in the second content interaction system or terminals that have logged into news and information browsing websites in the second content interaction system, etc.

[0060] The first content interaction system and the second content interaction system are heterogeneous systems. That is, the first content interaction system and the second content interaction system can be two independent systems that cannot communicate directly with each other and need to access data through interfaces or other means.

[0061] For example, the first content interaction system can receive content from the second content interaction system or send content to the second content interaction system through an interface. For instance, a user can use an interface to repost content published on the second content interaction system to the first content interaction system, or vice versa.

[0062] In one embodiment, the description will be from the perspective of a content publishing device, which can be integrated into the content management terminal of the system, wherein the content management terminal manages the content publishing system.

[0063] like Figure 2a As shown, a content publishing method is provided, which can be executed by a processor in a terminal such as a content management terminal. The specific process is as follows:

[0064] 101. Obtain the content to be published in the first content interaction system. The content to be published includes theme information and main body information.

[0065] The first content interaction system may include computer hardware and software, as well as network systems, for content presentation and interaction. For example, the first content interaction system may include various news and information applications and websites, or various news and information mini-programs, etc. For example, the first content interaction system may include apps like Kandian. Through the first content interaction system, users can not only browse content on the system, but also leave messages to interact with other users, and so on.

[0066] The content to be published can include content that is prepared for publication in the first content interaction system. For example, the content to be published can include current events, interesting stories, trending topics, and so on.

[0067] In one embodiment, such as Figure 2b As shown, the content to be published may include the main information, the main body information, the time to be published, the author of the publication, etc.

[0068] The main idea information can include content that represents the main idea of ​​the content to be published. For example, the main idea information can be the title of the content to be published. Or, the main idea information can be the feature content of the content after feature extraction, and so on.

[0069] For example, such as Figure 2b As shown, the main message of the content to be published could be: "In winter, it's better to eat less meat and fish and more of this. It's delicious and goes well with rice, and it'll be gone as soon as it's on the table."

[0070] The main information can include the content of the main body of the content to be published. For example, the main information could be the body of the content to be published, etc. Figure 2b As shown, the main information can be the body text of the content to be published.

[0071] In one embodiment, the content of the subject information may include multiple types. For example, the subject information may include one or more types of content such as text information and image information, and this application embodiment does not limit this.

[0072] The text information can include content composed of characters. These characters can include various symbols. For example, characters can be scripts used in different countries, relationship symbols defining the relationship between scripts, or even self-created symbols. Examples include Chinese characters, English letters, Arabic numerals, punctuation marks, Japanese characters, and so on.

[0073] Image information can include various content composed of images. For example, image information can include pictures, videos, animations, and so on.

[0074] In one embodiment, the content to be published in the first content interaction system can be obtained from a terminal or server. For example, when a user transfers content from the second content interaction system to the first content interaction system, the content to be published can be obtained from the server corresponding to the second content interaction system. As another example, when a user uploads content from their terminal to the first content interaction system, the content to be published can be obtained from the user's terminal.

[0075] 102. Based on the main idea information, obtain a set of reference content from the second content interaction system. The set of reference content includes at least one set of reference content that matches the main idea information. The reference content includes reference subject information.

[0076] The second content interaction system may include computer hardware and software, as well as network systems, for content presentation and interaction. For example, the second content interaction system may include various news and information applications and websites, or various news and information mini-programs, etc.

[0077] The second content interaction system and the first content interaction system are heterogeneous systems. That is, the first content interaction system and the second content interaction system can be two independent systems. They cannot communicate directly with each other and need to access data through interfaces or other means.

[0078] For example, the first content interaction system is a news and information APP, and the second content interaction system can be a collection of news and information APPs and web pages other than the news and information APP.

[0079] For example, the first content interaction system could be a collection of various news and information apps, web pages, and mini-programs under a certain technology company, while the second content interaction system could be a collection of various news and information apps, web pages, and mini-programs not under that technology company, and so on.

[0080] The reference content set is a whole consisting of at least one reference content. That is, the reference content set may include one reference content or more than one reference content.

[0081] Reference content refers to the objects that the content to be published references. By examining the reference content, it can be determined whether there are similar or identical reference objects for the content to be published in the second content interaction system. Reference content can include content already published in the second content interaction system. For example, reference content could be current events, interesting stories, trending topics, and other content published by the same content publisher in the second content interaction system.

[0082] In one embodiment, the reference content may include reference theme information and reference subject information, publication time and author, etc.

[0083] The reference body information can include the main text of the referenced content, etc. The reference theme information can include content that represents the main idea of ​​the referenced content. For example, the reference theme information can be the title of the referenced content. Or, for example, the reference theme information can be the feature content of the referenced content after feature extraction, etc.

[0084] In one embodiment, content in the second content interaction system can be retrieved based on the theme information to obtain a set of reference content. Specifically, the step "obtaining a set of reference content from the second content interaction system based on the theme information" includes:

[0085] Based on the main idea information, the content in the second content interaction system is retrieved to obtain an initial reference content set, which includes at least one reference content that matches the main idea information.

[0086] Calculate the number of references in the initial reference content set;

[0087] The reference content in the initial reference content set is filtered based on the number of reference content items to obtain the final reference content set.

[0088] The initial reference content set can include a collection of all reference content that matches the main theme information of the content to be published, obtained by retrieving content from the second content interaction system. For example, if 100 reference content items are obtained after retrieving content from the second content interaction system based on the main theme information of the content to be published, then the initial reference content set could be a set containing 100 reference content items, and so on.

[0089] In one embodiment, content in the second content interaction system can be retrieved using various methods. For example, existing search engines can be used to retrieve content in the second content interaction system. Alternatively, a custom-developed search tool can be used to retrieve content in the second content interaction system, and so on.

[0090] After retrieving content from the second content interaction system and obtaining an initial reference content set, the number of reference contents in the initial reference content set can be calculated, and then the reference content set can be determined based on the number of reference contents.

[0091] For example, a pre-defined calculation function can be used to calculate the number of references in the initial reference content set. Another example is using loop statements in a programming language to calculate the number of references in the reference content set, and so on.

[0092] In one embodiment, the number of reference contents in the initial reference content set can be compared with a preset number threshold. When the number of reference contents in the initial reference content set is less than or equal to the preset number threshold, the initial reference content set can be determined as a reference content set. When the number of reference contents is greater than the preset number threshold, the reference contents in the initial reference content set can be filtered to obtain the reference content set.

[0093] For example, a preset quantity threshold can be set to 50. When the number of references in the initial reference content set is less than or equal to 50, the initial reference content set is determined as the reference content set. When the number of references is greater than 50, the references in the initial reference content set are filtered to obtain the reference content set.

[0094] For example, when the preset quantity threshold is set to 50 and the number of reference contents in the initial reference content set is 30, the initial reference content set can be determined as the reference content set since the number of reference contents is less than 50.

[0095] For example, when the preset quantity threshold is set to 50 and the number of references in the initial reference content set is 80, since the number of references is greater than 50, it is necessary to filter the references in the initial reference content set to obtain the reference content set.

[0096] In one embodiment, when the number of reference contents in the initial reference content set is less than or equal to a preset number threshold, the initial reference content set can be determined as the reference content set, and information such as the subject information, publication time, and author of the reference contents in the reference content set can be obtained.

[0097] For example, when the reference content is content from a news browsing webpage, the reference content can be a Hyper Text Markup Language (HTML) document. In this case, information such as the main body of the reference content, the publication time, and the author can be obtained by parsing the HTML tags in the HTML document.

[0098] For example, when the reference content is from a news app, you can request information such as the subject, publication time, and author of the reference content from the app's server.

[0099] In one embodiment, when the number of reference content items exceeds a preset threshold, the reference theme information of the reference content can be obtained. Then, the similarity between the reference theme information and the theme information of the content to be published is calculated. The reference content is then processed based on the theme information similarity to obtain a reference content set. Specifically, the step "filtering the reference content in the initial reference content set according to the number of reference content items to obtain a reference content set" includes:

[0100] When the number of references exceeds a preset threshold, obtain the reference themes of the references.

[0101] Calculate the similarity of main idea information between the reference main idea information and the main idea information;

[0102] The similarity of the main idea is compared with the similarity of the preset main idea information to obtain the comparison results;

[0103] The reference content is processed based on the comparison results to obtain a set of reference content.

[0104] Among them, the similarity of the main information can include the similarity between the reference main information and the main information. That is, through the similarity of the main information, the similarity degree between the main information and the reference main information can be judged.

[0105] Among them, there are various ways to obtain the reference main information of the reference content. For example, when the reference content is web content, the reference content can be an HTML document. At this time, the reference main information of the reference content can be obtained by parsing the HTML tags in the HTML document.

[0106] In one embodiment, in addition to obtaining the reference main information of the reference content by parsing the HTML tags in the HTML document, the reference main body information, release time, release author and other information of the reference content can also be obtained by parsing the HTML tags in the HTML document.

[0107] Another example is that when the reference content is the content on a certain news information APP, the reference main information of the reference content can be obtained by requesting the server of the APP, etc.

[0108] Among them, there are various methods for calculating the similarity of the main information between the reference main information of each reference content and the main information. For example, the Minimum Edit Distance (MED) method, the Locality Sensitive Hashing (SimHash) method, the cosine distance method, the Siamese Network method, the word vector algorithm, etc. can be used.

[0109] For example, when calculating the similarity of the main information using the edit distance, the edit distance can refer to the minimum number of content editing operations required to convert one content to another between two contents. Among them, the content editing operations can include operations such as insertion, deletion, and replacement.

[0110] For example, the reference main information is "a courtyard with a story", and the main information of the content to be published is "a yard with a story". Then the edit distance can be the minimum number of content editing operations required to convert "a courtyard with a story" to "a yard with a story". Among them, converting "a courtyard with a story" to "a yard with a story" requires 2 deletion content editing operations, that is, deleting "one" and "piece", and also requires 1 replacement content editing operation, that is, replacing "courtyard" with "yard". Therefore, the minimum number of content editing operations is 3 times.

[0111] After obtaining the similarity between the reference theme information of each reference content and the theme information of the content to be published, the similarity can be compared with the preset similarity. Then, the reference content is processed according to the comparison results to obtain a set of reference content.

[0112] For example, a preset similarity score for main idea information can be set to 3. When the similarity score is greater than 3, the reference content corresponding to that score can be deleted from the initial reference content set. When the similarity score is less than or equal to 3, the reference content corresponding to that score can be retained in the initial reference content set. The initial reference content set obtained after processing all the reference content can be determined as the reference content set.

[0113] For example, setting the preset main idea similarity to 3, the initial reference content set includes reference content 1, reference content 2, and reference content 3. Reference content 1 has a main idea similarity of 2, reference content 2 has a main idea similarity of 1, and reference content 3 has a main idea similarity of 4. Since the main idea similarity of reference content 1 and reference content 2 is less than the preset main idea similarity, reference content 1 and reference content 2 are retained in the initial reference content set. Since the main idea similarity of reference content 3 is greater than the preset main idea similarity, reference content 3 is deleted from the initial reference content set. After the above processing, the initial reference content set can be determined as a reference content set, which includes reference content 1 and reference content 2, whose main idea similarity is less than the preset main idea similarity.

[0114] For example, an empty set can be created. When the similarity of the main content is less than or equal to a preset similarity, the reference content corresponding to that similarity can be added to the empty set, resulting in the added set. When the similarity is greater than the preset similarity, the reference content corresponding to that similarity is not added to the empty set. After all the reference content has been processed, the added set can be determined as the reference content set, and so on.

[0115] 103. Determine the target reference content from the reference content set based on the similarity between the subject information and the reference subject information.

[0116] Among them, subject information similarity can include the similarity between reference subject information and subject information. That is, subject information similarity can be used to determine the degree of similarity between subject information and reference subject information.

[0117] The target reference content can include the reference content with the highest subject similarity to the content to be published.

[0118] In one embodiment, target reference content can be determined from a set of reference content based on the similarity between the subject information and the reference subject information.

[0119] For example, the similarity between the subject information of the content to be published and the subject information of the reference content can be compared with a preset subject information similarity threshold, and then the target reference content can be determined based on the comparison result.

[0120] For example, when the similarity between the reference subject information and the subject information is greater than the preset subject information similarity, it means that there is no reference content in the reference content set that is the same as or similar to the content to be published. In other words, there is no similar or identical content in the second content interaction system, so the content to be published can be published.

[0121] For example, when the similarity between reference subject information and subject information is less than or equal to a preset subject information similarity, a candidate reference content set can be determined from the reference content set based on the subject information similarity. The candidate reference content set includes at least one reference content with a subject information similarity less than or equal to the preset subject information similarity. Then, the target reference content is determined from the candidate reference content set. Specifically, the step "determine the target reference content from the reference content set based on the subject information similarity between the subject information and the reference subject information" includes:

[0122] Calculate the similarity of subject information between the subject information and the reference subject information;

[0123] A candidate reference content set is determined from the reference content set based on the subject information similarity. The candidate reference content set includes at least one reference content whose subject information similarity is less than or equal to the preset subject information similarity.

[0124] Identify the target reference content from the set of candidate reference content.

[0125] In one embodiment, the subject information may include one or more of text information and image information; similarly, the reference subject information may also include one or more of text information and image information.

[0126] For example, the main subject information might only include text information, while the reference subject information might include both reference text information and reference image information. Or, for another example, the main subject information might include both text information and image information, while the reference subject information might only include reference image information. Because the main subject information and the reference subject information contain different types of content, there are different methods for calculating the similarity of subject information. Specifically, the step "calculating the subject information similarity between the main subject information and the reference subject information" includes:

[0127] Identify the content type of the main information and the reference content type of the reference main information;

[0128] Select the similarity calculation method based on the content type and the reference content type;

[0129] Calculate the similarity between the subject information and the reference subject information.

[0130] Here, the subject information similarity refers to the subject information similarity between the subject information and the reference subject information under this similarity type.

[0131] For example, when the identified subject information only includes text information, while the reference subject information includes reference text information and reference image information, the similarity calculation method can include calculating the text information similarity between the subject information and the reference text information of the reference subject information, and using the text information similarity as the subject information similarity, etc.

[0132] For example, when the identified subject information includes text information and image information, and the reference subject information includes reference text information and reference image information, the similarity calculation method may include calculating the text information similarity between the text information and the reference text information, and the image information similarity between the image information and the reference image information, and combining the text information similarity and the image information similarity as the subject information similarity, etc. Specifically, the step "calculating the subject information similarity between the subject information and the reference subject information" includes:

[0133] Calculate the text information similarity between the stated text information and the reference text information;

[0134] Calculate the image information similarity between the image information and the reference image information;

[0135] The similarity between text information and image information is calculated to obtain the subject information similarity.

[0136] Among them, text information similarity can include the similarity between reference text information and text information. That is, by using text information similarity, we can determine the degree of similarity between text information and reference text information.

[0137] Image information similarity can include the similarity between reference image information and image information. In other words, image information similarity can be used to determine the degree of similarity between image information and reference image information.

[0138] In one instance, there are various methods for calculating the text similarity between text information and reference text information. For example, methods such as the MED method, SimHash method, cosine distance method, Siamese network method, and word vector algorithm can be used to calculate the text similarity between text information and reference text information.

[0139] For example, when using the SimHash method to calculate the text similarity between text information and reference text information, we can first extract the feature information of the text information and the reference feature information of the reference text information. Then, we process the feature information and the reference feature information to obtain processed feature information and processed reference feature information. Finally, we calculate the similarity between the transformed feature information and the transformed reference feature information to obtain the text similarity. Specifically, the step "calculating the text similarity between text information and reference text information" includes:

[0140] Extract text feature information from text information and reference text feature information from reference text information;

[0141] The text feature information and the reference text feature information are processed to obtain the processed feature information and the processed reference feature information.

[0142] The text information similarity is obtained by calculating the similarity between the processed feature information and the processed reference feature information.

[0143] Feature information can include essential, basic information that best reflects the characteristics of things. For example, for textual information, textual feature information can be information that reflects the characteristics of the textual information, such as words, sentences, etc. Similarly, for reference textual information, reference textual feature information can be information that reflects the characteristics of the reference textual information, such as words, sentences, etc.

[0144] In one embodiment, feature information can be extracted by segmenting the text information and the reference text information into words, thereby obtaining effective feature information.

[0145] For example, the text information includes "the author of the user blog structure method algorithm method July", and the segmented text information includes "user (4) blog (5) structure (3) of (1) method (2) algorithm (3) of (1) way (2) of (1) author (5) July (5)". The numbers in parentheses represent the importance of the segmented word in the text information, i.e., its weight. A larger number indicates greater importance, or a smaller number indicates greater importance.

[0146] After obtaining the text feature information and the reference text feature information, these can be processed to obtain processed feature information and processed reference feature information. Processing can include hash mapping, weighting, merging, dimensionality reduction, and other techniques.

[0147] For example, such as Figure 2c As shown, the textual features of the text information include a1, a2, a3, etc. The weight corresponding to a1 is w1, the weight corresponding to a2 is w2, the weight corresponding to a3 is w3, and so on.

[0148] After obtaining the text feature information, a hash mapping can be performed on the text feature information. This hash mapping can include the process of mapping one value to another using a function.

[0149] For example, such as Figure 2c As shown, hashing is the process by which textual features such as a1, a2, and a3 are mapped to numbers. For example, a1 is hashed to "100110", a2 is hashed to "110000", and so on.

[0150] After hashing text feature information into hash mapping values, the hashed text feature information can be weighted. This weighting can include multiplying the hash mapping value by a weight.

[0151] For example, such as Figure 2c As shown, the hash mapping value "100110" corresponding to a1 is multiplied by the weight w1. When encountering a 1, the hash mapping value and the weight are multiplied positively; when encountering a 0, they are multiplied negatively. Therefore, the weighted hash mapping value of a1 is "w1-w1-w1w1w1-w1". Similarly, the weighted hash mapping value of a2 is "w2w2-w2-w2-w2", the weighted hash mapping value of a3 is "w3-w3w3-w3-w3w3", and so on.

[0152] After weighting the text feature information, the weighted results of each text feature can be combined. For example, the weighted results of each text feature can be summed to form a sequence string.

[0153] For example, the sequence values ​​at corresponding positions in “w1-w1-w1w1w1-w1”, “w2w2-w2-w2-w2”, and “w3-w3w3-w3-w3w3” can be summed to obtain a sequence string, namely “b1=w1+w2+w3”, “b2=–w1+w2-w3”, “b3=-w1-w2+w3”, “b4=w1-w2-w3”, “b5=w1-w2-w3”, and “b6=–w1-w2+w3”, thus obtaining “b1b2b3b4b5b6”.

[0154] Finally, the merged values ​​of the text feature information can be reduced in dimensionality to obtain the processed text feature information. This dimensionality reduction process can involve setting the merged values ​​of the text feature information to 1 if they are greater than 0, and 0 otherwise, thus obtaining the processed text feature information.

[0155] For example, for the merged value "b1b2b3b4b5b6", if "b1" is greater than 0, it is set to 1; otherwise, it is set to 0. The same applies to other values.

[0156] After the above processing, we can obtain the processed text feature information and the processed reference text feature information. Then, we calculate the similarity between the processed text feature information and the processed reference text feature information to obtain the text information similarity.

[0157] For example, the Hamming distance between the processed text features and the processed reference text features can be calculated, and the calculated Hamming distance can be used as the text information similarity. For example, if the Hamming distance between the processed text features and the processed reference text features is 2, then the text information similarity is 2.

[0158] In one embodiment, there are various methods for calculating the image information similarity between image information and reference image information. For example, differential hashing, average hashing, perceptual hashing, twin network methods, etc.

[0159] For example, when using the difference hashing method to calculate the image information similarity between image information and reference image information, the image information and reference image information can be preprocessed to obtain processed image information and processed reference image information. Then, image feature information of the processed image information and reference image feature information of the processed reference image information are extracted. Finally, the similarity between the image feature information and the reference image feature information is calculated to obtain the image information similarity. Specifically, the step "calculating the image information similarity between image information and reference image information" includes:

[0160] Preprocessing is performed on the image information and the reference image information to obtain the processed image information and the processed reference image information;

[0161] Extract image feature information from the processed image information and reference image feature information from the processed reference image information;

[0162] The similarity between image feature information and reference image feature information is calculated to obtain the image information similarity.

[0163] Image feature information can include information that reflects the characteristics of an image. These characteristics include color features, texture features, shape features, and spatial relationship features. For example, image feature information can be information about the image's color features, texture features, and so on.

[0164] Preprocessing the image information and reference image information may include reducing the size of the image information and reference image information, and then simplifying the colors of the resized image information.

[0165] For example, when the image information is a picture, the image size can be reduced, and then the image colors can be simplified. Reducing the image size can include reducing the length and width of the image, where the length and width can be in pixels or centimeters.

[0166] A pixel is a small square that makes up an image. Each square has a specific location and a assigned color value. The color value and location of the squares determine how the image is presented.

[0167] For example, when the image size is in pixels, it can be scaled to a size of x*y. Here, x*y means that the image is divided into y rows, with x pixels in each row, meaning the image consists of x*y pixels.

[0168] For example, when the image size is measured in pixels, it can be scaled down to a 9x8 size. Here, 9x8 means the image is divided into 8 rows, with 9 pixels in each row, resulting in an image consisting of 72 pixels.

[0169] After reducing the size of the image information, the colors of the reduced-size image can be simplified. This simplification can include converting a color image to a grayscale image, etc.

[0170] After preprocessing the image information and the reference image information, image feature information of the processed image information and reference image feature information of the processed reference image information can be extracted. Specifically, image feature information can be obtained by comparing the pixel differences between pixels in the image information. Similarly, reference image feature information can be obtained by comparing the pixel differences between pixels in the reference image information.

[0171] For example, when the image information is a picture, the grayscale values of adjacent pixels in each row of the grayscale picture can be compared. Among them, if in each row of pixels, if the value of the latter pixel is greater than the value of the previous pixel, it is recorded as 1, otherwise it is recorded as 0.

[0172] For example, the grayscale value of the current pixel s can be p[s]. Suppose the grayscale value of the next pixel s + 1 of the current pixel can be p[s + 1], and the difference value between the current pixel s and the next pixel s + 1 is k. If p[s] < p[s + 1], then k can be set to 1, otherwise it is 0. By comparing the grayscale values between pixels, the difference value can be determined, thereby obtaining the image feature information.

[0173] For example, in a picture with a size of 9 * 8, the image is divided into 8 rows, where each row has 9 pixels. Among the 9 pixels, a total of 8 different difference values are generated, that is, 64 difference values. Then, the sizes between adjacent pixels are compared respectively. By comparing the sizes between adjacent pixels, the difference value is determined, thereby obtaining the image feature information.

[0174] [[ID=--]]Finally, calculate the similarity between the image feature information and the reference image feature information, thereby obtaining the image information similarity. For example, the Hamming distance between the image feature information and the reference image feature information can be calculated, and the calculated Hamming distance can be the image information similarity. For example, if the calculated Hamming distance is 5, then the image information similarity can be 5.

[0175] In one embodiment, after obtaining the text information similarity and the image information similarity, calculation processing can be performed between the text information similarity and the image information similarity, thereby obtaining the subject information similarity. <--

[0176] Among them, the calculation processing between the text information similarity and the image information similarity can include various operation methods. For example, the text information similarity and the image information similarity can be added to obtain the subject information similarity. Another example is that after adding the text information similarity and the image information similarity and then dividing by a coefficient, the subject information similarity can be obtained, and so on. [[ID=--]]

[0177] For example, if the obtained text information similarity is 2 and the obtained image information similarity is 5, the text information similarity and the image information similarity can be added to obtain the subject information similarity of 7, and so on.

[0178] In one embodiment, after obtaining the subject information similarity, a candidate reference content set can be determined from the reference content set according to the subject information similarity. Among them, there are various ways to determine the candidate reference content set from the reference content set according to the subject information similarity.

[0179] For example, when the similarity of subject information is less than or equal to the preset similarity of subject information, the reference content corresponding to the similarity of subject information can be retained in the reference content set; while when the similarity of subject information is greater than the preset similarity of subject information, the reference content corresponding to the similarity of subject information can be deleted from the reference content; when the similarity of subject information corresponding to the reference content is compared with the preset similarity of subject information, and the reference content is processed according to the comparison results, the resulting reference content set can be a candidate reference content set.

[0180] For example, the reference content set includes reference content 1, reference content 2, reference content 3, and reference content 4. Reference content 1 has a subject information similarity of 3, reference content 2 has a similarity of 5, reference content 3 has a similarity of 7, and reference content 4 has a similarity of 8, while the preset subject information similarity is 6. Since the subject information similarity of reference content 1 and reference content 2 is less than the preset similarity, reference content 1 and reference content 2 will be retained in the reference content set. However, the subject information similarity of reference content 3 and reference content 4 is less than the preset similarity, so reference content 3 and reference content 4 will be deleted from the reference content set. At this point, the reference content set can be determined as a candidate reference content set, where the candidate reference content set includes reference content 1 and reference content 2, whose subject information similarity is less than or equal to the preset subject information similarity.

[0181] For example, an empty set can be created. When the similarity of the subject information is less than or equal to the preset subject information similarity, the reference content corresponding to the subject information similarity can be added to the new set, resulting in the added set; while when the similarity of the subject information is greater than the preset subject information similarity, the reference content corresponding to the subject information similarity is not added to the newly created empty set; when the subject information similarity corresponding to all reference content is compared with the preset subject information similarity, and the reference content that meets the conditions is added to the empty set according to the comparison results, the resulting added set can be a candidate reference content set.

[0182] For example, the reference content set includes reference content 1, reference content 2, reference content 3, and reference content 4. Reference content 1 has a subject information similarity of 3, reference content 2 has a similarity of 5, reference content 3 has a similarity of 7, and reference content 4 has a similarity of 8, while the preset subject information similarity is 6. Since the subject information similarity of reference content 1 and reference content 2 is less than the preset similarity, reference content 1 and reference content 2 will be added to an empty set, resulting in the added set. However, the subject information similarity of reference content 3 and reference content 4 is less than the preset subject information similarity, so reference content 3 and reference content 4 will not be added to the empty set. In this case, the added set can be determined as the candidate reference content set, which includes reference content 1 and reference content 2 whose subject information similarity is less than or equal to the preset subject information similarity.

[0183] After determining the candidate content set, you can identify the reference content that is most similar to or identical to the content to be published within the candidate content set as the target reference content. For example, you can select the reference content with the highest or lowest similarity in main information as the target reference content.

[0184] For example, the candidate reference content set includes reference content 1 and reference content 2, where reference content 1 has a subject information similarity of 3, and reference content 2 has a subject information similarity of 5. When selecting the reference content with the lowest subject information similarity as the target reference content, reference content 1 can be used as the target reference content.

[0185] Among them, when there are multiple reference contents in the candidate reference content set with the same subject information similarity, and the subject information similarity is the largest or smallest subject information similarity, one reference content can be randomly selected as the target reference content.

[0186] For example, the candidate reference content set includes reference content 1, reference content 2, and reference content 3. Reference content 1 has a subject information similarity of 3, reference content 2 has a subject information similarity of 3, and reference content 3 has a subject information similarity of 5. When selecting the reference content with the lowest subject information similarity as the target reference content, either reference content 1 or reference content 2 can be chosen as the target reference content.

[0187] 104. When the target time interval is less than the preset time interval, the content to be published is published. The target time interval is the time interval between the publication time of the content to be published and the publication time of the target reference content.

[0188] After determining the target reference content, the time interval between the release time of the content to be released and the release time of the target reference content can be calculated.

[0189] For example, the publication time of the target reference content can be used as the reference time to calculate the time difference between the publication time of the content to be published and the reference time. The time difference can be the time interval between the publication time and the publication time.

[0190] For example, the publication date of the target reference content could be "February 1, 2020", and the publication date of the content to be published could be "February 29, 2020". Taking the publication date of the target reference content as the reference time, the time difference between the publication date and the reference time is 28 days. Therefore, the time interval between the publication date and the publication date is 28 days.

[0191] For example, a reference time can be selected for the publication time of the target reference content and the publication time of the content to be published. Then, the reference time interval and the time interval of the publication time and the reference time interval can be calculated respectively. Finally, the time difference between the time interval and the reference time interval can be calculated to obtain the time interval between the publication time and the publication time.

[0192] For example, the publication date of the target reference content could be "February 1, 2020," and the publication date of the content to be published could be "February 29, 2020." Alternatively, "January 1, 2020" could be selected as the reference date. In this case, the reference time interval between the publication date and the reference date would be 30 days, and the time interval between the content to be published and the reference date would be 58 days. Then, the time difference between the time interval and the reference time interval would be calculated, resulting in a time interval of 28 days between the content to be published and the publication date.

[0193] After calculating the time interval between the release time of the content to be published and the release time of the target reference content, this time interval can be compared with a preset time interval.

[0194] When the time interval is greater than or equal to the preset time interval, it means that the content to be published is outdated and can be left unpublished.

[0195] When the time interval is less than the preset time interval, it indicates that the content to be published is time-sensitive, and therefore the content can be published.

[0196] For example, the time interval between the release date of the content to be released and the release date of the reference content is 28 days, while the preset time interval is 30 days. Since the time interval is less than the preset time interval, the content to be released is not outdated, so the content to be released is published.

[0197] In one embodiment, the content to be published can be published to a first content interaction system. For example, the content to be published can be published to various apps, web pages, mini-programs, etc., within the first content interaction system, so that terminals connected to the first content interaction system can browse the content to be published, and so on.

[0198] In one embodiment, in addition to publishing the content to be published to the first content interaction system, the content to be published can also be published to the second content interaction system through the interaction between the first and second content interaction systems, and so on. For example, the content to be published can be published to apps, web pages, etc., in the second content interaction system that have not yet published the content to be published.

[0199] The content publishing method described above includes: acquiring content to be published from a first content interaction system, the content to be published including theme information and subject information; then, based on the theme information, acquiring a set of reference content from a second content interaction system, wherein the set of reference content includes at least one reference content that matches the theme information; next, determining target reference content from the set of reference content based on the subject information similarity between the subject information and the reference subject information; publishing the content to be published when the target time interval is less than a preset time interval, wherein the target time interval is the time interval between the publication time of the content to be published and the publication time of the target reference content; by filtering reference content similar to the content to be published from the second content interaction system, and based on the subject information similarity between the reference subject information and the subject information in the reference content, it can be determined from the second content system whether there is content similar to the content to be published. If there is content similar to the content to be published in the second content system, it can be determined whether the content to be published is time-sensitive based on the time interval between the publication time of the content to be published and the target reference content. By judging the duplication rate and timeliness of content to be published from three dimensions—thematic information, main information, and time—it is possible to effectively prevent the publication of duplicate and outdated content, thereby improving the quality of published content in the content interaction system.

[0200] Based on the method described in the above embodiments, the following examples will provide further detailed explanations.

[0201] like Figure 3a As shown, a content publishing method has the following specific process:

[0202] 201. The terminal obtains the content to be published in the first content interaction system. The content to be published includes the main idea and the main body.

[0203] For example, when a user uploads content to the first content interaction system from their own terminal, the system can retrieve the content to be published from the user's terminal.

[0204] The first content interaction system could be a news and information app.

[0205] The content to be published can include content that is prepared for publication in the first content interaction system. For example, the content to be published can include current events, interesting stories, trending topics, and so on.

[0206] In one embodiment, the content to be published may include topic information, subject information, publication time, author, etc.

[0207] For example, such as Figure 3b As shown, the main information of the content to be published is "The Way of a Certain Algorithm!", the publication time is "xx year xx month xx day 06:33", the author is "XX News", and the main information is the text content.

[0208] 202. Based on the main idea information, the terminal obtains a set of reference content from the second content interaction system. The set of reference content includes at least one set of reference content that matches the main idea information. The reference content includes reference subject information.

[0209] Among them, the first content interaction system and the second content interaction system are heterogeneous systems.

[0210] For example, when the first content interaction system is a news and information app, the second content interaction system can be a collection of various news and information apps and web pages other than that news and information app.

[0211] In one embodiment, content in the second content interaction system can be retrieved based on the theme information to obtain an initial reference content set, which includes at least one reference content that matches the theme information. Then, the number of reference content items in the initial reference content set is calculated, and the reference content in the initial reference content set is filtered based on the number of reference content items to obtain a final reference content set.

[0212] Specifically, the number of reference contents in the initial reference content set can be compared with a preset number threshold. When the number of reference contents in the initial reference content set is less than or equal to the preset number threshold, the initial reference content set can be determined as a reference content set. When the number of reference contents is greater than the preset number threshold, the reference contents in the initial reference content set can be filtered to obtain the reference content set.

[0213] For example, based on the main message "The Way of a Certain Algorithm!", one can use an existing search engine to retrieve content from the second content interaction system, thereby obtaining an initial set of reference content.

[0214] Then, it is calculated that there are 100 references in the initial reference content set, while the preset quantity threshold is 50. Since the number of references is greater than 50, it is necessary to filter the references in the initial reference content set to obtain the reference content set.

[0215] For example, one can obtain the reference main idea information of the reference content, then calculate the similarity between the reference main idea information and the main idea information of the content to be published, and process the reference content according to the main idea information similarity to obtain a set of reference content.

[0216] The similarity between the reference topic information and the topic information of each reference content can be calculated using the edit distance method.

[0217] For example, the reference topic information is "The Way of Algorithm X". Changing "The Way of Algorithm X!" to "The Way of Algorithm Y" requires at least two content editing operations. Therefore, the similarity between the reference topic information of the referenced content and the topic information of the content to be published is 2.

[0218] After obtaining the similarity between the reference theme information of each reference content and the theme information of the content to be published, the similarity can be compared with the preset similarity. Then, the reference content is processed according to the comparison results to obtain a set of reference content.

[0219] For example, a preset similarity score for main idea information can be set to 3. When the similarity score is greater than 3, the reference content corresponding to that score can be deleted from the initial reference content set. When the similarity score is less than or equal to 3, the reference content corresponding to that score can be retained in the initial reference content set. The initial reference content set obtained after processing all the reference content can be determined as the reference content set.

[0220] For example, setting the preset main idea similarity to 3, the initial reference content set includes reference content 1, reference content 2, and reference content 3. Reference content 1 has a main idea similarity of 2, reference content 2 has a main idea similarity of 1, and reference content 3 has a main idea similarity of 4. Since the main idea similarity of reference content 1 and reference content 2 is less than the preset main idea similarity, reference content 1 and reference content 2 are retained in the initial reference content set. Since the main idea similarity of reference content 3 is greater than the preset main idea similarity, reference content 3 is deleted from the initial reference content set. After the above processing, the initial reference content set can be determined as a reference content set, which includes reference content 1 and reference content 2, whose main idea similarity is less than the preset main idea similarity.

[0221] 203. Similarity of subject information between terminal computing subject information and reference subject information.

[0222] For example, when the identified subject information includes text information and image information, and the reference subject information includes reference text information and reference image information, the terminal can calculate the text information similarity between the text information and the reference text information and the image information similarity between the image information and the reference image information, and combine the text information similarity and the image information similarity as the subject information similarity.

[0223] The SimHash method can be used to calculate the text similarity between the text information and the reference text information.

[0224] Specifically, taking text information as an example, the SimHash method can include the following steps:

[0225] (1) Word segmentation

[0226] This involves segmenting the text information into words to obtain text feature information.

[0227] For example, such as Figure 3b As shown, the text information includes "the author of the user blog structure algorithm July", and the segmented text information includes "the (1) author (5) July (5) of the user (4) blog (5) structure (3) of (1) method (2) algorithm (3) of (1) way (2)". The numbers in parentheses represent the importance of the segmented word in the text information, i.e., the weight.

[0228] (2) Hash Map

[0229] After obtaining the textual features of the text information, a hash mapping can be performed on these features to obtain a hash mapping value. This hash mapping can involve using a function to map one value to another.

[0230] After obtaining the text feature information "user(4)blog(5)structure(3)of(1)method(2)algorithm(3)of(1)way(2)of(1)author(5)July(5)", we can first perform a hash mapping on the text feature information. For example, hash mapping "user" to "100101" and hash mapping "blog" to "101011", so that the text feature information, which was originally a string, can be converted into a series of numbers.

[0231] (3) Weighted

[0232] After hashing text feature information into hash mapping values, the hashed text feature information can be weighted. This weighting can include multiplying the hash mapping value by a weight.

[0233] For example, the weighted hash mapping value of "user" is "4-4-44-44", the weighted hash mapping value of "blog" is "5-55-555", and so on.

[0234] (4) Accumulation

[0235] After weighting the text feature information, the weighted results of each text feature can be combined. For example, the weighted results of each text feature can be summed to form a sequence string.

[0236] For example, the weighted value of "user" "4-4-44-44" and the weighted value of "blog" "5-55-555" can be added together, i.e. "4+5-4+-5-4+54+-5-4+54+5", to get the combined value "9-91-119".

[0237] (5) Dimensionality reduction

[0238] Finally, the merged values ​​of the text feature information can be reduced in dimensionality to obtain the processed text feature information. This dimensionality reduction process can involve setting the merged values ​​of the text feature information to 1 if they are greater than 0, and 0 otherwise, thus obtaining the processed text feature information.

[0239] For example, the processed text feature information obtained after dimensionality reduction of the merged value "9-91-119" is "101011".

[0240] One method is to use differential hashing to calculate the similarity between image information and reference image information. Specifically, the application steps of differential hashing are as follows:

[0241] (1) Image preprocessing

[0242] Image preprocessing includes two steps: resizing and color simplification.

[0243] For example, when the image information is a picture, the image size can be reduced, and then the colors can be simplified. For instance, the image can be scaled to 9x8 pixels, and then a color image can be converted to grayscale.

[0244] (2) Extracting feature information from the image

[0245] For example, when the image information is an image, the grayscale values ​​of adjacent pixels in each row of the grayscale image can be compared. Specifically, if the value of a subsequent pixel in a row is greater than the value of the preceding pixel, it is recorded as 1; otherwise, it is recorded as 0.

[0246] For example, the grayscale value of the current pixel s can be p[s]. Suppose the grayscale value of the next pixel s + 1 of the current pixel can be p[s + 1], and the difference value between the current pixel s and the next pixel s + 1 is k. If p[s] < p[s + 1], then k can be set to 1, otherwise 0. By comparing the grayscale values between pixels, the difference value can be determined, thereby obtaining the image feature information.

[0247] 204. The terminal determines the target reference content from the reference content set according to the similarity of the subject information.

[0248] For example, when the similarity of the reference subject information and the subject information is less than or equal to the preset subject information similarity, the candidate reference content set can be determined from the reference content set. The candidate reference content set includes at least one reference content whose similarity of the subject information is less than or equal to the preset subject information similarity, and then the target reference content is determined from the candidate reference content set.

[0249] For example, when the similarity of the subject information is less than or equal to the preset subject information similarity, the reference content corresponding to the similarity of the subject information can be retained in the reference content set; when the similarity of the subject information is greater than the preset subject information similarity, the reference content corresponding to the similarity of the subject information can be deleted from the reference content; when the similarity of the subject information corresponding to the reference content is compared with the preset subject information similarity, and the reference content set obtained after processing the reference content according to the comparison result can be the candidate reference content set.

[0250] For example, the reference content set includes reference content 1, reference content 2, reference content 3, and reference content 4. Among them, the similarity of the subject information corresponding to reference content 1 is 3, the similarity of the subject information corresponding to reference content 2 is 5, the similarity of the subject information corresponding to reference content 3 is 7, the similarity of the subject information corresponding to reference content 4 is 8, and the preset subject information similarity is 6. Since the similarity of the subject information corresponding to reference content 1 and reference content 2 is less than the preset content similarity, reference content 1 and reference content 2 will be retained in the reference content set. And the similarity of the subject information corresponding to reference content 3 and reference content 4 is less than the preset subject information similarity, so reference content 3 and reference content 4 will be deleted from the reference content set. At this time, the reference content set can be determined as the candidate reference content set, where the candidate reference content set includes reference content 1 and reference content 2 whose similarity of the subject information is less than or equal to the preset subject information similarity.

[0251] After determining the candidate content set, you can identify the reference content that is most similar to or identical to the content to be published within the candidate content set as the target reference content. For example, you can select the reference content with the highest or lowest similarity in main information as the target reference content.

[0252] For example, the candidate reference content set includes reference content 1 and reference content 2, where reference content 1 has a subject information similarity of 3, and reference content 2 has a subject information similarity of 5. When selecting the reference content with the lowest subject information similarity as the target reference content, reference content 1 can be used as the target reference content.

[0253] 205. The terminal calculates the time interval between the time the content to be published is to be published and the time the target reference content is to be published.

[0254] After determining the target reference content, the time interval between the release time of the content to be released and the release time of the target reference content can be calculated.

[0255] For example, the publication date of the target reference content could be "February 1, 2020", and the publication date of the content to be published could be "February 29, 2020". Taking the publication date of the target reference content as the reference time, the time difference between the publication date and the reference time is 28 days. Therefore, the time interval between the publication date and the publication date is 28 days.

[0256] 206. When the time interval is less than the preset time interval, the terminal publishes the content to be published.

[0257] For example, the time interval between the release date of the content to be released and the release date of the reference content is 28 days, while the preset time interval is 30 days. Since the time interval is less than the preset time interval, the content to be released is not outdated, so the content to be released is published.

[0258] In one embodiment, the content to be published can be published to a first content interaction system. For example, the content to be published can be published to various apps, web pages, mini-programs, etc., within the first content interaction system, so that terminals connected to the first content interaction system can browse the content to be published, and so on.

[0259] The content publishing method described above involves the terminal acquiring content to be published from a first content interaction system. This content includes main body information and theme information. Then, based on the theme information, the terminal obtains a set of reference content from a second content interaction system. This set includes at least one reference content matching the theme information, and each reference content includes reference main body information. The terminal then calculates the similarity between the main body information and the reference main body information, and determines a target reference content from the set of reference content based on this similarity. Finally, the terminal calculates the time interval between the publication time of the content to be published and the publication time of the target reference content. If this time interval is less than a preset time interval, the terminal publishes the content to be published. This content publishing method, provided by the embodiments of this application, can combat outdated content, thereby improving the quality of content published in the content interaction system.

[0260] To better implement the content publishing method provided in this application, one embodiment also provides a content publishing device, which can be integrated into a terminal or server. The meanings of the terms are the same as in the web page content processing method described above, and specific implementation details can be found in the description of the method embodiments.

[0261] In one embodiment, a content publishing device is provided, which can be specifically integrated into a terminal such as a content publishing terminal. Figure 4a As shown, the content publishing device includes: a first acquisition unit 301, a second acquisition unit 302, a determination unit 303, and a publishing unit 304, as detailed below:

[0262] The first acquisition unit 301 is used to acquire content to be published in the first content interaction system, the content to be published including theme information and subject information;

[0263] The second acquisition unit 302 is used to acquire a set of reference content from the second content interaction system according to the theme information. The set of reference content includes at least one set of reference content that matches the theme information. The reference content includes reference subject information.

[0264] The determining unit 303 is used to determine target reference content from the reference content set based on the similarity between the subject information and the reference subject information;

[0265] The publishing unit 304 is used to publish the content to be published when the target time interval is less than the preset time interval, wherein the target time interval is the time interval between the time to be published of the content to be published and the time to be published of the target reference content.

[0266] In one embodiment, such as Figure 4b As shown, the second acquisition unit 302 includes:

[0267] The retrieval subunit 3021 is used to retrieve the content in the second content interaction system according to the theme information to obtain an initial reference content set, wherein the initial reference content set includes at least one reference content that matches the theme information.

[0268] The calculation subunit 3022 is used to calculate the number of reference contents in the initial reference content set;

[0269] The filtering subunit 3023 is used to filter the reference content in the initial reference content set according to the number of reference content, so as to obtain a reference content set.

[0270] In one embodiment, the filtering subunit 3023 can be used to:

[0271] When the number of reference contents exceeds a preset threshold, the reference theme information of the reference contents is obtained;

[0272] Calculate the similarity between the reference topic information and the topic information.

[0273] The main idea similarity is compared with the preset main idea similarity to obtain the comparison result;

[0274] The reference content is processed based on the comparison results to obtain a set of reference content.

[0275] In one embodiment, such as Figure 4b As shown, the determining unit 303 includes:

[0276] The calculation subunit 3031 is used to calculate the similarity between the subject information and the reference subject information;

[0277] The first determining subunit 3032 is used to determine a candidate reference content set from the reference content set based on the subject information similarity, wherein the candidate reference content set includes at least one reference content whose subject information similarity is less than or equal to a preset subject information similarity.

[0278] The second determining subunit 3033 is used to determine the target reference content from the candidate reference content set.

[0279] In one embodiment, the computing subunit 3031 can be used for:

[0280] Identify the content type of the subject information and the reference content type of the reference subject information;

[0281] Select a similarity calculation method based on the content type and the reference content type;

[0282] Calculate the subject information similarity between the subject information and the reference subject information.

[0283] In practice, each of the above units can be implemented as an independent entity or can be arbitrarily combined to be implemented as the same or several entities. For the specific implementation of each of the above units, please refer to the previous method embodiments, which will not be repeated here.

[0284] The aforementioned content publishing device can effectively combat duplicate and outdated content, thereby improving the quality of published content in the content interaction system.

[0285] This application also provides a computer device, which may include a terminal or a server. For example, the computer device can serve as a content publishing terminal, such as a mobile phone, tablet computer, etc.; or it can serve as a server, such as a content publishing server. Figure 5 As shown, it illustrates the structural diagram of the terminal involved in the embodiments of this application, specifically:

[0286] The computer device may include components such as a processor 401 with one or more processing cores, a memory 402 with one or more computer-readable storage media, a power supply 403, and an input unit 404. Those skilled in the art will understand that... Figure 5 The computer device structure shown does not constitute a limitation on the computer device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein:

[0287] The processor 401 is the control center of the computer device. It connects various parts of the computer device via various interfaces and lines, and performs various functions and processes data by running or executing software programs and / or modules stored in the memory 402, and by calling data stored in the memory 402, thereby providing overall monitoring of the computer device. Optionally, the processor 401 may include one or more processing cores; preferably, the processor 401 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 401.

[0288] The memory 402 can be used to store software programs and modules. The processor 401 executes various functional applications and data processing by running the software programs and modules stored in the memory 402. The memory 402 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 402 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 402 may also include a memory controller to provide the processor 401 with access to the memory 402.

[0289] The computer device also includes a power supply 403 that supplies power to the various components. Preferably, the power supply 403 can be logically connected to the processor 401 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 403 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0290] The computer device may also include an input unit 404, which can be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0291] Although not shown, the computer device may also include a display unit, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 401 in the computer device loads the executable files corresponding to the processes of one or more applications into the memory 402 according to the following instructions, and the processor 401 runs the applications stored in the memory 402 to realize various functions, as follows:

[0292] Obtain content to be published from the first content interaction system, the content to be published including theme information and main body information;

[0293] Based on the main theme information, a set of reference content is obtained from the second content interaction system. The set of reference content includes at least one reference content that matches the main theme information, and the reference content includes reference subject information.

[0294] Calculate the subject information similarity between the subject information and the reference subject information;

[0295] Target reference content is determined from the reference content set based on the similarity of the subject information;

[0296] Calculate the time interval between the publication time of the content to be published and the publication time of the target reference content;

[0297] When the time interval is less than the preset time interval, the content to be published is published.

[0298] As can be seen from the above, the terminal provided in this application embodiment can improve the quality of content published in the content interaction system.

[0299] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0300] Therefore, embodiments of this application provide a storage medium storing multiple instructions that can be loaded by a processor to execute steps in any of the content publishing methods provided in embodiments of this application.

[0301] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0302] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0303] Since the instructions stored in the storage medium can execute the steps of any of the content publishing methods provided in the embodiments of this application, the beneficial effects that any of the content publishing methods provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0304] Furthermore, embodiments of this application provide a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various alternative embodiments described above.

[0305] For example, the computer program can perform the following steps:

[0306] Obtain content to be published from the first content interaction system, the content to be published including theme information and main body information;

[0307] Based on the main theme information, a set of reference content is obtained from the second content interaction system. The set of reference content includes at least one reference content that matches the main theme information, and the reference content includes reference subject information.

[0308] Target reference content is determined from the set of reference content based on the similarity between the subject information and the reference subject information;

[0309] When the target time interval is less than the preset time interval, the content to be published is published. The target time interval is the time interval between the time when the content to be published is to be published and the time when the target reference content is published.

[0310] The quality of content published in a content interaction system can be improved through the aforementioned computer program products or programs.

[0311] The above provides a detailed description of a content publishing method, apparatus, electronic device, and storage medium provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A content publishing method, characterized in that, include: Obtain content to be published from the first content interaction system, the content to be published including theme information and main body information; Based on the main theme information, a set of reference content is obtained from the second content interaction system. The set of reference content includes at least one reference content that matches the main theme information. The reference content includes reference subject information. The first content interaction system and the second content interaction system are heterogeneous systems. When the reference content type is that the main information only includes text information, and the reference main information includes reference text information and reference image information, the text information similarity between the text information of the main information and the reference text information of the reference main information is calculated, and the text information similarity is used as the main information similarity. When the reference content type is that the main information includes text information and image information, and the reference main information includes reference text information and reference image information, the text information similarity between the text information and the reference text information, and the image information similarity between the image information and the reference image information are calculated, and the text information similarity and the image information similarity are combined as the main information similarity; Target reference content is determined from the set of reference content based on the similarity between the subject information and the reference subject information; The publication time of the target reference content is obtained by parsing the Hypertext Markup Language tags in the target reference content; When the target time interval is not less than the preset time interval, the content to be published is determined to be outdated and will not be published. When the target time interval is less than the preset time interval, the content to be published is published. The target time interval is the time interval between the time of publication of the content to be published and the publication time of the target reference content.

2. The content publishing method as described in claim 1, characterized in that, Determining target reference content from the reference content set based on the subject information similarity between the subject information and the reference subject information includes: Calculate the subject information similarity between the subject information and the reference subject information; Based on the subject information similarity, a candidate reference content set is determined from the reference content set, wherein the candidate reference content set includes at least one reference content whose subject information similarity is less than or equal to a preset subject information similarity. The target reference content is determined from the set of candidate reference content.

3. The content publishing method as described in claim 2, characterized in that, Calculating the subject information similarity between the subject information and the reference subject information includes: Identify the content type of the subject information and the reference content type of the reference subject information; Select a similarity calculation method based on the content type and the reference content type; Calculate the subject information similarity between the subject information and the reference subject information.

4. The content publishing method as described in claim 3, characterized in that, The subject information includes text information and image information, and the reference subject information includes reference text information and reference image information; Calculating the subject information similarity between the subject information and the reference subject information includes: Calculate the text information similarity between the stated text information and the reference text information; Calculate the image information similarity between the image information and the reference image information; The similarity between the text information and the image information is calculated to obtain the subject information similarity.

5. The content publishing method as described in claim 4, characterized in that, Calculating the image information similarity between the image information and the reference image information includes: The image information and the reference image information are preprocessed to obtain processed image information and processed reference image information; Extract the image feature information of the processed image information and the reference image feature information of the processed reference image information; The similarity between the image feature information and the reference image feature information is calculated to obtain the image information similarity.

6. The content publishing method as described in claim 1, characterized in that, Based on the aforementioned thematic information, a set of reference content is obtained from the second content interaction system, including: Based on the main theme information, the content in the second content interaction system is retrieved to obtain an initial reference content set, which includes at least one reference content that matches the main theme information. Calculate the number of reference contents in the initial reference content set; The reference content in the initial reference content set is filtered based on the number of reference content items to obtain a reference content set.

7. The content publishing method as described in claim 6, characterized in that, The reference content in the initial reference content set is filtered based on the number of reference content items to obtain a reference content set, including: When the number of reference contents exceeds a preset threshold, the reference theme information of the reference contents is obtained; Calculate the similarity between the reference topic information and the topic information; The main idea similarity is compared with the preset main idea similarity to obtain the comparison result; The reference content is processed based on the comparison results to obtain a set of reference content.

8. A content publishing device, characterized in that, include: The first acquisition unit is used to acquire content to be published in the first content interaction system, the content to be published including theme information and main body information; The second acquisition unit is used to acquire a set of reference content from the second content interaction system according to the theme information. The set of reference content includes at least one reference content that matches the theme information. The reference content includes reference subject information. The first content interaction system and the second content interaction system are heterogeneous systems. When the reference content type is that the main information only includes text information, and the reference main information includes reference text information and reference image information, the text information similarity between the text information of the main information and the reference text information of the reference main information is calculated, and the text information similarity is used as the main information similarity. When the reference content type is that the main information includes text information and image information, and the reference main information includes reference text information and reference image information, the text information similarity between the text information and the reference text information, and the image information similarity between the image information and the reference image information are calculated, and the text information similarity and the image information similarity are combined as the main information similarity; A determining unit is configured to determine target reference content from the set of reference content based on the similarity between the subject information and the reference subject information; Target reference content is determined from the set of reference content based on the similarity between the subject information and the reference subject information; The publication time of the target reference content is obtained by parsing the Hypertext Markup Language tags in the target reference content; When the target time interval is not less than the preset time interval, the content to be published is determined to be outdated and will not be published. The publishing unit is used to publish the content to be published when the target time interval is less than a preset time interval, wherein the target time interval is the time interval between the time to be published of the content to be published and the time to be published of the target reference content.

9. An electronic device, characterized in that, It includes a memory and a processor; the memory stores an application program, and the processor is used to run the application program within the memory to perform the operations in the content publishing method according to any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a plurality of instructions adapted for loading by a processor to execute the steps of the content publishing method according to any one of claims 1 to 7.

11. A computer program product comprising computer instructions stored in a computer-readable storage medium; a processor of an electronic device reading the computer instructions from the computer-readable storage medium, the processor executing the computer instructions to perform the steps of the content publishing method according to any one of claims 1 to 7.

Images