Search optimization method and device, electronic equipment, storage medium and program product
Through semantic analysis and hot topic lexicon matching, and combining with the question-and-answer model, the search tags are automatically updated, which solves the problems of low efficiency and relying on manual operations in the existing technology, and achieves efficient and personalized search engine optimization effects.
Patent Information
- Application Number
- CN202410992102.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-26
AI Technical Summary
The existing search engine optimization technology is inefficient and relies on manual operations, making it difficult to cope with large-scale website needs and cannot achieve full automation and personalized optimization effects.
By extracting text information of the content to be optimized for semantic analysis, semantic keywords and characteristics are determined, target keywords are determined based on hot keywords and semantic keywords in the hot vocabulary, and optimization information is obtained using the question-and-answer model, and search tags are automatically updated.
It significantly improves the search engine visibility and user relevance of content, reduces manual operations, and improves optimization efficiency.
Smart Images

Figure CN119025619A_ABST
Abstract
Description
Technical Field
[0001] Embodiments of the present invention relate to the field of computer technology, and in particular to a search optimization method, device, electronic device, storage medium, and program product. Background Art
[0002] In the modern Internet era, SEO is one of the key factors for websites to gain traffic and increase user engagement. SEO aims to improve the ranking of websites in search engine results pages, by optimizing the website content and structure, thereby increasing the chances of being discovered and clicked by users.
[0003] Traditional manual SEO optimization is labor-intensive and prone to omissions. It relies on personal experience and expertise, and its optimization efficiency is difficult to meet the needs of large-scale websites. SEO tools can provide more general optimization suggestions, but in scenarios that require in-depth analysis and personalized promotion, the optimization effect needs to be improved.
[0004] Therefore, there is an urgent need for a search optimization method that can improve optimization efficiency and optimization effect. Summary of the invention
[0005] In view of this, an embodiment of the present invention provides a search optimization method. One or more embodiments of the present invention also relate to a search optimization device, a computing device, a computer-readable storage medium and a computer program product to solve the technical defects existing in the prior art.
[0006] According to a first aspect of an embodiment of the present invention, a search optimization method is provided, comprising:
[0007] Extract the text information of the content to be optimized, and perform semantic analysis on the text information to determine the corresponding semantic keywords and semantic features;
[0008] Determine the target keyword based on the hot keywords and semantic keywords in the hot word library, wherein the hot word library is maintained based on the search popularity of the search keyword;
[0009] Obtain optimization information based on target keywords and semantic features, and update the search tags of the content to be optimized based on the optimization information.
[0010] According to a second aspect of an embodiment of the present invention, there is provided a search optimization device, comprising:
[0011] An extraction module is configured to extract text information of the content to be optimized, and perform semantic analysis on the text information to determine corresponding semantic keywords and semantic features;
[0012] A determination module is configured to determine a target keyword based on hot keywords and semantic keywords in a hot word library, wherein the hot word library is maintained based on the search heat of the search keyword;
[0013] The optimization module is configured to obtain optimization information based on target keywords and semantic features, and update the search tags of the content to be optimized based on the optimization information.
[0014] According to a third aspect of an embodiment of the present invention, there is provided a computing device, including:
[0015] Memory and processor;
[0016] The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, the steps of the above-mentioned search optimization method are implemented.
[0017] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, which stores a computer program / instruction, and the computer program / instruction implements the steps of the above-mentioned search optimization method when executed by a processor.
[0018] According to a fifth aspect of an embodiment of the present invention, there is provided a computer program product, comprising a computer program / instruction, which implements the steps of the above-mentioned search optimization method when executed by a processor.
[0019] An embodiment of the present invention realizes the extraction of text information of the content to be optimized, and performs semantic analysis on the text information to determine the corresponding semantic keywords and semantic features; determines the target keyword based on the hot keywords and semantic keywords in the hot word library, wherein the hot word library is maintained based on the search popularity of the search keyword; obtains optimization information based on the target keyword and the semantic features, and updates the search tag of the content to be optimized based on the optimization information. The semantic core of the content to be optimized is extracted through semantic analysis, high-heat words in the hot word library are accurately matched, and the search tags are dynamically adjusted, which significantly enhances the search engine visibility and user relevance of the content. At the same time, manual operations are reduced through automated processing, greatly improving the optimization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 is an interactive schematic diagram of a search optimization system provided by an embodiment of the present invention;
[0021] Figure 2 is a flow chart of a search optimization method provided by an embodiment of the present invention;
[0022] Figure 3 It is a process flow chart of a method for optimizing website article search provided by one embodiment of the present invention;
[0023] Figure 4 It is a structural schematic diagram of a search optimization device provided by an embodiment of the present invention;
[0024] Figure 5 It is a structural block diagram of a computing device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0025] Many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present invention, so the present invention is not limited to the specific implementation disclosed below.
[0026] The terms used in one or more embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of the present invention. The singular forms of "a", "said" and "the" used in one or more embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of the present invention refers to and includes any or all possible combinations of one or more associated listed items.
[0027] It should be understood that, although the terms first, second, etc. may be used to describe various information in one or more embodiments of the present invention, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of the present invention, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".
[0028] In addition, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of the present invention are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.
[0029] First, the terms involved in one or more embodiments of the present invention are explained.
[0030] SEO: Search Engine Optimization. It is a technique and process to improve the ranking of a website in search engines by optimizing its content and structure. Its purpose is to make the website more visible in the search results page and attract more visitors and traffic. SEO covers many aspects, including keyword optimization, content optimization, technical optimization (such as website speed and structure), user experience optimization, etc. These methods can increase the visibility of the website and the natural traffic of search engines, thereby enhancing the commercial value and competitiveness of the website.
[0031] CMS: Content Management System. A software application or system used to create, manage and publish digital content. It allows users to manage website content through a simplified interface, including documents, pictures, videos, audio and other forms of data. CMS usually provides functions such as user permission management, version control, workflow management, etc., so that multiple users can work together and manage the content updates of the website. CMS can also expand its functions through plug-ins or modules, such as adding search engine optimization (SEO) plug-ins, security plug-ins, multi-language support, etc., to meet different user needs and business requirements. Common CMS include WordPress, Jooml a, Drupal, etc.
[0032] TDK: refers to the title, description and keywords of a web page. These elements play an important role in SEO. Title: The title of a web page is the main text displayed in the search results, which directly affects the click-through rate and search engine ranking. An optimized title can include keywords and attractive language to help users understand the content of the page. Description: The description is a short summary displayed in the search results, usually below the title. An optimized description can attract users to click while informing them of the main content and value of the page. Keywords Keywords are words or phrases that identify the topic of the page.
[0033] In the modern Internet era, search engine optimization is one of the key factors for websites to obtain traffic and increase user engagement. Search engine optimization aims to improve the ranking of websites on search engine results pages, and to increase the chances of being discovered and clicked by users by optimizing website content and structure. Among the related technologies, SEO optimization is mainly carried out through manual optimization, SEO tools, CMS plug-in optimization, and optimization based on user feedback.
[0034] Manual SEO optimization is a traditional content optimization method, usually performed by SEO experts or content creators through manual inspection and adjustment of the SEO properties of web pages (such as TDK). Optimization technicians need to review the content of each page in detail, ensure that the use of keywords is in line with best practices, and optimize the structure and tags of the page. However, manual review and optimization requires a lot of time and effort, especially for large websites, it may take weeks or even months to optimize each page. At the same time, due to the complexity and tediousness of manual operations, omissions or errors are prone to occur during the manual optimization process, resulting in some pages not being effectively optimized. The effect of manual SEO optimization is highly dependent on the experience and knowledge level of SEO experts, and the optimization results of different experts may vary greatly. For example, a large e-commerce website publishes hundreds of new articles every day. The SEO team needs to check these articles one by one, determine the most appropriate keywords and optimize them, which is not only time-consuming but also prone to omissions due to fatigue and repetitive work.
[0035] In order to improve the efficiency of SEO optimization, many SEO tools have appeared on the market. These tools can provide rich data and analysis functions to help users conduct keyword analysis, competitor analysis and content suggestions. Although the tools can provide optimization suggestions, the final content adjustment and optimization still need to be done manually and cannot be fully automated. At the same time, some tools have complex functions and require users to have a certain technical background and operating experience. It is difficult for ordinary users to fully master and use them. In addition, high-end SEO tools are usually expensive, which puts a lot of cost pressure on small and medium-sized enterprises and individual webmasters. For example, a small blog site uses a certain SEO tool for SEO optimization. Although the tool provides a detailed analysis report, bloggers need to spend a lot of time learning and using the various functions of the tool and manually modifying the SEO settings of each article.
[0036] CMS, such as WordPress, provides built-in SEO plugins that can help users with basic SEO optimization. These plugins usually provide keyword suggestions, content analysis, and optimization tips, and users can make corresponding adjustments based on the tips. The optimization suggestions provided by CMS plugins are usually basic and lack in-depth analysis and personalized suggestions for specific content and target audiences. The suggestions provided by the plugins need to be implemented manually by users, and a fully automated optimization process cannot be achieved. For example, a news website using WordPress is optimized through an SEO plugin. Although the plugin provides some basic suggestions, the effect is not ideal for long reports that require in-depth analysis and personalized adjustments.
[0037] User feedback analysis is to improve content by collecting and analyzing user usage data and feedback. This method can help understand user needs and improve content relevance and user experience. However, collecting comprehensive and accurate user feedback data requires the design and implementation of an effective feedback mechanism, and the data collection process is complicated. At the same time, analyzing a large amount of user feedback data requires professional data analysis tools and techniques. The analysis process is time-consuming and technically demanding. It takes a certain amount of time to improve and optimize content based on the analysis results, and it is impossible to respond quickly to market changes and user needs. For example, a travel website improves content through user comments and feedback data. Although it can improve user satisfaction, the collection and analysis of feedback data takes a lot of time and cannot make adjustments quickly.
[0038] In response to the above problems, a search optimization method is provided in the present invention. The present invention also relates to a search optimization system, a search optimization device, a computing device, a computer-readable storage medium, and a computer program product, which are described in detail one by one in the following embodiments.
[0039] See also Figure 1 , Figure 1 It is an interactive schematic diagram of a search optimization system provided by an embodiment of the present invention, specifically including: a user end and an execution end, wherein the user end is used to receive the content to be optimized input by the user and send the content to be optimized to the execution end; the execution end is used to extract the text information of the content to be optimized, and perform semantic analysis on the text information, determine the corresponding semantic keywords and semantic features, and determine the target keywords based on the hot keywords and semantic keywords in the hot word library, wherein the hot word library is maintained based on the search popularity of the search keyword, optimization information is obtained based on the target keyword and semantic features, and the search tag of the content to be optimized is updated based on the optimization information. It can be understood that the user end and the execution end in the search optimization system can be located on different devices or on the same device.
[0040] Applied to the search optimization system, the text information of the content to be optimized is extracted, and the text information is semantically analyzed to determine the corresponding semantic keywords and semantic features; the target keywords are determined based on the hot keywords and semantic keywords in the hot word library, where the hot word library is maintained based on the search popularity of the search keyword; optimization information is obtained based on the target keyword and semantic features, and the search tags of the content to be optimized are updated based on the optimization information. The semantic core of the content to be optimized is extracted through semantic analysis, high-heat words in the hot word library are accurately matched, and search tags are dynamically adjusted, which significantly enhances the search engine visibility and user relevance of the content. At the same time, manual operations are reduced through automated processing, greatly improving optimization efficiency.
[0041] See also Figure 2 , Figure 2It is a flowchart of a search optimization method provided by an embodiment of the present invention, which specifically includes the following steps.
[0042] Step 202: extract text information of the content to be optimized, and perform semantic analysis on the text information to determine corresponding semantic keywords and semantic features.
[0043] Among them, the content to be optimized refers to any information carrier that needs to improve its ranking in search engine results. It includes but is not limited to: text, such as articles, news reports, product descriptions, blogs, comments, etc.; tables, such as data tables, statistical information, comparison lists, etc., which may be contained in web pages, documents or databases; pictures, graphics, photos, charts, etc., which usually have descriptive text or metadata; documents, such as PDF files, Word documents, PPT presentations, etc., which may contain a large amount of text information; videos and audios, although it is more complicated to extract text information directly from these media, they usually have auxiliary texts such as titles, descriptions and scripts.
[0044] Semantic keywords are words or phrases that reflect the core theme or key information of the content to be optimized. They are not limited to literal repetitions, but focus more on understanding the deeper meaning of the text, that is, "semantics". For example, in an article about environmental protection, words such as "sustainable development", "green energy", and "ecological balance" may be important semantic keywords because they convey the central idea of the article.
[0045] Semantic features include a series of elements that describe the characteristics of text content, style, emotional tendency, subject scope, etc. Semantic features can help to more fully understand the meaning and context of the text, such as: subject scope, the subject areas covered by the content, such as science and technology, health, education, etc.; emotional color, positive, negative or neutral emotional tendency; professional terminology, professional vocabulary in a specific industry or field; structural features, the organization of the text, such as narrative, explanatory, debate and other styles; target audience, the type of readers expected by the content, such as professionals, the general public, children, etc.
[0046] In practical applications, extracting text information involves extracting text data from the content to be optimized. For pure text content, this step is relatively simple, and it can be directly read and stored. However, for content containing non-text elements (such as tables, pictures, videos), corresponding technologies are needed to extract the text information therein. For tables and pictures, optical character recognition technology or other parsing tools are used to convert the text in the table into text format. For videos and audio, speech recognition technology is used to transcribe the audio into text, or text information is extracted from video descriptions and subtitles. Before performing semantic analysis, the extracted text needs to be preprocessed to remove noise and standardize the data: for example, deleting common meaningless words (such as "de", "he", "shi", etc.), which contribute little to semantic analysis, and removing punctuation marks that affect the analysis.
[0047] Segmenting the text into individual words or phrases and determining the grammatical functions of each word, such as nouns, verbs, adjectives, etc., helps to understand the sentence structure and meaning. Identifying proper nouns such as names of people, places, and organizations in the text, which often have important semantic values. Analyzing the relationships between words and understanding the sentence structure helps to identify key components such as the subject, predicate, and object. Evaluating the emotional tendency expressed in the text, such as positive, negative, or neutral. Using algorithms to automatically identify the main topic or theme of the text helps to refine semantic keywords. Based on the above analysis results, select the words that contribute the most to the meaning of the text. This usually includes topic words, high-frequency words, named entities, etc. Considering information such as part of speech, emotion, and topic distribution comprehensively, construct a multi-dimensional feature set that describes the characteristics of the text. For example, if the text mainly discusses technological innovation, then "technology", "innovation", "future", etc. can be used as semantic keywords, while "technology category", "positive emotion", "dense in technical terms" are semantic features.
[0048] Furthermore, perform semantic analysis on the text information to determine the corresponding semantic keywords and semantic features, including: segmenting the text information into at least one word unit; inputting the text information into a pre-trained semantic analysis model to determine the semantic features corresponding to the text information; determining semantic keywords in at least one word unit based on the semantic features.
[0049] In practical applications, text information is broken down into the smallest language units, the word units. This step may be done differently in different languages. For example, in English, words are usually naturally separated by spaces; in Chinese, specialized word segmentation tools such as Jieba word segmentation are needed to identify word boundaries. Each word unit is converted into a numerical vector so that the computer model can process it. This is usually achieved through word embedding technology, which can capture the semantic relationship between words and form a vector representation in a high-dimensional space. The vectorized representation of the text information is input into a pre-trained semantic analysis model. These models may include recurrent neural networks, long short-term memory networks, or large language models. The model extracts a series of semantic features by analyzing the word vectors and their positions in the text. The extracted semantic features are output in a structured form, such as using vectors, matrices, or specific data structures for subsequent processing and analysis. Based on the semantic features, each word unit is scored for importance. This can be achieved by calculating word frequency, inverse document frequency, semantic weight of the word, or probability scores predicted by the model. The word units with the highest scores are selected as semantic keywords. Usually, these keywords can represent the main theme or key information points of the text.
[0050] In a specific embodiment of the present invention, there is an article about "the application of artificial intelligence in medical diagnosis", and the article text is divided into word units such as "artificial intelligence", "medical", "diagnosis", and "application". Use a pre-trained word embedding model to convert each word unit into a vector representation. The word vector is input into a large language model to analyze the semantic structure of the text, and extract thematic features such as "medical diagnosis", "artificial intelligence technology", and "case analysis", as well as sentiment features such as "positive" and "professional". Based on the extracted semantic features, each word unit is scored for importance, and finally "artificial intelligence", "medical diagnosis", and "application" are selected as semantic keywords.
[0051] Based on this, using a pre-trained semantic analysis model to parse the semantic information of the text and determine the most representative and important semantic keywords can help understand and analyze the text content.
[0052] Step 204: Determine the target keyword based on the hot keywords and semantic keywords in the hot word library, wherein the hot word library is maintained based on the search popularity of the search keyword.
[0053] Among them, the hot word library is a database that collects and manages popular, hot or high-profile keywords on the current Internet. This database is usually maintained by search engines, social media platforms, content management systems or other Internet service providers in order to track and predict user search behavior and interest trends. The entries in the hot word library will be updated based on real-time or regular search data to ensure that the keywords contained in it reflect the latest Internet hot spots and popular topics.
[0054] Search keywords are words or phrases that users enter into search engines to find relevant information. They can be specific entity names, product descriptions, questions, events, concepts, etc., covering almost all possible query needs. In the context of hot word libraries, search keywords specifically refer to those terms that have been searched by a large number of users within a certain period of time. They have a high search popularity and represent the current focus of attention on the Internet.
[0055] Hot keywords are keywords in search keywords that match semantic keywords. For example, if the semantic keyword obtained by semantic analysis is "electric vehicle", in a hot word library that maintains a large number of search keywords, the hot keyword can be "new energy vehicle".
[0056] In actual applications, user search records from search engines, social media, news websites and other platforms are collected over a period of time, and statistical analysis is performed on the collected keywords. Indicators such as search frequency and click-through rate of each keyword are calculated to quantify the popularity of the keyword. A cycle (such as daily, weekly or monthly) is set to update the vocabulary, remove outdated keywords, add new hot words, and maintain the timeliness and accuracy of the vocabulary. Semantic keywords are used as query conditions to search for hot word databases with high popularity. The hot words found are scored for relevance to ensure that they are not only popular but also closely related to the original content, avoiding misleading information caused by irrelevant high-hot words. Several keywords that are most consistent with the content theme and have the highest search popularity are selected from the matching hot words as target keywords. Several keywords that are most consistent with the content theme and have the highest search popularity can also be selected from the matching hot words and combined with semantic keywords to obtain target keywords.
[0057] In a specific embodiment of the present invention, there is an article about "home fitness equipment buying guide", and several core keywords such as "home fitness", "fitness equipment", and "buying guide" have been determined through semantic analysis. "Home fitness" and "fitness equipment" are used as query words, and it is determined that "home gym equipment" and "fitness equipment recommendation" are very popular in the hot word library and are closely related to the topic of the article. Therefore, "home gym equipment" and "fitness equipment recommendation" are used as target keywords because they are not only popular, but also can attract target readers who are interested in home fitness.
[0058] In another specific embodiment of the present invention, there is an article about "Electric Vehicle Sales Trends", and through semantic analysis, the semantic keyword is determined to be "electric vehicle", and "new energy vehicle" is matched as the hot keyword in the search keywords in the hot word library. The hot keyword "new energy vehicle" is combined with the semantic keyword "electric vehicle" to obtain "new energy electric vehicle" as the target keyword.
[0059] Further, determining the target keyword based on the hot keywords and semantic keywords in the hot word library includes: calculating the semantic similarity between the search keyword and the semantic keyword in the hot word library; and determining the target keyword in the search keyword based on the semantic similarity.
[0060] In practical applications, semantic keywords and search keywords in the hot word library are converted into vector representations, and the similarity between the semantic keyword vector and each search keyword vector in the hot word library is calculated using cosine similarity, Euclidean distance or other similarity measurement methods. One possible way to determine the target keyword is to set a semantic similarity threshold based on actual conditions. Only search keywords that exceed this threshold will be considered as target keywords. Another possible way is to sort the calculated similarities and give priority to the search keywords with the highest similarity. Combined with the search popularity of the keywords, the most relevant and popular words are selected as target keywords.
[0061] In a specific embodiment of the present invention, known semantic keywords include "global warming", "sea level rise", and "extreme weather". Use the model to convert "global warming", "sea level rise", "extreme weather" and all search keywords in the hot word library into vector representations. Calculate the cosine similarity between the semantic keyword and each search keyword in the hot word library. Set the similarity threshold to 0.7, and find keywords with a similarity higher than 0.7 and high search popularity from the hot word library, such as "sea level change", "climate forecast", "global warming", etc.
[0062] Based on this, the pre-built hot word library and semantic similarity calculation are used to help find the target keywords that are most relevant to the input semantic keywords. These steps help to more effectively select suitable keywords in information retrieval or content optimization to improve the visibility and attractiveness of the content.
[0063] Furthermore, there are multiple semantic keywords; the semantic similarity between the search keywords in the hot word library and the semantic keywords is calculated, including: for any semantic keyword, the semantic similarity between each search keyword in the hot word library and any semantic keyword is calculated; accordingly, based on the semantic similarity, the target keyword is determined in the search keywords, including: the search keyword whose semantic similarity is greater than a preset threshold is determined as the target keyword corresponding to any semantic keyword.
[0064] In actual applications, a similarity matrix is created for each semantic keyword. The rows represent search keywords in the hot word library, and the columns represent semantic keywords. This matrix is used to store the calculated semantic similarity values. For each semantic keyword, all search keywords in the hot word library are traversed, and the semantic similarity between them is calculated using the word embedding model. A reasonable semantic similarity threshold is set according to the content relevance and hot trend. This threshold can be adjusted based on experiments and experience to achieve the best keyword matching effect. For each semantic keyword, search keywords with a similarity greater than the preset threshold are screened out and identified as target keywords. This step may generate multiple target keyword lists, each corresponding to a semantic keyword. Compare the target keyword lists generated by different semantic keywords to find co-occurring search keywords. These keywords are usually highly correlated with multiple semantic keywords and are preferred candidates. Considering the popularity of search keywords in the hot word library, the co-occurring search keywords can be further weighted according to their popularity, and keywords with higher popularity and stronger relevance to the content are selected as the final target keywords.
[0065] Step 206: Obtain optimization information based on the target keyword and the semantic feature, and update the search tag of the content to be optimized based on the optimization information.
[0066] Among them, optimization information is TDK. Search tags are a series of keywords or phrases used to describe the content of a web page. They help search engines understand and classify web pages, and also allow users to quickly locate relevant content by entering specific keywords.
[0067] Q&A big models or other Q&A systems refer to intelligent systems based on advanced natural language processing technology that can understand and generate natural language to answer users' questions. Q&A big models such as ChatGPT and BERT, etc., have been trained with a large amount of data and have deep language understanding and generation capabilities. They can handle complex questions and provide accurate and relevant answers. Other Q&A systems may include Q&A robots in specific fields or intelligent answering functions of search engines. They analyze user input through algorithms, retrieve relevant information and generate useful replies to help users quickly find the information they need or solve problems.
[0068] In actual application, through the Q&A big model or other Q&A systems, combine the target keywords and the core theme of the content to create a title that can attract users to click and accurately reflect the content. Write a description containing the target keywords, outlining the main ideas or selling points of the content, and encouraging users to learn more. Create a series of keyword tags based on the target keywords for the classification and indexing of the content. Integrate the optimized title, description, and keyword tags into the meta information of the content, ensuring that they are reflected in the appropriate location of the web page code. Apply the optimized information to the actual content. If it is new content, apply it directly when publishing; if it is existing content, make necessary updates. After publishing or updating, continue to monitor the changes in the content's ranking in search engines and user interaction data to evaluate the effectiveness of the optimization, and make necessary adjustments based on feedback.
[0069] In a specific embodiment of the present invention, there is a blog article about "Environmentally Friendly Life Tips". It is determined that the article mainly revolves around themes such as environmental protection, lifestyle, energy conservation and emission reduction, and "Environmentally Friendly Life" and "Energy Conservation and Emission Reduction" are preliminarily determined as keywords. The hot word library is queried to determine that "Green Life" is a hot word and is closely related to the theme of the article. Title optimization: "Green Living: Practical Tips for Energy Conservation and Emission Reduction", description optimization: "Explore simple and effective environmentally friendly life tips, learn how to save energy and reduce emissions, and create a greener living environment." Keyword tag optimization: "Green Living", "Energy Conservation and Emission Reduction". Update the title, description and keyword tags in the original information of the article to ensure that they are consistent with the optimized information. Publish the optimized article to the website, and use SEO tools to monitor its performance in search engines, collect user feedback, and continuously optimize.
[0070] Furthermore, before determining the target keyword based on the hot keywords and semantic keywords in the hot word library, it also includes: obtaining candidate keywords and corresponding search heat every preset time period; updating the candidate keywords whose search heat reaches the preset heat threshold in each candidate keyword to the hot word library.
[0071] Furthermore, before determining the target keyword based on the hot keywords and semantic keywords in the hot word library, it also includes: obtaining the search popularity of the search keywords in the hot word library every preset time period; and deleting the search keywords whose search popularity does not meet the standard in the hot word library.
[0072] Among them, search popularity is an indicator that measures the frequency of keyword searches in a specific period of time. It reflects the degree of user interest in a topic or product.
[0073] In actual applications, a preset period is defined, such as daily, weekly, or monthly, for regularly updating the hot word library. The choice of this time interval should be determined based on the frequency of keyword popularity changes and business needs. At the beginning of the preset period, the latest search data is captured from multiple data sources (such as search engines, social media, news websites, etc.). This data usually contains the keywords searched by users and their search times or popularity indicators. From the captured data, all searched keywords are extracted to form a candidate keyword list. For each candidate keyword, the number of searches or popularity indicators within the preset period are counted, which will serve as a quantitative standard for keyword popularity. According to historical data and business needs, a preset popularity threshold is set. This threshold is used to distinguish which keywords are hot enough to be included in the hot word library. From the candidate keyword list, filter out keywords whose search popularity reaches or exceeds the preset popularity threshold. These keywords are regarded as hot words in the current period. Add the filtered keywords to the hot word library, and remove old keywords whose popularity no longer meets the threshold.
[0074] In a specific embodiment of the present invention, taking the preset period as once a week and the preset heat threshold as 100 average daily search times as an example, starting from Monday of each week, the search data of all keywords in the past week are captured from each data platform. The total number of searches for each keyword in the past week is counted, and the average daily search times are calculated as the heat index. Keywords with an average daily search time of more than 100 times are screened out, and they are added to the heat word library, and old keywords that no longer meet the conditions are removed.
[0075] Based on this, by regularly updating the hot word library and combining it with semantic similarity calculation, the system can automatically discover and select the target keywords that best match specific semantic keywords, thereby optimizing the content's search engine performance and user retrieval experience.
[0076] Furthermore, optimization information is obtained based on target keywords and semantic features, including: inputting the target keywords and semantic features into a question-answering model to obtain subject information and description information of the content to be optimized, wherein the question-answering model is trained based on sample keywords, sample semantic features, label subject information and label description information; and constructing optimization information based on the target keywords, subject information and description information.
[0077] In practical applications, first, collect a large number of known sample keywords, corresponding semantic features, and expert-marked topic information and description information. These samples can come from relevant standard documents of SEO, successful optimization case studies, or knowledge bases in professional fields. Clean the collected data to remove irrelevant or erroneous information to ensure data quality and consistency. Then, preprocess the text data, such as word segmentation, stop word removal, stem extraction, etc., to prepare the input of the machine learning model. Train the model using sample keywords, sample semantic features, label topic information, and label description information. Adjust the model parameters and optimize the loss function until the model can achieve satisfactory performance on the training set. Evaluate the generalization ability of the model on the validation set to ensure that it not only performs well on the training data, but also accurately predicts unseen data. Pass the target keywords and semantic features as input to the trained question-answering model. The model will output the topic information and description information most relevant to the target keyword. The topic information summarizes the core topic of the content, while the description information provides a detailed description of the topic. Finally, the target keywords, the obtained topic information, and the description information are combined to construct the optimization information.
[0078] Based on this, the ability of the question-answering model is utilized to automatically analyze and optimize the subject and description information of the content by inputting target keywords and semantic features. This method can effectively improve the visibility and relevance of content in search engines, thereby optimizing the user's search experience and the attractiveness of the content.
[0079] Furthermore, the target keywords and semantic features are input into the question-answering model to obtain the subject information and description information of the content to be optimized, including: obtaining input prompt information, wherein the input prompt information is used to constrain the output format of the question-answering model; based on the input prompt information, the target keywords and semantic features are input into the question-answering model to obtain the subject information and description information of the content to be optimized.
[0080] The input prompt is usually a guiding text that tells the model the expected output format and type. For example, if you want the model to generate a description, the prompt may be "Please describe the details about..."
[0081] In practical applications, the target keywords, semantic features, and input prompt information are integrated into the model's input format, that is, keywords and features are incorporated into the prompt information, or they are provided to the model as additional input fields. The constructed input is passed to the model to perform the inference process. The model will generate corresponding topic information and description information based on the input.
[0082] In a specific embodiment of the present invention, an example of inputting prompt information is as follows:
[0083] “You are an SEO expert responsible for generating optimized titles, descriptions, and keywords for an article about {article title}. The current hot keywords are: {hot keyword list}. Please generate optimized titles, descriptions, and keywords based on these keywords.
[0084] Article content: {Article content summary}
[0085] Please generate:
[0086] title: A sentence that summarizes the article content and includes the article keywords and search engine hot keywords.
[0087] Description: Use two to three sentences to describe the article content in detail to attract readers to click, and combine it with search engine hot keywords to generate optimized titles, descriptions and keywords based on these keywords.
[0088] keywords: List the keywords related to the article content, including article keywords and search engine hot keywords.
[0089] Output format:
[0090] tit le:[generated tit le]
[0091] description:[generated description]
[0092] keywords:[generated keywords]”
[0093] Based on this, by combining input prompt information, target keywords and semantic features, the question-answering model is used to generate optimized content topics and description information, thereby improving the relevance and quality of the content.
[0094] Furthermore, after obtaining the optimization information based on the target keywords and semantic features, it also includes: responding to the correction instruction of the front end, correcting the optimization information to obtain the optimized correction information; training the question-answering model based on the target keywords, semantic features and the optimized correction information until the training stop condition is met.
[0095] Among them, correction instructions are guiding commands issued by users, administrators, or another system component. Such instructions are intended to instruct the system how to improve its output in order to achieve goals more accurately and effectively.
[0096] In actual applications, users or the system front end may propose corrections to the preliminary optimization information based on actual conditions, user preferences or other factors. The back-end system receives these correction instructions and understands the specific content and direction that needs to be modified. According to the correction instructions provided by the front end, the preliminary optimization information is modified, which may include adding or deleting content, rewriting sentences, adjusting styles, etc. After the correction is completed, the final optimization correction information is obtained, which is closer to actual needs and has higher quality.
[0097] The target keywords, semantic features, and optimization correction information are organized into training data sets, each of which contains input (sample keywords and sample semantic features) and expected output (optimization correction information). These data sets are used to train the question-answering model, with the goal of allowing the model to learn how to generate high-quality, compliant optimization information given the keywords and features. The training process may involve multiple cycles, and the model performance is evaluated after each training, and the model parameters are adjusted based on the evaluation results until the preset stop condition is reached.
[0098] Based on this, by responding to correction instructions, generating correction information, and retraining the question-answering model based on the corrected optimization information, the model's ability and accuracy in generating optimized content can be continuously improved. This cyclic training method can continuously optimize and update the system to better serve search engine optimization and user needs.
[0099] Applied to the search optimization method, the text information of the content to be optimized is extracted, and the text information is semantically analyzed to determine the corresponding semantic keywords and semantic features; the target keywords are determined based on the hot keywords and semantic keywords in the hot word library, where the hot word library is maintained based on the search popularity of the search keyword; optimization information is obtained based on the target keyword and semantic features, and the search tags of the content to be optimized are updated based on the optimization information. The semantic core of the content to be optimized is extracted through semantic analysis, high-heat words in the hot word library are accurately matched, and search tags are dynamically adjusted, which significantly enhances the search engine visibility and user relevance of the content. At the same time, manual operations are reduced through automated processing, greatly improving the optimization efficiency.
[0100] The following combination Figure 3 Taking the application of the search optimization method provided by the present invention in the SEO of website text as an example, the search optimization method is further explained. Figure 3 It is a process flow chart of a method for searching and optimizing website articles provided by an embodiment of the present invention, which specifically includes the following steps.
[0101] Step 302: The user enters the article to be optimized on the website editing page.
[0102] Step 304: Perform semantic analysis on the article to be optimized using a large language model.
[0103] Specifically, after receiving the article, the large language model uses its natural language processing capabilities to conduct an in-depth analysis of the article and identify the core topics and key information of the article.
[0104] Step 306: Generate TDK based on the semantic analysis results.
[0105] Specifically, the topics and concepts identified in the article are matched with the hot keywords in the hot word library to find the most relevant keywords to ensure that the article content is consistent with user search intent and search engine trends.
[0106] For the hot word library, you can regularly crawl SEO-related websites, search engine result pages, social media platforms, etc. to obtain the latest hot keywords and topics. By setting the hot threshold, you can monitor the search volume, click-through rate and other data of keywords in real time, and automatically identify hot keywords that are above the threshold. Use the API interface to obtain real-time data from search engines, and combine the mention frequency and interaction volume of social media to comprehensively evaluate the popularity of keywords. Clean and pre-process the crawled data to remove irrelevant information and noise to ensure the accuracy and validity of the data. Use natural language processing technology to segment the text and extract keywords to extract valuable information, regularly add hot words with search popularity higher than the threshold to the word library, and remove outdated or no longer popular keywords to maintain the real-time and relevance of the word library.
[0107] Search popularity can be determined by combining multiple dimensions of data such as keyword search volume, click-through rate, and number of social media mentions.
[0108] Step 308: The user reviews the TDK.
[0109] Specifically, website editors can review the TDK to ensure that it meets the article topic and SEO goals. Editors can make secondary revisions to ensure flexibility and accuracy.
[0110] Step 310: Based on the TDK that has been reviewed and confirmed, the article to be optimized is search optimized and published to the website.
[0111] Specifically, there will be three TDK input boxes in the background editing page of the website. The TDK after review and confirmation will be input into the input box as additional data of the current article. At this moment, the data structure of article editing is mainly divided into two parts: the original data of the article, including the article title, author, update time, article content, etc., and TDK, including the search tag information of the article that only works on HTML (HyperText Markup Language). The search tag information will be directly searched by other users as a search anchor.
[0112] Corresponding to the above method embodiment, the present invention also provides a search optimization device embodiment, Figure 4 FIG. 1 is a schematic diagram of the structure of a search optimization device provided by an embodiment of the present invention. Figure 4 As shown, the device comprises:
[0113] The extraction module 402 is configured to extract text information of the content to be optimized, and perform semantic analysis on the text information to determine corresponding semantic keywords and semantic features;
[0114] The determination module 404 is configured to determine the target keyword based on the hot keywords and semantic keywords in the hot word library, wherein the hot word library is maintained based on the search heat of the search keyword;
[0115] The optimization module 406 is configured to obtain optimization information based on the target keyword and the semantic feature, and update the search tag of the content to be optimized based on the optimization information.
[0116] Optionally, the extraction module 402 is further configured to segment the text information into at least one word unit; input the text information into a pre-trained semantic analysis model to determine semantic features corresponding to the text information; and determine semantic keywords in at least one word unit based on the semantic features.
[0117] Optionally, the determination module 404 is further configured to calculate the semantic similarity between the search keywords and the semantic keywords in the hot word library; and determine the target keyword in the search keywords based on the semantic similarity.
[0118] Optionally, the determination module 404 is further configured to calculate the semantic similarity between each search keyword in the hot word library and any semantic keyword for any semantic keyword; and determine the search keyword with a semantic similarity greater than a preset threshold as the target keyword corresponding to any semantic keyword.
[0119] Optionally, the search optimization device also includes an updating module, which is configured to obtain candidate keywords and corresponding search heats at preset time intervals; and update the candidate keywords whose search heats reach a preset heat threshold among the candidate keywords to the heat word library.
[0120] Optionally, the search optimization device also includes a deletion module, which is configured to obtain the search popularity of search keywords in the hot word library every preset time period; and delete search keywords whose search popularity does not meet the requirements in the hot word library.
[0121] Optionally, the optimization module 406 is further configured to input the target keywords and semantic features into a question-answering model to obtain the subject information and description information of the content to be optimized, wherein the question-answering model is trained based on sample keywords, sample semantic features, label subject information and label description information; and the optimization information is constructed based on the target keywords, subject information and description information.
[0122] Optionally, the optimization module 406 is further configured to obtain input prompt information, wherein the input prompt information is used to constrain the output format of the question-answering model; based on the input prompt information, target keywords and semantic features, the question-answering model is input to obtain the topic information and description information of the content to be optimized.
[0123] Optionally, the search optimization device also includes a correction module, which is configured to correct the optimization information in response to the correction instruction of the front end to obtain optimized correction information; and train the question and answer model based on the target keywords, semantic features and the optimized correction information until the training stop condition is met.
[0124] Applied to the search optimization device, the extraction module 402 extracts the text information of the content to be optimized, and performs semantic analysis on the text information to determine the corresponding semantic keywords and semantic features; the determination module 404 determines the target keyword based on the hot keywords and semantic keywords in the hot word library, wherein the hot word library is maintained based on the search popularity of the search keyword; the optimization module 406 obtains optimization information based on the target keyword and the semantic features, and updates the search tag of the content to be optimized based on the optimization information. The semantic core of the content to be optimized is extracted through semantic analysis, the high-heat words in the hot word library are accurately matched, and the search tags are dynamically adjusted, which significantly enhances the search engine visibility and user relevance of the content, and reduces manual operations through automated processing, greatly improving the optimization efficiency.
[0125] The above is a schematic scheme of a search optimization device of this embodiment. It should be noted that the technical scheme of the search optimization device and the technical scheme of the above search optimization method belong to the same concept, and the details not described in detail in the technical scheme of the search optimization device can be found in the description of the technical scheme of the above search optimization method.
[0126] Figure 5 The block diagram of a computing device 500 provided according to an embodiment of the present invention is shown. The components of the computing device 500 include but are not limited to a memory 510 and a processor 520. The processor 520 is connected to the memory 510 via a bus 530, and the database 550 is used to store data.
[0127] The computing device 500 also includes an access device 540 that enables the computing device 500 to communicate via one or more networks 560. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 540 may include one or more of any type of network interface, wired or wireless (e.g., a network interface card (NIC)), such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a World Wide Interoperability for Microwave Access (Wi-MAX) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, and a Near Field Communication (NFC).
[0128] In one embodiment of the present invention, the above components of the computing device 500 and Figure 5 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Figure 5 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of the present invention. Those skilled in the art may add or replace other components as needed.
[0129] The computing device 500 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 500 may also be a mobile or stationary server.
[0130] The processor 520 is used to execute the following computer program / instructions, which implement the steps of the above-mentioned search optimization method when executed by the processor.
[0131] Each embodiment of the present invention is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computing device embodiment, since it is basically similar to the search optimization method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the search optimization method embodiment.
[0132] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program / instruction, which implements the steps of the above-mentioned search optimization method when executed by a processor.
[0133] Each embodiment of the present invention is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computer-readable storage medium embodiment, since it is basically similar to the search optimization method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the search optimization method embodiment.
[0134] An embodiment of the present invention further provides a computer program product, including a computer program / instruction, which implements the steps of the above-mentioned search optimization method when executed by a processor.
[0135] The above is a schematic scheme of a computer program product of this embodiment. It should be noted that the technical scheme of the computer program product and the technical scheme of the above-mentioned search optimization method belong to the same concept, and the details not described in detail in the technical scheme of the computer program product can be referred to the description of the technical scheme of the above-mentioned search optimization method.
[0136] The above describes specific embodiments of the present invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0137] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.
[0138] It should be noted that, for the convenience of description, the aforementioned method embodiments are all described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the described action sequence, because according to the embodiments of the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of the present invention.
[0139] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0140] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The optional embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of the present invention. The present invention selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.
Claims
1. A search optimization method, characterized in that: include: Extracting text information of the content to be optimized, and performing semantic analysis on the text information to determine corresponding semantic keywords and semantic features; Determining a target keyword based on hot keywords in a hot word library and the semantic keyword, wherein the hot word library is maintained based on the search popularity of the search keyword; Optimization information is acquired based on the target keyword and the semantic feature, and the search tag of the content to be optimized is updated based on the optimization information.
2. The method according to claim 1, characterized in that The performing semantic analysis on the text information to determine corresponding semantic keywords and semantic features includes: Segmenting the text information into at least one word unit; Inputting the text information into a pre-trained semantic analysis model to determine the semantic features corresponding to the text information; A semantic keyword is determined in the at least one word unit based on the semantic feature.
3. The method according to claim 1, characterized in that The step of searching the hot word library for a target keyword that matches the semantic keyword includes: Calculating the semantic similarity between the search keyword in the hot word library and the semantic keyword; Based on the semantic similarity, a target keyword is determined among the search keywords.
4. The method according to claim 3, characterized in that The number of the semantic keywords is multiple; the calculating of the semantic similarity between the search keywords in the hot word library and the semantic keywords includes: For any semantic keyword, calculating the semantic similarity between each search keyword in the hot word library and the any semantic keyword; Accordingly, determining a target keyword from the search keywords based on the semantic similarity includes: The search keyword whose semantic similarity is greater than a preset threshold is determined as a target keyword corresponding to any of the semantic keywords.
5. The method according to claim 1, characterized in that Before searching the hot word library for a target keyword matching the semantic keyword, the method further includes: At every preset time period, candidate keywords and corresponding search popularity are obtained; The candidate keywords whose search popularity reaches a preset popularity threshold among the candidate keywords are updated to the popularity word library.
6. The method according to claim 1, characterized in that Before searching the hot word library for a target keyword matching the semantic keyword, the method further includes: At every preset time period, the search popularity of the search keywords in the hot word library is obtained; The search keywords whose search popularity does not meet the requirement are deleted from the hot word library.
7. The method according to claim 1, characterized in that The obtaining of optimization information based on the target keyword and the semantic feature includes: Inputting the target keyword and the semantic feature into a question-answering model to obtain the topic information and description information of the content to be optimized, wherein the question-answering model is trained based on sample keywords, sample semantic features, label topic information, and label description information; The optimization information is constructed based on the target keyword, the subject information and the description information.
8. The method according to claim 7, characterized in that The step of inputting the target keyword and the semantic feature into a question-answering model to obtain the subject information and description information of the content to be optimized includes: Obtaining input prompt information, wherein the input prompt information is used to constrain the output format of the question-answering model; Based on the input prompt information, the target keyword and the semantic feature, the question-answering model is input to obtain the subject information and description information of the content to be optimized.
9. The method according to claim 1, characterized in that: After obtaining the optimization information based on the target keyword and the semantic feature, the method further includes: In response to the correction instruction of the front end, the optimization information is corrected to obtain the optimized correction information; The question-answering model is trained based on the target keywords, the semantic features and the optimization correction information until a training stop condition is reached.
10. A search optimization device, characterized in that: include: An extraction module is configured to extract text information of the content to be optimized, and perform semantic analysis on the text information to determine corresponding semantic keywords and semantic features; A determination module is configured to search a hot word library for a target keyword that matches the semantic keyword, wherein the hot word library is maintained based on the search popularity of the search keyword; The optimization module is configured to obtain optimization information based on the target keyword and the semantic feature, and update the search tag of the content to be optimized based on the optimization information.
11. A computing device, characterized in that: include: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer program / instructions are executed by the processor, the steps of the search optimization method described in any one of claims 1-9 are implemented.
12. A computer-readable storage medium, characterized in that: It stores a computer program / instruction, which, when executed by a processor, implements the steps of the search optimization method described in any one of claims 1-9.
13. A computer program product, characterized in that It includes a computer program / instruction, which, when executed by a processor, implements the steps of the search optimization method described in any one of claims 1 to 9.
Citation Information
Patent Citations
Search engine optimization method and device, equipment and storage medium
CN113449165A
Webpage content optimization method, device and equipment
CN115827951A
Search engine optimization method and system based on AI analysis
CN117312711A
Cited By
Real estate registration archive arrangement service method and system
CN120386901A
Real estate registration file sorting service method and system
CN120386901B
Knowledge base document searching method and device, electronic equipment and storage medium
CN120821830A
Content generation method and device based on keyword semantics and electronic equipment
CN120910231A
Data processing method, computing device, storage medium and program product
CN121029772A
