Keyword-based data association storage method and device

Abstract

The present application provides a data association storage method and device based on keywords, which relates to the field of data processing technology and solves the technical problem of low accuracy of association relationships for keywords. The method includes: performing AI interpretation on the entire text of each of the multiple texts to be stored through an AI text interpretation module to obtain the result of the entire text interpretation, and generating the text theme corresponding to the text to be stored based on the result of the entire text interpretation; determining multiple first texts to be stored with the same or similar text themes from the multiple texts to be stored; performing AI interpretation on the multiple first texts to be stored sentence by sentence through the AI ​​text interpretation module to obtain the result of the text sentence interpretation, and searching for target sentences with the same or similar sentence meanings in the multiple first texts to be stored based on the text sentence interpretation result, as well as multiple second texts to be stored containing the target sentences.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a keyword-based data association storage method and device. Background Art

[0002] The massive amount of information collected from social media, news websites and other platforms is usually stored in a database. Currently, keyword search is a common method for searching indexes during database query, where keywords can be words used by a single media when making and using indexes. For example, it can be the specific names of products, services and companies that visitors want to know about. Therefore, the associations between the above-mentioned massive amounts of information stored in the database are usually also associated with keywords. However, the accuracy of the associations stored with keywords is low. Summary of the invention

[0003] The object of the present invention is to provide a keyword-based data association storage method and device to solve the technical problem of low accuracy of association relationships for keywords.

[0004] In a first aspect, an embodiment of the present application provides a keyword-based data association storage method, the method comprising:

[0005] Get multiple texts to be stored;

[0006] Performing AI interpretation on the entire text of each of the multiple texts to be stored by an artificial intelligence (AI) text interpretation module to obtain an entire text interpretation result, and generating a text theme corresponding to the text to be stored according to the entire text interpretation result;

[0007] Determining a plurality of first texts to be stored having the same or similar text themes from the plurality of texts to be stored, and performing primary associative storage on the plurality of first texts to be stored based on the subject association relationship of the text themes;

[0008] The AI ​​text interpretation module performs AI interpretation sentence by sentence on the plurality of first texts to be stored to obtain text sentence interpretation results, and searches for target sentences with the same or similar meanings in the plurality of first texts to be stored according to the text sentence interpretation results, as well as a plurality of second texts to be stored that contain the target sentences;

[0009] The plurality of second texts to be stored are stored in a secondary association manner according to the sentence meaning association relationship of the target sentence; the association height of the secondary association storage is greater than the association height of the primary association storage;

[0010] The AI ​​text interpretation module performs AI interpretation of the plurality of second texts to be stored word by word to obtain text vocabulary interpretation results, and searches the plurality of second texts to be stored for target keywords having the same or similar meanings as the designated keywords, and a plurality of third texts to be stored containing the target keywords according to the text vocabulary interpretation results;

[0011] The plurality of third texts to be stored are stored in a three-level associative manner through the semantic association relationship between the designated keyword and the target keyword; the association height of the three-level associative storage is greater than the association height of the two-level associative storage.

[0012] In a possible implementation, after the step of searching the plurality of second texts to be stored for target keywords having the same meaning or a similar meaning as the designated keyword according to the text vocabulary interpretation result, the method further includes:

[0013] Determining a text to be analyzed and a first network publishing time of the text to be analyzed from the plurality of texts to be stored;

[0014] Based on the subject of the text, the meaning of the target sentence and the meaning of the target keyword, query the source text corresponding to the text to be analyzed through several second network publishing times earlier than the first network publishing time, and determine the text with the earliest second network publishing time among the multiple source texts as the root text;

[0015] Determine the publication objects of the tracing text and the root text, and generate a propagation path gradient map of the text to be analyzed based on the publication objects of the tracing text and the root text, the publication methods of the tracing text and the root text, the second network publication time corresponding to the tracing text, and the second network publication time corresponding to the root text.

[0016] In a possible implementation, the publishing method includes text reprint publishing, text editing publishing and text link publishing; the step of generating the propagation path gradient map of the text to be analyzed according to the publishing objects of the tracing text and the root text, the publishing methods of the tracing text and the root text, the second network publishing time corresponding to the tracing text, and the second network publishing time corresponding to the root text, includes:

[0017] According to the publication objects of the traceable text and the root text, the second network publication time corresponding to the traceable text, and the second network publication time corresponding to the root text, a preliminary propagation path gradient map of the text to be analyzed is generated; wherein the gradient first connection relationship between the text to be analyzed, the traceable text and the root text in the preliminary propagation path gradient map is a gradient association relationship based on the chronological order;

[0018] Determine the dissemination mode association relationship among the text to be analyzed, the source text and the root text according to the text reprint relationship, text link relationship and text editing duplication among the text to be analyzed, the source text and the root text; wherein the dissemination mode association relationship includes a reprint association relationship, a link association relationship and a copy association relationship;

[0019] The preliminary propagation pathway gradient map is optimized based on the propagation mode association relationship to generate a final propagation pathway gradient map of the text to be analyzed; wherein the gradient second connection relationship between the text to be analyzed, the source text and the root text in the final propagation pathway gradient map is a gradient association relationship based on the propagation mode.

[0020] In a possible implementation, after the step of performing AI interpretation on the entire text of each of the plurality of texts to be stored by the AI ​​text interpretation module to obtain the result of the entire text interpretation, the step further includes:

[0021] Identify the main idea of ​​the text to be stored according to the result of the whole text interpretation, and generate a summary and a subdirectory corresponding to the text to be stored based on the main idea of ​​the text and the main idea;

[0022] A general directory corresponding to the multiple texts to be stored is generated based on the sub-directories corresponding to the multiple texts to be stored, and a query graph of the texts to be stored is generated based on the abstract, the general directory and the texts to be stored; wherein the query graph contains a mapping relationship between the abstract, the text to be stored and each of the sub-directories in the general directory.

[0023] In a possible implementation, the query graph further includes a mapping relationship between the text to be stored and the domain name where the text is published, and the original text page corresponding to the text in the network;

[0024] After the step of generating a general catalogue corresponding to the plurality of texts to be stored based on the sub-catalogues corresponding to the plurality of texts to be stored, and generating a query graph of the texts to be stored according to the abstract, the general catalogue and the texts to be stored, the method further includes:

[0025] Determine the text to be searched from the plurality of texts to be stored, and display the general catalog and the plurality of sub-catalogs in the general catalog in a graphical user interface;

[0026] In response to a first selection instruction for a target subdirectory among the plurality of subdirectories, determining a target subdirectory corresponding to the to-be-queried text in the general directory according to the first selection instruction;

[0027] Based on the mapping relationship between the target subdirectory through the abstract and each of the subdirectories in the general directory, determine and display the abstract corresponding to the text to be queried in a graphical user interface;

[0028] In response to a second selection instruction for a target abstract among the multiple abstracts corresponding to the text to be queried, determining a target text to be stored corresponding to the text to be queried according to the target abstract through a mapping relationship between the abstract and the text to be stored;

[0029] According to the mapping relationship between the text to be stored and the text publishing domain name, and the original text page corresponding to the text in the network, the target text publishing domain name corresponding to the text to be queried and the target original text page corresponding to the text to be queried in the network are reversely queried from the query graph.

[0030] In a possible implementation, after the step of performing three-level associative storage on the plurality of third texts to be stored through the semantic association relationship between the designated keyword and the target keyword, the method further includes:

[0031] Based on the main theme of the text, the overall idea, the sentence meaning of the target sentence and the word meaning of the target keyword, the advanced association relationship between the multiple third texts to be stored is determined, and the multiple third texts to be stored are stored in a four-level association through the advanced association relationship; wherein the advanced association relationship includes any one or more of the text corroboration relationship, the text extension relationship, and the main theme sublimation relationship.

[0032] In a possible implementation, after the step of performing three-level associative storage on the plurality of third texts to be stored through the semantic association relationship between the designated keyword and the target keyword, the method further includes:

[0033] Determining the actual occurrence frequency of the same target keyword in each of the plurality of third texts to be stored;

[0034] If the actual occurrence frequency is lower than the preset occurrence frequency threshold, the plurality of third texts to be stored are stored in a moderately associated manner;

[0035] If the actual occurrence frequency is higher than or equal to the preset occurrence frequency threshold, the plurality of third texts to be stored are stored with high association.

[0036] In a second aspect, a keyword-based data association storage device is provided, comprising:

[0037] An acquisition module is used to acquire multiple texts to be stored;

[0038] A generating module, configured to perform AI interpretation on the entire text of each of the plurality of texts to be stored through an AI text interpretation module to obtain an interpretation result of the entire text, and generate a text theme corresponding to the text to be stored according to the interpretation result of the entire text;

[0039] A determination module, used for determining a plurality of first texts to be stored having the same or similar text subjects from the plurality of texts to be stored, and performing primary associative storage on the plurality of first texts to be stored according to the subject association relationship of the text subjects;

[0040] A first query module is used to perform AI interpretation sentence by sentence on the plurality of first texts to be stored through the AI ​​text interpretation module to obtain text sentence interpretation results, and to search for target sentences with the same or similar sentence meanings in the plurality of first texts to be stored, and a plurality of second texts to be stored containing the target sentences according to the text sentence interpretation results;

[0041] A first storage module is used to perform secondary association storage on the plurality of second texts to be stored according to the sentence meaning association relationship of the target sentence; the association height of the secondary association storage is greater than the association height of the primary association storage;

[0042] A second query module is used to perform AI interpretation of the plurality of second texts to be stored word by word through the AI ​​text interpretation module to obtain text vocabulary interpretation results, and search for target keywords with the same or similar meanings as the specified keyword in the plurality of second texts to be stored according to the text vocabulary interpretation results, and at least two third texts to be stored containing the target keywords;

[0043] The second storage module is used to perform three-level associative storage on the at least two third texts to be stored through the semantic association relationship between the designated keywords and the target keywords; the association height of the three-level associative storage is greater than the association height of the two-level associative storage.

[0044] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program that can be executed on the processor, and when the processor executes the computer program, the method described in the first aspect is implemented.

[0045] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to execute the method described in the first aspect above.

[0046] The embodiments of the present application bring the following beneficial effects:

[0047] The embodiment of the present application provides a keyword-based data association storage method and device, which can obtain multiple texts to be stored, perform AI interpretation on the entire text of each of the multiple texts to be stored through an AI text interpretation module to obtain a text interpretation result, and generate a text theme corresponding to the text to be stored according to the text interpretation result, determine multiple first texts to be stored with the same or similar text themes from the multiple texts to be stored, and perform primary association storage on the multiple first texts to be stored according to the theme association relationship of the text themes, perform sentence-by-sentence AI interpretation on the multiple first texts to be stored through the AI ​​text interpretation module to obtain a text sentence interpretation result, and query the multiple first texts to be stored according to the text sentence interpretation result for target sentences with the same or similar sentence meanings and multiple second texts to be stored containing the target sentences, perform secondary association storage on the multiple second texts to be stored according to the sentence meaning association relationship of the target sentences, the association height of the secondary association storage is greater than the association height of the primary association storage, and perform word-by-word AI interpretation on the multiple second texts to be stored through the AI ​​text interpretation module , obtain the text vocabulary interpretation result, and according to the text vocabulary interpretation result, search for the target keyword with the same or similar meaning as the designated keyword in the plurality of second to-be-stored texts, and the plurality of third to-be-stored texts containing the target keyword, and perform three-level association storage on the plurality of third to-be-stored texts through the semantic association relationship between the designated keyword and the target keyword, the association height of the three-level association storage is greater than the association height of the two-level association storage, in this scheme, by firstly associating through the similar theme of the whole interpreted article, and then further associating through the similar sentences in the more subdivided interpreted article, and then further associating through the words with similar meanings to the keywords that are more finely interpreted in the sentences, a gradually fine association process is realized from the general aspect, i.e. the overall theme of the article, to the specific aspect, i.e. the sentences and even more specific key words, so as to avoid the problem that the association basis is too one-sided due to only associating with the keywords, resulting in the occurrence of the problem of low accuracy of the association relationship, and the association basis is made more comprehensive through the theme, sentence meaning and word meaning, not just limited to the keywords, thereby improving the accuracy of the association relationship, and solving the technical problem of low accuracy of the association relationship for the keywords.

[0048] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0050] Figure 1 A flowchart of a keyword-based data association storage method provided in an embodiment of the present application;

[0051] Figure 2 An example of a propagation path gradient map in the keyword-based data association storage method provided in an embodiment of the present application;

[0052] Figure 3 A schematic diagram of the structure of a keyword-based data association storage device provided in an embodiment of the present application;

[0053] Figure 4 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution of the present application will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

[0055] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0056] The massive amounts of information collected from social media, news websites and other platforms are usually stored in databases. Currently, the associations between these massive amounts of information stored in databases are usually also associated with keywords. However, the accuracy of the associations stored with keywords is low.

[0057] Based on this, an embodiment of the present application provides a keyword-based data association storage method and device, through which the technical problem of low accuracy of association relationships for keywords can be solved, and the method can be applied to database management systems.

[0058] The embodiments of the present invention are further described below in conjunction with the accompanying drawings.

[0059] Figure 1 A flowchart of a keyword-based data association storage method provided in an embodiment of the present application. Figure 1 As shown, the method includes:

[0060] Step S110, obtaining multiple texts to be stored.

[0061] As a possible implementation, the text to be stored may be in various text forms, for example, the text to be stored may be a published opinion article, academic paper, research result, etc. In this step, at least two articles may be obtained.

[0062] Step S120, performing AI interpretation on the entire text of each of the multiple texts to be stored through the AI ​​text interpretation module to obtain the entire text interpretation result, and generating the text theme corresponding to the text to be stored according to the entire text interpretation result.

[0063] In an optional implementation, for each article, an artificial intelligence (AI) text interpretation is performed on the entire article to obtain an interpretation result of the entire article, and then a general theme of the article is generated based on the interpretation result of the entire article.

[0064] For the specific implementation process of this step, illustratively, data preparation is performed first, that is, first ensure that all the texts to be stored have been collected and are ready for processing. Then text preprocessing is performed, that is, each text to be stored is preprocessed, including but not limited to removing punctuation marks, stop word filtering, stem extraction and other steps, so as to improve the accuracy of subsequent AI interpretation. After that, AI text interpretation is performed, that is, each text to be stored is processed using the AI ​​text interpretation module. This step usually involves natural language processing techniques, such as sentiment analysis, topic modeling, etc., to obtain key information for each article. The specific steps can be described as follows: Load the AI ​​text interpretation module: Select a suitable AI model, such as a deep learning-based model (such as BERT, GPT series) or other machine learning models. Interpret each text: Input the text into the AI ​​model one by one, and output the interpretation results of each text.

[0065] Then, the interpretation results are summarized, that is, the interpretation results of each text are summarized to form a structured data set. These results may contain information such as keywords and topic distribution. Then the text theme is generated, that is, the text theme is generated according to the interpretation results of each text. This step can be completed in the following ways: keyword extraction: extract the most relevant keywords from the interpretation results as the text theme; theme summary: use automatic summary technology to generate a short theme summary; manual review: in some cases, manual intervention may be required to review whether the generated text theme is accurate; result verification: verify the generated text theme to ensure that they accurately reflect the main content of each text to be stored.

[0066] As an example, assuming that there is already a pre-trained AI text interpretation module, the specific steps can be implemented through the following code example:

[0067] # Load the AI ​​text interpretation module

[0068] ai_interpreter = load_ai_interpreter_module()

[0069] # List of texts to be processed

[0070] texts_to_store = ["Text A", "Text B", "Text C"]

[0071] # Dictionary to store interpretation results and text main points

[0072] interpretation_results = {}

[0073] text_summaries = {}

[0074] # Traverse each text

[0075] for text in texts_to_store:

[0076] # Text preprocessing

[0077] preprocessed_text = preprocess_text(text)

[0078] # AI Text Interpretation

[0079] interpretation_result = ai_interpreter(preprocessed_text)

[0080] #Store the interpretation results

[0081] interpretation_results[text] = interpretation_result

[0082] # Generate text subject

[0083] summary = generate_summary(interpretation_result)

[0084] text_summaries[text] = summary

[0085] # Output results

[0086] print("Interpretation results:", interpretation_results)

[0087] print("Text Summary:", text_summaries)

[0088] In the process corresponding to the above code, preprocess_text is a text preprocessing function, load_ai_interpreter_module is a function for loading the AI ​​text interpretation module, and generate_summary is a function for generating the main idea of ​​the text based on the interpretation results.

[0089] Step S130 , determining a plurality of first texts to be stored having the same or similar text subjects from the plurality of texts to be stored, and performing primary associative storage on the plurality of first texts to be stored based on the subject association relationship of the text subjects.

[0090] As an optional implementation, after the general theme of each article is identified, articles with the same or similar themes can be low-level associated, which is also convenient for subsequent searches for articles with the same or similar themes.

[0091] For the specific implementation process of this step, illustratively, the main theme of the text is determined first, that is, the main theme of each text to be stored needs to be determined first, which mainly includes using the AI ​​text interpretation module to process the text and extract the main theme information from it. Next, it is necessary to compare the main theme similarity between different texts. Common methods include using cosine similarity, Jaccard similarity, etc. to measure the similarity between two text main theme vectors. Then, determine the texts with the same or similar themes, and determine which texts have the same or very similar themes based on the results of the main theme similarity calculation. Here, it can be determined which texts should be classified together based on a certain threshold. Finally, a first-level associated storage is performed, and for those texts with the same or similar themes, they are stored in a first-level associated manner. This means that in the storage system, these texts will be marked as having a certain relevance, and all related texts can be retrieved by querying a certain theme.

[0092] As an example, assume that there are the following three texts to be stored: Text A: "How to Learn Programming Effectively"; Text B: "Sharing Programming Learning Skills"; Text C: "Efficient Reading Guide".

[0093] The first step is to determine the main theme of the text. Use the AI ​​text interpretation module to process the above text and obtain the following main themes: Text A theme: "Programming learning methods"; Text B theme: "Programming learning skills"; Text C theme: "Efficient reading skills".

[0094] The second step is to calculate the subject similarity. Suppose we use cosine similarity to calculate the similarity between text subjects: the similarity between text A and text B = 0.9; the similarity between text A and text C = 0.3; the similarity between text B and text C = 0.2.

[0095] The third step is to determine the texts with the same or similar themes and set a similarity threshold, for example 0.8, so that only the thematic similarity between text A and text B reaches this standard.

[0096] The fourth step is to perform primary associative storage: marking text A and text B as texts with the same or similar subject matter, and establishing a primary associative relationship between them in the storage system.

[0097] Step S140, using the AI ​​text interpretation module to perform AI interpretation sentence by sentence on the multiple first texts to be stored, obtain text sentence interpretation results, and search for target sentences with the same or similar meanings in the multiple first texts to be stored, as well as multiple second texts to be stored containing the target sentences based on the text sentence interpretation results.

[0098] For each article with the same or similar theme as mentioned above, AI text interpretation is performed on each sentence in the entire article to obtain the sentence interpretation results of each sentence in the article. Then, based on the sentence interpretation results of each article, target sentences with the same or similar meanings are determined in each article, that is, target sentences with the same or similar meanings are searched sentence by sentence from each article.

[0099] As an example, the AI ​​text interpretation module is first used. The AI ​​text interpretation module needs to be used to perform AI interpretation sentence by sentence on multiple first texts to be stored, so as to obtain the sentence interpretation results of each sentence. Then, a query with the same or similar sentence meaning is performed, that is, based on these interpretation results, the target sentence with the same or similar sentence meaning is queried. This usually involves comparing the semantic similarity of the sentences. Next, the text containing the target sentence is screened, that is, multiple second texts to be stored containing the target sentence are screened out.

[0100] For example, suppose there are three texts to be stored: Text A: "How to effectively learn programming"; Text B: "Sharing programming learning skills"; Text C: "Efficient reading guide". The first step is to perform AI text interpretation, and use the AI ​​text interpretation module to process the above texts to obtain the following interpretation results:

[0101] Text A interpretation results: "How to effectively learn programming" → "Introducing effective programming learning methods"; Text B interpretation results: "Sharing programming learning skills" → "Sharing programming learning skills"; Text C interpretation results: "Efficient reading guide" → "Providing a guide to efficient reading". Since these texts are titles, it can be assumed that they have only one sentence. If each text has multiple sentences, each sentence will go through a similar interpretation process.

[0102] The second step is to query for the same or similar sentence meanings: suppose you want to query content related to "effective learning methods", that is, to find sentences with similar sentence meanings. The next third step is to filter the text containing the target sentence: "How to effectively learn programming" in text A has a high similarity with "effective learning methods". "Sharing programming learning skills" in text B is also associated with "effective learning methods" because skills often mean effective methods. Text C has no direct correlation with "effective learning methods". Therefore, the multiple second texts to be stored that are finally filtered out are text A and text B.

[0103] Sentences with similarity to the query keyword "effective learning methods" that are higher than a given threshold will be found, and the texts in which they are located will be marked as relevant texts. This method can effectively organize and manage texts with the same or similar meanings, making it easier for users to search and browse based on specific meanings.

[0104] Step S150 , performing secondary association storage on the plurality of second texts to be stored based on the sentence meaning association relationship of the target sentence.

[0105] Among them, the association height of the secondary association storage is greater than the association height of the primary association storage. As a possible implementation method, after identifying the target sentences with the same or similar meanings in each article with the same or similar subject, these multiple articles with at least similar subject and sentence meaning can be moderately associated, which is also convenient for subsequent searches for articles with similar subject and sentence meaning.

[0106] As an optional implementation method, first, determine the target sentences: all target sentences with the same or similar sentence meanings found in the first step. Then, analyze the sentence meaning association relationship: analyze whether there is a certain association relationship (such as causal relationship, progressive relationship, contrast relationship, etc.) between these target sentences. Then establish a secondary association: based on the sentence meaning association relationship, establish a deeper level of association between the second texts to be stored. Next, implement secondary association storage: organize and store the second texts to be stored according to this association relationship.

[0107] Exemplarily, assume that there are already the following second texts to be stored: Text D: "How to learn programming effectively"; Text E: "Sharing programming learning skills"; Text F: "Ten tips to improve programming efficiency". The first step is to determine the target sentence. The known sentences related to "effective learning methods" include: "How to learn programming effectively" in Text D; "Sharing programming learning skills" in Text E. The second step is to analyze the sentence association relationship: Text D: "How to learn programming effectively" → provides specific methods on how to learn programming; Text E: "Sharing programming learning skills" → shares some programming learning skills; Although Text F is not directly marked as related, its title "Ten tips to improve programming efficiency" may also contain content related to effective learning methods. The content of Text F can be further analyzed to determine whether it is also related to "effective learning methods".

[0108] The third step is to establish a secondary association: the association between text D and text E is direct because they both provide information related to learning programming effectively. Although text F does not directly mention learning methods, the tips for improving programming efficiency can also be considered a learning method. Therefore, text F can be indirectly associated with text D and text E.

[0109] The final fourth step is to implement a secondary associative storage: In practice, you can create an index or database structure to represent these associations. For example, you can create an association graph where nodes represent texts and edges represent the associations between texts.

[0110] For example, assuming that there is already a storage and indexing system for the above texts, the following are specific implementation steps: define the type of association: such as "directly related", "indirectly related", etc.; establish an associated index: according to the type of association, an index is established between texts; query the associated text: the user can query the related text according to the keyword or sentence meaning. Through this method, the second text to be stored can be better managed and organized, so that the user can query the related text collection according to the specific sentence meaning, thereby improving the efficiency and accuracy of information retrieval.

[0111] Step S160, using the AI ​​text interpretation module to perform AI interpretation word by word on the multiple second texts to be stored, obtain text vocabulary interpretation results, and search for target keywords with the same or similar meanings as the specified keywords in the multiple second texts to be stored, as well as multiple third texts to be stored containing the target keywords based on the text vocabulary interpretation results.

[0112] In a possible implementation, for each article having similarities in the above-mentioned main themes and sentence meanings, AI text interpretation is performed on each word in the entire article to obtain the vocabulary and sentence interpretation results of each word in the article, and then based on the vocabulary interpretation results of each article, it is queried in each article whether there is a target keyword with the same or similar meaning as the specified keyword, that is, the target keyword with the same or similar meaning as the specified keyword is searched word by word from each article.

[0113] In actual applications, the above-mentioned designated keywords and target keywords can be the words used by a single media when making and using indexes. Keyword search is a method of network search indexing. For example, it can be the specific names of products, services, companies, etc. that visitors want to know about.

[0114] Step S170 , performing three-level associative storage on the plurality of third texts to be stored by using the semantic association relationship between the designated keywords and the target keywords.

[0115] Among them, the correlation height of the third-level correlation storage is greater than the correlation height of the second-level correlation storage. In a possible implementation, after identifying the target keyword with the same or similar meaning as the designated keyword in each article with at least similarities in both subject and sentence meaning, the multiple articles with at least similar relations in subject, sentence meaning and word meaning can be highly correlated, which is also convenient for subsequent querying of articles with similar subject, sentence meaning and word meaning according to the designated keyword or target keyword.

[0116] In the embodiment of the present application, association is first performed through similar themes throughout the interpreted article, and then further association is performed through similar sentences in the more subdivided interpreted article, and then further association is performed through words with similar meanings to the keywords in the more finely interpreted sentences, thereby achieving a gradually fine association process from the general aspect, i.e., the overall theme of the article, to the specific aspect, i.e., the sentences and even more specific keywords, thereby avoiding the problem of too one-sided basis of association caused by association only for keywords, resulting in low accuracy of the association relationship. The basis of association is made more comprehensive through the theme, sentence meaning and word meaning, and is not limited to keywords, thereby improving the accuracy of the association relationship and optimizing the association storage results of the data.

[0117] The above steps are described in detail below.

[0118] In some embodiments, after step S160, the method may further include the following steps:

[0119] Determining a text to be analyzed and a first network publishing time of the text to be analyzed among a plurality of texts to be stored;

[0120] Based on the main idea of ​​the text, the meaning of the target sentence and the meaning of the target keyword, the source text corresponding to the text to be analyzed is searched through several second network publishing times earlier than the first network publishing time, and the text with the earliest second network publishing time among the multiple source texts is determined as the root text;

[0121] Determine the publication objects of the tracing text and the root text, and generate a propagation path gradient map of the text to be analyzed based on the publication objects of the tracing text and the root text, the publication methods of the tracing text and the root text, the second network publication time corresponding to the tracing text, and the second network publication time corresponding to the root text.

[0122] As an optional implementation, the source of articles and keywords can be traced. For example, some articles are generated based on other earlier articles, such as reprinted articles, re-edited and republished original articles, etc. The source article of the current article can be traced back through the article publication time, and the articles corresponding to the earlier publication time of the same or similar article content (query for the article's main idea, main meaning, sentence meaning, and multiple articles with the same meaning as the keyword) are used as the tracing articles. By continuously tracing back to the articles with earlier publication time, the root object that first knew and published the article is found, thereby finding the gradient formation of the distribution and propagation of the article, as well as the path of propagation and divergence. The propagation path gradient graph can be shown as follows Figure 2 As shown in the figure, the propagation path gradient map can accurately show the exact propagation process of the article.

[0123] In some embodiments, the publishing methods include text reprint publishing, text editing publishing, and text link publishing; the above-mentioned step of generating a propagation path gradient map of the text to be analyzed according to the publishing objects of the traceable text and the root text, the publishing methods of the traceable text and the root text, the second network publishing time corresponding to the traceable text, and the second network publishing time corresponding to the root text may include the following steps:

[0124] According to the publication objects of the traceable text and the root text, the second network publication time corresponding to the traceable text, and the second network publication time corresponding to the root text, a preliminary propagation path gradient map of the text to be analyzed is generated; wherein the gradient first connection relationship between the text to be analyzed, the traceable text and the root text in the preliminary propagation path gradient map is a gradient association relationship based on the time sequence;

[0125] According to the text reprint relationship, text link relationship and text editing duplication between the text to be analyzed, the traceable text and the root text, determine the communication mode association relationship between the text to be analyzed, the traceable text and the root text; wherein the communication mode association relationship includes a reprint association relationship, a link association relationship and a copy association relationship;

[0126] The preliminary propagation pathway gradient map is optimized based on the correlation relationship of propagation modes to generate a final propagation pathway gradient map of the text to be analyzed; wherein the gradient second connection relationship between the text to be analyzed, the traceable text and the root text in the final propagation pathway gradient map is a gradient correlation relationship based on propagation modes.

[0127] In the embodiment of the present application, a propagation path gradient diagram including specific propagation methods such as text reprinting and publishing, text editing and publishing, and text link publishing can more comprehensively display the precise propagation process of the article.

[0128] In some embodiments, after the step of performing AI interpretation on the entire text of each of the multiple texts to be stored by the AI ​​text interpretation module to obtain the result of the entire text interpretation, the method may further include the following steps:

[0129] Identify the main idea of ​​the text to be stored according to the result of the whole text interpretation, and generate a summary and sub-directory corresponding to the text to be stored based on the main idea of ​​the text and the main idea of ​​the text;

[0130] A general directory corresponding to multiple texts to be stored is generated based on the subdirectories corresponding to the multiple texts to be stored, and a query graph of the texts to be stored is generated based on the abstract, the general directory and the texts to be stored; wherein the query graph contains a mapping relationship between the abstract, the text to be stored and each subdirectory in the general directory.

[0131] In the embodiment of the present application, by including a query chart of the mapping relationship between the text to be stored and its abstract and sub-directories, the user can more accurately search for multiple articles with similar overall meaning, general ideas, and theme.

[0132] In some embodiments, the query graph further includes a mapping relationship between the text to be stored and the domain name of the text publication, and the original text page corresponding to the text in the network; after the above steps of generating a general directory corresponding to multiple texts to be stored based on the subdirectories corresponding to the multiple texts to be stored, and generating a query graph of the text to be stored based on the abstract, the general directory and the text to be stored, the method may further include the following steps:

[0133] Determine a text to be searched from multiple texts to be stored, and display a general directory and multiple subdirectories in the general directory in a graphical user interface;

[0134] In response to a first selection instruction for a target subdirectory among the multiple subdirectories, determining a target subdirectory corresponding to the to-be-queried text in the general directory according to the first selection instruction;

[0135] Based on the mapping relationship between the target subdirectory through the abstract and each subdirectory in the general directory, determine and display the abstract corresponding to the text to be queried in the graphical user interface;

[0136] In response to a second selection instruction for a target abstract among the multiple abstracts corresponding to the text to be queried, determining a target text to be stored corresponding to the text to be queried according to the mapping relationship between the abstract and the text to be stored;

[0137] According to the mapping relationship between the text to be stored and the text publishing domain name, as well as the original text page corresponding to the text on the network, a reverse query is performed from the query graph to find the target text publishing domain name corresponding to the text to be queried, as well as the target original text page corresponding to the text to be queried on the network.

[0138] By including a query chart that includes the mapping relationship between the text to be stored, its abstract and sub-directories, users can more easily identify the abstract after easily identifying the directory, allowing users to more easily search for multiple articles with similar general meanings, general ideas and themes. It also enables reverse query of the original pages, networks, domain names, etc. corresponding to these articles from the generated chart, making the query results more comprehensive.

[0139] In some embodiments, after the above step of performing three-level associative storage of the plurality of third texts to be stored by specifying keywords and the semantic association relationship of the target keywords, the method may further include the following steps:

[0140] Based on the main idea of ​​the text, the overall meaning, the meaning of the target sentence and the meaning of the target keyword, the advanced association relationship between the multiple third texts to be stored is determined, and the multiple third texts to be stored are stored in a four-level association through the advanced association relationship.

[0141] Among them, the advanced association relationship includes any one or more of the text corroboration relationship, the text extension relationship, and the theme sublimation relationship. As an example, for two articles with related association points, one article can be the theme extended by the other article, the two articles can be the relationship in which the themes can corroborate each other, and the two articles can also be controversial, contradictory and divergent themes, making the advanced association relationship stored for multiple third texts to be stored more refined.

[0142] In some embodiments, after the above step of performing three-level associative storage of the plurality of third texts to be stored by specifying keywords and the semantic association relationship of the target keywords, the method may further include the following steps:

[0143] Determining the actual occurrence frequency of the same target keyword in each of the plurality of third texts to be stored;

[0144] If the actual occurrence frequency is lower than the preset occurrence frequency threshold, the plurality of third texts to be stored are stored in a moderately associated manner;

[0145] If the actual occurrence frequency is higher than or equal to the preset occurrence frequency threshold, the plurality of third texts to be stored are stored with high association.

[0146] As an optional implementation, the two articles can be associated based on the frequency of the same keywords in the two articles. Exemplarily, for the frequency of the same keywords: if it is lower than the first threshold (such as 5 times) and higher than the second threshold (such as 2 times), such as 3 times, then there is a moderate association; if the frequency of the same keywords in the two articles is lower than the second threshold (such as 2 times) such as 1 time, then there is a low association; if the frequency of the same keywords in the two articles is higher than the first threshold (such as 5 times), such as 7 times, then there is a high association. By associating based on the frequency of keywords in the two articles, it is possible to accurately subdivide the various degrees of association of the association relationship.

[0147] Figure 3 A schematic diagram of a data association storage device based on keywords is provided. Figure 3 As shown, the keyword-based data association storage device 300 includes:

[0148] An acquisition module 301 is used to acquire multiple texts to be stored;

[0149] The first generating module 302 is used to perform AI interpretation on the entire text of each of the multiple texts to be stored through the AI ​​text interpretation module to obtain the entire text interpretation result, and generate the text theme corresponding to the text to be stored according to the entire text interpretation result;

[0150] A first determination module 303 is used to determine a plurality of first texts to be stored having the same or similar text themes from the plurality of texts to be stored, and to perform primary association storage on the plurality of first texts to be stored based on the subject association relationship of the text themes;

[0151] The first query module 304 is used to perform AI interpretation sentence by sentence on the plurality of first texts to be stored through the AI ​​text interpretation module to obtain text sentence interpretation results, and to search for target sentences with the same or similar sentence meanings in the plurality of first texts to be stored, and a plurality of second texts to be stored containing the target sentences according to the text sentence interpretation results;

[0152] The first storage module 305 is used to perform secondary association storage on the plurality of second texts to be stored according to the sentence meaning association relationship of the target sentence; the association height of the secondary association storage is greater than the association height of the primary association storage;

[0153] The second query module 306 is used to perform AI interpretation of the plurality of second to-be-stored texts word by word through the AI ​​text interpretation module to obtain text vocabulary interpretation results, and search the plurality of second to-be-stored texts for target keywords having the same or similar meanings as the designated keywords, and at least two third to-be-stored texts containing the target keywords according to the text vocabulary interpretation results;

[0154] The second storage module 307 is used to perform three-level associative storage on the at least two third texts to be stored through the semantic association relationship between the designated keyword and the target keyword; the association height of the three-level associative storage is greater than the association height of the two-level associative storage.

[0155] In some embodiments, the apparatus further comprises:

[0156] A second determination module is used to determine a text to be analyzed and a first network publishing time of the text to be analyzed in the plurality of second texts to be stored, after searching for a target keyword having the same or similar meaning as the designated keyword in the plurality of second texts to be stored according to the text vocabulary interpretation result;

[0157] A third query module is used to query the source text corresponding to the text to be analyzed through a plurality of second network publishing times earlier than the first network publishing time based on the text subject, the sentence meaning of the target sentence and the word meaning of the target keyword, and determine the text with the earliest second network publishing time among the plurality of source texts as the root text;

[0158] The second generating module is used to determine the publishing objects of the tracing text and the root text, and generate a propagation path gradient map of the text to be analyzed according to the publishing objects of the tracing text and the root text, the publishing methods of the tracing text and the root text, the second network publishing time corresponding to the tracing text, and the second network publishing time corresponding to the root text.

[0159] In some embodiments, the publishing methods include text reprint publishing, text editing publishing, and text link publishing; the generation module is specifically used to:

[0160] According to the publication objects of the traceable text and the root text, the second network publication time corresponding to the traceable text, and the second network publication time corresponding to the root text, a preliminary propagation path gradient map of the text to be analyzed is generated; wherein the gradient first connection relationship between the text to be analyzed, the traceable text and the root text in the preliminary propagation path gradient map is a gradient association relationship based on the chronological order;

[0161] Determine the dissemination mode association relationship among the text to be analyzed, the source text and the root text according to the text reprint relationship, text link relationship and text editing duplication among the text to be analyzed, the source text and the root text; wherein the dissemination mode association relationship includes a reprint association relationship, a link association relationship and a copy association relationship;

[0162] The preliminary propagation pathway gradient map is optimized based on the propagation mode association relationship to generate a final propagation pathway gradient map of the text to be analyzed; wherein the gradient second connection relationship between the text to be analyzed, the source text and the root text in the final propagation pathway gradient map is a gradient association relationship based on the propagation mode.

[0163] In some embodiments, the apparatus further comprises:

[0164] The recognition module is used to perform AI interpretation on the entire text of each of the multiple texts to be stored through the AI ​​text interpretation module, and after obtaining the result of the entire text interpretation, identify the main idea of ​​the text to be stored according to the result of the entire text interpretation, and generate a summary and a subdirectory corresponding to the text to be stored based on the main idea of ​​the text and the main idea;

[0165] The third generation module is used to generate a general directory corresponding to the multiple texts to be stored based on the sub-directories corresponding to the multiple texts to be stored, and generate a query chart for the texts to be stored based on the abstract, the general directory and the texts to be stored; wherein the query chart contains the mapping relationship between the abstract, the text to be stored and each of the sub-directories in the general directory.

[0166] In some embodiments, the query graph further includes a mapping relationship between the text to be stored and the domain name where the text is published, and the original page corresponding to the text in the network; the device further includes:

[0167] A first display module is used to determine a text to be queried from the plurality of texts to be stored, and display the general catalog and the plurality of sub-catalogs in the general catalog in a graphical user interface after generating a general catalog corresponding to the plurality of texts to be stored based on the sub-catalogs corresponding to the plurality of texts to be stored, and generating a query chart of the texts to be stored according to the abstract, the general catalog and the texts to be stored;

[0168] A third determining module, configured to respond to a first selection instruction for a target subdirectory among the plurality of subdirectories and determine, in the general directory, a target subdirectory corresponding to the text to be queried according to the first selection instruction;

[0169] A second display module, configured to determine and display in a graphical user interface the abstract corresponding to the text to be queried based on the mapping relationship between the target subdirectory through the abstract and each of the subdirectories in the general directory;

[0170] A fourth determination module is used to determine, in response to a second selection instruction for a target abstract among the multiple abstracts corresponding to the text to be queried, a target text to be stored corresponding to the text to be queried according to the target abstract through a mapping relationship between the abstract and the text to be stored;

[0171] The fourth query module is used to reversely query the target text publishing domain name corresponding to the text to be queried and the target original page corresponding to the text to be queried in the network from the query chart based on the mapping relationship between the text to be stored and the text publishing domain name, and the original page corresponding to the text in the network.

[0172] In some embodiments, the apparatus further comprises:

[0173] The third storage module is used to determine the advanced association relationship between the multiple third texts to be stored based on the main theme of the text, the general idea, the sentence meaning of the target sentence and the meaning of the target keyword after the multiple third texts to be stored are stored in a third-level association manner through the semantic association relationship between the designated keywords and the target keywords, and to perform a fourth-level association storage on the multiple third texts to be stored through the advanced association relationship; wherein the advanced association relationship includes any one or more of a text corroboration relationship, a text extension relationship, and a main theme sublimation relationship.

[0174] In some embodiments, the apparatus further comprises:

[0175] A fifth determination module, for determining the actual occurrence frequency of the same target keyword in each of the plurality of third texts to be stored, after performing three-level association storage on the plurality of third texts to be stored through the semantic association relationship between the designated keyword and the target keyword;

[0176] A fourth storage module, configured to store the plurality of third texts to be stored in a moderately associated manner if the actual occurrence frequency is lower than a preset occurrence frequency threshold;

[0177] The fifth storage module is used to store the plurality of third texts to be stored with high correlation if the actual occurrence frequency is higher than or equal to the preset occurrence frequency threshold.

[0178] The keyword-based data association storage device provided in the embodiment of the present application has the same technical features as the keyword-based data association storage method provided in the above embodiment, so it can also solve the same technical problems and achieve the same technical effects.

[0179] An electronic device provided in an embodiment of the present application is Figure 4 As shown, the electronic device 400 includes a processor 402 and a memory 401, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the method provided in the above embodiment when executing the computer program.

[0180] See also Figure 4 The electronic device further includes: a bus 403 and a communication interface 404, a processor 402, a communication interface 404 and a memory 401 are connected via the bus 403; the processor 402 is used to execute an executable module stored in the memory 401, such as a computer program.

[0181] The memory 401 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk storage. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 404 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0182] The bus 403 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0183] Among them, the memory 401 is used to store programs, and the processor 402 executes the program after receiving the execution instruction. The method executed by the device defined by the process disclosed in any embodiment of the present application can be applied to the processor 402 or implemented by the processor 402.

[0184] The processor 402 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 402. The above processor 402 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor to execute, or the hardware and software modules in the decoding processor are combined to execute. The software module may be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 401, and the processor 402 reads the information in the memory 401 and completes the steps of the above method in combination with its hardware.

[0185] Corresponding to the above-mentioned keyword-based data association storage method, an embodiment of the present application also provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to execute the steps of the above-mentioned keyword-based data association storage method.

[0186] The keyword-based data association storage device provided in the embodiment of the present application can be specific hardware on the device or software or firmware installed on the device. The device provided in the embodiment of the present application, its implementation principle and the technical effect produced are the same as those in the aforementioned method embodiment. For the sake of brief description, the parts not mentioned in the device embodiment can refer to the corresponding contents in the aforementioned method embodiment. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above can all refer to the corresponding processes in the aforementioned method embodiment, and will not be repeated here.

[0187] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0188] For another example, the flowchart and block diagram in the accompanying drawings show the possible architecture, function and operation of the device, method and computer program product according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of the boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0189] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0190] In addition, each functional unit in the embodiments provided in the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0191] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the keyword-based data association storage method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store program codes.

[0192] It should be noted that similar numbers and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are only used to distinguish the description and are not to be understood as indicating or implying relative importance.

[0193] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solution of the present application, rather than to limit it. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the aforementioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solution recorded in the aforementioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solution deviate from the scope of the technical solution of the embodiment of the present application. They should all be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A keyword-based data association storage method, characterized in that: The method comprises: Get multiple texts to be stored; Performing AI interpretation on the entire text of each of the multiple texts to be stored by the AI ​​text interpretation module to obtain an interpretation result of the entire text, and generating a text theme corresponding to the text to be stored according to the interpretation result of the entire text; Determining a plurality of first texts to be stored having the same or similar text themes from the plurality of texts to be stored, and performing primary associative storage on the plurality of first texts to be stored based on the subject association relationship of the text themes; The AI ​​text interpretation module performs AI interpretation sentence by sentence on the plurality of first texts to be stored to obtain text sentence interpretation results, and searches for target sentences with the same or similar meanings in the plurality of first texts to be stored according to the text sentence interpretation results, as well as a plurality of second texts to be stored that contain the target sentences; The plurality of second texts to be stored are stored in a secondary association manner according to the sentence meaning association relationship of the target sentence; the association height of the secondary association storage is greater than the association height of the primary association storage; The AI ​​text interpretation module performs AI interpretation of the plurality of second texts to be stored word by word to obtain text vocabulary interpretation results, and searches the plurality of second texts to be stored for target keywords having the same or similar meanings as the designated keywords, and a plurality of third texts to be stored containing the target keywords according to the text vocabulary interpretation results; The plurality of third texts to be stored are stored in a three-level associative manner through the semantic association relationship between the designated keyword and the target keyword; the association height of the three-level associative storage is greater than the association height of the two-level associative storage.

2. The method according to claim 1, characterized in that After the step of searching the plurality of second texts to be stored for target keywords having the same or similar meanings as the designated keywords according to the text vocabulary interpretation results, the method further includes: Determining a text to be analyzed and a first network publishing time of the text to be analyzed from the plurality of texts to be stored; Based on the subject of the text, the meaning of the target sentence and the meaning of the target keyword, query the source text corresponding to the text to be analyzed through several second network publishing times earlier than the first network publishing time, and determine the text with the earliest second network publishing time among the multiple source texts as the root text; Determine the publication objects of the tracing text and the root text, and generate a propagation path gradient map of the text to be analyzed based on the publication objects of the tracing text and the root text, the publication methods of the tracing text and the root text, the second network publication time corresponding to the tracing text, and the second network publication time corresponding to the root text.

3. The method according to claim 2, characterized in that The publishing methods include text reprint publishing, text editing publishing and text link publishing; the step of generating the propagation path gradient map of the text to be analyzed according to the publishing objects of the tracing text and the root text, the publishing methods of the tracing text and the root text, the second network publishing time corresponding to the tracing text, and the second network publishing time corresponding to the root text, includes: According to the publication objects of the traceable text and the root text, the second network publication time corresponding to the traceable text, and the second network publication time corresponding to the root text, a preliminary propagation path gradient map of the text to be analyzed is generated; wherein the gradient first connection relationship between the text to be analyzed, the traceable text and the root text in the preliminary propagation path gradient map is a gradient association relationship based on the chronological order; Determine the dissemination mode association relationship among the text to be analyzed, the source text and the root text according to the text reprint relationship, text link relationship and text editing duplication among the text to be analyzed, the source text and the root text; wherein the dissemination mode association relationship includes a reprint association relationship, a link association relationship and a copy association relationship; The preliminary propagation pathway gradient map is optimized based on the propagation mode association relationship to generate a final propagation pathway gradient map of the text to be analyzed; wherein the gradient second connection relationship between the text to be analyzed, the source text and the root text in the final propagation pathway gradient map is a gradient association relationship based on the propagation mode.

4. The method according to claim 1, characterized in that: After the step of performing AI interpretation on the entire text of each of the plurality of texts to be stored by the AI ​​text interpretation module to obtain the result of the entire text interpretation, the method further includes: Identify the main idea of ​​the text to be stored according to the result of the whole text interpretation, and generate a summary and a subdirectory corresponding to the text to be stored based on the main idea of ​​the text and the main idea; A general directory corresponding to the multiple texts to be stored is generated based on the sub-directories corresponding to the multiple texts to be stored, and a query graph of the texts to be stored is generated based on the abstract, the general directory and the texts to be stored; wherein the query graph contains a mapping relationship between the abstract, the text to be stored and each of the sub-directories in the general directory.

5. The method according to claim 4, characterized in that The query graph also includes a mapping relationship between the text to be stored and the domain name where the text is published, and the original text page corresponding to the text in the network; After the step of generating a general catalogue corresponding to the plurality of texts to be stored based on the sub-catalogues corresponding to the plurality of texts to be stored, and generating a query graph of the texts to be stored according to the abstract, the general catalogue and the texts to be stored, the method further includes: Determine the text to be searched from the plurality of texts to be stored, and display the general catalog and the plurality of sub-catalogs in the general catalog in a graphical user interface; In response to a first selection instruction for a target subdirectory among the plurality of subdirectories, determining a target subdirectory corresponding to the to-be-queried text in the general directory according to the first selection instruction; Based on the mapping relationship between the target subdirectory through the abstract and each of the subdirectories in the general directory, determine and display the abstract corresponding to the text to be queried in a graphical user interface; In response to a second selection instruction for a target abstract among the multiple abstracts corresponding to the text to be queried, determining a target text to be stored corresponding to the text to be queried according to the target abstract through a mapping relationship between the abstract and the text to be stored; According to the mapping relationship between the text to be stored and the text publishing domain name, and the original text page corresponding to the text in the network, the target text publishing domain name corresponding to the text to be queried and the target original text page corresponding to the text to be queried in the network are reversely queried from the query graph.

6. The method according to claim 4, characterized in that After the step of performing three-level associative storage on the plurality of third texts to be stored through the semantic association relationship between the designated keywords and the target keywords, the method further includes: Based on the main theme of the text, the overall idea, the sentence meaning of the target sentence and the word meaning of the target keyword, the advanced association relationship between the multiple third texts to be stored is determined, and the multiple third texts to be stored are stored in a four-level association through the advanced association relationship; wherein the advanced association relationship includes any one or more of the text corroboration relationship, the text extension relationship, and the main theme sublimation relationship.

7. The method according to claim 1, characterized in that After the step of performing three-level associative storage on the plurality of third texts to be stored through the semantic association relationship between the designated keywords and the target keywords, the method further includes: Determining the actual occurrence frequency of the same target keyword in each of the plurality of third texts to be stored; If the actual occurrence frequency is lower than the preset occurrence frequency threshold, the plurality of third texts to be stored are stored in a moderately associated manner; If the actual occurrence frequency is higher than or equal to the preset occurrence frequency threshold, the plurality of third texts to be stored are stored with high association.

8. A data association storage device based on keywords, characterized in that: include: An acquisition module is used to acquire multiple texts to be stored; A generating module, configured to perform AI interpretation on the entire text of each of the plurality of texts to be stored through an AI text interpretation module to obtain an interpretation result of the entire text, and generate a text theme corresponding to the text to be stored according to the interpretation result of the entire text; A determination module, used for determining a plurality of first texts to be stored having the same or similar text subjects from the plurality of texts to be stored, and performing primary associative storage on the plurality of first texts to be stored according to the subject association relationship of the text subjects; A first query module is used to perform AI interpretation sentence by sentence on the plurality of first texts to be stored through the AI ​​text interpretation module to obtain text sentence interpretation results, and to search for target sentences with the same or similar sentence meanings in the plurality of first texts to be stored, and a plurality of second texts to be stored containing the target sentences according to the text sentence interpretation results; A first storage module is used to perform secondary association storage on the plurality of second texts to be stored according to the sentence meaning association relationship of the target sentence; the association height of the secondary association storage is greater than the association height of the primary association storage; A second query module is used to perform AI interpretation of the plurality of second texts to be stored word by word through the AI ​​text interpretation module to obtain text vocabulary interpretation results, and search for target keywords with the same or similar meanings as the specified keyword in the plurality of second texts to be stored according to the text vocabulary interpretation results, and at least two third texts to be stored containing the target keywords; A second storage module is used to perform three-level association storage on the at least two third texts to be stored through the semantic association relationship between the designated keyword and the target keyword; The association height of the third-level association storage is greater than the association height of the second-level association storage.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to execute the method according to any one of claims 1 to 7.

Images