Data processing method, device and electronic equipment based on artificial intelligence

Abstract

The present application provides a data processing method, device and electronic device based on artificial intelligence, the method comprising: performing category analysis on course catalog data received from a user input according to a preset course catalog system to obtain a category analysis result including a first category; establishing a multi-level folder according to a first course catalog determined from the preset course catalog system and matching the first category and the category analysis result; inputting the acquired course resource data into a preset data classification model; mapping and storing the course resource data according to a second category and the multi-level folder output by the preset data classification model; determining multiple data access paths matching the first query information input by the user from a target database; and adjusting the multi-level folder into a first folder and a second folder according to one or more second query matching rates that meet a preset period screened out from multiple first query matching rates corresponding to the multiple data access paths.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence data processing technology, and in particular to an artificial intelligence-based data processing method, device and electronic equipment. Background Art

[0002] With the rapid development of information technology, the field of education is welcoming new development opportunities. Especially driven by intelligent teaching, traditional teaching methods are gradually moving towards a new stage of digitalization and intelligence. Intelligent teaching not only optimizes the teaching quality, but also brings students a more personalized and flexible learning path, making the learning experience richer and more efficient.

[0003] In the related art, in a modern education environment, after class, students not only need to complete course assignments covering a variety of content categories, which include but are not limited to PPT, Word, Excel and other formats, but also correspond to electronic courseware materials for each course, such as handouts, videos, audios, etc. At present, these diverse course assignments and courseware materials are mainly managed manually by students, that is, the course-related teaching materials obtained through various channels are manually transferred to personal computers or cloud storage services. When a specific material is needed later, it is manually screened from a personal computer or cloud storage service to find the specific material.

[0004] However, as teaching materials continue to accumulate, more teaching materials may need to be stored, which may require more time to manually retrieve a specific teaching courseware from a huge amount of data, resulting in low data retrieval efficiency in related technologies. Summary of the invention

[0005] The present application provides an artificial intelligence-based data processing method, device, and electronic device for improving data retrieval efficiency.

[0006] In a first aspect, the present application provides an artificial intelligence-based data processing method, which is applied to the above-mentioned electronic device, and the method includes: when receiving course catalog data input by a user, classifying the course catalog data according to a preset course catalog system to obtain a class parsing result, wherein the course catalog data is a collection of first course information, and the class parsing result includes a first category of the first course information; determining a first course catalog matching the first category from the preset course catalog system, and establishing a multi-level folder based on the first course catalog and the class parsing result; obtaining course resource data, and inputting the course resource data into a preset data classification model; obtaining a second category output by the preset data classification model, and mapping and storing the course resource data according to the second category and the multi-level folder, wherein the second category is the preset data classification model for classifying the course resource data The first classification result is output after processing; receiving the first query information input by the user, and determining multiple data access paths matching the first query information from the target database; determining the first keyword from the first query information, and retrieving the user's preset query behavior from the target database, wherein the preset query behavior includes a query matching rate, and the query matching rate is the ratio of the number of keyword queries to the actual number of accesses to the data access path; determining the first query matching rate corresponding to each data access path included in the multiple data access paths according to the first keyword and the preset query behavior to obtain multiple first query matching rates, wherein the query matching rate includes multiple first query matching rates; screening out one or more second query matching rates that meet the preset period from the multiple first query matching rates, and adjusting the multi-level folders into first folders and second folders according to the one or more second query matching rates.

[0007] By adopting the above technical scheme, the course catalog data input by the received user is subjected to category analysis according to the preset course target system, and the category analysis result of the first category including the first course information is obtained, which can ensure that the course catalog data can be accurately classified. The first course catalog matching the first category is determined from the preset course catalog system, so that a multi-level folder can be established according to the first course catalog and the category analysis result, and the structured processing of the course catalog data is realized, so that the structure of the course catalog data is clearly hierarchical, so that the data under the specific hierarchical folder can be quickly accessed. The acquired course resource data is input into the preset data classification model, so that the classification of the course resource data is more accurate. The course resource data is mapped and stored according to the second category and the multi-level folder output by the acquired preset classification model, so that the mapping storage of the course resource data is more accurate, so that the data can be accessed more quickly, and the efficiency of data access is further improved. Multiple data access paths matching the first query information input by the received user are determined from the target database to comprehensively determine the data access path matching the first query information, so as to avoid missing the matching data access path and causing the data retrieval accuracy and retrieval efficiency to decrease. According to the first keyword determined from the first query information and the preset query behavior including the query matching rate (i.e. the ratio of the number of keyword queries to the actual number of accesses to the data access path) retrieved from the target database, the first query matching rate (i.e. the ratio of the number of queries of the first keyword to the actual number of accesses to each data access path included in the multiple data access paths) corresponding to each data access path included in the multiple data access paths is determined, which can accurately reflect the access frequency of each data access path corresponding to the first keyword. The method of the present application is adopted to establish a multi-level folder by matching the first course catalog and the category parsing result matched with the first category from the preset course catalog system, and then mapping and storing the course resource data according to the second category and the multi-level folder output by the preset data classification model, so as to realize the accurate mapping and storage of the course resource data and the multi-level folder, so as to realize the classification adjustment of the multi-level folder, so as to make the data access more efficient and orderly, thereby solving the technical problem of low data retrieval efficiency in the related technology, and achieving the technical effect of improving the data retrieval efficiency.

[0008] Optionally, the first folder is used to store the first course resource data under the data access path corresponding to a third query matching rate that is greater than a preset query matching rate, and the second folder is used to store the second course resource data under the data access path corresponding to a fourth query matching rate that is less than or equal to the preset query matching rate, one or more second query matching rates include the third query matching rate and the fourth query matching rate, and the course resource data include the first course resource data and the second course resource data.

[0009] By adopting the above-mentioned technical scheme, according to one or more second query matching rates that meet the preset period screened out from multiple first query matching rates, the multi-level folders are adjusted to a first folder for storing the first course resource data under the data access path corresponding to the third query matching rate that is greater than the preset query matching rate, and a second folder for storing the second course resource data under the data access path corresponding to the fourth query matching rate that is less than or equal to the preset query matching rate. This makes it possible to provide two single-level or multi-level folders (i.e., the first folder and the second folder) that meet the different query requirements of the user within the preset period, so as to achieve data access that can quickly respond to different requirements, further improving the data retrieval efficiency.

[0010] Optionally, after screening out one or more second query matching rates that meet a preset period from a plurality of first query matching rates, and adjusting the multi-level folders into first folders and second folders according to the one or more second query matching rates, the method further includes: receiving second query information input by a user, and determining a second keyword from the second query information; determining a fifth query matching rate corresponding to the second keyword from a preset query behavior; comparing the fifth query matching rate with the preset query matching rate to obtain a comparison result; and when the comparison result shows that the fifth query matching rate is greater than the preset query matching rate, performing data access processing on the first folder according to the fifth query matching rate to obtain a data access result. result; when the data access result shows that the first data access path corresponding to the fifth query matching rate exists in the first folder, the third course resource data under the first data access path in the first folder is sent to the user; or, when the data access result shows that the first data access path corresponding to the fifth query matching rate does not exist in the first folder, path query processing is performed on the second folder according to the first data access path to obtain a path query result; when the path query result shows that the first data access path exists in the second folder, the third course resource data under the first data access path in the second folder is sent to the user, and the third course resource data is stored in the first folder.

[0011] By adopting the above technical solution, the second keyword is determined from the second query information received from the user input, and then the fifth query matching rate corresponding to the second keyword is determined from the preset query behavior, so that the second keyword can be ensured to be highly corresponding to the fifth query matching rate, and the reliability of the second keyword and the fifth query matching rate is further guaranteed. The fifth query matching rate is compared with the preset query matching rate to obtain a comparison result, and the fifth query matching rate is determined by comparison to determine which query requirement of the user is met, thereby providing the user with a faster data retrieval response and a more accurate retrieval matching result. When the comparison result shows that the fifth query matching rate is greater than the preset query matching rate, it is determined that the query requirement of the fifth query matching rate is consistent with the first folder, so that the first folder can be processed for data access according to the fifth query matching rate to obtain a data access result, and then when the data access result shows that there is a first data access path corresponding to the fifth query matching rate in the first folder, the third course resource data under the first data access path in the first folder is sent to the user, so as to retrieve the third course resource data required by the user from a specific folder that meets the query requirement, effectively saving data retrieval time, and further improving data retrieval efficiency. When the data access result shows that the first data access path corresponding to the fifth query matching rate does not exist in the first folder, the second folder is subjected to path query processing according to the first data access path to obtain the path query result, thereby ensuring that the data retrieval will not be interrupted due to the non-existence of the first data access path in a folder (i.e., the first folder), but the query will continue in other related folders (i.e., the second folder), thereby ensuring the continuity of data retrieval. In addition, when the first data access path exists in the second folder, the third course resource data under the first data access path in the second folder will also be stored in the first folder, so that the user can retrieve the required data more quickly when performing similar queries in the future, thereby further improving the data retrieval efficiency.

[0012] Optionally, a first course directory matching the first category is determined from a preset course directory system, and a multi-level folder is established based on the first course directory and the category parsing results, specifically including: name-dividing the first course directory according to the first category to obtain a second course directory; establishing a first-level folder of the first category under the preset directory, wherein the preset directory is a preset second data access path, the second data access path is used to map and store course resource data, and the name of the first-level folder includes the first code corresponding to the first category; establishing a second-level folder of the second course directory under the first-level folder to obtain a multi-level folder, wherein the name of the second-level folder includes the second code corresponding to the second course directory.

[0013] By adopting the above technical solution, the first course catalog is divided into names according to the first category to obtain the second course catalog, which further clarifies the classification attribute of the second course catalog, and establishes corresponding folders through the second course catalog to ensure the accuracy of folder establishment. A first-level folder of the first category is established under the preset directory, and then a second-level folder of the second course catalog is established under the first-level folder to obtain a multi-level folder, which can realize the structured processing of course resource data, making the course resource data highly manageable and scalable.

[0014] Optionally, before obtaining course resource data and inputting the course resource data into a preset data classification model, the method further includes: obtaining sample data, wherein the sample data is a collection of multiple second course information; constructing a knowledge graph based on the sample data, wherein the knowledge graph includes multiple course catalog nodes and multiple course name nodes and multiple groups of associated attribute data, each course catalog node included in the multiple course catalog nodes is associated with a corresponding course name node included in the multiple course name nodes through a directed edge, and the multiple groups of associated attribute data are used to represent the associated attributes between the multiple course name nodes; constructing a graph neural network model based on the knowledge graph; connecting the graph neural network model to a classifier to obtain an initial data classification model; training the initial data classification model based on the sample data to obtain a preset data classification model.

[0015] By adopting the above technical solution, a knowledge graph including multiple course catalog nodes, multiple course name nodes and multiple groups of associated attribute data is constructed based on the acquired sample data, which can more accurately clarify the association between multiple course catalog nodes and multiple course name nodes and multiple course name nodes, laying the foundation for improving the classification accuracy of sample data. By constructing a graph neural network model through the knowledge graph, connecting the constructed graph neural network model with the classifier to obtain an initial data classification model, and then using the sample data to train the initial data classification model to obtain the required preset data classification model, the classification accuracy of sample data can be further improved.

[0016] Optionally, the initial data classification model is trained based on the sample data to obtain a preset data classification model, specifically including: inputting the sample data into the initial data classification model to obtain a second classification result; determining a sample label from the sample data; and adjusting the initial data classification model based on the second classification result and the sample label to obtain a preset data classification model.

[0017] By adopting the above technical solution, by inputting the sample data into the initial data classification model to obtain the second classification result, it can be ensured that the initial data classification model has been fully tested and verified before actual application. Sample labels are key indicators for evaluating model performance. By determining sample labels from sample data, an accurate reference standard can be provided for comparison with the classification results of the initial data classification model, making the adjustment of the initial data classification model more targeted, thereby obtaining the desired preset data classification model by improving the classification accuracy of the initial data classification model.

[0018] Optionally, obtain the second category output by the preset data classification model, and map and store the course resource data according to the second category and the multi-level folders, specifically including: matching a third course directory that meets the second category from the multi-level folders; determining a third folder corresponding to the third course directory, wherein the multi-level folder includes the third folder; mapping and storing the course resource data in the third folder, and assigning a version number to the mapped and stored course resource data.

[0019] By adopting the above technical solution, by matching the third course directory that meets the specific category (i.e., the second category) from the multi-level folders, it is possible to ensure that the course resource data matching the required category is quickly and accurately located, which greatly simplifies the process of manual search by users in a large number of folders, improves the retrieval efficiency of data, and matches according to clear categories, which can reduce the possibility of mismatching, and further improve the retrieval accuracy of data. Determining the third folder corresponding to the third course directory makes the storage structure of the course resource data clearer and more orderly. By associating the course resource data with a specific folder, not only can the classification and storage of the course resource data be refined, but also the course resource data can be conveniently managed and maintained, making the management of the course resource data more flexible and efficient. The course resource data is mapped and stored in a specific folder, and a version number is assigned to it, which ensures the integrity and traceability of the course resource data. By mapping the storage, it can ensure that a clear association relationship is established between the course resource data and its corresponding folder, which is convenient for subsequent search and use. By assigning a version number, the version of the course resource data can be managed and controlled, avoiding problems caused by version conflicts or incorrect use of old version data. The update history of the course resource data can also be recorded through the version number, which provides convenience for the maintenance and management of the course resource data.

[0020] In a second aspect, an embodiment of the present application provides an artificial intelligence-based data processing device, which includes: a parsing module, which is used to perform category parsing on the course catalog data according to a preset course catalog system upon receiving the course catalog data input by the user, so as to obtain a category parsing result, wherein the course catalog data is a collection of first course information, and the category parsing result includes a first category of the first course information; an establishment module, which is used to determine a first course catalog matching the first category from the preset course catalog system, and establish a multi-level folder based on the first course catalog and the category parsing result; an input module, which is used to obtain course resource data and input the course resource data into a preset data classification model; a mapping storage module, which is used to obtain a second category output by the preset data classification model, and map and store the course resource data according to the second category and the multi-level folder, wherein the second category is the first classification result output by the preset data classification model after the course resource data is classified.

[0021] In a third aspect, an embodiment of the present application provides an electronic device, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the electronic device to perform the method described in the first aspect and any possible implementation method of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, comprising instructions. When the instructions are executed on an electronic device, the electronic device executes the method described in the first aspect and any possible implementation manner of the first aspect.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0024] 1. A multi-level folder is established by matching the first course catalog and the category parsing results that match the first category from the preset course catalog system, and then mapping and storing the course resource data according to the second category and the multi-level folder output by the preset data classification model to achieve accurate mapping and storage of course resource data and multi-level folders, thereby achieving classification adjustment of multi-level folders, making data access more efficient and orderly, thereby solving the technical problem of low data retrieval efficiency in related technologies and achieving the technical effect of improving data retrieval efficiency.

[0025] 2. According to one or more second query matching rates that meet the preset period screened out from multiple first query matching rates, the multi-level folders are adjusted to include a first folder for storing the first course resource data under the data access path corresponding to the third query matching rate that is greater than the preset query matching rate, and a second folder for storing the second course resource data under the data access path corresponding to the fourth query matching rate that is less than or equal to the preset query matching rate, so that two single-level or multi-level folders (i.e., the first folder and the second folder) that meet the different query requirements of the user can be provided within the preset period, so as to realize data access that can quickly respond to different requirements, thereby further improving the data retrieval efficiency.

[0026] 3. When the data access result shows that the first data access path corresponding to the fifth query matching rate does not exist in the first folder, path query processing is performed on the second folder according to the first data access path to obtain a path query result, thereby ensuring that data retrieval will not be interrupted due to the non-existence of the first data access path in a folder (i.e., the first folder), but will continue to be queried in other related folders (i.e., the second folder), thereby ensuring the continuity of data retrieval. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flowchart of a data processing method based on artificial intelligence in an embodiment of the present application;

[0028] Figure 2 It is a structural block diagram of a data processing device based on artificial intelligence provided by an embodiment of the present application;

[0029] Figure 3 It is a schematic diagram of a physical device structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0030] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations comprising one or more listed items.

[0031] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0032] This application provides a data processing method based on artificial intelligence, see Figure 1 , Figure 1 This is a flow chart of a data processing method based on artificial intelligence in an embodiment of the present application, comprising the following steps:

[0033] Step S101, when receiving course catalog data input by a user, classifying and parsing the course catalog data according to a preset course catalog system to obtain a classifying and parsing result, wherein the course catalog data is a collection of first course information, and the classifying and parsing result includes a first class of the first course information;

[0034] In the above embodiment, the course catalog data refers to a collection of a series of course information, such as course name, course number, course description, etc. The preset course catalog system is a predefined course classification structure used to classify and parse the course catalog data. The preset course catalog system may contain multiple levels, each level representing a different course category.

[0035] In the above embodiment, a preset course catalog system can be constructed from multiple dimensions, including but not limited to disciplines, grades, semesters, difficulty levels, etc. It is assumed that a discipline catalog system has been preset, including but not limited to natural disciplines, experimental sciences, social disciplines, humanities, etc. The specific implementation steps are: after receiving the mathematics, physics, chemistry and other course information input by the user, the mathematics, physics, chemistry and other course information can be classified according to the preset discipline catalog system, that is, the course information input by the user is parsed, the name of each course is identified, and the name of each course is matched with the preset discipline classification system to find the most suitable first category. For mathematics and physics, since they are both disciplines that study natural laws and basic theories, they will be classified into the "natural discipline" category. For chemistry, although it also involves some basic theories and laws, it is more about verifying these theories or exploring unknown areas through experiments, so it will be classified into the "experimental discipline" category. In addition to subject classification, course information can also be classified more finely according to other dimensions. For example, course information can be divided into elementary school, junior high school, high school, etc. according to grade, and can be divided into upper semester, lower semester, etc. according to semester. Course information can also be divided into elementary, intermediate, advanced, etc. according to difficulty. Through classification in multiple dimensions, we can better meet the different needs of users and provide more precise services.

[0036] Step S102, determining a first course catalog that matches the first category from a preset course catalog system, and creating a multi-level folder according to the first course catalog and the category parsing result;

[0037] In the above embodiment, the first course catalog is a catalog item in the preset course catalog system that matches the first category parsed from the course catalog data. This catalog item may be a specific subject, grade, semester, etc. A multi-level folder is a directory structure with multiple levels. Data can be stored in different folders according to different categories or attributes to achieve orderly storage and access of data.

[0038] In the above embodiment, corresponding multi-level folders can be established according to the two first categories of "natural science" and "experimental science". For example, a first-level folder named "natural science" is established, and a second-level folder "mathematics" and a second-level folder "physics" are created under the first-level folder. The second-level folder "mathematics" can be used to store mathematics course resource data, and the second-level folder "physics" can be used to store physics course resource data. The classification can be further refined, that is, a three-level folder "algebra", a three-level folder "geometry", a three-level folder "calculus", etc. are established in the second-level folder "mathematics" to store different mathematics branch course resource data, and a three-level folder "mechanics", a three-level folder "electromagnetism", a three-level folder "quantum physics", etc. are established in the second-level folder "physics" to store different physics branch course resource data. It is also possible to establish a corresponding multi-level folder according to the first category of "mathematics". For example, a first-level folder named "mathematics" is established, and a second-level folder "applied mathematics" and a second-level folder "pure mathematics" are created under the first-level folder, and so on.

[0039] Step S103, obtaining course resource data, and inputting the course resource data into a preset data classification model;

[0040] In the above embodiment, the course resource data refers to various data resources related to educational and teaching activities, including but not limited to teaching materials, lesson plans, courseware, teaching videos, audio, pictures, test questions, student homework, learning progress information, etc. The preset data classification model is a built deep learning model, which is used to automatically classify the input course resource data more accurately.

[0041] In the above embodiment, it is assumed that the preset data classification model can not only identify traditional subject categories such as mathematics, physics, and chemistry, but also identify more subdivided subject areas or subcategories. For example, under the mathematics category, the preset data classification model can further distinguish subcategories such as algebra, geometry, calculus, and statistics. Similarly, under the physics category, the preset data classification model can distinguish subcategories such as mechanics, electromagnetism, optics, and quantum mechanics. When a teaching article on "The Application of Calculus in Physics" is obtained, the preset data classification model will not only recognize that this article belongs to the mathematics category, but will also further classify it into the "Calculus" subcategory. At the same time, since this article involves the application of physics, the preset data classification model may also associate it with relevant subcategories in the "Physics" category (for example, "Mechanics" or "Electromagnetism"). Such fine classification can more accurately understand the content and attributes of course resource data, thereby providing users with more personalized and accurate storage services and recommendation services, etc. For example, according to the user's subject preferences and learning progress, calculus or physics course resources that match their current learning stage and interest points can be recommended. In addition, by associating resource data from different subject categories and subcategories, we can also discover intersections and connections between different disciplines, providing support for interdisciplinary learning and research. For example, we can recommend interdisciplinary courses or research projects involving both mathematics and physics to help users broaden their horizons and deepen their understanding.

[0042] Step S104, obtaining the second category output by the preset data classification model, and mapping and storing the course resource data according to the second category and the multi-level folder, wherein the second category is the first classification result output after the preset data classification model classifies the course resource data;

[0043] In the above embodiment, the second category is a label or identifier for a more fine-grained classification of course resource data. For example, the second category may be "algebra", "geometry", "calculus", etc. The determination of the second category is based on the understanding and analysis of the course resource content by a preset data classification model.

[0044] In the above-described embodiment, the second category is utilized to carry out the mapping storage process of course resource data, and the management accuracy and management efficiency of course resource data can be improved. By storing the course resource data according to more refined classification, the retrievability and accessibility of course resource data can be enhanced, so that the user can find required teaching materials faster. In the present embodiment, assuming that a university course database contains a large number of teaching videos of different subjects and themes, the specific content in the teaching video can be automatically identified and classified into the second category by the preset data classification model, for example, the preset data classification model classifies the teaching video into calculus, if there is a storage path of "natural science / mathematics / calculus" in the multi-level folder, the video can be directly mapped to "natural science / mathematics / calculus" for storage, if there is only a storage path of "natural science / mathematics" in the multi-level folder, the calculus file can be automatically generated under the "natural science / mathematics" folder, and the video is mapped to "natural science / mathematics / calculus" for storage, so that the user can quickly access the video.

[0045] In the above embodiment, in addition to receiving the course catalog data input by the user, a cross-platform folder can also be created in the multi-level folder. No matter on which platform or on which device the user puts the course catalog data and / or course resource data into the cross-platform folder, the course catalog data and / or course resource data can be classified and mapped for storage. For example, when the user puts the course catalog data into the cross-platform folder through a computer terminal (e.g., a desktop computer) and puts the course resource data into the cross-platform folder through a mobile terminal (e.g., a mobile phone, a laptop computer, a smart bracelet, etc.), the course catalog data and course resource data in the cross-platform folder can be directly classified, mapped, stored, etc., thereby avoiding the situation where the user transfers data on different platforms and causes omissions and re-transfers.

[0046] Step S105, receiving first query information input by a user, and determining multiple data access paths matching the first query information from a target database;

[0047] In the above embodiment, the target database may be a relational database or a non-relational database. The user inputs the first query information, which may be one or more keywords, conditions or specifications. By querying the target database, multiple data access paths that match the first query information are screened out. The data access path can be understood as a set of operation steps determined according to the first query information to obtain and process data from the target database. By executing the data access path, data matching the first query information can be extracted from the target database for subsequent processing and presentation. In short, the target database is a collection of structured data, and by executing a specific query path, data related to the user's query information is retrieved and matched from the target database.

[0048] Step S106, determining a first keyword from the first query information, and retrieving a preset query behavior of the user from the target database, wherein the preset query behavior includes a query matching rate, and the query matching rate is a ratio of the number of keyword queries to the number of actual accesses to the data access path;

[0049] In the above embodiment, the preset query behavior is the user's historical query behavior pre-stored in the target database, and the historical query behavior refers to the historical records and behavior patterns of the user's query for one or some keywords. The query matching rate is the number of keyword queries and the actual number of accesses to the data access path. The query matching rate reflects the frequency of the user actually accessing the data access path after querying a certain keyword.

[0050] Step S107, determining a first query matching rate corresponding to each data access path included in the multiple data access paths according to the first keyword and the preset query behavior, to obtain multiple first query matching rates, wherein the query matching rate includes multiple first query matching rates;

[0051] In the above embodiment, the first query matching rate is used to measure the matching degree between a specific data access path and a query input by a user (usually based on a keyword or a group of keywords). The matching degree can be determined based on multiple factors, including but not limited to the number of results returned by the data access path, the relevance of the results, the efficiency of query execution, etc.

[0052] Step S108, selecting one or more second query matching rates that meet a preset period from the plurality of first query matching rates, and adjusting the multi-level folders into first folders and second folders according to the one or more second query matching rates.

[0053] In the above embodiment, the preset period is a pre-set time period. The preset period can be a fixed time period, such as a week, a month, a quarter, etc., or a dynamic time period, such as the last 10 queries, the last 30 days, etc. Assuming that the preset period is the last month, all query behavior data of the user within this month will be collected, and the query matching rate of each keyword will be calculated. The structure of the multi-level folders is optimized based on these query matching rates, and the data paths that users frequently access are placed in a more accessible location.

[0054] Through the above steps, the course catalog data input by the received user is subjected to category analysis according to the preset course target system, and the category analysis result of the first category including the first course information is obtained, which can ensure that the course catalog data can be accurately classified. The first course catalog matching the first category is determined from the preset course catalog system, so that a multi-level folder can be established according to the first course catalog and the category analysis result, and the structured processing of the course catalog data is realized, so that the structure of the course catalog data is clearly hierarchical, so that the data under the specific hierarchical folder can be quickly accessed. The acquired course resource data is input into the preset data classification model, so that the classification of the course resource data is more accurate. The course resource data is mapped and stored according to the second category and the multi-level folder output by the acquired preset classification model, so that the mapping storage of the course resource data is more accurate, so that the data can be accessed more quickly, and the efficiency of data access is further improved. By adopting the method of the present application, a multi-level folder is established by matching the first course catalog and category parsing results that match the first category from the preset course catalog system, and then the course resource data is mapped and stored according to the second category and the multi-level folder output by the preset data classification model to achieve accurate mapping and storage of course resource data and multi-level folders, so as to achieve classification adjustment of multi-level folders, making data access more efficient and orderly, thereby solving the technical problem of low data retrieval efficiency in related technologies and achieving the technical effect of improving data retrieval efficiency.

[0055] Among them, the executor of the above steps can be a system with classification mapping storage function, or a device with classification mapping storage function, or a controller or processor in the device or system, or a separate controller or processor, or other processing devices or processing units with similar processing functions, etc., but not limited to this.

[0056] In an optional embodiment, the first folder is used to store the first course resource data under the data access path corresponding to the third query matching rate that is greater than the preset query matching rate, and the second folder is used to store the second course resource data under the data access path corresponding to the fourth query matching rate that is less than or equal to the preset query matching rate, one or more second query matching rates include the third query matching rate and the fourth query matching rate, and the course resource data include the first course resource data and the second course resource data.

[0057] In the above embodiment, the preset query matching rate can be preset, and after the preset query matching rate is preset, the preset query matching rate can be adjusted according to actual application requirements, etc., which is not limited here.

[0058] In the above embodiment, whenever a user makes a query, query information is recorded, including query keywords, query time, and the data access path finally selected, etc. The query matching rate is calculated based on the keywords queried by the user and the data access path finally selected. The query matching rate is defined as the ratio of the number of keyword queries to the number of actual accesses to the data access path. As time goes by, more and more user query data is collected, and the query matching rate corresponding to each keyword is continuously updated. The preset query behavior is a dynamic data set, which is continuously updated and optimized according to the actual query behavior of the user. The query history of the user is analyzed to identify query patterns such as query frequency and query time distribution. Based on the analysis of the user query pattern, a suitable preset period can be determined to filter the query matching rates that meet the conditions. The preset period may be a fixed time period or a dynamic time period based on the user query pattern. If it is found that the user's query pattern has changed (for example, the query frequency increases or decreases, the query time distribution changes, etc.), the preset period may be adjusted accordingly. The update of the preset period is intended to ensure that the data access path that meets the conditions can be more accurately filtered out according to the latest user query behavior, and the corresponding folder adjustment is performed.

[0059] In an optional embodiment, after selecting one or more second query matching rates that meet a preset period from a plurality of first query matching rates, and adjusting the multi-level folders into first folders and second folders according to the one or more second query matching rates, the method further includes: receiving second query information input by a user, and determining a second keyword from the second query information; determining a fifth query matching rate corresponding to the second keyword from a preset query behavior; comparing the fifth query matching rate with the preset query matching rate to obtain a comparison result; and when the comparison result shows that the fifth query matching rate is greater than the preset query matching rate, performing data access processing on the first folder according to the fifth query matching rate to obtain a data According to the access result; when the data access result shows that the first data access path corresponding to the fifth query matching rate exists in the first folder, the third course resource data under the first data access path in the first folder is sent to the user; or, when the data access result shows that the first data access path corresponding to the fifth query matching rate does not exist in the first folder, path query processing is performed on the second folder according to the first data access path to obtain a path query result; when the path query result shows that the first data access path exists in the second folder, the third course resource data under the first data access path in the second folder is sent to the user, and the third course resource data is stored in the first folder.

[0060] In the above embodiment, it is assumed that there is an online education platform that stores a large number of course resources, which are mapped and stored in multi-level folders. In order to quickly respond to user query requests and provide relevant teaching materials, the following steps can be implemented to determine the query matching rate of each keyword based on historical query data and set periodic updates. When receiving the query information input by the user, for example, the student queries "Basics of Calculus", the query information is parsed, the keyword "Calculus" is extracted, and the query matching rate related to "Calculus" is searched from the historical query data (assuming it is 0.75, of course it can also be other values, which are not limited here), and the query matching rate is compared with the current preset query matching rate (assuming it is 0.70, of course it can also be other values, which are not limited here). Because 0.75 is greater than 0.70, the relevant teaching materials are searched in the first folder according to the higher query matching rate. If the relevant teaching materials are found in the first folder (for example, Introduction to Calculus.pdf, Application of Calculus.pdf, Calculus and Derivatives.pdf, etc., which are not limited here), the teaching materials are sent to the user. If the relevant teaching materials are not found in the first folder, they can be searched in the second folder. After the relevant teaching materials are found in the second folder, they are sent to the user, and the teaching materials are stored in the first folder to improve the subsequent access speed.

[0061] In an optional embodiment, a first course directory matching the first category is determined from a preset course directory system, and a multi-level folder is established based on the first course directory and the category parsing result, specifically including: performing name division processing on the first course directory according to the first category to obtain a second course directory; establishing a first-level folder of the first category under the preset directory, wherein the preset directory is a preset second data access path, the second data access path is used to map and store course resource data, and the name of the first-level folder includes the first code corresponding to the first category; establishing a second-level folder of the second course directory under the first-level folder to obtain a multi-level folder, wherein the name of the second-level folder includes the second code corresponding to the second course directory.

[0062] In the above embodiment, it is assumed that there is a preset course catalog system, which includes courses of multiple categories, such as "mathematics", "physics", "chemistry", etc. Each course category contains multiple specific course catalogs, for example, there may be sub-catalogs such as "basic mathematics", "advanced mathematics", and "applied mathematics" under the "mathematics" course category. Assuming that the first category to be processed is determined to be "mathematics", the first course catalog that matches the "mathematics" category is found from the preset course catalog system. Assuming that the "mathematics" category contains three sub-catalogs of "basic mathematics", "advanced mathematics" and "applied mathematics", these sub-catalogs are name-divided to obtain the second course catalog. Under the preset directory (for example, the "course data" folder of the D disk), a layer of folders is established for the first category (i.e., "mathematics"), and the preset directory is a pre-set second data access path for mapping and storing course resource data. The layer of file name corresponding to the "mathematics" course category can be "math" (i.e., the first code, directly corresponding to the first category "mathematics"). Under the first-level folder "math", a second-level folder is created for the second course catalog. The name of each second-level folder includes the second code of the corresponding second course catalog. For "Basic Mathematics", a second-level folder named "basic" (i.e., the second code of "Basic Mathematics") can be created. For "Advanced Mathematics", a second-level folder named "advanced" (i.e., the second code of "Advanced Mathematics") can be created. For "Applied Mathematics", a second-level folder named "applied" (i.e., the second code of "Applied Mathematics") can be created. The first code directly corresponds to the code of the first category (i.e., "mathematics") and is used to identify the entire mathematics course category. It can be fixed or dynamically adjusted based on user preferences. The second code is generated based on the name of the second course catalog (i.e., "Basic Mathematics", "Advanced Mathematics", "Applied Mathematics", etc.) and specific coding rules, and is used to identify different course sub-catalogs.

[0063] In an optional embodiment, before obtaining course resource data and inputting the course resource data into a preset data classification model, the method further includes: obtaining sample data, wherein the sample data is a collection of multiple second course information; constructing a knowledge graph based on the sample data, wherein the knowledge graph includes multiple course catalog nodes and multiple course name nodes and multiple groups of associated attribute data, each course catalog node included in the multiple course catalog nodes is associated with a corresponding course name node included in the multiple course name nodes through a directed edge, and the multiple groups of associated attribute data are used to represent the associated attributes between the multiple course name nodes; constructing a graph neural network model based on the knowledge graph; connecting the graph neural network model to a classifier to obtain an initial data classification model; training the initial data classification model based on the sample data to obtain a preset data classification model.

[0064] In the above embodiment, by first constructing the knowledge graph and graph neural network model, the model can more deeply understand the association and hierarchical relationship between the course resource data, thereby improving the accuracy of classification. The use of the graph neural network model can process unstructured and semi-structured data, and has good adaptability to complex and diverse course resource data. The knowledge graph and graph neural network model can be continuously expanded and updated with the addition of new course resource data, maintaining the timeliness and accuracy of the data classification model.

[0065] In the above embodiment, sample data is collected, and the sample data includes a set of multiple second course information, and the second course information includes but is not limited to course catalog, course name, course content, etc. A knowledge graph is constructed based on the sample data. In the knowledge graph, each course catalog node and the corresponding course name node are associated through directed edges. At the same time, multiple sets of associated attribute data are added to represent the associated attributes between the course name nodes, such as dependencies and similarities between courses. A graph neural network model is constructed based on the knowledge graph, and then the graph neural network model is used to capture the complex relationship between nodes. The graph neural network model can generate embedded representations of nodes by learning the features of nodes and edges, and these embedded representations can be used for subsequent data classification tasks. The graph neural network model is connected to the classifier to obtain an initial data classification model, and the initial data classification model is trained using sample data to adjust the parameters of the initial data classification model and optimize its performance. The training process can be implemented by an iterative optimization algorithm, such as a gradient descent algorithm. After the training is completed, the performance of the training model is evaluated and verified. If the performance of the training model meets the requirements, it can be applied to the actual course resource data classification task.

[0066] In an optional embodiment, the initial data classification model is trained based on the sample data to obtain a preset data classification model, specifically including: inputting the sample data into the initial data classification model to obtain a second classification result; determining a sample label from the sample data; and adjusting the initial data classification model based on the second classification result and the sample label to obtain the preset data classification model.

[0067] In the above embodiment, the sample label is a category corresponding to the predetermined sample data. Before the sample data is input into the initial data classification model, the sample data needs to be preprocessed (for example, denoising, etc.), and the preprocessed data is divided into a training data set, a verification data set, and a test data set according to a preset ratio. The training data set is used in the training process of the initial data classification model. During the training process, the initial data classification model learns how to extract features from the sample data and predict the output based on the extracted features. Through multiple iterations and adjustment of model parameters, the model performance is gradually optimized until a certain accuracy is achieved on the training data set. During the training process, lower numerical precision (for example, half-precision floating point numbers) can also be used for training to first speed up the training speed and reduce resource consumption, and then dynamically adjust the learning rate according to the performance changes of the model during the training process to obtain better training results. The validation dataset is used to evaluate the performance of the model during the training process and adjust the model's hyperparameters or structure accordingly. At each stage of model training, the model is evaluated on the validation dataset and returns indicators such as accuracy and loss. These indicators are used to determine whether the model is overfitting or underfitting and adjust the model accordingly. For example, when the performance of the model on the validation dataset begins to decline, the training is automatically stopped to prevent overfitting. Use the validation dataset to select the best model, that is, save multiple models during the training process and select the model with the best performance on the validation dataset as the final model. You can also integrate multiple models with good performance on the validation dataset to form a more powerful classification model. The test dataset is used to evaluate the generalization ability of the model after the model training is completed. Unlike the validation dataset, the test dataset is invisible during the model training process to ensure the fairness and reliability of the evaluation results. Through the test dataset, you can understand the performance of the model on unseen data and judge whether the model is suitable for application in actual tasks. By using the training dataset, validation dataset, and test dataset reasonably, you can effectively train and optimize the data classification model and evaluate its performance in actual tasks.

[0068] In an optional embodiment, the second category output by the preset data classification model is obtained, and the course resource data is mapped and stored according to the second category and the multi-level folders, specifically including: matching a third course directory that meets the second category from the multi-level folders; determining a third folder corresponding to the third course directory, wherein the multi-level folder includes the third folder; mapping and storing the course resource data in the third folder, and assigning a version number to the mapped and stored course resource data.

[0069] In the above embodiments, the version number allocation method includes but is not limited to automatic sequential numbering, timestamp, hash value, semantic version control, and custom numbering. Among them, the automatic sequential numbering is to increase a version number in sequence every time a new course resource data mapping is stored in a folder. The timestamp is the time (accurate to seconds, milliseconds, etc.) stored using the course resource data mapping as the version number, which can clearly reflect the update time of the course resource data. A hash value is calculated based on the content of the course resource data and used as the version number, which can be used to verify whether the course resource data content has been tampered with. Semantic version control is that the version number usually adopts the format of "major version number. Minor version number. Revision number" (for example, 1.0.0), and the increase of each number represents a different level of change. The major version number represents a major update, which may not be compatible with the old version of the resource. The minor version number represents a minor update or an additional function, but it is still compatible with the previous version. The revision number is usually used for small changes, for example, fixing errors, which does not affect the main function. Custom numbering is to allocate versions according to specific business rules or user input. It should be noted that the above-mentioned example of the allocation method of the version number is only an exemplary embodiment, and the above-mentioned allocation method of the version number is not limited to the above-mentioned example.

[0070] Obviously, the above-described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application will be described in detail below in conjunction with specific embodiments.

[0071] The present application embodiment provides a data processing process based on artificial intelligence, which includes the following steps:

[0072] S1, importing course catalog data, wherein the course catalog data may be data such as images, tables, documents, etc. containing a course schedule.

[0073] S2, parse the course catalog data, establish each course category (corresponding to the first category mentioned above), and create a corresponding data folder for each course category (corresponding to the multi-level folder mentioned above), wherein the path and name of the data folder support pre-configured unified management format.

[0074] Preferably, a course catalog system (corresponding to the above-mentioned preset course catalog system) can be pre-set, and the course catalog under each course catalog system (corresponding to the above-mentioned first course catalog) can be further divided into specific course names. When creating corresponding data folders for each course category, a first-level folder (corresponding to the above-mentioned first-level folder) of the course catalog (corresponding to the above-mentioned second course catalog) to which the course category belongs can be first established under the preset directory, and the name of the first-level folder can include the code corresponding to the course catalog (corresponding to the above-mentioned first code). Afterwards, a second-level folder (corresponding to the above-mentioned second-level folder) of the corresponding course category is established under the first-level folder, and the name of the second-level folder can include the code corresponding to the course category (corresponding to the above-mentioned second code). Among them, natural language processing and regular expressions can be used, combined with a preset course catalog system, to deeply analyze the imported course catalog data and establish various course categories.

[0075] S3, receiving each data to be processed (corresponding to the above-mentioned course resource data) in batches, inputting the data to be processed into a pre-trained data classification model (corresponding to the above-mentioned preset data classification model), and determining the data category that matches each data to be processed, wherein the data to be processed includes data of multiple courses, multiple content categories, and multiple formats, such as electronic courseware materials, electronic course homework materials, etc.

[0076] Among them, the data classification model can be trained based on the following steps:

[0077] Obtain sample data, and build a knowledge graph based on the sample data. Each node of the knowledge graph may include each course catalog node, each course name node under each course catalog node (a relationship is established between the course catalog node and the course name node through a directed edge), and attribute data associated with each course name node (historical courseware, historical homework data, etc.);

[0078] Construct a graph neural network model based on the knowledge graph, connect the graph neural network model to the classifier, and build an initial data classification model (corresponding to the above initial data classification model); input the sample data to be classified (courseware and homework material data to be classified) into the initial data classification model so that the data classification model outputs its classification results, and iterate and optimize the model based on the classification results and sample labels to obtain the final data classification model.

[0079] S4, performing data mapping management according to the data category (corresponding to the above-mentioned second category) matched with each data to be processed and the data folder (corresponding to the above-mentioned multi-level folder) corresponding to each course category, and storing each data to be processed in the corresponding folder.

[0080] Preferably, a version control mechanism can also be introduced to track and manage data modifications and updates to ensure data integrity and consistency.

[0081] S5, after storing each to-be-processed data in the corresponding folder, the multi-level folder can be adjusted into a relatively frequently accessed folder and a relatively infrequently accessed folder in the backend according to the user's access frequency, so as to provide faster and more accurate access services according to different access requirements. In the human-computer interaction interface, the multi-level folder is still presented to the user in a complete hierarchical structure.

[0082] S6 provides a human-computer interaction interface, receives the query content input by the user in text / voice, performs semantic understanding and analysis on the query content, determines the matching folder path and the corresponding file name, and opens the file with the specified file name in the specified folder. Among them, deep learning can be used to build a semantic understanding model to accurately capture the user's query intention and convert it into accurate retrieval instructions to perform file queries and realize fast file opening. In addition, it combines user feedback and behavior data to continuously optimize the query algorithm and result display method to improve user satisfaction and query efficiency.

[0083] Through the embodiments of the present application, a multi-level folder is established by matching the first course catalog and category parsing results that match the first category from the preset course catalog system, and then mapping and storing the course resource data according to the second category and the multi-level folder output by the preset data classification model to achieve accurate mapping and storage of course resource data and multi-level folders, thereby achieving classification adjustment of multi-level folders, making data access more efficient and orderly, thereby solving the technical problem of low data retrieval efficiency in related technologies and achieving the technical effect of improving data retrieval efficiency.

[0084] The above describes the data processing method based on artificial intelligence in the embodiment of the present application. The following describes in detail the data processing device based on artificial intelligence in the embodiment of the present application in combination with the above data processing method based on artificial intelligence. Figure 2 , Figure 2 : is a structural block diagram of an artificial intelligence-based data processing device provided in an embodiment of the present application, the device comprising:

[0085] The parsing module 201 is used to perform category parsing on the course catalog data according to a preset course catalog system when receiving the course catalog data input by the user, so as to obtain a category parsing result, wherein the course catalog data is a collection of first course information, and the category parsing result includes a first category of the first course information;

[0086] Establishing module 202, for determining a first course catalog that matches the first category from a preset course catalog system, and establishing a multi-level folder according to the first course catalog and the category parsing result;

[0087] An input module 203 is used to obtain course resource data and input the course resource data into a preset data classification model;

[0088] The mapping storage module 204 is used to obtain the second category output by the preset data classification model, and map and store the course resource data according to the second category and the multi-level folder, wherein the second category is the first classification result output after the preset data classification model classifies the course resource data;

[0089] The determination module 205 is used to receive first query information input by a user and determine multiple data access paths matching the first query information from the target database;

[0090] The first processing module 206 is used to determine the first keyword from the first query information, and retrieve the user's preset query behavior from the target database, wherein the preset query behavior includes a query matching rate, and the query matching rate is a ratio of the number of keyword queries to the number of actual accesses to the data access path;

[0091] The second processing module 207 is used to determine a first query matching rate corresponding to each data access path included in the multiple data access paths according to the first keyword and the preset query behavior, so as to obtain multiple first query matching rates, wherein the query matching rate includes multiple first query matching rates;

[0092] The adjustment module 208 is used to select one or more second query matching rates that meet a preset period from the multiple first query matching rates, and adjust the multi-level folders into the first folder and the second folder according to the one or more second query matching rates.

[0093] The electronic device in the embodiment of the present invention is described below from the perspective of hardware processing. Figure 3 , Figure 3 It is a schematic diagram of a physical device structure of an electronic device in an embodiment of the present application.

[0094] It should be noted that Figure 3 The structure of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0095] like Figure 3As shown, the electronic device includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 to the random access memory (RAM) 303, such as executing the method described in the above embodiment. In the RAM 303,

[0096] There are various programs and data required for system operation. The CPU 301 , the ROM 302 , and the RAM 303 are connected to each other via a bus 304 . An input / output (I / O) interface 305 is also connected to the bus 304 .

[0097] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, a button switch, etc.; an output section 307 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read therefrom is installed into the storage section 308 as needed.

[0098] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 309, and / or installed from a removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present invention are performed.

[0099] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0100] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0101] Specifically, the electronic device of this embodiment includes a processor and a memory, and a computer program is stored in the memory. When the computer program is executed by the processor, the artificial intelligence-based data processing method provided in the above embodiment is implemented.

[0102] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the electronic device, the electronic device implements the data processing method based on artificial intelligence provided in the above embodiment.

[0103] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0104] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. A data processing method based on artificial intelligence, characterized in that: include: When receiving the course catalog data input by the user, performing category parsing on the course catalog data according to a preset course catalog system to obtain a category parsing result, wherein the course catalog data is a collection of first course information, and the category parsing result includes a first category of the first course information; Determining a first course catalog matching the first category from the preset course catalog system, and establishing a multi-level folder according to the first course catalog and the category parsing result, specifically includes: performing name division processing on the first course catalog according to the first category to obtain a second course catalog; establishing a first-level folder of the first category under the preset directory, wherein the preset directory is a preset second data access path, the second data access path is used to map and store the course resource data, and the name of the first-level folder includes the first code corresponding to the first category; establishing a second-level folder of the second course catalog under the first-level folder to obtain the multi-level folder, wherein the name of the second-level folder includes the second code corresponding to the second course catalog; Acquire course resource data, and input the course resource data into a preset data classification model; Obtaining the second category output by the preset data classification model, and mapping and storing the course resource data according to the second category and the multi-level folder, wherein the second category is the first classification result output after the preset data classification model classifies the course resource data; Receiving first query information input by the user, and determining a plurality of data access paths matching the first query information from a target database; Determine a first keyword from the first query information, and retrieve the preset query behavior of the user from the target database, wherein the preset query behavior includes a query matching rate, and the query matching rate is a ratio of the number of keyword queries to the number of actual accesses to the data access path; Determine a first query matching rate corresponding to each data access path included in the plurality of data access paths according to the first keyword and the preset query behavior, so as to obtain a plurality of first query matching rates, wherein the query matching rate includes the plurality of first query matching rates; One or more second query matching rates that meet a preset period are screened out from the multiple first query matching rates, and the multi-level folders are adjusted into first folders and second folders according to the one or more second query matching rates; the first folder is used to store the first course resource data under the data access path corresponding to the third query matching rate that is greater than the preset query matching rate, and the second folder is used to store the second course resource data under the data access path corresponding to the fourth query matching rate that is less than or equal to the preset query matching rate, the one or more second query matching rates include the third query matching rate and the fourth query matching rate, and the course resource data include the first course resource data and the second course resource data.

2. The method according to claim 1, characterized in that After selecting one or more second query matching rates that meet a preset period from the plurality of first query matching rates, and adjusting the multi-level folders into the first folder and the second folder according to the one or more second query matching rates, the method further includes: receiving second query information input by the user, and determining a second keyword from the second query information; Determining a fifth query matching rate corresponding to the second keyword from the preset query behavior; Comparing the fifth query matching rate with the preset query matching rate to obtain a comparison result; When the comparison result shows that the fifth query matching rate is greater than the preset query matching rate, performing data access processing on the first folder according to the fifth query matching rate to obtain a data access result; When the data access result shows that the first data access path corresponding to the fifth query matching rate exists in the first folder, the third course resource data under the first data access path in the first folder is sent to the user; or When the data access result shows that the first data access path corresponding to the fifth query matching rate does not exist in the first folder, path query processing is performed on the second folder according to the first data access path to obtain a path query result; when the path query result shows that the first data access path exists in the second folder, the third course resource data under the first data access path in the second folder is sent to the user, and the third course resource data is stored in the first folder.

3. The method according to claim 1, characterized in that Before obtaining the course resource data and inputting the course resource data into the preset data classification model, the method further includes: Acquire sample data, wherein the sample data is a collection of multiple second course information; Constructing a knowledge graph according to the sample data, wherein the knowledge graph includes a plurality of course catalog nodes and a plurality of course name nodes and a plurality of sets of associated attribute data, each course catalog node included in the plurality of course catalog nodes is associated with a corresponding course name node included in the plurality of course name nodes through a directed edge, and the plurality of sets of associated attribute data are used to represent the associated attributes between the plurality of course name nodes; Constructing a graph neural network model according to the knowledge graph; Connecting the graph neural network model to the classifier to obtain an initial data classification model; The initial data classification model is trained according to the sample data to obtain the preset data classification model.

4. The method according to claim 3, characterized in that The training of the initial data classification model according to the sample data to obtain the preset data classification model specifically includes: Inputting the sample data into the initial data classification model to obtain a second classification result; Determining a sample label from the sample data; The initial data classification model is adjusted according to the second classification result and the sample label to obtain the preset data classification model.

5. The method according to claim 1, characterized in that The obtaining of the second category output by the preset data classification model and mapping and storing the course resource data according to the second category and the multi-level folders specifically includes: Matching a third course catalogue satisfying the second category from the multi-level folders; Determine a third folder corresponding to the third course catalog, wherein the multi-level folders include the third folder; The course resource data is mapped and stored in the third folder, and a version number is assigned to the mapped and stored course resource data.

6. A data processing device based on artificial intelligence, characterized in that: include: A parsing module, configured to, upon receiving the course catalog data input by the user, perform category parsing on the course catalog data according to a preset course catalog system to obtain a category parsing result, wherein the course catalog data is a collection of first course information, and the category parsing result includes a first category of the first course information; A creation module is used to determine a first course catalog that matches the first category from the preset course catalog system, and to create a multi-level folder according to the first course catalog and the category parsing result; the creation module is specifically used to perform name division processing on the first course catalog according to the first category to obtain a second course catalog; to create a first-level folder of the first category under the preset directory, wherein the preset directory is a preset second data access path, the second data access path is used to map and store the course resource data, and the name of the first-level folder includes the first code corresponding to the first category; to create a second-level folder of the second course catalog under the first-level folder to obtain the multi-level folder, wherein the name of the second-level folder includes the second code corresponding to the second course catalog; An input module is used to obtain course resource data and input the course resource data into a preset data classification model; a mapping storage module is used to obtain a second category output by the preset data classification model, and map and store the course resource data according to the second category and the multi-level folder, wherein the second category is a first classification result output after the preset data classification model classifies the course resource data; A determination module, configured to receive first query information input by a user, and determine a plurality of data access paths matching the first query information from a target database; A first processing module is used to determine a first keyword from the first query information, and retrieve a preset query behavior of a user from a target database, wherein the preset query behavior includes a query matching rate, and the query matching rate is a ratio of the number of keyword queries to the number of actual accesses to the data access path; A second processing module is used to determine a first query matching rate corresponding to each data access path included in the multiple data access paths according to the first keyword and the preset query behavior, so as to obtain multiple first query matching rates, wherein the query matching rate includes multiple first query matching rates; An adjustment module is used to filter out one or more second query matching rates that meet a preset period from multiple first query matching rates, and adjust the multi-level folders into first folders and second folders according to the one or more second query matching rates; the first folder is used to store the first course resource data under the data access path corresponding to the third query matching rate that is greater than the preset query matching rate, and the second folder is used to store the second course resource data under the data access path corresponding to the fourth query matching rate that is less than or equal to the preset query matching rate, the one or more second query matching rates include the third query matching rate and the fourth query matching rate, and the course resource data includes the first course resource data and the second course resource data.

7. An electronic device, characterized in that: The electronic device comprises: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code comprises computer instructions, and the one or more processors call the computer instructions so that the electronic device executes the method as described in any one of claims 1-5.

8. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on an electronic device, the electronic device is caused to execute the method as claimed in any one of claims 1 to 5.

Images