File processing method and device based on AI

Abstract

The invention discloses a file processing method and device based on AI. The method comprises the following steps: executing a plurality of query statements to obtain a plurality of query results, storing the plurality of query results into different worksheets, and integrating the different worksheets into a single file; and generating a unique final file name based on the plurality of query results, and storing the single file according to the generated file name. The technical problems that the file name is not visual enough and the file content is difficult to quickly identify are solved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular, to an AI-based file processing method and apparatus. Background Art

[0002] There are some problems in the prior art in terms of data export. First, the exported file names are usually not intuitive, making it difficult for users to quickly identify the file content and time. Second, when multiple SQL queries are involved, the platform exports the data in the form of a compressed package, and users need to decompress and process the file, which increases the complexity of use. Finally, numbers with more than 15 digits are displayed in scientific notation during export, resulting in loss of data precision and inability to accurately save important data. Although some existing tools have tried to improve these problems, they still fail to provide a comprehensive solution, especially in terms of file naming and data precision maintenance.

[0003] In view of the above problems, no effective solution has been proposed yet. Summary of the Invention

[0004] Embodiments of the present invention provide an AI-based file processing method and apparatus to at least solve the technical problem that the file name is not intuitive enough and it is difficult to quickly identify the file content.

[0005] According to one aspect of the embodiments of the present invention, there is provided an AI-based file processing method, including: executing a plurality of query statements to obtain a plurality of query results, storing the plurality of query results in different worksheets, and integrating the different worksheets into a single file; generating a unique final file name based on the plurality of query results, and saving the single file with the generated file name.

[0006] According to another aspect of the embodiments of the present invention, there is also provided an AI-based file processing apparatus, including: a generating module configured to execute a plurality of query statements to obtain a plurality of query results, store the plurality of query results in different worksheets, and integrate the different worksheets into a single file; a storing module configured to generate a unique final file name based on the plurality of query results, and save the single file with the generated file name.

[0007] In the embodiments of the present invention, a plurality of query statements are executed to obtain a plurality of query results, the plurality of query results are stored in different worksheets, and the different worksheets are integrated into a single file; a unique final file name is generated based on the plurality of query results, and the single file is saved with the generated file name. By the above solution, the technical problems that the file name is not intuitive enough and it is difficult to quickly identify the file content are solved. Brief Description of the Drawings

[0008] The accompanying drawings described herein are used to provide a further understanding of the present invention and form a part of this application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0009] Figure 1 is a flowchart of an alternative AI-based file processing method according to an embodiment of the present invention;

[0010] Figure 2 is a flowchart of another alternative AI-based file processing method according to an embodiment of the present invention;

[0011] Figure 3 is a flowchart of an alternative method for generating a file name according to an embodiment of the present invention;

[0012] Figure 4 is a flowchart of yet another alternative AI-based file processing method according to an embodiment of the present invention;

[0013] Figure 5 is a flowchart of an alternative method for generating a file name based on AI according to an embodiment of the present invention;

[0014] Figure 6 is a schematic structural diagram of an alternative AI-based file processing apparatus according to an embodiment of the present invention;

[0015] Figure 7 shows a schematic structural diagram of an electronic device suitable for implementing the embodiments of the present disclosure. Detailed Embodiments

[0016] In order to enable those skilled in the art of the present technology to better understand the solution of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present invention.

[0017] It should be noted that the terms "first", "second", etc. in the description, claims and above-mentioned drawings of the present invention are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present invention described here can be implemented in an order different from those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0018] According to an embodiment of the present invention, there is provided a method embodiment of a file processing method based on AI. It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0019] In the prior art, when exporting files, the following problems exist: First, the naming method of the exported files is not intuitive enough, and it is difficult for users to quickly identify the file content; Second, when multiple SQL queries are involved, the exported data needs to be decompressed by the user, and the table names are not intuitive, increasing the complexity of use; Finally, numbers with more than 15 digits are displayed in scientific notation when exported, resulting in the loss of data precision. For file naming, the prior art usually only generates file names according to default rules and cannot provide clear identification according to the file content, which makes the efficiency of users in file management and subsequent processing relatively low. For the export of multiple query results, the prior art often uses the form of a compressed package to pack multiple query results into a single compressed file, and the user must manually decompress it. Finally, for data containing large numbers, the display method of scientific notation may cause the loss of digital precision and affect the accuracy of the data, especially in fields such as finance and scientific research where high data precision is required.

[0020] To solve the above problems, an embodiment of the present application provides a file processing method based on AI, as Figure 1 shown, the method includes the following steps:

[0021] Step S102, execute multiple query statements to obtain multiple query results, store the multiple query results in different worksheets, and integrate the different worksheets into a single file.

[0022] Among the multiple query results obtained by multiple query statements, if there are numbers with more digits than a preset threshold, the data needs to be specially processed. Specifically, redundant fields are generated. For example, the CAST function in the SQL statement is used to convert the number into a string form, thus avoiding the scientific notation problem when there are more than 15 digits, and ensuring that the data precision in the exported file is not lost. Using the redundant fields, the number is saved in the corresponding worksheet in string form, which can avoid the loss of data precision while retaining the integrity of the data.

[0023] Step S104, generate a unique final file name based on multiple query results and save a single file with the generated file name.

[0024] First, based on multiple query results, a preliminary file name is generated. For example, obtain the field names, data types, and data values in each query result among the multiple query results. Then, combine this information to determine the chart title for each query result. The chart title is a description of the data content and can be generated based on the field and data type information of the query result. For example, if the query result is sales data, the generated chart title can be "Sales Data Analysis". Based on the chart title of each query result, a preliminary file name is generated, such as "sales_data_analysis.csv".

[0025] Next, generate the final file name. Obtain the current time and generate a timestamp based on the current time in a predetermined format. The format of this timestamp can be %Y%m%d%H%i%s (i.e., year, month, day, hour, minute, second), ensuring that the file name contains accurate time information for users to identify the creation time of the file based on the time. Based on the timestamp and the preliminary file name, generate the final file name. For example, the final file name can be "20230101123456_sales_data_analysis.csv".

[0026] Finally, save the file with the final file name and directly export a single file instead of providing the exported file in the form of a compressed package. This enables users to directly use the exported file without decompression, improving the convenience and efficiency of the operation.

[0027] Through the above method, the present invention can effectively solve the problems of non-intuitive file naming, inconvenient export of multiple SQL queries, and loss of data precision in the prior art, thereby providing a more efficient and intuitive file export solution.

[0028] Figure 2Another AI-based file processing method according to an embodiment of the present application. To solve the problems of non-intuitive file naming, inconvenient multi-SQL query export, and data precision loss in the prior art, the present application provides another AI-based file processing method. This method improves file processing efficiency, ensures data integrity and accuracy, and enhances the user experience by optimizing file naming rules, improving the export method of multi-SQL queries, and solving data precision problems. As Figure 2 shown, the method includes the following steps:

[0029] Step S202, execute multiple query statements, obtain multiple query results, and integrate them into a single file.

[0030] In this embodiment, multiple SQL query statements need to be executed first to obtain multiple query results. Each query statement extracts a certain data set from the database. To ensure data tidiness and operability, all query results will be stored in different worksheets. In this step, the result of each query statement will be assigned to a separate worksheet, and each worksheet is named according to the content of the query, such as "Query 1", "Query 2", etc.

[0031] Among these query results, if some query results contain numbers with more digits than a preset threshold (for example, numbers with more than 15 digits), these data will be specially processed. Specifically, the CAST function in the SQL statement is used to convert these numbers into string format to avoid the display of numbers in scientific notation after exceeding 15 digits, thus causing data precision loss. In this way, users can ensure that all data retains full precision when exported, especially in scenarios with high-precision requirements such as financial data and statistical data, avoiding precision loss problems caused by scientific notation.

[0032] For example, if the query result contains a number "12345678901234567", which exceeds 15 digits, the traditional export method will convert it to scientific notation "1.2345678901234567E+16", which will cause data precision loss. However, in this embodiment, by converting the number to the string form "12345678901234567", the accuracy and integrity of the number are ensured.

[0033] Finally, the query results of these different worksheets will be integrated into a single file. For example, they will be integrated into an EXL file with multiple sheet pages for further processing and export. The integration process can be automatically executed programmatically, and users do not need to operate manually.

[0034] Step S204, generate a file name.

[0035] After completing data integration, a file name is generated. In this embodiment, a unique and intuitive file naming method helps users quickly identify the content of the file. Traditional file naming is usually automatically generated randomly by the system, making it difficult to reflect the actual content and creation time of the file. Therefore, this embodiment proposes a naming rule based on the query results. As Figure 3 shown, the method for generating the file name includes the following steps:

[0036] Step S2042, generate a preliminary file name.

[0037] First, extract relevant information from multiple query results, specifically including the field name, data type, and data value in each query result. By analyzing this information, a chart title can be generated for each query result. The chart title is a descriptive file name for the query result and is usually based on the query content. For example, if the query result is sales data for a certain period, the chart title can be "Sales Data Analysis"; if the query result is user registration data, the chart title can be "User Registration Status".

[0038] After generating the chart title for each query result, these titles can be used as part of the preliminary file name. For example, if Query 1 is "Sales Data Analysis" and Query 2 is "User Registration Status", the preliminary file name may be "sales_data_analysis_user_registration.csv".

[0039] Step S2044, generate the final file name.

[0040] After generating the preliminary file name, in this embodiment, the current timestamp is added to the preliminary file name to generate the final file name. The current timestamp is generated using a standard format, specifically %Y%m%d%H%i%s, which includes year, month, day, hour, minute, and second. The timestamp generated in this format can be accurate to the second, so that the file name contains the exact time when the file was created. In this way, users can directly judge the creation time of the file through the file name, avoiding confusion in file version management.

[0041] For example, if the current time is 12:34:56 on January 1, 2023, the timestamp is "20230101123456". Combining with the preliminary file name, "sales_data_analysis_user_registration.csv", the final file name will become "20230101123456_sales_data_analysis_user_registration.csv".

[0042] In some other embodiments, to meet different business requirements, a function for customizing file naming rules can be provided, allowing users to customize the file naming format according to actual needs. For example, users can choose whether to add a specific project number or customer information to the file name to better meet different file management requirements.

[0043] Step S2046, save the file and export it.

[0044] After the file name is generated, save the data as a file with the finally generated file name. The file format can be in various forms, such as CSV, XLSX, etc., and the specific format is determined according to user requirements or system default settings. Importantly, these files will be directly exported without using the compressed package method. In this way, users can directly open the files without decompression, further improving the operation convenience. For example, when a user executes a query and generates a data file, the system will directly generate a file named "20230101123456_sales_data_analysis_user_registration.csv" and allow the user to download it with one click without going through the additional step of decompression. In some other embodiments, to increase the system compatibility, users can choose the data export format according to different needs, such as JSON, etc., for easy integration and data processing with other software or systems.

[0045] In some embodiments, before data export, a data preview function can also be provided to allow users to confirm the accuracy and format of the data. Users can check whether the data meets expectations before export to avoid exporting incorrect data.

[0046] In the embodiments of the present invention, by including the timestamp and chart title in the file name, users can more intuitively identify the file content and can quickly judge the creation time of the file based on the file name, improving the file management and identification efficiency. In addition, traditional multi-SQL query exports usually use the compressed package form, and users need to manually decompress and process multiple files, while the present invention saves multiple query results in the same file in the multi-Sheet form, making it more convenient and efficient for users to process multiple query results. Also, by converting numbers with more than 15 digits into string form, the present invention effectively avoids the problem of data precision loss caused by scientific notation display, ensuring the accuracy and integrity of the data, which has important application value especially in the fields of finance, scientific research, etc. Finally, by simplifying the data export process, users can directly download and use the exported files, avoiding the cumbersome decompression process and improving the overall operation experience.

[0047] This application also provides another AI-based file processing method, which is different from the above embodiments in that AI is used to generate the file name. Specifically, asFigure 4 As shown, it includes the following steps:

[0048] Step S402: Execute multiple query statements to obtain multiple query results, store the multiple query results in different worksheets, and integrate the different worksheets into a single file.

[0049] It is the same as the step of generating a single file in the above embodiment and will not be elaborated here.

[0050] Step S404: Generate a file name based on AI.

[0051] In this step, the system automatically generates a file name associated with the file using AI based on the context information of the fields in the source data table. Different from the traditional method that relies on manual rules or fixed-pattern file naming, this application adopts an intelligent method based on technologies such as machine learning and natural language processing (NLP). According to the semantics, data structure, and business scenarios of the fields, it automatically generates a suitable and meaningful file name for the file. Specifically, as Figure 5 shown, it includes the following steps:

[0052] Step S4042: Extract features and conduct analysis.

[0053] Extract and analyze based on the multi-dimensional feature information of the fields in the source data table, and gradually optimize the semantic expression of the fields to provide an accurate basis for generating the file name. The specific process is as follows:

[0054] Conduct preliminary metadata extraction for each field in the source data table, including information such as field name, field type, field value range, field description, and its position in the table. Obtain the metadata of each field through an automated tool and store it in a structured data format (such as JSON or a database table) for convenient subsequent calculation.

[0055] Specifically, extract and enhance field features through machine learning, including the extraction of global features and local features. Global features include the basic information of the field (such as data type, the table to which the field belongs, etc.), while local features focus on the relationship between the field and other fields. For example, the relationship between the fields "order_date" and "order_id" reveals their structural dependencies in the database. Use a convolutional neural network (CNN) to extract the semantic information of the field name, use a graph neural network (GNN) to model the relationship between fields, and apply a recurrent neural network (RNN) to capture the time series changes of the field values.

[0056] In local feature analysis, the context relationship of fields is further deeply analyzed through machine learning algorithms to enhance the correlation between fields. For example, through a generative adversarial network (GAN) model, the dependency relationship between fields is enhanced, making the performance of the business role of each field more accurate in a specific context. For example, analyze the context information between fields to determine information such as the dependency relationship, foreign key constraint, and index structure between fields. Use the metadata of the relational database (such as foreign key constraints and index relationships) to perform joins between fields to determine the relevance between fields. By analyzing the logical structure of fields and their relative positions in the data table, determine the actual business role of the fields. For example, the field "order_date" may be closely related to "order_id" and "customer_id". Analyze the associations between these fields to understand their roles in actual business. In this way, not only can the basic meaning of the fields be identified, but also the semantics of the fields can be dynamically adjusted according to business requirements. For example, the semantics of the "creation time" field will be associated with the "modification time" field to improve the accuracy of field descriptions.

[0057] Since there may be redundant information and noisy data between fields, noise suppression algorithms in machine learning can also be applied to filter out irrelevant or low-quality field information. In particular, algorithms such as Euclidean distance and cosine similarity are used to evaluate the similarity between fields, and similar fields are grouped through clustering algorithms to reduce data redundancy. After optimizing the global features, the most representative feature set will be selected. In this way, unnecessary redundant fields can be removed, improving the accuracy of subsequent file name generation.

[0058] After extracting the global features and local features, these two types of features are fused. Through a multi-layer fusion algorithm, the feature information from different sources and levels is integrated to generate a more comprehensive and accurate semantic description for each field. Specifically, the global features and local features are integrated through weighted fusion to ensure that the finally generated feature description covers both the basic attributes of the fields and can accurately express their roles and functions in the data table and business process.

[0059] After completing all feature extraction and fusion, semantic inference will be performed on each field based on the semantic information and feature description of the fields, providing a basis for file name generation according to the inference results. For example, for fields containing "customer ID" and "order amount", it may be speculated that the file is related to customer orders, so the file name may be in a format such as "customer order_amount_20230101.xlsx". Flexible adjustments will be made according to the requirements of the business scenario to ensure that the file name can accurately reflect the business application background of the fields.

[0060] To enhance the accuracy and semantic consistency of file name generation, this application adds an entropy regularization term to the loss function to further optimize the file name generation process. Specifically, an entropy penalty term is added to the loss function. By penalizing the uncertainty generated by the model, the file names generated by the system are made more accurate and have a business context. The entropy regularization term can control the generated uncertainty based on the probability distribution P(y t |y<t) and the historical label y<t.

[0061] The improved loss function can be expressed as:

[0062]

[0063] Among them, L prediction is the standard prediction loss function, Hw represents the weighted entropy, β is the dynamic weight at time step t, λ is a hyperparameter used to control the intensity of the entropy penalty, and T is the total number of time steps. is the predicted value of the i-th sample, y i is the true value of the i-th sample, and N is the number of samples.

[0064] The above formula combines the traditional loss function and the entropy regularization term. By penalizing predictions with excessive uncertainty generated by the model, it can effectively improve the accuracy of predictions and the generalization ability of the model. The entropy regularization term enables the model to control uncertainty during the optimization process, avoid being overly confident in fuzzy predictions, thereby reducing overfitting and improving the model's performance on new data. At the same time, when the model processes time series data, it can better capture long-term and short-term dependencies and optimize the output distribution to ensure more reasonable predictions. By adjusting the hyperparameter λ, the model can flexibly balance the control of accuracy and uncertainty, further improving the adaptability and stability of the model.

[0065] Through the above steps, rich feature information can be gradually extracted from the fields of the source data table and deeply analyzed, providing precise semantic support for subsequent file name generation. Finally, based on the context information, business semantics, and data relationships of the fields, file names that meet business requirements and have high accuracy can be automatically generated.

[0066] Step S4044, generate candidate solutions for file names.

[0067] Based on the extracted field context information and semantics, multiple candidate file names are generated through an AI algorithm. The candidate file names not only reflect the physical attributes of the fields but also embody the actual applications of the fields in the business process. For example, for a file containing fields such as "order date" and "customer information", a candidate file name like "order details_20230101_customer information.xlsx" is generated. When generating candidate file names, dynamic adjustments are also made according to the business relevance and context of the fields, making the file names have stronger business relevance and readability.

[0068] Step S4046., clustering analysis and optimizing file names.

[0069] The generated candidate file names are optimized using the clustering analysis method. By clustering similar fields, fields with similar functions or semantic similarities are identified, thereby further screening out the most suitable file names. For example, for a file containing a timestamp and an amount, a file name like "transaction record_2023_year_amount_report.xlsx" is automatically generated through clustering analysis. Clustering analysis ensures the consistency of file names within the same category and improves the accuracy of file name generation.

[0070] Step S4048, file name confirmation and feedback.

[0071] After generating multiple candidate file names, the user is allowed to confirm or modify the automatically generated file names. The user can modify the file names according to actual needs or select the most suitable candidate file name. The system will use the user's feedback as new training data to further optimize the AI model and improve the accuracy and business adaptability of file name generation.

[0072] Through continuous iterative optimization and training, the AI model can gradually improve the accuracy of generating file names. Each time the user selects a file name or provides feedback, the system uses the feedback information to further optimize the model parameters, making the file names generated in the future more in line with actual needs. This process can ensure that as the data volume increases and the business scenarios change, the system can always generate high-quality file names that meet the user's needs.

[0073] This application also provides an AI-based file processing device, as Figure 6 shown, including: a generation module 62, configured to execute multiple query statements to obtain multiple query results, store the multiple query results in different worksheets, and integrate the different worksheets into a single file; a storage module 64, configured to generate a unique final file name based on the multiple query results and save the single file with the generated file name.

[0074] It should be noted that: The AI-based document processing device provided in the above embodiments is only illustrated by dividing the above-mentioned functional modules. In practical applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the AI-based document processing device provided in the above embodiments and the embodiments of the AI-based document processing method belong to the same concept. For the specific implementation process, please refer to the method embodiments and will not be elaborated here.

[0075] Figure 7 FIG. shows a schematic structural diagram of an electronic device suitable for implementing the embodiments of the present disclosure. It should be noted that Figure 7 The electronic device shown is only an example and should not impose any limitations on the functions and usage scope of the embodiments of the present disclosure.

[0076] As Figure 7 shown, the electronic device includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 1002 or the program loaded from the storage section 1008 into the random access memory (RAM) 1003. In the RAM 1003, various programs and data required for system operations are also stored. The CPU 1001, ROM 1002, and RAM 1003 are connected to each other via a bus 1004. The input / output (I / O) interface 1005 is also connected to the bus 1004.

[0077] The following components are connected to the I / O interface 1005: an input section 1006 including a keyboard, a mouse, etc.; an output section 1007 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section 1008 including a hard disk, etc.; and a communication section 1009 including a network interface card such as a LAN card, a modem, etc. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as needed. A removable medium 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1010 as needed so that a computer program read from it can be installed into the storage section 1008 as needed.

[0078] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. A file processing method based on AI, characterized in that: include: Execute multiple query statements to obtain multiple query results, store the multiple query results in different worksheets, and integrate the different worksheets into a single file; A unique final file name is generated based on the multiple query results, and the single file is saved with the generated final file name.

2. The method according to claim 1, characterized in that Generate a unique final file name based on the multiple query results, including: Based on the multiple query results, generate a preliminary file name, obtain a current time, and generate a timestamp in a predetermined format based on the current time; Based on the timestamp and the preliminary file name, the final file name is generated.

3. The method according to claim 2, characterized in that Based on the multiple query results, a preliminary file name is generated, including: Obtaining a field name, a data type, and a data value in each of the plurality of query results; A chart title of each query result is determined based on the field name, the data type and the data value, and the preliminary file name is generated based on the chart title of each query result.

4. The method according to claim 1, characterized in that: The multiple query results are stored in different worksheets, including: When the multiple query results obtained by the multiple query statements contain digits exceeding a preset threshold, generating a redundant field; The redundant fields are used to save the numbers in a character string format into a corresponding worksheet.

5. The method according to claim 4, characterized in that Generating redundant fields includes: generating redundant fields by using a CAST method.

6. The method according to claim 1, characterized in that After saving the single file with the generated file name, the method further includes: directly exporting the single file without providing the exported file in a compressed package form.

7. An AI-based file processing device, characterized in that: include: A generating module configured to execute multiple query statements to obtain multiple query results, store the multiple query results in different worksheets, and integrate the different worksheets into a single file; The storage module is configured to generate a unique final file name based on the multiple query results, and save the single file with the generated file name.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 6.

9. A computer device, characterized in that: include: Memory and processor, The memory stores a computer program; The processor is used to execute the computer program stored in the memory, and when the computer program is run, the processor is enabled to execute the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images