Metadata updating method and device for scientific data set, electronic equipment and storage medium

Abstract

The invention discloses a metadata updating method and device for a scientific data set, electronic equipment and a storage medium, and the method comprises the steps: obtaining a first updating object and a corresponding updating behavior in real time, generating an updating record, and comparing information differences before and after updating to generate an updating log; generating cue words based on a preset cue word template and the update log, extracting key information and semantic features thereof from the first update object, generating reference metadata, verifying the updated metadata, generating a metadata verification report, distributing the metadata verification report to a data set associated user, obtaining summarized feedback information, and generating an update rule; generating a metadata updating instruction, a version number and an updating key based on a mapping relationship between the updating rule and the updating record, and analyzing the updating instruction, the version number and the updating key to update the second updating object; the efficiency and the quality of data management are improved by acquiring updating behaviors in real time, generating updating records, extracting key information, generating reference metadata and performing metadata verification.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a metadata updating method, device, electronic device and storage medium for a scientific data set. Background Art

[0002] In modern scientific research, the accuracy and completeness of datasets are crucial, and metadata, as important information describing the content, structure, and quality of datasets, must be updated in a timely and accurate manner. However, existing metadata update methods often face problems such as low automation, insufficient ability to handle complex data changes, and imperfect authority management. These limitations make the update process cumbersome and error-prone. Summary of the invention

[0003] To solve at least one of the above-mentioned defects, the present invention application proposes a metadata update method, device, electronic device and storage medium for a scientific data set, which are used to solve the problems of low automation of metadata update, insufficient ability to handle complex data changes and imperfect authority management.

[0004] According to a first aspect of the present application, a method for updating metadata of a scientific dataset is provided, comprising:

[0005] Acquire the first update object and the corresponding update behavior in real time, and generate an update record;

[0006] When the updating behavior is finished, based on the update record, the information difference before and after the updating is compared to generate an update log;

[0007] Preset a prompt word template, generate prompt words based on the prompt word template and the update log, extract key information and key information semantic features from the first update object based on the prompt words, and generate reference metadata;

[0008] Verifying the updated metadata associated with the first update object based on the reference metadata, generating a metadata verification report and distributing it to users associated with the scientific dataset, obtaining and summarizing user feedback information, and generating an update rule based on the feedback information;

[0009] Based on the mapping relationship between the update rule and the update record, generate a metadata update instruction, a version number and an update key;

[0010] Parsing the update instruction, the version number and the update key to update the second update object;

[0011] The first update object includes a scientific data set and / or a matching information carrier matching the scientific data set and / or metadata associated with the scientific data set;

[0012] The second update object is the scientific data set, supporting information carrier, and metadata corresponding to the first update object.

[0013] Preferably, the real-time acquisition of the first update object and the corresponding update behavior and the generation of the update record specifically includes:

[0014] For each first updated object, obtaining metadata update records of each updated version thereof, extracting metadata features and update behavior features based on the metadata update records, and generating a correlation matrix between metadata features and update behavior features according to the metadata features and update behavior features;

[0015] Preset mapping rules of scientific data sets, supporting information carriers and metadata, and obtain metadata features associated with the update behavior features in real time based on the mapping rules;

[0016] Determine the current metadata state and update requirements based on the metadata features associated with the update behavior features and the association matrix;

[0017] Based on the current metadata state and update requirements, the associated update behavior characteristics are marked to generate an update record.

[0018] Preferably, the step of generating an update log by comparing the information difference before and after the update based on the update record specifically includes:

[0019] Obtain metadata features corresponding to scientific datasets and metadata in historical versions based on update records;

[0020] The information difference of the metadata features before and after the update behavior is compared, and the update type, update content, feature keywords of the update behavior and semantic features of the update behavior of the metadata are obtained to generate an update log.

[0021] Preferably, the preset prompt word template includes:

[0022] Get the overall task and subtasks of metadata update;

[0023] Determine the subtask execution steps based on the update target characteristics, update condition standards, update content basis and metadata characteristics associated with the update behavior characteristics corresponding to each subtask;

[0024] Set the subtask's start conditions, output format, and task completion criteria.

[0025] Preferably, the generating of prompt words based on the prompt word template and the update log includes:

[0026] Acquire the update type, update content, characteristic keywords of the update behavior and semantic features of the update behavior in the update log;

[0027] The characteristic keywords of the update behavior and the semantic features of the update behavior are filled into the prompt word template to generate a prompt word.

[0028] Preferably, extracting key information and key information semantic features from the first update object based on the prompt word to generate reference metadata specifically includes:

[0029] Extracting key information basis from the first update object based on the generated prompt word;

[0030] Extract key information fields, key information feature keywords and semantic features from key information basis as the latest metadata features;

[0031] Map the latest metadata features to predefined metadata fields, standardize the mapped metadata fields, and obtain structured metadata information;

[0032] Transform structured metadata information into reference metadata through predefined metadata rules or templates.

[0033] Preferably, the verifying the updated metadata associated with the first update object based on the reference metadata to generate a metadata verification report includes:

[0034] Preset verification rules, and identify difference features between reference metadata and updated metadata associated with the first update object based on the verification rules;

[0035] A metadata verification report for different users is generated based on the difference characteristics.

[0036] Preferably, the aggregating user feedback information and generating update rules based on the feedback information includes:

[0037] Marking key information fields in the feedback information based on characteristic keywords;

[0038] Extracting key semantic features and question types in the feedback information through semantic analysis;

[0039] The key semantic features and common features and differences of question types in feedback information of different users are obtained, and update rules are generated according to the common features and differences.

[0040] Preferably, generating a metadata update instruction, a version number and an update key based on the mapping relationship between the update rule and the update record includes:

[0041] Determining update instructions based on update rules and update record associations;

[0042] The update instructions are classified according to the metadata fields corresponding to the update instructions and the metadata field relevance, and the version number corresponding to each type of update instruction is generated according to the preset semantic version number rules;

[0043] Generate update keys based on metadata management permissions, user level and version number;

[0044] and / or,

[0045] The parsing update instruction, the version number and the update key to update the second update object includes:

[0046] After receiving the update instruction, parse the semantic information in the instruction to determine the specific content and steps of the update;

[0047] Update metadata version control information according to the version number rules, and verify the completion of the update task based on the version control information;

[0048] Parse the update key information and check the update permissions. If the user's permissions match the update behavior, perform the update operation. If the user's permissions do not match the update behavior, automatically feedback the update failure information and notify the superior user to perform the update.

[0049] According to a second aspect of the present application, a metadata updating device for a scientific data set is provided, comprising:

[0050] An update record generation module, used to obtain the first update object and the corresponding update behavior in real time and generate an update record;

[0051] An update log generation module, used to generate an update log by comparing the information difference before and after the update based on the update record when the update behavior is completed;

[0052] A prompt word generation module, used for presetting a prompt word template and generating a prompt word based on the prompt word template and the update log;

[0053] A reference metadata generation module, configured to extract key information and key information semantic features from the first update object based on the prompt word, and generate reference metadata;

[0054] A verification module, configured to verify the updated metadata associated with the first update object based on the reference metadata and generate a metadata verification report;

[0055] An update rule generation module is used to distribute metadata verification reports to users associated with scientific datasets, obtain and summarize user feedback information, and generate update rules based on the feedback information;

[0056] An update instruction generation module, used to generate a metadata update instruction, a version number and an update key based on a mapping relationship between the update rule and the update record;

[0057] An update module, used for parsing the update instruction, the version number and the update key to update the second update object;

[0058] The first update object includes a scientific data set and / or a matching information carrier matching the scientific data set and / or metadata associated with the scientific data set;

[0059] The second update object is the scientific data set, supporting information carrier, and metadata corresponding to the first update object.

[0060] According to a third aspect of the present application, an electronic device is provided, including:

[0061] a memory for storing one or more computer programs;

[0062] The processor implements the metadata updating method of a scientific dataset according to the first aspect when the one or more computer programs are executed by the processor.

[0063] According to a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement a metadata update method for a scientific dataset according to the first aspect when executed.

[0064] Based on any of the above aspects, the embodiments of the present application provide a metadata update method, device, electronic device and storage medium for a scientific data set, which obtains data update behavior in real time, automatically generates update records, intelligently extracts key information, generates reference metadata and performs metadata verification. At the same time, the update instructions are combined with version numbers and update keys to achieve metadata update permission management and version control, thereby ensuring the accuracy, security and integrity of metadata updates, thereby improving the efficiency and quality of data management. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0066] Figure 1 A flowchart of a metadata updating method for a scientific dataset provided in this application.

[0067] Figure 2-9A flowchart of the specific steps of a metadata update method for a scientific dataset provided in this application.

[0068] Fig.10 A schematic diagram of functional modules of a metadata updating device for a scientific dataset provided in an embodiment of the present application.

[0069] Fig.11 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0070] Fig.12 A schematic application scenario diagram of a metadata update method for a scientific dataset provided in an embodiment of the present application.

[0071] Icons: 100, server; 200, terminal; 1001, update record generation module; 1002, update log generation module; 1003, prompt word generation module; 1004, reference element generation module; 1005, verification module; 1006, update rule generation module; 1007, update instruction generation module; 1008, update module; 710, electronic device; 711, memory; 712, processor; 713, communication module; 714, input / output interface; 715, bus. DETAILED DESCRIPTION

[0072] The present invention provides a method for updating metadata of a scientific data set. To make the purpose, technical solution and effect of the present invention clearer and more specific, the present invention is further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0073] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0074] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with those in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as here.

[0075] The application content is further explained below through the description of embodiments in conjunction with the accompanying drawings.

[0076] The drawings of the present invention are only for illustrative purposes and should not be construed as limiting the present invention. In order to better illustrate the following embodiments, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product; it is understandable to those skilled in the art that some well-known structures and their descriptions in the drawings may be omitted.

[0077] On the one hand, the present application provides a method for updating metadata of a scientific data set, such as Figure 1 As shown, the technical solution provided by the present application is described in detail below in conjunction with specific embodiments.

[0078] S1. Acquire a first update object and a corresponding update behavior in real time, and generate an update record.

[0079] In this embodiment, the first update object includes a scientific data set and / or a matching information carrier matching the scientific data set and / or metadata associated with the scientific data set. In the specific implementation process, when the scientific data set, the matching information carrier matching the scientific data set, and the metadata associated with the scientific data set are updated, the metadata in the other two first update objects need to be updated accordingly.

[0080] The update behavior may involve any one or more objects of scientific datasets, supporting information carriers, or metadata.

[0081] Exemplarily, the supporting information carrier may be an experimental report, a scientific report, a project implementation report, a project task book, a data submission plan, a data management plan, a data quality report, a data-related paper, a data service case introduction, etc. that are compatible with the scientific data set.

[0082] Exemplarily: only a certain data record in the scientific dataset is updated; only the experimental report in the supporting information carrier is modified; and the scientific dataset and metadata are updated simultaneously (for example, data fields are modified and related description documents are updated).

[0083] Update behavior refers to specific operations performed on datasets, supporting information, or metadata. These update behaviors include but are not limited to:

[0084] (1) Scientific dataset updates:

[0085] Adding data records: adding new data items to the data set.

[0086] Modification of data records: updating the value of an existing data item, such as adjusting the numerical value of an experimental result.

[0087] Deletion of data records: Deleting data items that are no longer needed from a data set.

[0088] (2) Update of supporting information carriers:

[0089] Document editing: Modify the content of the experimental report or data quality description document.

[0090] New additions to the documentation: Add new supplementary documentation, such as instructions for using the dataset.

[0091] Deletion of documents: Delete documents that are outdated or no longer applicable.

[0092] (3) Metadata update:

[0093] Adding fields: Add new description fields to the data set, such as adding an "experimental date" field.

[0094] Field modification: Change the description or data type of an existing field, such as modifying the description of the "Experimental Results" field.

[0095] Field Deletion: Remove metadata fields that are no longer needed.

[0096] All of these updates affect the metadata fields.

[0097] S2. When the updating behavior is finished, based on the update record, the information difference before and after the updating is compared to generate an update log.

[0098] In a specific embodiment, step S2 may include the following specific steps:

[0099] S21, for each first updated object, obtaining metadata update records of each updated version thereof, extracting metadata features and update behavior features based on the metadata update records, and generating a correlation matrix of metadata features and update behavior features according to the metadata features and update behavior features;

[0100] For example, in the "Climate Change Monitoring Project", the scientific data set is "2024 temperature observation data", and there is a data management plan (DMP) as a supporting information carrier, which details the methods of data collection, storage and processing. Currently, it is necessary to update the metadata of "2024 temperature observation data" and generate an update record.

[0101] The historical version records of this dataset include:

[0102] Version 1 (initial version): Contains basic fields such as "Observation time" (field type: date and time, field description: record observation time).

[0103] Version 2 (adding data source description): Adding a new field "Data Source" and modifying the field description of "Observation Time" to "record the specific time of temperature observation".

[0104] Version 3 (updated observation location information): Added a new field "Observation location" and updated the "Data source" field with a detailed description.

[0105] When generating the association matrix between metadata features and update behavior features, the following can be done:

[0106] Extract metadata features: Analyze historical version records and extract metadata features, such as field name, field type, and field description. For example, extract the "observation time" field information from version 1, and then extract the "data source" and "observation location" field information from versions 2 and 3.

[0107] Extract update behavior features: Extract the corresponding update behavior features based on the changes in each version. For example, from version 1 to version 2, the modification behavior of field descriptions ("modify field descriptions") and the addition behavior of new fields ("add new data fields") should be recorded. Similarly, from version 2 to version 3, the changes in new fields and field descriptions (for example, the addition of "observation location" and the update of "data source") should also be captured. Construct an association matrix: Based on the extracted metadata features and update behavior features, generate a feature association matrix containing the update trajectories of historical version records.

[0108] S22, presetting a mapping rule between a scientific data set, a supporting information carrier and metadata, and acquiring metadata features associated with the update behavior features in real time based on the mapping rule;

[0109] The establishment of mapping rules requires the clarification of the specific types of scientific datasets and their related attributes. There are different definitions for scientific datasets in different fields. Mapping rules are formulated based on the attributes of the scientific datasets and supporting information carriers determined above. Mapping rules may include but are not limited to field mapping rules, format and standardization rules, and update rules. For example, the field mapping rules may determine that the dataset name in the scientific dataset corresponds to the data description in the data submission plan as the supporting information carrier, the data creation date in the scientific dataset corresponds to the data acquisition date in the data quality report as the supporting information carrier, and the data source as the supporting information carrier corresponds to the data source in the data management plan as the supporting information carrier. Update rules may specify how to update between scientific datasets, supporting information carriers, and metadata. For example, when a data file or metadata field in a scientific dataset is updated, the relevant fields in the supporting information carrier (such as the data quality report) must also be updated synchronously.

[0110] Exemplarily, a mapping rule defines the relationship between "2024 temperature observation data" as a scientific data set and the metadata fields in the data management plan as a supporting information carrier. For example, the description and format requirements of the "observation time" field are specified in the data management plan. When an update to the "observation time" field is detected, relevant metadata features are extracted according to the mapping rules, such as the field's data type ("date time"), description ("specific time to record temperature observations"), and format requirements ("YYYY-MM-DD HH:MM").

[0111] S23, determining the current metadata state and update requirements based on the metadata features associated with the update behavior features and the association matrix;

[0112] For example, the current version of the scientific dataset "2024 Temperature Observation Data" contains an "Observation Time" field, which is described as "Record observation time", but no format requirements are specified. Based on the association matrix and metadata features, it is identified that the description of the "Observation Time" field needs to be updated to include format requirements. For example, update from "Observation Time" to "Record the specific time of temperature observation, in the format of YYYY-MM-DD HH:MM.

[0113] S24: Annotate the associated update behavior characteristics based on the current metadata state and update requirements, and generate an update record.

[0114] S25. Obtain metadata features corresponding to scientific data sets and metadata in historical versions based on the update records, compare information differences of metadata features before and after the update behavior, obtain metadata update type, update content, feature keywords of the update behavior, and semantic features of the update behavior to generate an update log.

[0115] S3. Preset a prompt word template, generate prompt words based on the prompt word template and the update log, extract key information and key information semantic features from the first update object based on the prompt words, and generate reference metadata.

[0116] The preset prompt word template defines how to describe and identify data updates. The template may include the following fields: task description "data fields that need to be updated and their descriptions", update target "specific requirements for updated content", update conditions "such as field value range or format requirements", and associated fields "fields related to the updated content".

[0117] In a specific embodiment, the preset prompt word template may include the following specific steps:

[0118] S311, obtaining the overall task and subtasks of metadata update;

[0119] Exemplarily, the overall metadata update task is to update the field descriptions in the scientific dataset "2024 temperature observation data" to ensure that the metadata of the scientific dataset can accurately reflect actual application needs and data submission plans, including: updating data domain field descriptions based on data application cases; updating the "observation time" field description based on the data submission plan.

[0120] The overall task is divided into several subtasks according to semantic features and task data dependencies, such as:

[0121] (1) Data domain field description update: Update the description of the data domain field according to the domain requirements defined in the data application case as the supporting information carrier.

[0122] (2) Update of the description of the observation time field: Update the description of the “observation time” field according to the standards specified in the data submission scheme as the supporting information carrier.

[0123] (3) Verify updates: Ensure that all updates meet the requirements of the application case and submission plan and are verified.

[0124] S312, determining the subtask execution steps based on the update target characteristics, update condition standards, update content basis and metadata characteristics associated with the update behavior characteristics corresponding to each subtask;

[0125] For example, analyzing the characteristics of the "Data Domain Field Description" update subtask yields:

[0126] (1) Update target features: Update the description of data domain fields to make them meet the requirements of the latest data application cases.

[0127] (2) Update condition standards: Field descriptions need to be consistent with the definitions in the data application case, including the accuracy of terminology and description.

[0128] (3) Update content based on: field definitions and requirements in the data application case document.

[0129] (4) Associated metadata features: original field description, field data type, and usage scenarios.

[0130] Get the task execution steps:

[0131] (1) Consult the data application case documentation to understand the specific description requirements of the data domain fields.

[0132] (2) Compare the current field description with the requirements in the application case and determine the parts that need to be updated.

[0133] (3) Modify the field description to ensure that the description accurately reflects the requirements in the application case.

[0134] (4) Save and record the changes and prepare for verification.

[0135] S313: Set the subtask start conditions, output format and task completion judgment criteria.

[0136] According to the intersection of the starting conditions and output results between subtasks, it can be split into several parallel prompt projects and apply heuristic search, for example:

[0137]

[0138] The driving method and heuristic search implementation of subtask 2 are as follows:

[0139] Subtask 1:

[0140] Input the latest data application case documents;

[0141] Use the prompt word template to process the data application case documents and extract all relevant field description features;

[0142] Conduct frequency and correlation analysis on the extracted features to identify the main update needs;

[0143] Output fields describe a list of update requirements, sorted by feature frequency and relevance.

[0144] Driver subtask 2:

[0145] Input the sorted feature list (extracted from the output of subtask 1);

[0146] According to the sorted feature list, information related to these features is retrieved from a supporting information carrier (such as a data submission scheme);

[0147] Describe the update requirements for each observation time field related to feature extraction;

[0148] Update the observation time field description based on the extracted information;

[0149] Outputs the observation time field update record in XML format.

[0150] By using a heuristic search method, the output of subtask 1 (the field description update record in JSON format) is used as the driving information to provide the update start condition for subtask 2. The system sorts the features according to their frequency and relevance, automatically extracts the relevant observation time field description update requirements from the supporting information carrier, and processes them. This method optimizes the order and efficiency of task execution, ensuring that each task can be processed based on the most relevant information.

[0151] In a specific embodiment, generating a prompt word based on the prompt word template and the update log may include the following specific steps:

[0152] S321, obtaining the update type, update content, characteristic keywords of the update behavior and semantic features of the update behavior in the update log;

[0153] For example,

[0154] Update type: Modify metadata fields and add supporting information carriers

[0155] Update content: The sharing mode field is changed from "unconditional sharing" to "conditional sharing", and new characteristic keywords for data classification report update behavior are added: "sharing mode", "conditional sharing", "data classification report"

[0156] Semantic characteristics of update behavior: Sharing policy changes impose stricter sharing controls on certain types of data. Historical update records show that changes in data sharing policies usually lead to multiple updates of fields.

[0157] S322: Fill the characteristic keywords of the update behavior and the semantic features of the update behavior into a prompt word template to generate a prompt word.

[0158] For example,

[0159] Task 1: Find the data classification elements corresponding to "conditional sharing" in the reference of the "XX" standard specification document;

[0160] Task 2: Extract the content of the "Basis for Sharing Method Settings" section / paragraph in the data classification report;

[0161] Task 3: Whether the content generated in Task 2 involves changes in sharing policies, if so, output 1, if not, output 0

[0162] Task 4: When the output of Task 3 is 0, compare the classification factors obtained in Task 1 with the "basis for setting the sharing method" and output the key information of the "basis for setting the sharing method"

[0163] In a specific embodiment, extracting key information and key information semantic features from the first update object based on the prompt word and generating reference metadata may include the following specific steps:

[0164] S331: Extract key information basis from the first update object based on the generated prompt word.

[0165] For example, when extracting key information from a scientific data set, you can focus on data field definitions, actual data values, data formats, and data quality indicators. For a meteorological observation data set, key information includes the definitions of the "observation time" and "observation location" fields, the recorded temperature and humidity values, the time format, and the missing rate of data. This information is directly related to the structure and content of the metadata, and it is necessary to ensure that the metadata accurately reflects the actual information in the dataset.

[0166] It is also important to extract key information from supporting information carriers, such as data management plans and data quality reports. The data management plan provides data storage location and backup strategies, and the data quality report contains data verification standards and quality inspection processes. This information helps determine the storage location, quality control methods, and usage agreements in the metadata, ensuring that the metadata accurately reflects the management and quality standards of the data during the update process.

[0167] S332, extracting key information fields, characteristic keywords and semantic features of the key information in the key information basis as the latest metadata features;

[0168] Exemplarily, key information fields include the core field names, field types, and field formats in the dataset. For example, the "Observation Time" field may have a date format, while the "Temperature" field is a floating point type. These fields define the basic structure and storage method of the data, which directly affects the organization and understanding of metadata. Feature keywords cover vocabulary that describes the content of the data, such as "temperature", "humidity", etc., as well as classification information such as "meteorological data". These keywords help identify and describe the specific content of the data, and provide classification information for data items, which helps to accurately reflect the nature and purpose of the data in the metadata.

[0169] Semantic features involve the semantic logic of keywords extracted from language elements such as sentence structure and part-of-speech features in data descriptions. These features can reveal how data is collected, where the data comes from, and the correlation between data features. For example, by analyzing the semantic features of data fields, it is possible to determine the data collection method (such as real-time collection or batch processing) of the "observation time" field, the source of the data (such as satellite observation or ground station), and the logical relationship with other fields such as "temperature" and "humidity". This in-depth semantic analysis helps to build more accurate metadata descriptions that can more comprehensively reflect the actual application and context of the data.

[0170] S333, mapping the latest metadata features to predefined metadata fields, and performing standardization processing on the mapped metadata fields to obtain structured metadata information;

[0171] S334. Convert the structured metadata information into reference metadata using predefined metadata rules or templates.

[0172] Exemplarily, predefined metadata rules or templates are specifications used to ensure metadata standardization and consistency. For example, predefined rules may include field format requirements (such as dates must meet preset standards), data value ranges (such as temperature values ​​should be between -50 and 50 degrees Celsius), and other consistency check criteria. A metadata template is a structured table or format that presets the location, type, and content requirements of each field to ensure that all data conforms to the same format standards.

[0173] The conversion process involves applying these rules and templates to structured data (such as JSON or XML format) to generate standardized reference metadata. Specifically, the system adjusts the data format according to the rules, checks whether the field value conforms to the specified range, and then organizes the processed data into a format that conforms to the preset template, and finally generates standardized reference metadata. This process ensures the consistency, accuracy and availability of metadata.

[0174] S4. Verify the updated metadata associated with the first update object based on the reference metadata, generate a metadata verification report and distribute it to users associated with the scientific dataset, obtain and summarize user feedback information, and generate update rules based on the feedback information.

[0175] In a specific embodiment, verifying the updated metadata associated with the first update object based on the reference metadata and generating a metadata verification report may include the following specific steps:

[0176] S411, presetting a verification rule, and identifying difference features between the reference metadata and the updated metadata associated with the first update object based on the verification rule;

[0177] The sorting of differential features is based on multiple factors. First, the severity of the differential features is sorted. For example, features that affect the integrity and accuracy of field values ​​are usually given high priority. Second, the scope of impact is considered. Features that widely affect the entire dataset will be given higher priority. In addition, the difficulty of repair will also affect the sorting, and priority will be given to those issues that are relatively simple to repair or have a large impact but are easy to repair. This sorting method ensures that critical issues are prioritized during the metadata update process to improve overall data quality and consistency.

[0178] To facilitate analysis and processing, difference features are grouped into different categories. These categories include field types (such as numeric, date, and text), difference types (such as new fields, missing fields, format errors, and value range out of limit), and data properties (such as static data and dynamic data). This classification method helps the system quickly identify and focus on difference features of a specific category, thereby improving the efficiency of verification and repair. For example, for features in the "format error" category, the system can automatically apply format matching rules to correct errors, while for features in the "new field" category, the system may need to determine whether additional associated fields need to be added based on field dependency rules.

[0179] After the difference features are sorted and classified, the system associates these features with the preset metadata rules. Mapping rules include format matching, value range matching, and field dependency. For example, the system will associate the format error of the "Observation Time" field with the defined format matching rules and automatically adjust the format. At the same time, for features with temperature value ranges exceeding the limit, the system will remind users to check the data according to the value range matching rules. In addition, for the newly added field "humidity", the system will determine whether the user needs to provide additional information based on the field dependency rules. Through these steps, the system ensures the consistency and accuracy of metadata, and provides users with detailed metadata verification reports to guide their subsequent updates and corrections.

[0180] S412, generating metadata verification reports for different users based on the difference characteristics;

[0181] In this embodiment, when generating the metadata verification report for different users, the user type and the metadata management authority are also considered.

[0182] Exemplarily, the user type and metadata management authority specifically include:

[0183] (1) Data producer (data submitter):

[0184] Permissions: Ability to update fields related to data content (such as data format fields, data domain fields, and associated results fields), but for data identification fields, shared type fields, etc., only update suggestions can be submitted.

[0185] Verification report generation: The large language model generates a detailed report based on the verification rules, indicating the data content fields that need to be updated, as well as suggestions for identification fields and shared type fields. The report also reminds data producers which fields cannot be modified directly and need to submit suggestions or applications.

[0186] (2) Data reviewer:

[0187] Permissions: Can review and approve or reject update suggestions submitted by data producers, and have the final say over the data usage field and sharing type field.

[0188] Verification report generation: The verification report provides detailed audit suggestions for the data auditor's permissions, including modification suggestions for data usage and sharing types, emphasizing data consistency and sharing requirements. The report also lists all update suggestions to be reviewed to help auditors make decisions quickly.

[0189] (3) Data users:

[0190] Permissions: Ability to submit update suggestions for data usage fields, but cannot modify core metadata fields such as data identification and sharing type.

[0191] Verification report generation: For data users, the report lists in detail the current status and historical changes of the data usage fields, and encourages data users to submit suggestions. The large language model will provide targeted suggestions to users based on their usage and domain knowledge to improve data usage efficiency.

[0192] (4) System Administrator:

[0193] Permissions: Full access and modification permissions to all metadata fields, responsible for overall data quality and system operation.

[0194] Verification report generation: The report provides system administrators with comprehensive metadata status and quality analysis, including update suggestions and audit records submitted by all users. The system administrator's report focuses on system performance optimization and data consistency maintenance.

[0195] According to the characteristics of user type and metadata management authority, the corresponding difference feature sorting and mapping are obtained, including:

[0196] (1) High priority features may include:

[0197] Data Producers: Field format differences and field content differences are considered high priority as these directly impact the validity and usability of the data.

[0198] Data Reviewer: Data usage discrepancies and data identification discrepancies are given high priority to ensure the actual application and uniqueness of the data.

[0199] Data users: Differentiation of data usage is a high priority to ensure the applicability and accuracy of the data.

[0200] (2) Medium priority features can include

[0201] Data producers: Differences in sharing types affect data sharing agreements and access rights.

[0202] Data reviewer: Differences in data usage (medium priority), which affects the application scenarios of the data but is secondary to data identification.

[0203] Data users: Specific details of the proposed changes to the use of the data.

[0204] (3) Low priority features can include

[0205] System Administrator: A comprehensive review of all differential features. As a global perspective for system optimization and management, all differential features are included in the comprehensive review, but the priority of each feature depends on the specific issue.

[0206] Generate reports based on user type and permissions, specifically:

[0207] (1) The data producer (submitter) report may include:

[0208] Content: The report will highlight the differences in fields that data producers can modify (such as fields related to data content) and provide format and content correction suggestions. In addition, the report will also include a suggested submission process for fields that data producers do not have permission to modify.

[0209] Purpose: To guide data producers to efficiently complete metadata updates within their authority, and to provide recommended submission paths for fields that they cannot modify directly.

[0210] (2) The data reviewer report may include:

[0211] Content: The report lists in detail the update suggestions to be reviewed, including the difference characteristics of fields such as sharing type and data usage. Reviewers can approve or reject the suggestions of data producers based on this information. In addition, the report will provide the review criteria and basis for field modifications.

[0212] Purpose: To help data reviewers quickly and efficiently review update suggestions from data producers to ensure that all modifications comply with data management specifications.

[0213] (3) Data user reports may include:

[0214] Content: The report will focus on the current status and suggested changes to the data usage fields, including information such as data usage frequency and scenarios. In addition, the report will also prompt data users on how to submit suggestions for data usage updates.

[0215] Purpose: To provide data users with update guidance on data usage and encourage them to actively participate in submitting update suggestions for the data usage field.

[0216] (4) System administrator reports may include:

[0217] Content: The system administrator's report provides a comprehensive metadata status and quality analysis, including the sorting and mapping of all differential features. The report will describe in detail the modification suggestions and expected impact of each differential feature.

[0218] Purpose: To provide system administrators with comprehensive metadata management information to help them optimize system performance and data consistency.

[0219] At the same time, heuristic search is used to assist report generation, including the use of heuristic search algorithms by large language models to generate the start conditions of the next task based on the output of each subtask. The report generation process will set different heuristic search strategies based on user type and permissions to ensure that each user obtains report content that matches their permissions and roles.

[0220] Specifically, it is assumed that different types of users (data producers, data auditors, data users, and system administrators) need to generate metadata verification reports for the same dataset, and there are overlaps and data dependencies between report generation tasks.

[0221] For example, through task identification, we can know that data producers submit update suggestions, data reviewers review and generate verification reports, data users submit data usage modification suggestions, and platform administrators generate comprehensive reports. Determine the dependencies. The tasks of data reviewers depend on the updates of data producers, and the administrator's report needs to integrate the feedback of data producers and data reviewers.

[0222] Evaluate the distance from the current state to the target state, which can be defined as:

[0223] h(n)=w1i+w2j

[0224] Among them, n represents the current task and task status, including the task progress of all users in the current system and the dependencies between tasks. w1 and w2 are weight coefficients that reflect the importance of the number of tasks and dependencies. i represents the number of unprocessed tasks, and j represents the number of task dependencies.

[0225] For example, if the current state n indicates that there are three pending tasks (format correction by the data producer, review by the data reviewer, and modification of the purpose by the data user), and the task of the data reviewer depends on the correction by the data producer, this means:

[0226] The number of unprocessed tasks is 3.

[0227] The number of task dependencies is 1 (data auditor depends on data producer).

[0228] h(n) will reflect the complexity of the current state, which in turn affects the processing order of the priority queue and optimizes the entire report generation process.

[0229] Use priority queues to process tasks based on the heuristic value h(n) to ensure that high-priority tasks are completed first. Check all pending update suggestions and difference features, and for each user's task, analyze its intersection and dependency with other tasks to ensure that related tasks can be processed synchronously.

[0230] Exemplarily, the cost g(n) and the heuristic value h(n) are calculated for each extended state, and the state with the lowest cost is selected for the next step of processing, which is specifically expressed as:

[0231]

[0232] Where t(n) is the task processing time; L is the system load factor, which reflects the current system load (such as the utilization rate of CPU, memory and other resources). It can be d(n): the number of task dependency nodes, indicating the number of other tasks that the current task depends on; h(n): the historical average number of dependency nodes of similar tasks, used to compare the dependency complexity of the current task with historical data.

[0233] The extended state computation cost g(n) represents the computational overhead when executing metadata update tasks, which mainly reflects the amount of resources consumed by the system when executing these tasks. The calculation method of g(n) may be adjusted according to the dependencies, resource requirements and computational complexity of different tasks.

[0234] Task processing time is a core factor in computational cost. The processing time required for different metadata update tasks may vary greatly, depending on the complexity of the task and the size of the data. The processing time can be estimated using historical task execution records or prediction models.

[0235] For routine tasks (such as metadata field updates): make predictions based on the average processing time of similar tasks in the past.

[0236] For complex tasks (such as adding new fields, batch updating, etc.): classify the tasks according to their complexity and use the estimation model to calculate the expected execution time.

[0237] For example, if a metadata field update task involves modifying the field type, the processing time may be shorter; but if the task involves tracing historical metadata, the processing time will increase significantly.

[0238] The current load of the system is also an important factor affecting the calculation of g(n). If the system is under high load, the execution cost of even simple tasks will increase. For example, the usage of resources such as CPU, memory, and I / O will affect the execution efficiency of the task.

[0239] The number of dependent nodes reflects the degree of dependence of a task on other tasks during execution. By comparing historical data, we can determine whether the dependence of the current task is abnormal or exceeds expectations, and thus adjust the computing cost.

[0240] In other implementation schemes, heuristic search strategies can be specifically designed to assist report generation according to specific generation tasks, and intelligent heuristic search strategies can be designed through multi-dimensional evaluation and adaptive optimization.

[0241] For example, the weights in the heuristic function can be adjusted dynamically, based on historical task feedback, giving priority to high-frequency and high-impact metadata update tasks, and automatically optimizing parameters in combination with machine learning models. At the same time, multi-dimensional evaluation indicators such as task priority, user needs, and data quality are introduced, and a combined evaluation method (such as the analytic hierarchy process AHP) is used for comprehensive evaluation. Secondly, a multi-level task dependency graph is constructed to analyze direct and indirect dependencies, and task scheduling is optimized using graph algorithms. In addition, shared data nodes are established in the graph to reduce redundant calculations. Thirdly, dynamic scheduling is performed through priority queues, and the execution order is adjusted in real time to achieve load balancing and ensure efficient resource utilization. At the same time, data is processed in parallel with the help of a distributed computing framework to improve the speed of task execution. In addition, based on user historical behavior learning preferences, personalized report generation strategies are provided, supporting diversified report formats and custom template functions to meet the specific needs of different users.

[0242] In a specific embodiment, distributing the metadata verification report to users associated with the scientific dataset, obtaining and summarizing user feedback information, and generating update rules based on the feedback information specifically include:

[0243] S421, marking key information fields in the feedback information based on characteristic keywords;

[0244] Feature keywords are terms or phrases that frequently appear in user feedback and are related to metadata updates. These keywords can help identify specific issues or suggestions mentioned in user feedback. For example, in scientific dataset management, common feature keywords may include "field format", "data type", "update frequency", "data integrity", etc.

[0245] The system will automatically scan user feedback information and mark the corresponding key information fields according to the position of the characteristic keywords in the text. Key information fields refer to metadata fields that need attention or may need to be modified as mentioned in user feedback. The integrity of the context and the semantic position of the keywords need to be considered when marking. For example, in a sentence, if relevant descriptions appear before and after the keyword "update frequency" (such as "hope to increase" or "need to decrease"), the system needs to mark "update frequency" together with the context before and after it to maintain the integrity and accuracy of the information.

[0246] S422, extracting key semantic features and question types in the feedback information through semantic analysis;

[0247] S423: Acquire common features and differences of key semantic features and question types in feedback information from different users, and generate update rules according to the common features and differences.

[0248] For example, when a user gives feedback, "We found some problems when using metadata fields, especially in the 'reference format' and 'data source' fields. First, the 'reference format' field currently only contains the elements {data producer name} {data set name} {data creation agency} {data generation date} {data publishing agency} {data link}, but for our internal project management and external reference needs, more information needs to be added, such as {associated project name} {associated project number}, etc. Secondly, the 'data source' field currently only marks the basic original data acquisition method, but we hope to be able to explain in more detail, including the data collection device, collection frequency, collection method and specific data processing steps, so as to ensure the integrity and traceability of the data. In addition, we found that the reference format of some data does not meet our reference standards. This may be because the standards of different departments are inconsistent. We hope to have an automated way to verify and prompt these inconsistencies. In addition, for the reference of different data sets, we need to be able to automatically generate the appropriate reference format according to the type of data set to facilitate unified management and output. These problems may have a certain impact on the use of data in different application scenarios, and we hope to solve them through system optimization.

[0249] First, stop words are filtered for meaningless, non-scientific professional vocabulary involved in data management, such as "we", "hope", "include", etc.; a corresponding keyword list is created according to the characteristics of the metadata field, and the corresponding key information and named entities in the user feedback information are extracted according to keyword matching, such as [metadata field, reference format][associated project name, information field]; based on the dependency syntax, key semantic information "Add → [more information, reference format, associated project name, associated project number (modified)]" can be extracted; sentiment analysis of user feedback information is performed to obtain "neutral attitude, including improvement suggestions", and further obtain the feedback priority "general matters"; finally, the update requirement and the corresponding rule mapping are generated "Update requirement: the reference format field needs to be expanded to include associated project information Rule mapping: according to the 'reference format field expansion rule', an update instruction containing the associated project name and number is generated."

[0250] S5. Generate a metadata update instruction, a version number and an update key based on the mapping relationship between the update rule and the update record.

[0251] In this embodiment, based on the mapping relationship between the update rule and the update record, generating the metadata update instruction, the version number and the update key specifically includes:

[0252] S51, determining an update instruction based on the metadata update rule and the feature association of the update record;

[0253] Update record features need to be extracted according to the update rule feature fields. For example, the update rule is "Update field: data source; Trigger condition: when the data source field is empty or contains incomplete information; Update operation: automatically extract data source related information from scientific data sets and supporting information carriers, and update the field content". The update record is "The data source field lacks observation site information, and an add operation is performed on the data source field. The content before the update is 'Guangzhou, October 2022', and the content after the update is 'Guangzhou, October 2022-Shiziyang Waters Station'". The associated feature matching obtains [the data source field contains incomplete information, and the data source field lacks observation site information], and generates an update instruction "Metadata field: data source; Update operation: add field content; new information: Shiziyang Waters Station".

[0254] S52, classifying the update instructions according to the metadata fields corresponding to the update instructions and the metadata field relevance, and generating a version number corresponding to each type of update instruction according to a preset semantic version number rule;

[0255] For example, there is a metadata update requirement that involves the modification of multiple fields. The update instructions for this requirement include: updating the privacy fields of patient data (such as name and address); formatting medical imaging data fields; modifying the fields of patient treatment plans; updating the fields of patient medication information; and supplementing the fields of patient follow-up records. A version identifier is generated for each update instruction based on the preset feature identification rules, for example: privacy field update (privacy feature): PRIV; medical imaging data formatting (imaging feature): IMG.

[0256] Furthermore, for update rule features or metadata features corresponding to the update behavior, multiple sub-version identifiers are merged to generate an update batch number. For example, the update of the treatment plan field and the medication information update can be merged into an update batch number. When all update instructions for an update are completed, a version number is generated. If an update instruction cannot be completed, the update batch number is applied. The version number or batch number can preset encoding rules so that the update features can be obtained based on rule parsing, such as "CN (country) GDCC (creation organization) 20240815 (update date) V1 (dataset version number) THR-MED (update features: treatment plan update, medication information update)".

[0257] S53: Generate an update key based on the metadata management authority, user level and version number.

[0258] S6. Parse the update instruction, version number and update key to update the second update object.

[0259] In a specific embodiment, the parsing update instruction, the version number and the update key to update the second update object specifically includes:

[0260] S61, after receiving the update instruction, parsing the semantic information in the instruction to determine the specific content and steps of the update;

[0261] S62, updating metadata version control information according to the version number rule, and verifying the completion of the update task based on the version control information;

[0262] S63, parse the update key information, check the update authority, if the user's authority matches the update behavior, execute the update operation, if the user's authority does not match the update behavior, automatically feedback the update failure information, and notify the superior user to execute the update.

[0263] On the other hand, an embodiment of the present application provides a scientific data set data updating device, comprising:

[0264] An update record generating module 1001 is used to obtain a first update object and a corresponding update behavior in real time and generate an update record;

[0265] An update log generating module 1002 is used to generate an update log by comparing the information difference before and after the update based on the update record when the update behavior is completed;

[0266] A prompt word generating module 1003, used for presetting a prompt word template, and generating a prompt word based on the prompt word template and the update log;

[0267] A reference metadata generation module 1004 is used to extract key information and key information semantic features from the first update object based on the prompt word to generate reference metadata;

[0268] A verification module 1005, configured to verify the updated metadata associated with the first update object based on the reference metadata and generate a metadata verification report;

[0269] An update rule generation module 1006 is used to distribute the metadata verification report to users associated with the scientific dataset, obtain and summarize user feedback information, and generate update rules based on the feedback information;

[0270] An update instruction generation module 1007, used to generate a metadata update instruction, a version number and an update key based on the mapping relationship between the update rule and the update record;

[0271] The update module 1008 is used to parse the update instruction, version number and update key to update the second update object.

[0272] It can be understood that the above-mentioned system embodiment and the above-mentioned method embodiment can correspond to each other, and similar descriptions of the system embodiment can refer to the method embodiment. To avoid repetition, it will not be repeated here. A metadata update device for a scientific data set provided in an embodiment of the present application can execute a metadata update method for a scientific data set provided in any embodiment of the present application, and has functional modules and beneficial effects corresponding to the execution method. The functional modules of the metadata update device for a scientific data set can be implemented in the form of hardware, can be implemented by instructions in the form of software, and can also be implemented by a combination of hardware and software modules.

[0273] Optionally, the software module may be located in a random access memory, a read-only memory, a programmable read-only memory, a flash memory, an electrically erasable programmable memory, a register, or other storage media. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps in the above method embodiment in combination with its hardware.

[0274] The present application embodiment provides an electronic device 710, whose structure is as follows: Fig.11 The electronic device 710 may be an embodiment of the present application Fig.12 The server 100 or the terminal 200 is shown.

[0275] like Fig.11 As shown, the electronic device 710 includes a memory 711, a processor 712, a communication module 713 and an input / output interface 714, etc. Optionally, the memory 711, the processor 712, the communication module 713 and the input / output interface 714 can be connected and communicated through a bus 715.

[0276] The memory 711 is used to store one or more computer programs and transmit the code of the computer program to the processor 712; when the one or more computer programs are executed by the processor 711, a metadata update method for a scientific data set in an embodiment of the present application is implemented.

[0277] Optionally, the electronic device 710 can be connected to a network via a communication module 713 to communicate with other devices, such as a terminal or a server, through the network to achieve data interaction. The electronic device 710 can be various forms of digital computers, such as desktop computers, servers, workstations, mainframe computers or other types of computers. The electronic device 710 can also be various forms of mobile terminals, such as smart phones, tablet computers, wearable devices (such as helmets, glasses, watches, etc.) and other similar mobile terminals.

[0278] Optionally, the electronic device 710 can be connected to a required input / output device, such as a keyboard, a display device, etc., through the input / output interface 714. The electronic device 710 itself can have a display device, and can also be connected to other external display devices through the input / output interface 714. Optionally, a storage device, such as a hard disk, can be connected through the input / output interface 714, so that data in the electronic device 710 can be stored in the storage device, or data in the storage device can be read, and data in the storage device can also be stored in the memory 711.

[0279] It is understandable that the input / output interface 714 may be a wired interface or a wireless interface. Depending on the actual application scenario, the device connected to the input / output interface 714 may be a component of the electronic device 710 or an external device connected to the electronic device 710 when needed.

[0280] Optionally, the memory 711 may be a volatile memory and / or a non-volatile memory, the volatile memory may be a random access memory, etc., and the non-volatile memory may be a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory or a flash memory, etc.

[0281] Optionally, the computer program stored in the processor 711 may be divided into one or more modules, which are stored in the memory 711 and executed by the processor 712 to complete the method provided in the embodiment. The one or more modules may be a series of computer program instruction segments capable of completing specific functions, and the computer program instruction segments are used to describe the execution process of the computer program in the electronic device 710.

[0282] Optionally, the processor 712 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 712 include, but are not limited to, a central processing unit, a graphics processing unit, a digital signal processor, various special artificial intelligence computing chips, various processors that run machine learning model algorithms, and may also be any appropriate controller, microcontroller, processor, etc. The processor 712 executes various methods and processes of the embodiments of the present application, illustratively, such as a metadata update method for a scientific data set in an embodiment of the present application.

[0283] Optionally, the bus 715 may include a path for transmitting information. The bus 715 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. According to different functions, the bus 715 may be divided into an address bus, a data bus, a control bus, etc.

[0284] In an optional implementation, the embodiment of the present application further provides a computer storage medium on which a computer program is stored, and when the computer program is executed by a computer, the computer can perform the method of the above method embodiment. Part or all of the computer program can be loaded and / or installed on the memory 711 of the electronic device 710. When the computer program is executed by the processor 712, one or more steps of the metadata update method of a scientific data set in the embodiment of the present application can be performed.

[0285] Optionally, the computer-readable storage medium may be a random access memory, a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory, or the like.

[0286] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the technical solution of the present invention, and are not intended to limit the specific implementation methods of the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the claims of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A method for updating metadata of a scientific dataset, characterized in that: include: Acquire the first update object and the corresponding update behavior in real time, and generate an update record; When the updating behavior is finished, based on the update record, the information difference before and after the updating is compared to generate an update log; Preset a prompt word template, generate prompt words based on the prompt word template and the update log, extract key information and key information semantic features from the first update object based on the prompt words, and generate reference metadata; Verifying the updated metadata associated with the first update object based on the reference metadata, generating a metadata verification report and distributing it to users associated with the scientific dataset, obtaining and summarizing user feedback information, and generating update rules based on the feedback information; Based on the mapping relationship between the update rule and the update record, generate a metadata update instruction, a version number and an update key; Parsing the update instruction, the version number and the update key to update the second update object; The first update object includes a scientific data set and / or a matching information carrier matching the scientific data set and / or metadata associated with the scientific data set; The second update object is the scientific data set, supporting information carrier, and metadata corresponding to the first update object.

2. A method for updating metadata of a scientific dataset according to claim 1, characterized in that: The real-time acquisition of the first update object and the corresponding update behavior and the generation of the update record specifically includes: For each first updated object, obtaining metadata update records of each updated version thereof, extracting metadata features and update behavior features based on the metadata update records, and generating a correlation matrix between metadata features and update behavior features according to the metadata features and update behavior features; Preset mapping rules of scientific data sets, supporting information carriers and metadata, and obtain metadata features associated with the update behavior features in real time based on the mapping rules; Determine the current metadata state and update requirements based on the metadata features associated with the update behavior features and the association matrix; Based on the current metadata state and update requirements, the associated update behavior characteristics are marked to generate an update record.

3. A method for updating metadata of a scientific dataset according to claim 2, characterized in that: The step of generating an update log by comparing the information difference before and after the update based on the update record specifically includes: Obtain metadata features corresponding to scientific datasets and metadata in historical versions based on update records; The information difference of the metadata features before and after the update behavior is compared, and the update type, update content, feature keywords of the update behavior and semantic features of the update behavior of the metadata are obtained to generate an update log.

4. The method for updating metadata of a scientific dataset according to claim 1, characterized in that: The preset prompt word template includes: Get the overall task and subtasks of metadata update; Determine the subtask execution steps based on the update target characteristics, update condition standards, update content basis and metadata characteristics associated with the update behavior characteristics corresponding to each subtask; Set the subtask's start conditions, output format, and task completion criteria.

5. The method for updating metadata of a scientific dataset according to claim 3, characterized in that: Generating a prompt word based on the prompt word template and the update log includes: Acquire the update type, update content, characteristic keywords of the update behavior and semantic features of the update behavior in the update log; The characteristic keywords of the update behavior and the semantic features of the update behavior are filled into the prompt word template to generate a prompt word.

6. A method for updating metadata of a scientific dataset according to claim 1, characterized in that: Extracting key information and key information semantic features from the first update object based on the prompt word and generating reference metadata specifically includes: Extracting key information basis from the first update object based on the generated prompt word; Extract key information fields, key information feature keywords and semantic features from key information basis as the latest metadata features; Map the latest metadata features to predefined metadata fields, standardize the mapped metadata fields, and obtain structured metadata information; Transform structured metadata information into reference metadata through predefined metadata rules or templates.

7. A method for updating metadata of a scientific dataset according to claim 6, characterized in that: The verifying the updated metadata associated with the first update object based on the reference metadata to generate a metadata verification report includes: Preset verification rules, and identify difference features between reference metadata and updated metadata associated with the first update object based on the verification rules; A metadata verification report for different users is generated based on the difference characteristics.

8. A method for updating metadata of a scientific dataset according to claim 7, characterized in that: The aggregating user feedback information and generating update rules based on the feedback information includes: Marking key information fields in the feedback information based on characteristic keywords; Extracting key semantic features and question types in the feedback information through semantic analysis; The key semantic features and common features and differences of question types in feedback information of different users are obtained, and update rules are generated according to the common features and differences.

9. The method for updating metadata of a scientific dataset according to claim 1, characterized in that: Based on the mapping relationship between the update rule and the update record, a metadata update instruction, a version number and an update key are generated, including: Determining update instructions based on update rules and update record associations; The update instructions are classified according to the metadata fields corresponding to the update instructions and the metadata field relevance, and the version number corresponding to each type of update instruction is generated according to the preset semantic version number rules; Generate update keys based on metadata management permissions, user level and version number; and / or, The parsing update instruction, the version number and the update key to update the second update object includes: After receiving the update instruction, parse the semantic information in the instruction to determine the specific content and steps of the update; Update metadata version control information according to the version number rules, and verify the completion of the update task based on the version control information; Parse the update key information and check the update permissions. If the user's permissions match the update behavior, perform the update operation. If the user's permissions do not match the update behavior, automatically feedback the update failure information and notify the superior user to perform the update.

10. A metadata updating device for a scientific data set, characterized in that: include, An update record generation module, used to obtain the first update object and the corresponding update behavior in real time and generate an update record; An update log generation module, used to generate an update log by comparing the information difference before and after the update based on the update record when the update behavior is completed; A prompt word generation module, used for presetting a prompt word template and generating a prompt word based on the prompt word template and the update log; A reference metadata generation module, configured to extract key information and key information semantic features from the first update object based on the prompt word, and generate reference metadata; A verification module, configured to verify the updated metadata associated with the first update object based on the reference metadata and generate a metadata verification report; An update rule generation module is used to distribute metadata verification reports to users associated with scientific datasets, obtain and summarize user feedback information, and generate update rules based on the feedback information; An update instruction generation module, used to generate a metadata update instruction, a version number and an update key based on a mapping relationship between the update rule and the update record; An update module, used for parsing the update instruction, the version number and the update key to update the second update object; The first update object includes a scientific data set and / or a matching information carrier matching the scientific data set and / or metadata associated with the scientific data set; The second update object is the scientific data set, supporting information carrier, and metadata corresponding to the first update object.

Images