Automatic updating method and device for metadata of scientific data set, electronic equipment and storage medium

Abstract

The invention discloses an automatic updating method and device for metadata of a scientific data set, electronic equipment and a storage medium. The method comprises the steps that a scientific data set portrait is constructed according to information submitted by a data producer, user behavior group image features are extracted according to user data interaction behavior records associated with a scientific data set, and scientific data set sharing use features are extracted according to scientific data set sharing use records; generating a feature change index based on the user behavior group image feature and the sharing use feature, performing variable difference analysis based on the feature change index and the corresponding scientific data set portrait, generating a metadata update demand index and generating a metadata update trigger condition, and automatically updating the metadata when the metadata update trigger condition is met. Through an automatic metadata updating mechanism, in combination with scientific data set portrait construction, user behavior group image feature extraction and shared use feature analysis, the intelligent level and the use efficiency of data management are improved.

Description

Technical Field

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

[0002] Scientific data sets usually include raw data and its metadata. Metadata describes the content, structure and background information of the data and is essential for data retrieval, management and sharing. However, as data is constantly used and changed, the original metadata may gradually fail to meet new needs, resulting in reduced data usage efficiency and even affecting the accuracy and effectiveness of scientific research. Summary of the invention

[0003] To solve at least one of the above-mentioned defects, the present invention proposes a method, device, electronic device and storage medium for automatically updating metadata of scientific data sets, so as to improve the management efficiency of metadata of scientific data sets and better promote the dissemination and sharing of scientific data.

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

[0005] Build scientific dataset profiles based on structured or unstructured information submitted by scientific data producers;

[0006] Extract user behavior group features based on user data interaction behavior records associated with scientific data sets;

[0007] Extract the shared use characteristics of scientific datasets based on the shared use records of scientific datasets;

[0008] Generate a feature change index based on the user behavior group portrait feature and the shared usage feature, perform variable difference analysis based on the feature change index and its corresponding scientific data set portrait, generate a metadata update demand index, and generate a metadata update trigger condition based on the metadata update demand index;

[0009] When the metadata update triggering condition is met, the metadata is automatically updated.

[0010] Furthermore, the structured or unstructured information submitted by the scientific data producer includes the original data, metadata and associated supporting information carriers of the scientific data set;

[0011] The scientific data set portrait includes a scientific data set description portrait and a scientific data set feature portrait;

[0012] The scientific data set image is constructed based on the structured or unstructured information submitted by the scientific data producer. The specific process includes:

[0013] Classify the raw data, metadata and associated supporting information carriers of scientific datasets according to the domain type of scientific datasets;

[0014] Extract the naming structure and multi-dimensional features of metadata of various types of raw data, where the naming structure includes the naming field content, field order, folder naming hierarchy and naming relationship;

[0015] Extract data resource description information from various supporting information carriers, obtain corresponding data resource description information based on the naming field content of the original data, and construct description portraits of various scientific data sets according to the naming hierarchy;

[0016] Based on the multi-dimensional features of the metadata, establishing a portrait generation task through a portrait generation engine;

[0017] Based on the portrait constituent elements and metadata fields in the portrait generation task, the associated data resources in the scientific dataset are searched, and a feature portrait of the scientific dataset is generated.

[0018] Furthermore, extracting user behavior group image features based on user data interaction behavior records associated with the scientific data set includes:

[0019] Obtain the data access, query, download, and processing behavior logic and behavior association semantics of users associated with scientific datasets, and build a user behavior dataset based on user interaction behavior;

[0020] Based on the user behavior data set, construct a profile variable of the user behavior characteristics, wherein the behavior characteristics include behavior type characteristics, behavior frequency characteristics, behavior pattern characteristics and behavior preference characteristics;

[0021] Perform cluster analysis and feature extraction on different types of portrait variables to obtain the user behavior group portrait characteristics of each type of user group;

[0022] Furthermore, the extracting of the scientific dataset sharing and use features according to the scientific dataset sharing and use records includes:

[0023] Acquire similar interactive behavior data associated with scientific data sets, extract the scientific data set propagation sharing path, and generate a user node sharing network diagram, wherein the user node sharing network diagram includes a main node and nodes, wherein the main node is a main user or scientific data set, and the node is a user related to the main node;

[0024] Obtain multi-user shared samples in the scientific dataset sharing and use records, extract the scientific dataset sharing path in the multi-user shared samples as a subgraph of the main node of the user node sharing network graph;

[0025] The shared usage features of scientific datasets are extracted by performing graph network analysis on the user node shared network graph and the subgraph of the main node of the shared network graph.

[0026] Furthermore, generating a feature change index based on the user behavior group portrait feature and the shared usage feature, performing variable difference analysis based on the feature change index and its corresponding scientific data set portrait, generating a metadata update demand index, and generating a metadata update trigger condition based on the metadata update demand index, includes:

[0027] Obtaining the user behavior group image features associated with the scientific data set and the shared usage features, and generating a feature change index;

[0028] Perform difference analysis based on the feature change index and its corresponding scientific data set portrait variable to generate an associated metadata update demand index;

[0029] A metadata update trigger condition is generated according to a preset metadata update rule, a threshold value and the metadata update demand index.

[0030] Furthermore, when the metadata update triggering condition is met, the metadata is automatically updated, including:

[0031] generating a metadata update instruction including a priority and a metadata field hierarchy based on the metadata update trigger condition;

[0032] Execute the update instruction, monitor the completion of the update task node in real time, and generate metadata update records;

[0033] A metadata update report is generated based on a preset prompt word template, the metadata update record and a metadata update trigger condition corresponding to each update.

[0034] Furthermore, the method further includes deploying a metadata update demand-aware network, and deploying the metadata update demand-aware network according to the scientific dataset profile and the metadata update trigger condition, specifically including:

[0035] Constructing an update feature association matrix, the update feature association matrix matches the image of the scientific dataset with the metadata update trigger condition, and records the correspondence and semantic relevance between the image of the scientific dataset and the metadata update trigger condition,

[0036] Building an update requirement perception model for identifying scientific dataset update requirements based on the update feature association matrix;

[0037] Based on the update demand perception model, variable analysis is performed on the scientific data set portrait to generate update requirements. The update requirements are compared with the metadata update trigger conditions to obtain difference features. Based on the difference features, adjusted update instructions are generated. The adjusted update instructions are used to update erroneous metadata.

[0038] Furthermore, the method further comprises deploying a metadata update feedback network, and the deployment process of the metadata update feedback network comprises:

[0039] When the metadata update instruction is executed, the metadata update instruction is synchronously sent to each associated user to obtain feedback on the metadata update triggering conditions;

[0040] When the metadata update is completed, the metadata update record is sent to the users associated with each scientific dataset to obtain metadata effect evaluation opinions;

[0041] Based on the user metadata management authority and user characteristics, key information in the feedback opinions as well as common features and individual features of the key information are extracted, and the metadata update demand perception model is modified according to the common features and the individual features.

[0042] According to a second aspect of the present application, a device for automatically updating metadata of a scientific data set is provided, comprising:

[0043] A scientific data set portrait construction module is used to construct a scientific data set portrait based on the structured or unstructured information submitted by scientific data producers;

[0044] A user behavior group image feature extraction module associated with a scientific data set is used to extract user behavior group image features based on user data interaction behavior records associated with a scientific data set;

[0045] A scientific dataset sharing and use feature extraction module is used to extract scientific dataset sharing and use features based on scientific dataset sharing and use records;

[0046] A metadata automatic update trigger module, used to generate a feature change index based on the user behavior group portrait feature and the shared use feature, perform variable difference analysis based on the feature change index and its corresponding scientific data set portrait, generate a metadata update demand index, and generate a metadata update trigger condition based on the metadata update demand index;

[0047] The metadata automatic updating module is used to automatically update the metadata when the metadata update triggering condition is met.

[0048] According to a third aspect of the present application, an electronic device is provided, including:

[0049] a memory for storing one or more computer programs;

[0050] The processor implements the method for automatically updating metadata of a scientific dataset according to the first aspect when the one or more computer programs are executed by the processor.

[0051] 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 method for automatically updating metadata of a scientific dataset according to the first aspect above when executed.

[0052] Based on any of the above aspects, the embodiments of the present application provide a method, device, electronic device and storage medium for automatically updating metadata of a scientific data set, which realizes efficient and accurate updating of metadata through an automated metadata update mechanism, combined with the construction of scientific data set portraits, extraction of user behavior group features and analysis of shared usage features. By using a dynamic perception network and a feedback analysis network, the method can identify and respond to changes in data and user needs in real time, ensuring that metadata is always consistent with the actual data status and user needs, greatly improving the intelligence level and usage efficiency of data management. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] 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.

[0054] Figure 1 A flowchart of a method for automatically updating metadata of a scientific dataset provided in this application.

[0055] Figure 2-6 A flowchart of the specific steps of a method for automatically updating metadata of a scientific dataset provided in this application.

[0056] Figure 7 A flowchart of the specific steps for deploying a metadata update demand-aware network provided in this application.

[0057] Figure 8 A flowchart of the specific steps for deploying a metadata update feedback network provided in this application.

[0058] Fig. 9 A schematic diagram of functional modules of a device for automatically updating metadata of a scientific dataset provided in an embodiment of the present application.

[0059] Fig.10 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0060] Fig.11 A schematic application scenario diagram of a method for automatically updating metadata of a scientific dataset provided in an embodiment of the present application.

[0061] Icons: 100, server; 200, terminal; 901, scientific data set portrait construction module; 902, scientific data set associated user behavior group portrait feature extraction module; 903, scientific data set shared use feature extraction module; 904, metadata automatic update trigger module; 905, metadata automatic update module; 710, electronic device; 711, memory; 712, processor; 713, communication module; 714, input / output interface; 715, bus. DETAILED DESCRIPTION

[0062] The present invention provides a method for automatically updating metadata of a scientific data set. In order 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.

[0063] 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.

[0064] 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.

[0065] The application content is further explained below through the description of embodiments in conjunction with the accompanying drawings.

[0066] 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.

[0067] On the one hand, the present application provides a method for automatically 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.

[0068] S1. Build a scientific dataset profile based on the structured or unstructured information submitted by scientific data producers.

[0069] The core elements of scientific data set portraits include data description information (such as data name, structure and source, and detailed description of the basic attributes of the data), data quality information (involving data accuracy, completeness, reliability and update frequency, ensuring that data quality meets expected standards), data context information (describing the application scenarios of the data, related scientific research projects and their background information, and helping to understand the actual application value of the data) and metadata (such as field definitions, data relationships, data standards and constraints, and providing detailed technical specifications of the data). The structured information submitted by scientific data producers usually includes data resource lists, metadata tables, field definitions, data dictionaries, and proper noun tables. This information is presented in a standardized format to provide clear data structure and attributes. Unstructured information includes scientific research papers, technical reports, project specifications, data management plans, etc. These documents provide background, instructions for use, and other supplementary information for the data.

[0070] In a specific embodiment, step S1 may include the following specific steps:

[0071] S11. Classify the original data, metadata and associated supporting information carriers of scientific data sets according to the domain type of scientific data sets, extract the naming structure of the original data and the multi-dimensional features of the metadata,

[0072] The naming structure includes the naming field content, field order, folder naming hierarchy and naming relationship;

[0073] For example, we are dealing with a scientific dataset on rice variety selection. This scientific dataset contains raw data, metadata, and related supporting information carriers at different breeding stages. The raw data include seed genome sequencing data, phenotypic observation data (such as plant height, leaf width, yield, etc.), disease resistance test data, etc. Metadata include data collection time, location, collection personnel, experimental conditions (such as temperature, humidity, soil conditions, etc.), measurement equipment, etc. A scientific dataset contains multiple data folders, each folder has subfolders, and each folder has several data files. The naming structure refers to the naming field content, field order, and naming hierarchy and naming relationship of the data file or data folder. For example, the first-level folder Phenotypic_Data stores data related to phenotypes, such as field trial results and laboratory analysis reports. The second-level folder Field_Trial_Results has a naming structure of Location_TrialYear_ResultType.csv, and the field content includes the trial location, year, and result type, which clearly distinguishes different trial data. These naming information clarifies the nature, source, and generation time of the data through the field order and content. In the process of building a scientific dataset portrait, feature keywords are extracted from the naming structure of the scientific dataset. By parsing named fields (such as sample ID, date, location, etc.), scientific data set portraits containing multi-dimensional features such as genetic background, phenotypic performance and environmental adaptability can be automatically generated. These portraits provide detailed data information support for the analysis and decision-making of crop variety breeding. For example, based on the feature keywords extracted from Phenotypic_Data / Field_Trial_Results / Location_TrialYear_ResultType.csv, a detailed portrait of the trial results at a specific location and year can be generated to help researchers understand the performance of the variety under different environmental conditions.

[0074] S12. Extract the data resource description information from various supporting information carriers, obtain the corresponding data description according to the naming fields of the original data, and construct description portraits of various scientific data sets according to the naming hierarchy;

[0075] Corresponding to the multi-dimensional features of metadata extracted from the naming structure, the descriptive information of the data resources is further extracted from the supporting information carriers. For rice variety breeding data, the supporting information carriers may include breeding plans, experimental design documents, standard operating procedures (SOPs), breeding goal descriptions and other documents. Data resource description information is extracted based on these documents, such as breeding goals: improving disease resistance, increasing yield, and optimizing growth period. Breeding methods: genome editing, hybrid breeding, stress resistance testing.

[0076] This information is matched with the naming field characteristics of the original data, and a description portrait of the scientific dataset is constructed based on the naming hierarchy of these fields (such as "genome->chromosome->gene locus"), which shows the data and detailed description of each stage in the rice variety breeding process.

[0077] S13, establishing a portrait generation task according to the portrait generation engine corresponding to each dimension feature of the metadata;

[0078] Generate a portrait generation task based on the characteristics of each dimension of metadata (such as the time dimension: different seasons of breeding; the experimental condition dimension: experimental data under different soil conditions, etc.). For example: in the time dimension, generate data portraits containing different breeding stages. In the experimental condition dimension, generate yield and disease resistance data portraits under different experimental conditions.

[0079] S14. Search for related data resources based on the portrait components and metadata fields in the portrait generation task, and generate a feature portrait of the scientific dataset.

[0080] Through the portrait elements and metadata fields in the portrait generation task, relevant data resources are automatically found and associated. For example, all relevant experimental data will be found based on the metadata field "disease resistance test data" and integrated into the scientific data set portrait. The final portrait shows in detail the raw data, metadata and their relationships at each breeding stage, including genomic information, phenotypic characteristics, disease resistance test results, and their association with breeding goals.

[0081] S2. Extract user behavior group features based on user data interaction behavior records associated with scientific data sets.

[0082] The associated users of a scientific dataset are researchers or institutions that use or access the dataset. For example, the associated users of a genomic dataset may include researchers engaged in genomics research, drug development companies, and bioinformatics laboratories. These users interact with scientific datasets through different channels (such as data download, query, analysis, etc.).

[0083] User behavior group characteristics refer to the common characteristics of user groups derived from user behavior and demand analysis. For example, in the use of genomic data sets, user behavior group characteristics may include research fields, data usage patterns (such as download frequency distribution, data analysis method characteristics, data tool software characteristics), data requirements (such as data format, data update frequency, data quality, etc.), and other characteristics.

[0084] Interaction behavior characteristics refer to access patterns, data usage purposes, data demand changes and other characteristics obtained based on user interaction behavior records and interaction-related data characteristics. For example, access patterns include "some users may access data every day, while other users may only access it during specific projects"; data usage purposes include "data usage patterns for different purposes such as basic research, clinical trials, and drug development, based on user data applications"; data demand changes include "user demands for data quality, format, and update frequency change over time. For example, emerging research fields may require higher resolution data or more frequent updates."

[0085] In a specific embodiment, step S2 may include the following specific steps:

[0086] S21. Obtain the data access, query, download, and processing behavior logic and behavior association semantics of users associated with scientific datasets, and construct a user behavior dataset based on user interaction behavior;

[0087] Behavioral logic refers to the behavioral process and steps of users when accessing, querying, downloading or processing scientific data sets. For example, in a scientific data set, a user may first search (query) for data, download the data of interest after finding it, and then use specific software to analyze the data. Behavioral logic describes the order and conditions of this series of operations. Behavioral logic can be modeled through behavioral time series analysis, behavioral event feature logic analysis, and behavioral path analysis.

[0088] Behavior association semantics involves the relationship and background information between user behaviors. For example, if a user frequently uses a specific data analysis method after downloading data, or tends to use a specific tool during data processing, these behaviors can reflect the user's data needs, data usage purposes, and data processing methods. Behavior association semantics helps explain the connection between these behaviors and the user's potential intentions. For behavior association semantics, association rule mining or cluster analysis can be used to identify user data usage tendencies or shared usage tendencies of data associations.

[0089] Based on the above analysis and mining, a user behavior dataset is established containing fields such as user ID, behavior type, timestamp, behavior logic and semantic information.

[0090] S22. Based on the user behavior data set, construct a profile variable containing different user behavior level characteristics;

[0091] Different user behavior level features refer to the features extracted from the user behavior dataset that can describe different dimensions and levels of user behavior. These features include behavior type features, such as user access, query, download and other specific operations; behavior frequency features, such as the frequency and time interval of user access to scientific datasets; behavior pattern features, such as user operation paths and rules; and behavior preference features, such as user preferences for data formats or analysis tools.

[0092] Profile variables are built based on these hierarchical features and are used to describe specific variables of user behavior. For example, access frequency variables record the number and frequency of user visits, download preference variables describe the data type or format that users prefer, behavior path variables represent the user's operation sequence and path, and tool variables record the type of tools or software used by users. These profile variables provide detailed basic data for the construction of user profiles.

[0093] The feature extraction step identifies the user's operation type, frequency, mode, and preference, and maps these features to specific profile variables. Based on the extracted profile variables, a user profile is generated, including the user's behavioral characteristics and demand characteristics.

[0094] S23, performing cluster analysis and feature extraction on different types of portrait variables respectively, to obtain user behavior group portrait features of each type of user group;

[0095] Use clustering algorithms to perform cluster analysis on the preprocessed user behavior profile variables. Clustering algorithms group users with similar behavior characteristics into the same group, thereby identifying user groups with common characteristics. Each cluster represents a user behavior group, such as a user group that frequently downloads data, a user group that prefers a specific analysis tool, etc.

[0096] After clustering is completed, each user group is analyzed in depth to extract the key features of the group. This includes counting the average behavioral characteristics of each group, identifying the commonality and preferences of behavioral patterns within the group, and exploring the special needs of the group. For example, a user group may show high-frequency data access behavior and a preference for a specific data format.

[0097] Combine the cluster analysis results and the extracted features to generate user behavior group features. These features include the behavior patterns, access habits, tool usage preferences, etc. of user groups, reflecting the behavior characteristics and needs of different user groups. Through these group features, we can better understand and serve different user groups and implement personalized data management and support strategies.

[0098] S24. Perform cluster analysis and feature extraction on all user behavior profile variables to obtain common characteristics of user behavior.

[0099] Use a hierarchical clustering algorithm to cluster all user behavior profile variables and identify common patterns or features among all users. Analyze the behavioral features in each cluster and extract common features that are widely applicable to most users. For example, you may find that certain behavioral patterns, such as high frequency of data access and tool usage preferences, exist in most users. These common features can help further optimize the management of data sets and user service strategies.

[0100] S3. Extract the shared use characteristics of scientific datasets based on the shared use records of scientific datasets.

[0101] Sharing and use characteristics refer to various characteristics related to the sharing and use of scientific datasets, which reflect the dissemination and utilization of scientific datasets in actual use. These characteristics can include sharing paths (the data sharing and use agreement can limit the use to third parties or stipulate that it can be used by project partners. For example, a scientific dataset may be shared from the original provider to several research groups, and then these groups share the scientific dataset with partners), usage records (such as usage methods, citation methods, usage scenarios, data processing, etc.), feedback records (scientific dataset evaluation suggestions, derived data, data-related results or application cases fed back or submitted by users after using the data), and data utilization patterns (the utilization of scientific datasets in different fields or projects, such as whether scientific datasets are used in a variety of different research fields, or in-depth use in a specific field).

[0102] In a specific embodiment, step 3 may include the following specific steps:

[0103] S31, obtaining similar interactive behavior data associated with the scientific data set, extracting the scientific data set propagation sharing path, and generating a user node sharing network diagram;

[0104] S32, obtaining a multi-user shared sample in the scientific dataset sharing and use record, extracting a scientific dataset sharing path in the multi-user shared sample as a subgraph of a main node of the shared network graph;

[0105] S33. Extract shared usage features through graph network analysis.

[0106] Suppose we have a set of animal and plant genome scientific datasets, involving gene sequence information of some animal and plant resources. Due to the multifaceted application value of genome data, we need to evaluate the interdisciplinary, cross-organizational and cross-field scalability of the dataset, and update the metadata content or add metadata fields in a targeted manner. By analyzing the interactive behavior data, we can extract the dissemination and sharing paths of scientific datasets. This includes identifying which users (such as researchers, laboratories or institutions) have downloaded scientific datasets, how scientific datasets are used in different projects, and how scientific datasets are shared with other users. We will build a user node sharing network graph, where each node represents a user or organization, and each connection represents the dissemination or sharing path of scientific datasets. Through the generated user node sharing network graph, we analyze the roles of different users (such as project leaders, data managers), data interaction behavior tendencies (such as high-frequency downloaders, data users), data sharing tendencies, and user fields (such as agricultural science, bioinformatics, etc.). Based on these analyses, we evaluate the scalability of scientific datasets in different fields and organizations, and determine how to improve the applicability of scientific datasets in multiple fields by supplementing and updating metadata fields. For example, if the network diagram shows that the dataset is mainly used by researchers in the field of agricultural science, you can consider adding specific field information related to agriculture to enhance the functionality and cross-field application capabilities of the scientific dataset.

[0107] In addition, multi-user shared samples refer to the situation where the same scientific dataset is used by multiple users in different studies or projects. These samples record the use of scientific datasets, including but not limited to the access, analysis, modification or sharing of scientific datasets by different users. This may involve different project teams, laboratories or institutions using the same scientific dataset in different research projects, thus forming a data interaction and sharing network. A subgraph is a local graph in the entire sharing network graph, which extracts a part of the data from the network graph and focuses on showing the behavior of a specific sharing path or user group. A subgraph usually consists of the main node of the network graph (i.e., the main user or scientific dataset) and its directly connected nodes (i.e., the user or institution related to the main node). A subgraph can help to deeply analyze the sharing and usage patterns of a specific scientific dataset or user. Specifically, the data sharing and use agreement can clarify the scope of data sharing and who can obtain and use the data. For example, project group A applied for a scientific dataset. Due to the data management capability of project group A, it was clearly stated in the data sharing authorization agreement that it was not allowed to share with third parties. At this time, project group B, as a cooperative organization of project group A, also needs to apply for the scientific dataset when conducting collaborative research. When project group B applies for the scientific dataset, the data sharing authorization agreement clearly states that the scientific dataset can be authorized to project group C, a cooperative organization of project group B. Then we can draw the complete sharing path of the scientific dataset in the subgraph.

[0108] S4. Generate a feature change index based on the user behavior group portrait feature and the shared usage feature, perform variable difference analysis based on the feature change index and its corresponding scientific data set portrait, generate a metadata update demand index, and generate a metadata update trigger condition based on the metadata update demand index.

[0109] The characteristic changes of scientific data sets are generally based on the description information of scientific data sets, such as metadata, data description documents, and updates to the data set content. For example, there are additions, deletions, or modifications to fields in scientific data sets, changes in data types or formats, additions, reductions, or modifications to data entries, significant changes in the size of scientific data sets (such as when the amount of data reaches a certain threshold), and updates to data collection tools, methods, or sources.

[0110] Changes in user group characteristics refer to changes in the usage patterns of user groups. For example, a certain type of user group begins to frequently use a new analysis tool or method, which may require updating metadata to reflect these new usage patterns, or changes in user needs based on user behavior cluster analysis. If the user group's requirements for data format or data quality change, this change will trigger updates to related metadata fields, such as fields that describe data format or data quality standards.

[0111] Changes in shared usage characteristics may include changes in the sharing path of scientific datasets. For example, scientific datasets are shared to new research fields or associated with new user groups, which may trigger metadata updates for the sharing path or associated users, or related users submit derivative data or related results. The metadata information associated with these contents needs to be recorded and updated.

[0112] The automatic update trigger condition determination may include feature change detection, priority setting, or feedback loop. For feature change detection, a feature change threshold can be set, and metadata updates are triggered when the threshold is exceeded. For priority setting, metadata update priorities can be classified and graded according to the degree of impact of changes in associated information of scientific datasets. For example, metadata changes affected by changes in data rights, data classification, and data standards can be set to high priority and regularly scanned and checked for updates. For feedback loops, dynamic adjustments and optimizations are made through feedback from user groups and evaluation results after data use.

[0113] In a specific embodiment, step S4 may include the following specific steps:

[0114] S41, obtaining user behavior group features and shared usage features associated with the scientific data set, and generating a feature change index;

[0115] The feature change index is calculated based on the feature change, such as the change in access frequency (calculating the change in access volume in the recent period, using incremental ratio or percentage change), and the change in the shared network graph (using graph analysis methods to calculate the spread of the data set among users, using the change in nodes and edges to measure). In the calculation of the feature change index, different calculation methods need to be used according to specific circumstances. For example, when a major research project is established and implemented, or when important industry standards and specifications are released and implemented, or when key software tools and methods are applied, significant changes in time series data can be identified through change point detection. For another example, a geographic or earth resource or a species genome data set has a high frequency of viewing, downloading, and use, and the associated user behavior is relatively complex. It is necessary to detect the abnormality of access frequency and interactive behavior to obtain the real dynamic sharing and use pattern of data. The feature change index can be calculated by clustering and grouping user behaviors and using machine learning algorithms. For another example, if a certain clinical medical data has been obtained and used by several types of user groups for a long time, the feature change index can be calculated by a dynamic weighting algorithm based on the changes in user behavior group characteristics and shared use characteristics to ensure that metadata updates can reflect the latest needs of special users in a targeted manner.

[0116] S42, performing variable difference analysis based on the feature change index and its corresponding scientific data set portrait, and generating an associated metadata update demand index;

[0117] The update demand index is calculated by combining the results of the feature change index and the variable difference analysis using a weighted model. For example, if the feature change index indicates that users have increased demand for data accuracy, and the variable difference analysis shows the gap between the accuracy field and the current demand, the demand index will increase accordingly. The weighted sum of the feature change and variable difference is considered comprehensively to obtain an update demand index value that reflects the urgency of metadata update.

[0118] S43: Generate metadata update trigger conditions according to preset metadata update rules, thresholds, and metadata update demand indexes.

[0119] The preset metadata update rules and thresholds can be adjusted dynamically. The setting of rule thresholds needs to consider a variety of specific situations, such as the identification of key data management needs. The legal requirements for data privacy protection and data integrity protection may need to be considered in the key data management needs of genomic datasets. For example, the definition of shared usage scenarios, the sharing needs of scientific datasets in emergency situations (such as sudden epidemic research) are different from the usage needs in daily research. These scenarios will determine different metadata update rules. For emergency sharing needs, a quick update rule can be set, that is, when the changes in user behavior group characteristics or shared usage characteristics are detected to exceed the threshold, the metadata update is triggered immediately.

[0120] The preset update threshold needs to define the update threshold of different metadata fields according to the needs of specific scientific data sets or even specific scientific data fields. It is also necessary to consider different time points and specific needs and dynamically adjust the threshold. For example, during the peak period of new agricultural variety breeding, a lower trigger threshold can be set to update metadata more frequently to ensure the timeliness and accuracy of the data. It is also necessary to consider the feedback cycle of scientific data use. For example, the feedback cycle of agricultural and forestry scientific data use may be longer. If user feedback is not timely enough, the trigger threshold can be appropriately lowered to increase the update frequency.

[0121] S5, metadata automatic update module, used to automatically update and verify metadata, specifically including the following steps:

[0122] S51, generating a metadata update instruction including a priority and a metadata field hierarchy based on the metadata update trigger condition;

[0123] In research projects with long research cycles, metadata needs to be updated regularly to ensure data availability and relevance due to the continuous accumulation of research data and changes in research needs. For example, in crop breeding projects, the frequency and importance of crop genome data usage will increase significantly during the critical stages of the crop growing season. When a significant increase in data download frequency is detected, and access from multiple multinational research institutions is added. According to the preset metadata update rules (such as high-frequency access and high priority corresponding to the critical crop growth period), a metadata update instruction is generated, which contains:

[0124] Priority: Set as high priority because the data has urgent use needs at key nodes in agricultural breeding.

[0125] Metadata field hierarchy: The directive requires updating the description fields in the genomic dataset, such as gene annotation, data version, data source, etc., giving priority to fields closely related to the crop growth cycle;

[0126] S52, execute the update instruction, monitor the completion of the update task node in real time, and generate a metadata update record;

[0127] After the metadata update instruction is generated, it is necessary to ensure that the update task can be executed accurately and timely, and monitor the progress of the task in real time to avoid any data inconsistency or delay.

[0128] Start the metadata update process, giving priority to high-priority metadata fields. For example, first update the description information of the genomic data, and update the key fields such as the latest gene variation, research results, and usage permissions to the metadata of the dataset. During the execution process, each update task node is monitored in real time. For example, monitor the update progress of each data field to check whether there are delays or errors. Once a task node is abnormal (such as data synchronization failure), an alarm is immediately issued, and an attempt is automatically made to re-execute or report to the administrator. During the execution process, each update task node is monitored in real time. For example, monitor the update progress of each data field to check whether there are delays or errors. Once a task node is abnormal (such as data synchronization failure), an alarm is immediately issued, and an attempt is automatically made to re-execute or report to the administrator;

[0129] S53, generating a metadata update report based on a preset prompt word template, metadata update records, and metadata update trigger conditions corresponding to each update;

[0130] The prompt word template is preset according to the scientific dataset management requirements and feature change index, mainly considering scientific dataset updates, data field modifications, group feature changes, changes in usage sharing patterns, important user feedback, etc. For example, for the aforementioned genomic scientific datasets required for agricultural quality breeding, this may involve, for example, newly added gene variation information, updated descriptions of crop breeding results, or data standard adjustments in international cooperation. The necessary components of the update report include updated metadata fields, reasons and triggering conditions for the update, update priority, update process records, update execution summary, and update impact analysis. According to the differences in actual update scenarios, the content of the prompt word template will be automatically filled in and adjusted. Some parts of the template may be dynamically generated according to the specific circumstances of the update. For example, the updated content, description of the triggering conditions, and summary of the execution process will be automatically filled in according to the actual update situation.

[0131] In this embodiment, the method further includes deploying the metadata update demand awareness network according to the scientific data set profile variable and the metadata update trigger condition.

[0132] On the one hand, in the daily work of scientific data management in the scientific data center, the management team is responsible for data sets in multiple different fields (such as medicine, meteorology, ecology, etc.). Each scientific data set has its own unique image indicators (such as data type, format, update frequency, metadata structure, etc.). To ensure that metadata is updated in a timely manner, a demand collection network is deployed. The network automatically identifies situations where metadata may need to be updated by monitoring these image variables and users' access, download, and query behaviors in real time.

[0133] On the other hand, at a specific time point or in an emergency data usage scenario, it is necessary to dynamically monitor the profile indicators of this type of data set, and automatically identify the metadata of potential needs that need to be updated in combination with the behavioral data of the user group (such as download frequency, access time, etc.). For example, if the frequency of use of a gene tag increases dramatically, it will be necessary to meet the change in demand and prompt the possible update of related metadata fields or the addition of new descriptive information.

[0134] The specific steps to deploy a metadata update demand-aware network are as follows:

[0135] Z1, construct and update feature association matrix;

[0136] In order to ensure that the metadata of the dataset always meets the current scientific research needs, a project team established a "feature association matrix". The matrix matches the image indicators of the scientific dataset with the defined metadata update trigger conditions (such as field content changes, increased data access, etc.), and records the correspondence between variables and trigger conditions and their semantic relevance, such as the impact that variable changes may have on the generated analytical data.

[0137] Z2, build an update demand perception model based on the update feature association matrix;

[0138] The feature association matrix is ​​trained through machine learning to build a demand model. The model analyzes historical data and current data sequences to identify data fields or descriptions that may need to be updated. For example, if the collection frequency of a data field suddenly increases, the model can predict that the field description may need to be updated to reflect the new data collection pattern.

[0139] Z3. Generate an update demand signal and update the metadata that needs to be updated;

[0140] Based on the update demand perception model, variable analysis is performed on the scientific data set portrait to generate update requirements, the update requirements are compared with the metadata update trigger conditions to obtain difference features, and adjusted update instructions are generated based on the difference features.

[0141] Compare the output analysis results of the demand collection model with the preset update trigger conditions to modify the specific update content. For example, if the model detects that the description of certain fields deviates from the actual usage, it will generate instructions to update the description information of the field to ensure that it accurately reflects the current data status and user needs. These update instructions will be transmitted to the data management platform to execute the corresponding metadata update tasks and record relevant information during the update process.

[0142] In a specific embodiment, the update feature association matrix may be partially customized, especially in the initial stage, the user or system administrator may customize the basic associations between metadata features according to existing business requirements or the organization of metadata.

[0143] Customized content may include:

[0144] Which metadata fields are strongly correlated with trigger conditions (such as data collection time and data update frequency).

[0145] Which fields' changes are most often related to update requirements (e.g., data usage and user feedback and quality standards).

[0146] As the system runs, it can continuously adjust and optimize the structure of the matrix by automatically learning the usage patterns of the dataset, the metadata update history and its results. Therefore, the updated feature association matrix can be automatically generated and dynamically adjusted by the system, reflecting more data-driven characteristics in the model.

[0147] The construction of the update feature association matrix needs to consider the relationship between the profile variables in the scientific dataset (such as the description field, data type, data source, usage frequency, etc. of the dataset) and the metadata update trigger conditions (such as new dataset upload, user demand changes, statutory update cycle, technology upgrade, etc.). Exemplarily, the construction process of the update feature association matrix is:

[0148] 1. The system extracts all relevant metadata fields from the scientific dataset portrait, such as data type, data size, data collection tool, etc. Each field corresponds to a column or row in the matrix.

[0149] 2. Identify and define a series of conditions that trigger metadata updates, such as data update cycles, data sharing behaviors, technology updates, legal requirements, etc. The triggering conditions correspond to another column or row in the matrix.

[0150] 3. Determine which profile variables have semantic associations with update triggers by analyzing historical usage of the dataset, user behavior patterns, update logs, and metadata requirements in specific scenarios. For example, when users frequently request a certain type of data or tool upgrade, it may trigger the need to update data quality metadata. In the matrix, the associations between these variables can be presented in the form of numerical values ​​or weights.

[0151] 4. The system can continuously optimize the matrix based on feedback and new data. For example, when a specific metadata field is frequently updated, the system can automatically improve the correlation between the field and a specific trigger condition.

[0152] 5. Semantic relevance means that when analyzing metadata fields and trigger conditions, not only the literal matching relationship is considered, but also the semantic similarity between fields is identified through natural language processing technology (NLP). For example, although "data update cycle" and "timeliness mandatory requirements" may seem different, they may have a high degree of semantic relevance in the update feature association matrix because both point to similar metadata update requirements.

[0153] In this embodiment, the method further deploys a metadata update feedback network, and the specific process is as follows:

[0154] A1. When the metadata update instruction is executed, the update instruction is synchronously sent to each associated user to obtain feedback on the metadata update triggering conditions;

[0155] A2. When the metadata update is completed, the metadata update record is sent to each associated user to obtain metadata effect evaluation opinions;

[0156] A3. Extract key information in the feedback opinions and common features and individual features of the key information based on the user metadata management authority and user characteristics, and modify the metadata update demand perception model according to the common features and individual features.

[0157] Exemplarily, the process of extracting key information from feedback and the common features and individual features of the key information is as follows:

[0158] When the metadata update instruction is executed or the metadata update is completed, the relevant instructions and records are sent to each associated user, and the feedback information is classified according to each user's metadata management authority and role. Different users may pay attention to different aspects of feedback according to their authority and functions, such as:

[0159] Data managers may be concerned with the structural integrity and consistency of metadata updates.

[0160] Data users may pay more attention to the availability and accuracy of data.

[0161] Bind user feedback to their permissions to ensure that different feedback opinions have different weights and analysis priorities in the system.

[0162] Extract key information: Extract key information from user feedback through natural language processing (NLP) and text analysis technology. Key information is specific content related to the decision of metadata update, such as:

[0163] Positive feedback: A metadata field is updated to better meet usage requirements.

[0164] Negative feedback: Updates to metadata fields create barriers to the use of existing scientific datasets.

[0165] For example, suppose a user's feedback is: "The newly added 'Data Collection Equipment' field is very helpful for analysis, but the description of the field is not intuitive, and I hope it can be expressed in more popular terms." You can extract the positive feedback of the "Data Collection Equipment" field, and at the same time extract the improvement suggestion "The description is not intuitive" as key information.

[0166] Extraction of common features: Common features refer to similar information or requirements that appear repeatedly in feedback from multiple users. By calculating the semantic similarity of feedback texts through cluster analysis, topic extraction, and other methods, similar issues or requirements mentioned by different users can be identified. For example, multiple users have mentioned the "field naming" problem. Although the expressions are different, their semantic commonalities can be identified; the frequency of key information that appears in feedback from different users is counted. For example, if multiple users object to the description of a metadata field, the need to modify the field can be regarded as a common feature;

[0167] By extracting keywords from feedback and clustering them, the system can find out which features are highly common among different user groups. For example, "field naming" and "inconsistent data format" may be high-frequency common problems.

[0168] Extraction of personality traits: Extract feedback specific to certain users or specific permissions as personality traits, which reflect the needs or problems of certain specific roles or user groups. Although personality traits only involve a small number of users, they may be very critical, especially in specific application scenarios. For example, during the cycle of certain emergencies, certain metadata fields only have an impact on users with specific permissions (such as data quality managers), so relevant personality feedback should be given high attention.

[0169] By analyzing user roles or permissions, identify feedback related to their specific functions and specific needs. For example, data managers may pay special attention to field permissions and field type settings, while data users may be more concerned about the ease of use and compatibility of data sets. Categorize feedback information and distinguish individual characteristics from common characteristics.

[0170] On the other hand, an embodiment of the present application provides a device for automatically updating metadata data of a scientific data set, comprising:

[0171] A scientific data set portrait construction module 901 is used to construct a scientific data set portrait based on structured or unstructured information submitted by a scientific data producer;

[0172] A user behavior group image feature extraction module 902 associated with a scientific data set is used to extract user behavior group image features according to user data interaction behavior records associated with a scientific data set;

[0173] A scientific dataset sharing and use feature extraction module 903 is used to extract scientific dataset sharing and use features according to scientific dataset sharing and use records;

[0174] The metadata automatic update trigger module 904 is used to generate a feature change index based on the user behavior group portrait feature and the shared use feature, perform variable difference analysis based on the feature change index and its corresponding scientific data set portrait, generate a metadata update demand index, and generate a metadata update trigger condition based on the metadata update demand index;

[0175] The metadata automatic updating module 905 is used to automatically update and verify the metadata when the metadata update triggering condition is met.

[0176] It can be understood that the above-mentioned embodiments and the above-mentioned method embodiments can correspond to each other, and descriptions similar to the embodiments can refer to the method embodiments. To avoid repetition, it will not be repeated here. A metadata automatic update device for a scientific data set provided in an embodiment of the present application can execute a metadata automatic 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 automatic 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.

[0177] 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.

[0178] The present application embodiment provides an electronic device 710, whose structure is as follows: Fig.10 The electronic device 710 may be an embodiment of the present application Fig.11 The server 100 or the terminal 200 is shown.

[0179] like Fig.10 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.

[0180] 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 method for automatically updating metadata of a scientific data set in an embodiment of the present application is implemented.

[0181] 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.

[0182] 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.

[0183] 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.

[0184] 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.

[0185] 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.

[0186] 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 running 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 method for automatically updating metadata of a scientific data set in an embodiment of the present application.

[0187] 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.

[0188] 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 method for automatically updating metadata of a scientific data set in the embodiment of the present application can be performed.

[0189] 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.

[0190] 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 automatically updating metadata of a scientific dataset, characterized in that: include: Build scientific dataset profiles based on structured or unstructured information submitted by scientific data producers; Extract user behavior group features based on user data interaction behavior records associated with scientific data sets; Extract the shared use characteristics of scientific datasets based on the shared use records of scientific datasets; Generate a feature change index based on the user behavior group portrait feature and the shared usage feature, perform variable difference analysis based on the feature change index and its corresponding scientific data set portrait, generate a metadata update demand index, and generate a metadata update trigger condition based on the metadata update demand index; When the metadata update triggering condition is met, the metadata is automatically updated.

2. The method for automatically updating metadata of a scientific data set according to claim 1, characterized in that: The structured or unstructured information submitted by the scientific data producer includes the original data, metadata and associated supporting information carriers of the scientific data set; The scientific data set portrait includes a scientific data set description portrait and a scientific data set feature portrait; The scientific data set image is constructed based on the structured or unstructured information submitted by the scientific data producer. The specific process includes: Classify the raw data, metadata and associated supporting information carriers of scientific datasets according to the domain type of scientific datasets; Extract the naming structure and multi-dimensional features of metadata of various types of raw data, where the naming structure includes naming field content, field order, folder naming hierarchy and naming relationship; Extract data resource description information from various supporting information carriers, obtain corresponding data resource description information based on the naming field content of the original data, and construct description portraits of various scientific data sets according to the naming hierarchy; Based on the multi-dimensional features of the metadata, establishing a portrait generation task through a portrait generation engine; Based on the portrait constituent elements and metadata fields in the portrait generation task, the associated data resources in the scientific dataset are searched, and a feature portrait of the scientific dataset is generated.

3. The method for automatically updating metadata of a scientific dataset according to claim 1, characterized in that: The extracting of user behavior group image features based on user data interaction behavior records associated with the scientific data set includes: Obtain the data access, query, download, and processing behavior logic and behavior association semantics of users associated with scientific datasets, and build a user behavior dataset based on user interaction behavior; Based on the user behavior data set, construct a profile variable of the user behavior characteristics, wherein the behavior characteristics include behavior type characteristics, behavior frequency characteristics, behavior pattern characteristics, and behavior preference characteristics; Cluster analysis and feature extraction are performed on the portrait variables of different behavioral type characteristics to obtain the user behavior group characteristics of each type of user group.

4. The method for automatically updating metadata of a scientific dataset according to claim 1, characterized in that: The extracting of the scientific dataset sharing and use features according to the scientific dataset sharing and use records includes: Acquire similar interactive behavior data associated with scientific data sets, extract the scientific data set propagation sharing path, and generate a user node sharing network diagram, wherein the user node sharing network diagram includes a main node and nodes, wherein the main node is a main user or scientific data set, and the node is a user related to the main node; Obtain multi-user shared samples in the scientific dataset sharing and use records, extract the scientific dataset sharing path in the multi-user shared samples as a subgraph of the main node of the user node sharing network graph; The shared usage features of scientific datasets are extracted by performing graph network analysis on the user node shared network graph and the subgraph of the main node of the shared network graph.

5. The method for automatically updating metadata of a scientific data set according to claim 1, characterized in that: The generating of the feature change index based on the user behavior group feature and the shared usage feature, and the generating of the metadata update trigger condition based on the metadata update demand index, includes: A metadata update trigger condition is generated according to a preset metadata update rule, a threshold value and the metadata update demand index.

6. The method for automatically updating metadata of a scientific data set according to claim 1, characterized in that: When the metadata update triggering condition is met, automatically updating the metadata includes: generating a metadata update instruction including a priority and a metadata field hierarchy based on the metadata update trigger condition; Execute the update instruction, monitor the completion of the update task node in real time, and generate metadata update records; A metadata update report is generated based on a preset prompt word template, the metadata update record and a metadata update trigger condition corresponding to each update.

7. A method for automatically updating metadata of a scientific dataset according to claim 6, characterized in that: It also includes deploying a metadata update demand awareness network, and deploying the metadata update demand awareness network according to the scientific data set profile and the metadata update trigger condition. The specific process includes: Constructing an update feature association matrix, wherein the update feature association matrix matches the scientific dataset portrait with the metadata update trigger condition and records the correspondence and semantic correlation between the scientific dataset portrait and the metadata update trigger condition; Building an update requirement perception model for identifying scientific dataset update requirements based on the update feature association matrix; Based on the update demand perception model, variable analysis is performed on the scientific data set portrait to generate update requirements. The update requirements are compared with the metadata update trigger conditions to obtain difference features. Based on the difference features, adjusted update instructions are generated. The adjusted update instructions are used to update erroneous metadata.

8. A method for automatically updating metadata of a scientific dataset according to claim 6, characterized in that: The method further includes deploying a metadata update feedback network, wherein the deployment process of the metadata update feedback network includes: When the metadata update instruction is executed, the metadata update instruction is synchronously sent to each associated user to obtain feedback on the metadata update triggering conditions; When the metadata update is completed, the metadata update record is sent to the users associated with each scientific dataset to obtain metadata effect evaluation opinions; Based on the user metadata management authority and user characteristics, key information in the feedback opinions as well as the common characteristics and individual characteristics of the key information are extracted, and the metadata update demand perception model is modified according to the common characteristics and individual characteristics.

9. A device for automatically updating metadata of scientific data sets, characterized in that: include, A scientific data set portrait construction module is used to construct a scientific data set portrait based on the structured or unstructured information submitted by scientific data producers; A user behavior group image feature extraction module associated with a scientific data set is used to extract user behavior group image features based on user data interaction behavior records associated with a scientific data set; A scientific dataset sharing and use feature extraction module is used to extract scientific dataset sharing and use features based on scientific dataset sharing and use records; A metadata automatic update trigger module, used to generate a feature change index based on the user behavior group portrait feature and the shared use feature, perform variable difference analysis based on the feature change index and its corresponding scientific data set portrait, generate a metadata update demand index, and generate a metadata update trigger condition based on the metadata update demand index; The metadata automatic updating module is used to automatically update the metadata when the metadata update triggering condition is met.

10. An electronic device, characterized in that: include: a memory for storing one or more computer programs; A processor, when the one or more computer programs are executed by the processor, implements the method for automatically updating metadata of a scientific dataset as described in any one of claims 1 to 8.

Images