Biological safety data intelligent management system

Abstract

The invention discloses a biological safety data intelligent management system, and relates to the technical field of data management, and the system comprises the steps: achieving the automatic collection of experiment data, environment data and personnel data through the Internet of Things technology; automatically marking and classifying the unstructured data; performing real-time quality evaluation and anomaly detection on the collected data; a distributed storage technology is adopted, and data are stored in at least two nodes in a scattered mode; carrying out encrypted storage and evidence storage, and realizing fine-grained data access control and operation auditing based on a smart contract of a block chain; constructing a knowledge graph in the field of biological safety; early warning, prevention and control are carried out on biological safety risks; the analysis result is presented in the form of charts and reports; and a unified biological safety data sharing platform is established. Through the intelligent contract technology of the block chain, fine-grained data access control and operation auditing are realized, and the security and traceability of data storage are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to an intelligent management system for biosafety data. Background Art

[0002] In today's era of globalization, biosafety issues are becoming increasingly important. From laboratory safety management, disease prevention and control to ecological and environmental monitoring, biosafety covers many areas and is directly related to human health, ecological balance and social stability. With the rapid development of biotechnology, biosafety issues are becoming increasingly prominent and have become an important part of public health and the ecological environment.

[0003] In the field of biosafety, data collection, processing, storage and analysis are crucial. Traditional data management methods have many shortcomings, such as low data collection efficiency, low data standardization, insufficient data quality control, high data storage security risks, insufficient data analysis and utilization, etc. These problems seriously restrict the efficiency and accuracy of biosafety management and cannot meet the urgent needs of the current biosafety field. Summary of the invention

[0004] In order to solve the above technical problems, a biosafety data intelligent management system is provided. This technical solution solves the problems raised in the above background technology.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is: A biosafety data intelligent management system, comprising: A data acquisition module, which is used to connect laboratory equipment, sensors and monitoring systems using the Internet of Things technology to achieve automatic collection of experimental data, environmental data and personnel data; A data annotation module, which is used to automatically annotate and classify unstructured data to improve data standardization. The unstructured data includes experimental records and pictures; A quality control module, which is used to perform real-time quality assessment and anomaly detection on the collected data using a machine learning algorithm; A data storage module, wherein the data storage module is used to adopt a distributed storage technology to disperse and store data in at least two nodes to avoid single point failure and data loss risks; A data encryption module, which is used to encrypt and store data blocks marked as important data and to implement fine-grained data access control and operation auditing based on smart contracts on the blockchain; A graph construction module, which is used to construct a knowledge graph in the field of biosafety to achieve data association analysis and knowledge reasoning; A risk prediction module, which is used to analyze historical data, establish a risk prediction model, and conduct early warning and prevention and control of biosafety risks; A data visualization module, which presents the analysis results in the form of charts and reports to provide data-driven decision support for managers; The data sharing module is used to establish a unified biosafety data sharing platform for data sharing and collaboration among various institutions and departments.

[0006] Preferably, the data annotation module specifically includes: Based on the type and annotation requirements of unstructured data, natural language processing technology is selected as the automatic annotation tool; The annotation process is divided into keyword extraction, label classification and bounding box annotation, and representative samples are selected from the existing unstructured data for training the annotation model; Manually label representative samples to ensure the consistency of labeling results, and train the labeling model based on representative samples and corresponding manual labeling results; During the training process, adjust the model parameters of the annotation model and optimize the algorithm; Use the trained annotation model to automatically annotate new unstructured data. The annotation results include keywords, tags, and bounding boxes. Based on the annotation results, preliminarily classify the unstructured data according to the experiment type, sample type and image features; Manually review the results of automatic labeling and preliminary classification, and feed the results of manual review back to the labeling model to further optimize the model parameters and labeling algorithm of the labeling model; Based on business needs, formulate standardized specifications for biosafety data, including annotation formats and classification standards; Based on standardized specifications, the labeled and classified data are formatted to improve the standardization of biosafety data.

[0007] Preferably, the quality control module specifically includes: Clean the collected biosafety data to remove duplicate, missing and invalid data points; Determine the use of the local outlier factor algorithm for quality assessment and outlier detection of biosafety data; Based on the characteristics and requirements of biosafety data, the neighborhood initial value and anomaly score threshold of the local anomaly factor algorithm are set; Use the preprocessed data to train the local anomaly factor algorithm model, learn the normal distribution pattern of the data and identify abnormal data; The biosafety data collected in real time is input into the trained local anomaly factor algorithm, and the local anomaly factor algorithm model calculates the anomaly score of each data point based on the input data; Output the density difference of the current data point relative to other data points in its neighborhood as the anomaly score of the data point; Determine whether the anomaly score of each data point exceeds the anomaly score threshold. If so, mark the data point as abnormal data; if not, do not output it. Based on the results of anomaly detection, quality assessment is performed on the group of biosafety data; Determine whether the proportion of abnormal data is higher than the preset ratio threshold. If so, trigger the alarm mechanism and take measures to strengthen monitoring, adjust experimental conditions and isolate abnormal organisms. If not, no output is made; The performance of the local anomaly factor algorithm model is evaluated, and based on the results of the performance evaluation, the parameters of the local anomaly factor algorithm are fine-tuned.

[0008] Preferably, the data storage module specifically includes: Determine whether the number of bytes of the data set to be stored exceeds a preset byte number threshold. If so, output the data set to be stored as a large file and divide it into at least two data blocks for storage. If not, do not output it. The master control node is responsible for global scheduling, metadata management, and load balancing. It communicates with the working nodes through heartbeat packets, monitors the status of the working nodes, and schedules data migration tasks based on the load conditions. The master control node sends heartbeat packets to the working nodes at preset intervals to detect the status of the working nodes; Determine whether the working node responds normally to the heartbeat packet of the master control node. If yes, no output is made. If no, the master control node records that the working node has failed and issues an alarm. The working nodes send load information to the master control node through heartbeat packets, and the master control node identifies the nodes with the highest load and the nodes with the lowest load; The amount of data to be migrated is obtained by calculating half of the load difference between the nodes with the highest and lowest loads. A migration task is generated to migrate the data to be migrated to the machine with the lowest load, thus achieving automatic load balancing. Each data block is replicated at least twice to form corresponding copies, which are stored on at least two nodes to avoid single point failure and data loss risks; The primary replica provides read-write services, and the backup replica provides read-only services. When the primary replica fails, a backup replica is automatically elected as the new primary replica. The data is synchronized to at least two storage nodes through the replication protocol, and the data consistency between the copies is ensured. The user's write request is required to be output successfully when it is synchronized from the primary copy to the backup copy.

[0009] Preferably, the data encryption module specifically includes: Use symmetric key encryption to encrypt data blocks marked as important data; Record the hash value of the data, and use timestamp technology to record the generation time and existence time of the data, and combine the hash value, generation time and existence time to output as a digital signature; storing the encrypted data and the digital signature in a computer-readable storage medium; Use smart contract language to write fine-grained data access control rules and deploy smart contracts on the blockchain network; Define the mapping relationship between users, data objects and access rights in the smart contract; Based on the required business needs, set the access control policy to access control based on user roles, data attributes or context information; The smart contract verifies the user's identity and permissions, and determines whether the user is allowed to access the data based on the access control policy. If so, the smart contract performs the corresponding data decryption and access operations. If not, the user's request is rejected and an error message is returned; Write an audit log recording function in the smart contract to record relevant information about data access and operations, including user identity, access time, accessed data objects, and performed operations; The audit logs are recorded on the blockchain, and the blockchain's consensus mechanism and data structure are used to ensure the authenticity and integrity of the audit logs.

[0010] Preferably, the graph construction module specifically includes: Based on the field of biosafety, entities in the knowledge graph include organisms, pathogens, diseases, drugs, and experimental techniques; The relationships in the knowledge graph include interactions between organisms, associations between pathogens and diseases, and the inhibitory effects of drugs on pathogens; The attributes in the knowledge graph include classification information of organisms, transmission methods of pathogens, clinical manifestations of diseases, and disease prevention and control measures; Identify entities of biosafety data from text data based on the annotation results; The identified entities and relationships are mapped into the knowledge graph to form a grid structure of nodes and edges.

[0011] Preferably, the risk prediction module specifically includes: Extract biosafety risk-related features based on entities and relationships in the knowledge graph, including the spread rate of pathogens, the incidence of diseases, and the effectiveness of prevention and control measures; Using machine learning algorithms, combined with features and label information in historical data, to build and train risk prediction models; Apply the trained risk prediction model to real-time monitoring data to provide early warning of biosafety risks and sound an alarm when the model predicts potential risks; Based on the early warning results and the prevention and control measures information in the knowledge graph, formulate targeted prevention and control strategies, which include strengthening monitoring, raising prevention and control awareness, and taking isolation and disinfection measures.

[0012] Preferably, the data sharing module specifically includes: Based on platform planning and technical requirements, select the database technology stack to develop a biosafety data sharing platform; Develop a biosafety data sharing platform based on the functional modules of user management, data upload, data retrieval, data analysis and collaborative communication; Conduct unit testing, integration testing and system testing on the biosafety data sharing platform to confirm the working status of each functional module; Perform performance tests of load testing and stress testing on the biosafety data sharing platform, and record and output the number of users and data that the platform can carry.

[0013] Compared with the prior art, the present invention has the following beneficial effects: By integrating Internet of Things technology, machine learning algorithms, distributed storage technology, blockchain technology and knowledge graph construction technology, comprehensive, efficient and intelligent management of biosafety data is achieved. The system not only improves the efficiency and accuracy of data collection, but also improves the quality control and storage security of data. It also supports in-depth analysis and utilization of data, provides strong support for biosafety management, and improves the security and traceability of data storage through data access control and operation auditing. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is a flow chart of the biosafety data intelligent management method of the present invention; Figure 2 A flow chart of the method for automatically labeling and classifying unstructured data according to the present invention; Figure 3 A flow chart of a method for real-time quality assessment and anomaly detection of collected data according to the present invention; Figure 4 A flow chart of the method for avoiding single point failure and data loss risk of the present invention; Figure 5 A flowchart of the method for encrypting and storing evidence for the present invention and implementing fine-grained data access control and operation auditing based on smart contracts on blockchain; Figure 6 A flow chart of the method for constructing a knowledge graph in the field of biosafety and implementing data association analysis and knowledge reasoning for the present invention; Figure 7 A flow chart of the method for analyzing historical data and establishing a risk prediction model according to the present invention; Figure 8 The present invention provides a flow chart of the method for establishing a unified biosafety data sharing platform. DETAILED DESCRIPTION

[0015] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0016] Reference Figure 1 As shown, a biosafety data intelligent management system includes: A data acquisition module, which is used to connect laboratory equipment, sensors and monitoring systems using the Internet of Things technology to achieve automatic collection of experimental data, environmental data and personnel data; A data annotation module, which is used to automatically annotate and classify unstructured data to improve data standardization. The unstructured data includes experimental records and pictures; A quality control module, which is used to perform real-time quality assessment and anomaly detection on the collected data using a machine learning algorithm; A data storage module, wherein the data storage module is used to adopt a distributed storage technology to disperse and store data in at least two nodes to avoid single point failure and data loss risks; A data encryption module, which is used to encrypt and store data blocks marked as important data and to implement fine-grained data access control and operation auditing based on smart contracts on the blockchain; A graph construction module, which is used to construct a knowledge graph in the field of biosafety to achieve data association analysis and knowledge reasoning; A risk prediction module, which is used to analyze historical data, establish a risk prediction model, and conduct early warning and prevention and control of biosafety risks; A data visualization module, which presents the analysis results in the form of charts and reports to provide data-driven decision support for managers; The data sharing module is used to establish a unified biosafety data sharing platform for data sharing and collaboration among various institutions and departments.

[0017] Reference Figure 2 As shown in the figure, the data annotation module specifically includes: Based on the type and annotation requirements of unstructured data, natural language processing technology is selected as the automatic annotation tool; The annotation process is divided into keyword extraction, label classification and bounding box annotation, and representative samples are selected from the existing unstructured data for training the annotation model; Manually label representative samples to ensure the consistency of labeling results, and train the labeling model based on representative samples and corresponding manual labeling results; During the training process, adjust the model parameters of the annotation model and optimize the algorithm; Use the trained annotation model to automatically annotate new unstructured data. The annotation results include keywords, tags, and bounding boxes. Based on the annotation results, preliminarily classify the unstructured data according to the experiment type, sample type and image features; Manually review the results of automatic labeling and preliminary classification, and feed the results of manual review back to the labeling model to further optimize the model parameters and labeling algorithm of the labeling model; Based on business needs, formulate standardized specifications for biosafety data, including annotation formats and classification standards; Based on standardized specifications, the labeled and classified data are formatted to improve the standardization of biosafety data.

[0018] From the existing unstructured data, a certain number of representative samples are screened out based on the diversity, representativeness and annotation requirements of the data. These samples should cover various experimental types, sample types and image features to ensure the generalization ability of the annotation model.

[0019] Reference Figure 3 As shown, the quality control module specifically includes: Clean the collected biosafety data to remove duplicate, missing and invalid data points; Determine the use of the local outlier factor algorithm for quality assessment and outlier detection of biosafety data; Based on the characteristics and requirements of biosafety data, the neighborhood initial value and anomaly score threshold of the local anomaly factor algorithm are set; Use the preprocessed data to train the local anomaly factor algorithm model, learn the normal distribution pattern of the data and identify abnormal data; The biosafety data collected in real time is input into the trained local anomaly factor algorithm, and the local anomaly factor algorithm model calculates the anomaly score of each data point based on the input data; Output the density difference of the current data point relative to other data points in its neighborhood as the anomaly score of the data point; Determine whether the anomaly score of each data point exceeds the anomaly score threshold. If so, mark the data point as abnormal data; if not, do not output it. Based on the results of anomaly detection, quality assessment is performed on the group of biosafety data; Determine whether the proportion of abnormal data is higher than the preset ratio threshold. If so, trigger the alarm mechanism and take measures to strengthen monitoring, adjust experimental conditions and isolate abnormal organisms. If not, no output is made; The performance of the local anomaly factor algorithm model is evaluated, and based on the results of the performance evaluation, the parameters of the local anomaly factor algorithm are fine-tuned.

[0020] Remove duplicate rows by checking the uniqueness of the data; for missing data points, delete the rows with missing values ​​or fill in the missing values, such as using the mean, median, mode, or interpolation method, depending on the specific situation; identify and remove invalid data points, such as values ​​outside the reasonable range, non-numeric characters, etc., according to rules such as data range and data type.

[0021] Reference Figure 4 As shown, the data storage module specifically includes: Determine whether the number of bytes of the data set to be stored exceeds a preset byte number threshold. If so, output the data set to be stored as a large file and divide it into at least two data blocks for storage. If not, do not output it. The master control node is responsible for global scheduling, metadata management, and load balancing. It communicates with the working nodes through heartbeat packets, monitors the status of the working nodes, and schedules data migration tasks based on the load conditions. The master control node sends heartbeat packets to the working nodes at preset intervals to detect the status of the working nodes; Determine whether the working node responds to the heartbeat packet of the master control node normally. If yes, no output is made. If no, the master control node records that the working node has failed and issues an alarm. The working nodes send load information to the master control node through heartbeat packets, and the master control node identifies the nodes with the highest load and the nodes with the lowest load; The amount of data to be migrated is obtained by calculating half of the load difference between the nodes with the highest and lowest loads. A migration task is generated to migrate the data to be migrated to the machine with the lowest load, thus achieving automatic load balancing. Each data block is replicated at least twice to form corresponding copies, which are stored on at least two nodes to avoid single point failure and data loss risks; The primary replica provides read-write services, and the backup replica provides read-only services. When the primary replica fails, a backup replica is automatically elected as the new primary replica. The data is synchronized to at least two storage nodes through the replication protocol, and the data consistency between the copies is ensured. The user's write request is required to be output successfully when it is synchronized from the primary copy to the backup copy.

[0022] The user's write request is first sent to the primary replica. After the primary replica processes the write request, it synchronizes the changes to all backup replicas. When all backup replicas confirm that they have received and applied the changes, the primary replica returns a request success message to the user. This ensures data consistency between replicas and improves system reliability.

[0023] Reference Figure 5 As shown, the data encryption module specifically includes: Use symmetric key encryption to encrypt data blocks marked as important data; Record the hash value of the data, and use timestamp technology to record the generation time and existence time of the data, and combine the hash value, generation time and existence time to output as a digital signature; storing the encrypted data and the digital signature in a computer-readable storage medium; Use smart contract language to write fine-grained data access control rules and deploy smart contracts on the blockchain network; Define the mapping relationship between users, data objects and access rights in the smart contract; Based on the required business needs, set the access control policy to access control based on user roles, data attributes or context information; The smart contract verifies the user's identity and permissions, and determines whether the user is allowed to access the data based on the access control policy. If so, the smart contract performs the corresponding data decryption and access operations. If not, the user's request is rejected and an error message is returned; Write an audit log recording function in the smart contract to record relevant information about data access and operations, including user identity, access time, accessed data objects, and performed operations; The audit logs are recorded on the blockchain, and the blockchain's consensus mechanism and data structure are used to ensure the authenticity and integrity of the audit logs.

[0024] Use hash functions such as SHA-256 to calculate the hash value of the data block before encryption, and the hash value is used to verify the integrity and authenticity of the data; use timestamp technology to record the generation time and existence time of the data to ensure the timeliness and traceability of the data; merge the calculated hash value, generation time and existence time, and add necessary metadata such as data identification, version information, etc. to form a digital signature.

[0025] Reference Figure 6 As shown in the figure, the graph construction module specifically includes: Based on the field of biosafety, entities in the knowledge graph include organisms, pathogens, diseases, drugs, and experimental techniques; The relationships in the knowledge graph include interactions between organisms, associations between pathogens and diseases, and the inhibitory effects of drugs on pathogens; The attributes in the knowledge graph include classification information of organisms, transmission methods of pathogens, clinical manifestations of diseases, and disease prevention and control measures; Identify entities of biosafety data from text data based on the annotation results; The identified entities and relationships are mapped into the knowledge graph to form a grid structure of nodes and edges.

[0026] In the knowledge graph, potential associations and patterns in the field of biosafety can be discovered by querying and analyzing the relationships between nodes and edges. For example, it is possible to analyze which diseases a certain pathogen is associated with, or which drugs have an inhibitory effect on a certain pathogen. By using the known information in the knowledge graph, new knowledge can be discovered through reasoning algorithms. For example, based on the known classification information of organisms and the association between pathogens and diseases, the type of disease that a newly discovered pathogen may cause can be inferred. Common reasoning methods include rule-based reasoning, machine learning-based reasoning, etc.

[0027] Reference Figure 7 As shown in FIG. 1 , the risk prediction module specifically includes: Extract biosafety risk-related features based on entities and relationships in the knowledge graph, including the spread rate of pathogens, the incidence of diseases, and the effectiveness of prevention and control measures; Using machine learning algorithms, combined with features and label information in historical data, to build and train risk prediction models; Apply the trained risk prediction model to real-time monitoring data to provide early warning of biosafety risks and sound an alarm when the model predicts potential risks; Based on the early warning results and the prevention and control measures information in the knowledge graph, formulate targeted prevention and control strategies, which include strengthening monitoring, raising prevention and control awareness, and taking isolation and disinfection measures.

[0028] After receiving the warning information, first analyze the warning results to understand the type, degree and scope of impact of potential risks, use the prevention and control measures information in the knowledge graph to query the prevention and control measures that match the warning results. These measures should provide specific response strategies for different types of risks. Combined with the warning results and prevention and control measures information, formulate targeted prevention and control strategies, implement the formulated prevention and control strategies, and continuously monitor their effects. Adjust and optimize the strategies based on actual results and feedback.

[0029] Reference Figure 8As shown, the data sharing module specifically includes: Based on platform planning and technical requirements, select the database technology stack to develop a biosafety data sharing platform; Develop a biosafety data sharing platform based on the functional modules of user management, data upload, data retrieval, data analysis and collaborative communication; Conduct unit testing, integration testing and system testing on the biosafety data sharing platform to confirm the working status of each functional module; Perform performance tests of load testing and stress testing on the biosafety data sharing platform, and record and output the number of users and data that the platform can carry.

[0030] Have an in-depth understanding of platform planning and technical requirements, including data type, scale, access frequency, security requirements, etc., and determine key indicators such as the number of users, concurrent access volume, and data update frequency that the data sharing platform needs to support.

[0031] Furthermore, the present solution also proposes a computer-readable storage medium on which a computer-readable program is stored, and when the computer-readable program is called, the above-mentioned intelligent biosafety data management system is executed.

[0032] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0033] In summary, the advantages of the present invention are: by integrating Internet of Things technology, machine learning algorithms, distributed storage technology, blockchain technology and knowledge graph construction technology, comprehensive, efficient and intelligent management of biosafety data is achieved. The system not only improves the efficiency and accuracy of data collection, but also improves the quality control and storage security of data, while supporting in-depth analysis and utilization of data, providing strong support for biosafety management, and improving the security and traceability of data storage through data access control and operation auditing.

[0034] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A biosafety data intelligent management system, characterized in that: include: A data acquisition module, which is used to connect laboratory equipment, sensors and monitoring systems using the Internet of Things technology to achieve automatic collection of experimental data, environmental data and personnel data; A data annotation module, which is used to automatically annotate and classify unstructured data to improve data standardization. The unstructured data includes experimental records and pictures; A quality control module, which is used to perform real-time quality assessment and anomaly detection on the collected data using a machine learning algorithm; A data storage module, wherein the data storage module is used to adopt a distributed storage technology to disperse and store data in at least two nodes to avoid single point failure and data loss risks; A data encryption module, which is used to encrypt and store data blocks marked as important data and to implement fine-grained data access control and operation auditing based on smart contracts on the blockchain; A graph construction module, which is used to construct a knowledge graph in the field of biosafety to achieve data association analysis and knowledge reasoning; A risk prediction module, which is used to analyze historical data, establish a risk prediction model, and conduct early warning and prevention and control of biosafety risks; A data visualization module, which presents the analysis results in the form of charts and reports to provide data-driven decision support for managers; The data sharing module is used to establish a unified biosafety data sharing platform for data sharing and collaboration among various institutions and departments.

2. A biosafety data intelligent management system according to claim 1, characterized in that: The data annotation module specifically includes: Based on the type and annotation requirements of unstructured data, natural language processing technology is selected as the automatic annotation tool; The annotation process is divided into keyword extraction, label classification and bounding box annotation, and representative samples are selected from the existing unstructured data for training the annotation model; Manually label representative samples to ensure the consistency of labeling results, and train the labeling model based on representative samples and corresponding manual labeling results; During the training process, adjust the model parameters of the annotation model and optimize the algorithm; Use the trained annotation model to automatically annotate new unstructured data. The annotation results include keywords, tags, and bounding boxes. Based on the annotation results, preliminarily classify the unstructured data according to the experiment type, sample type and image features; Manually review the results of automatic labeling and preliminary classification, and feed the results of manual review back to the labeling model to further optimize the model parameters and labeling algorithm of the labeling model; Based on business needs, formulate standardized specifications for biosafety data, including annotation formats and classification standards; Based on standardized specifications, the labeled and classified data are formatted to improve the standardization of biosafety data.

3. A biosafety data intelligent management system according to claim 2, characterized in that: The quality control module specifically includes: Clean the collected biosafety data to remove duplicate, missing and invalid data points; Determine the use of the local outlier factor algorithm for quality assessment and outlier detection of biosafety data; Based on the characteristics and requirements of biosafety data, the neighborhood initial value and anomaly score threshold of the local anomaly factor algorithm are set; Use the preprocessed data to train the local anomaly factor algorithm model, learn the normal distribution pattern of the data and identify abnormal data; The biosafety data collected in real time is input into the trained local anomaly factor algorithm, and the local anomaly factor algorithm model calculates the anomaly score of each data point based on the input data; Output the density difference of the current data point relative to other data points in its neighborhood as the anomaly score of the data point; Determine whether the anomaly score of each data point exceeds the anomaly score threshold. If so, mark the data point as abnormal data; if not, do not output it. Based on the results of anomaly detection, quality assessment is performed on the group of biosafety data; Determine whether the proportion of abnormal data is higher than the preset ratio threshold. If so, trigger the alarm mechanism and take measures to strengthen monitoring, adjust experimental conditions and isolate abnormal organisms. If not, no output is made; The performance of the local anomaly factor algorithm model is evaluated, and based on the results of the performance evaluation, the parameters of the local anomaly factor algorithm are fine-tuned.

4. A biosafety data intelligent management system according to claim 3, characterized in that: The data storage module specifically includes: Determine whether the number of bytes of the data set to be stored exceeds a preset byte number threshold. If so, output the data set to be stored as a large file and divide it into at least two data blocks for storage. If not, do not output it. The master control node is responsible for global scheduling, metadata management, and load balancing. It communicates with the working nodes through heartbeat packets, monitors the status of the working nodes, and schedules data migration tasks based on the load conditions. The master control node sends heartbeat packets to the working nodes at preset intervals to detect the status of the working nodes; Determine whether the working node responds to the heartbeat packet of the master control node normally. If yes, no output is made. If no, the master control node records that the working node has failed and issues an alarm. The working nodes send load information to the master control node through heartbeat packets, and the master control node identifies the nodes with the highest load and the nodes with the lowest load; The amount of data to be migrated is obtained by calculating half of the load difference between the nodes with the highest and lowest loads. A migration task is generated to migrate the data to be migrated to the machine with the lowest load, thus achieving automatic load balancing. Each data block is replicated at least twice to form corresponding copies, which are stored on at least two nodes to avoid single point failure and data loss risks; The primary replica provides read-write services, and the backup replica provides read-only services. When the primary replica fails, a backup replica is automatically elected as the new primary replica. The data is synchronized to at least two storage nodes through the replication protocol, and the data consistency between the copies is ensured. The user's write request is required to be output successfully when it is synchronized from the primary copy to the backup copy.

5. The biosafety data intelligent management system according to claim 4, characterized in that: The data encryption module specifically includes: Use symmetric key encryption to encrypt data blocks marked as important data; Record the hash value of the data, and use timestamp technology to record the generation time and existence time of the data, and combine the hash value, generation time and existence time to output as a digital signature; storing the encrypted data and the digital signature in a computer-readable storage medium; Use smart contract language to write fine-grained data access control rules and deploy smart contracts on the blockchain network; Define the mapping relationship between users, data objects and access rights in the smart contract; Based on the required business needs, set the access control policy to access control based on user roles, data attributes or context information; The smart contract verifies the user's identity and permissions, and determines whether the user is allowed to access the data based on the access control policy. If so, the smart contract performs the corresponding data decryption and access operations. If not, the user's request is rejected and an error message is returned; Write an audit log recording function in the smart contract to record relevant information about data access and operations, including user identity, access time, accessed data objects, and performed operations; The audit logs are recorded on the blockchain, and the blockchain's consensus mechanism and data structure are used to ensure the authenticity and integrity of the audit logs.

6. A biosafety data intelligent management system according to claim 5, characterized in that: The graph construction module specifically includes: Based on the field of biosafety, entities in the knowledge graph include organisms, pathogens, diseases, drugs, and experimental techniques; The relationships in the knowledge graph include interactions between organisms, associations between pathogens and diseases, and the inhibitory effects of drugs on pathogens; The attributes in the knowledge graph include classification information of organisms, transmission methods of pathogens, clinical manifestations of diseases, and disease prevention and control measures; Identify entities of biosafety data from text data based on the annotation results; The identified entities and relationships are mapped into the knowledge graph to form a grid structure of nodes and edges.

7. The biosafety data intelligent management system according to claim 6, characterized in that: The risk prediction module specifically includes: Extract biosafety risk-related features based on entities and relationships in the knowledge graph, including the spread rate of pathogens, the incidence of diseases, and the effectiveness of prevention and control measures; Using machine learning algorithms, combined with features and label information in historical data, to build and train risk prediction models; Apply the trained risk prediction model to real-time monitoring data to provide early warning of biosafety risks and sound an alarm when the model predicts potential risks; Based on the early warning results and the prevention and control measures information in the knowledge graph, formulate targeted prevention and control strategies, which include strengthening monitoring, raising prevention and control awareness, and taking isolation and disinfection measures.

8. The biosafety data intelligent management system according to claim 7, characterized in that: The data sharing module specifically includes: Based on platform planning and technical requirements, select the database technology stack to develop a biosafety data sharing platform; Develop a biosafety data sharing platform based on the functional modules of user management, data upload, data retrieval, data analysis and collaborative communication; Conduct unit testing, integration testing and system testing on the biosafety data sharing platform to confirm the working status of each functional module; Perform performance tests of load testing and stress testing on the biosafety data sharing platform, and record and output the number of users and data that the platform can carry.

Images