Information management platform and method supporting multi-center data fusion

Abstract

The invention belongs to the technical field of data security, and particularly relates to an information management platform and method supporting multi-center data fusion, and the platform constructs a portrait monitoring node network based on different center user monitoring information and a single center hash index through a portrait index module by using a hash algorithm and a graph algorithm; the global fusion module generates a cross-center global portrait node network through an association matching algorithm by means of a security index hash table containing single and cross-center hash index mapping, and the cross-center hash index mapping is constructed based on a dynamic feature anchor point algorithm and effective time; the hierarchical permission module is combined with a hierarchical permission criterion and an access control algorithm to generate a global permission hierarchical access node network; the cross-center access module obtains an index intention information list according to the user access intention, the authority level and the historical intention access frequency in combination with the global authority network; according to the platform, multi-center data security fusion and collaborative access are realized, and data isolation and cross-domain sharing efficiency are guaranteed.

Description

Technical Field

[0001] The present invention belongs to the field of data security technology, and in particular relates to an information management platform and method supporting multi-center data fusion. Background Art

[0002] In existing medical information systems, centers for cancer prevention and treatment, chronic disease management, and other areas utilize independent information management systems, resulting in numerous technical deficiencies. Regarding data, decentralized storage and heterogeneous formats create data silos, leading to inefficient cross-center queries and difficulty communicating patient historical data. Regarding authority management, the traditional model is crude, posing risks of unauthorized access and privacy breaches. Business collaboration is difficult, with multidisciplinary consultation and referral processes relying on offline processes and inefficiencies. Management decisions lack multi-center data support, and patients must repeat operations, resulting in a fragmented experience. Current solutions, such as data interface docking, achieve only limited interoperability, facing drawbacks such as difficulty balancing data interoperability and business isolation, insufficient authority control, and a lack of intelligent analysis capabilities. More comprehensive technical solutions are urgently needed. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the present invention proposes an information management platform and method that supports multi-center data fusion. The portrait index module of the platform constructs a portrait monitoring node network based on user monitoring information from different centers and a single center hash index; the global fusion module uses a secure index hash table containing single and cross-center hash index mappings to generate a cross-center global portrait node network through an association matching algorithm, wherein the cross-center index mapping is based on a dynamic feature anchor algorithm and effective time construction; the hierarchical permission module combines hierarchical permission criteria with an access control algorithm to generate a global permission hierarchical access node network; the cross-center access module obtains a list of index intention information based on the global permission network according to user access intention, permission level and historical intention access frequency; this platform realizes multi-center data security fusion and collaborative access, ensuring data isolation and cross-domain sharing efficiency.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] An information management platform supporting multi-center data fusion, including:

[0006] The portrait index module is used to obtain user monitoring information from different centers, and based on the user monitoring information from different centers combined with the preset single center user hash index, obtain the monitoring node network of different center portrait information through graph algorithms;

[0007] The global fusion module builds a secure index hash table based on the profile information monitoring node network of different centers combined with all user hash indexes. Through the association matching algorithm, it obtains a cross-center global profile information monitoring node network. The secure index hash table includes a single-center user hash index and a cross-center user hash index mapping.

[0008] The hierarchical permission module is based on the cross-center global portrait information monitoring node network and the preset hierarchical permission criteria. Through the access control algorithm, it obtains the cross-center global permission hierarchical access node network. The hierarchical permission criteria are constructed based on the access object permissions combined with the information sensitivity level of the cross-center global portrait information monitoring node network and the cross-center information transmission security level.

[0009] The cross-center access module determines the user's access intention and permission level based on the access user information, and obtains the access user index intention information list based on the user's access intention and permission level and the access frequency of historical intention information combined with the cross-center global permission hierarchical access node network through the preset cross-center business isolation algorithm; the single-center user hash index is obtained through the hash algorithm based on the user's basic information combined with the user's corresponding center monitoring information; the cross-center user hash index mapping is constructed based on the user hash index of the same user in different centers combined with the dynamic feature anchor algorithm and the preset effective time.

[0010] Specifically, the portrait index module includes a portrait unit, a sensitivity evaluation unit, a single index unit, and a node construction unit;

[0011] The portrait unit obtains the single-center user portrait and the corresponding sensitive feature extraction space based on the basic information of a single user in different centers, multimodal monitoring information, and the preset multimodal portrait model;

[0012] The sensitivity evaluation unit obtains the hierarchical sensitivity information space corresponding to the single-center user information based on the single-center user portrait and the corresponding sensitive feature extraction space combined with the preset hierarchical sensitivity evaluation model;

[0013] A single index unit obtains a hierarchical hash permission index corresponding to a single central user through a hash algorithm based on the hierarchical sensitivity information space corresponding to the single central user information and the preset permission level;

[0014] The node construction unit obtains a single center portrait information monitoring node network based on the edge weights constructed by the single portrait node constructed according to the single center user portrait and the corresponding hierarchical hash authority index and the similarity of the monitoring information between the nodes.

[0015] Specifically, the global fusion module includes a global mapping unit and a cross-center mapping unit;

[0016] The global mapping unit obtains the cross-center mapping feature code of the corresponding user between different centers based on the hierarchical hash permission index corresponding to the same user in the portrait information monitoring node network of different centers combined with the dynamic feature anchor algorithm;

[0017] The cross-center mapping unit obtains the cross-center user hash index mapping based on the cross-center mapping feature codes of all users in the portrait information monitoring node network of different centers through the feature splicing algorithm combined with the preset random valid time interval, and obtains the cross-center global portrait information monitoring node network based on the cross-center user hash index mapping combined with the center portrait information monitoring node network through the topology algorithm.

[0018] Specifically, the hierarchical authority module includes a hierarchical authority unit and a hierarchical authority mapping unit;

[0019] The hierarchical permission unit obtains the hierarchical permission level corresponding to the access user based on the basic information of the access user and the preset sensitivity level range of the accessible information;

[0020] The hierarchical permission mapping unit simulates the cross-center permission information indexing through the access control algorithm combined with the simulation algorithm according to the hierarchical access permission level corresponding to the access user and the hierarchical sensitivity information space and cross-center mapping feature code corresponding to each user in the cross-center global portrait information monitoring node network, and obtains the cross-center global permission hierarchical access node network.

[0021] Specifically, the construction process of the cross-center user hash index mapping includes:

[0022] Input the basic information and multimodal monitoring information of each user in a single center into the multimodal feature extraction layer of the modal portrait model to obtain the basic feature space of a single user and a single monitoring state feature space in the current center;

[0023] The basic features of a single user are set to the highest sensitivity level. At the same time, the single monitoring state feature space is input into the sensitive keyword feature extraction layer with a pre-set sensitive word level library to obtain the single user's graded sensitive word sequence and the number of sensitive words at the corresponding level;

[0024] According to the graded sensitive word sequence of a single user and the number of sensitive words of the corresponding level, the comprehensive sensitivity level and single level sensitivity of the single user are obtained through the sensitivity level assessment layer.

[0025] Specifically, the construction process of the cross-center user hash index mapping also includes:

[0026] Based on the comprehensive sensitivity level and single-level sensitivity of a single user, a decision tree structure is used. The comprehensive sensitivity level is used as the trunk, and the basic characteristics of a single user are set as the root node. At the same time, based on the single-level sensitivity, the keyword information corresponding to each sensitivity level is stored in the bifurcation node of the corresponding sensitivity level on the trunk in a manner from near to far. This obtains a single user sensitive information decision tree.

[0027] Based on the information stored in each node of the decision tree for sensitive information of a single user, the label and distinguishing identification information corresponding to each node are obtained through a random interception algorithm;

[0028] Based on the label and distinguishing identification information corresponding to each node, a hierarchical hash index corresponding to each node of the single user sensitive information decision tree is obtained through a hash algorithm.

[0029] Specifically, the construction process of the cross-center user hash index mapping also includes:

[0030] According to the sensitivity level corresponding to the access rights of historical users in a single center, the access rights-sensitivity mapping function in a single center is obtained through support vector machine;

[0031] The access permission-sensitivity mapping function enables the accessing user to access the current user's sensitive information decision tree in a single center, and at the same time access the user information stored in the nodes of the same sensitivity level and below in the sensitive information decision tree of the associated user;

[0032] Based on the access permission-sensitivity mapping function within a single center and the hierarchical hash index corresponding to each node, the hierarchical hash permission index corresponding to each node is obtained and built into the node corresponding to the single user sensitive information decision tree to obtain a single user hierarchical permission decision tree;

[0033] Based on the information stored in the nodes of all single-user hierarchical permission decision trees in a single center and the corresponding hierarchical hash permission indexes, the association similarity between each single-user hierarchical permission decision tree and the node information of the same sensitivity level and the lower-level node information in the remaining single-user hierarchical permission decision trees is obtained through the lower-level association matching algorithm.

[0034] Specifically, the construction process of the cross-center user hash index mapping also includes:

[0035] Based on the correlation similarity between each single user hierarchical permission decision tree and the node information of the same sensitivity level and the lower level node information in the remaining single user hierarchical permission decision trees, a correlation similarity connection is constructed, and the hierarchical hash permission index corresponding to the two connected nodes is combined with the public key obtained by the zero-proof algorithm and embedded into the corresponding connection. At the same time, the generated private key is respectively configured into the two connected nodes to obtain a single central user hierarchical permission decision forest;

[0036] Based on the hierarchical hash permission indexes built into nodes of different sensitivity levels in the single user hierarchical permission decision tree for the same user in different centers, a dynamic anchor point algorithm is used to randomly intercept index subsequences of length N from the hierarchical hash permission indexes of different levels and then concatenate all intercepted index subsequences to obtain a cross-center fused index sequence.

[0037] At the same time, feature index information of length M is extracted from each index subsequence corresponding to the center containing the same user information, and based on the two feature index information extracted from any two centers containing the same user, the verifiable segmentation algorithm is used to obtain the corresponding splicing feature code between any two cross-center fusion index sequences, where N>M.

[0038] Specifically, the construction process of the cross-center user hash index mapping also includes:

[0039] Based on the corresponding splicing signature between any two cross-center fusion index sequences, the cross-center fusion index sequences corresponding to a single user in any two centers are spliced together, and the bidirectional mapping between the access levels of any two centers and the sensitivity levels of non-centers and the cross-center security index factor constructed by the cross-center data security level are built into the splicing signature corresponding to the spliced arbitrary two centers to obtain a single user cross-center user hash index mapping;

[0040] Based on the above single user cross-center user hash index mapping construction process, obtain the single user cross-center user hash index mapping corresponding to all users;

[0041] Based on the single-center user hierarchical permission decision forest combined with the single-user cross-center user hash index mapping corresponding to all users, the access control algorithm is combined with the simulation algorithm and the index exception function corresponding to the single-center user hierarchical permission decision forest and the global index exception function corresponding to the cross-center user hierarchical permission decision forest. The single-center sub-authority information index and the cross-center sub-authority information index training are performed to obtain the trained cross-center global permission hierarchical access node network.

[0042] An information management method supporting multi-center data fusion, comprising:

[0043] Obtain user monitoring information from different centers, and based on this information, combine it with the preset single user hash index and use graph algorithms to obtain a monitoring node network of different center portrait information.

[0044] Based on the security index hash table constructed by combining the portrait information monitoring node network of different centers with the hash index of all users, a cross-center global portrait information monitoring node network is obtained through an association matching algorithm; the security index hash table includes a single-center user hash index and a cross-center user hash index mapping;

[0045] Based on the cross-center global portrait information monitoring node network combined with the preset hierarchical authority criteria, a cross-center global authority hierarchical access node network is obtained through the access control algorithm; the hierarchical authority criteria are constructed based on the access object permissions combined with the information sensitivity level of the cross-center global portrait information monitoring node network and the cross-center information security level;

[0046] The single-center user hash index is obtained by combining the user's basic information with the monitoring information of the user's corresponding center through a hash algorithm; the cross-center user hash index mapping is constructed based on the user hash indexes of the same user in different centers, combined with the dynamic feature anchor algorithm and the preset effective time;

[0047] The user's access intention and permission level are determined based on the access user information, and the access user index intention information list is obtained based on the user's access intention and permission level and the access frequency of historical intention information combined with the cross-center global permission hierarchical access node network.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] In response to the deficiencies of the existing technology, the present invention realizes the secure fusion and intelligent access of multi-center data through the collaborative use of multi-module technology. Specifically, the portrait index module uses hash algorithms and graph algorithms to construct heterogeneous monitoring information into a semantic node network, thereby realizing structured association and integration of cross-center data; the global fusion module uses a secure index hash table with dynamic anchor mapping to complete the precise aggregation and identity uniqueness verification of cross-center data based on the association matching algorithm, thereby ensuring the consistency and integrity of multi-source data; the hierarchical permission module combines information sensitivity levels with cross-center security rules to construct a dynamic permission system, and realizes fine-grained cross-domain permission stratification through access control algorithms, thereby meeting compliance access requirements while ensuring data isolation; the cross-center access module generates precise index lists based on user intentions, permission levels and historical access characteristics, thereby improving the efficiency and pertinence of collaborative access to multi-center data, and achieving an organic balance between data fusion depth and security protection strength. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a module structure diagram of an information management platform supporting multi-center data fusion according to Example 1 of the present invention;

[0051] Figure 2 This is a simplified schematic diagram of a single-center user hierarchical authority decision forest according to Example 1 of the present invention;

[0052] Figure 3 This is a flow chart of an information management method supporting multi-center data fusion according to embodiment 2 of the present invention. DETAILED DESCRIPTION

[0053] Example 1

[0054] See also Figure 1The present invention provides an embodiment of an information management platform that supports multi-center data fusion, which is applied to cross-center index security protection of patient data in various hospital departments or multiple clinical centers, such as centers for cancer prevention and treatment, chronic disease management, minimally invasive intervention, and pain diagnosis and treatment, including:

[0055] The portrait index module is used to obtain user monitoring information from different centers, and based on the user monitoring information from different centers combined with the preset single center user hash index, obtain the monitoring node network of different center portrait information through graph algorithms;

[0056] The global fusion module builds a secure index hash table based on the profile information monitoring node network of different centers combined with all user hash indexes. Through the association matching algorithm, it obtains a cross-center global profile information monitoring node network. The secure index hash table includes a single-center user hash index and a cross-center user hash index mapping.

[0057] The hierarchical permission module is based on the cross-center global portrait information monitoring node network and the preset hierarchical permission criteria. Through the access control algorithm, it obtains the cross-center global permission hierarchical access node network. The hierarchical permission criteria are constructed based on the access object permissions combined with the information sensitivity level of the cross-center global portrait information monitoring node network and the cross-center information transmission security level.

[0058] The cross-center access module determines the user's access intention and permission level based on the access user information, and obtains the access user index intention information list through the preset cross-center business isolation algorithm based on the user's access intention and permission level and the access frequency of historical intention information combined with the cross-center global permission hierarchical access node network; the cross-center business isolation algorithm in this embodiment is used to isolate the query process of each business scenario, for example, queries from the chronic disease center channel are isolated from queries from the pain center to avoid business conflicts.

[0059] It should be further explained that if this embodiment is applied to a hospital, each department is considered as a separate center;

[0060] It should be further explained that, in this embodiment, the single-center user hash index is obtained through a hash algorithm based on the user's basic information combined with the user's corresponding center monitoring information; the cross-center user hash index mapping is constructed based on the user hash index of the same user in different centers combined with the dynamic feature anchor algorithm and the preset effective time.

[0061] Furthermore, the portrait indexing module includes a portrait unit, a sensitivity evaluation unit, a single indexing unit, and a node construction unit;

[0062] The portrait unit obtains the single-center user portrait and the corresponding sensitive feature extraction space based on the basic information of a single user in different centers, multimodal monitoring information, and the preset multimodal portrait model;

[0063] The sensitivity evaluation unit obtains the hierarchical sensitivity information space corresponding to the single-center user information based on the single-center user portrait and the corresponding sensitive feature extraction space combined with the preset hierarchical sensitivity evaluation model;

[0064] A single index unit obtains a hierarchical hash permission index corresponding to a single central user through a hash algorithm based on the hierarchical sensitivity information space corresponding to the single central user information and the preset permission level;

[0065] The node construction unit obtains a single center portrait information monitoring node network based on the edge weights constructed by the single portrait node constructed according to the single center user portrait and the corresponding hierarchical hash authority index and the similarity of the monitoring information between the nodes.

[0066] Furthermore, the global fusion module includes a global mapping unit and a cross-center mapping unit;

[0067] The global mapping unit obtains the cross-center mapping feature code of the corresponding user between different centers based on the hierarchical hash permission index corresponding to the same user in the portrait information monitoring node network of different centers combined with the dynamic feature anchor algorithm;

[0068] The cross-center mapping unit obtains the cross-center user hash index mapping based on the cross-center mapping feature codes of all users in the portrait information monitoring node network of different centers through the feature splicing algorithm combined with the preset random valid time interval, and obtains the cross-center global portrait information monitoring node network based on the cross-center user hash index mapping combined with the center portrait information monitoring node network through the topology algorithm.

[0069] Furthermore, the hierarchical authority module includes a hierarchical authority unit and a hierarchical authority mapping unit;

[0070] The hierarchical permission unit obtains the hierarchical access permission level corresponding to the access user based on the basic information of the access user and the preset sensitivity level range of the accessible information. It should be noted that the permission levels in this embodiment are defined by those skilled in the art based on the security protection requirements of specific data;

[0071] The hierarchical permission mapping unit simulates the cross-center permission information indexing based on the hierarchical access permission level corresponding to the accessing user and the hierarchical sensitivity information space and cross-center mapping feature code corresponding to each user in the cross-center global portrait information monitoring node network. This obtains a cross-center global permission hierarchical access node network. It should be further noted that the access control algorithm in this embodiment preferentially adopts an attribute-based access control algorithm and a role-based access control algorithm, specifically:

[0072] Based on the hierarchical access rights of accessing users, the hierarchical sensitivity information space of users in the cross-center global profile information monitoring node network, cross-center mapping feature codes, and information sensitivity level intervals, RBAC (role-based access control algorithm) is used to determine the user's basic role permissions. ABAC (attribute-based access control algorithm) is then used to refine the role permissions. A multidimensional access policy is constructed by combining user attributes (such as the center and position), environmental attributes (such as access time and device IP address), and data attributes (such as information sensitivity level and cross-center mapping feature codes). The policy engine parses policy rules, matches access requests, and resolves conflicts. For example, when a doctor at the oncology center accesses chronic disease data, the system dynamically adjusts the accessible fields (such as displaying only blood sugar trends instead of specific genetic data) based on their role permissions and data sensitivity levels through ABAC policies. The final output is a cross-center global permission hierarchical access node network. This network uses nodes to represent accessible data resources and edges to represent access paths based on permission levels, achieving fine-grained cross-center data access control and hierarchical protection of sensitive information. It should be further explained that the multi-dimensional access strategy in this embodiment is based on user attributes (such as the center to which they belong, role identity, and position level), environmental attributes (such as access time, device IP address, and network location), data attributes (such as information sensitivity level, data type, and cross-center mapping feature code), and operation attributes (such as read, modify, and delete operation types) as core dimensions. The attributes of each dimension are logically combined and weighted through a predefined policy rule engine. For example, the strategy can be defined as "When the user role is the attending physician of the tumor center, and accesses the chronic disease data with a medium sensitivity level in the cross-center global portrait through the hospital intranet device during working hours, only non-genetic related data such as blood glucose trends are allowed to be read." By matching the conditions and resolving conflicts of the attributes of each dimension, precise control of cross-center data access rights based on dynamic scenario adjustment is achieved.

[0073] The patient-side self-service module is used to provide end users with cross-system data query and business processing access.

[0074] The patient-side self-service module includes a unified service portal unit, a personalized data dashboard unit, and an intelligent semantic search unit;

[0075] Unified service portal unit, integrating multiple system service entrances, such as appointment, inquiry, and declaration, supporting single sign-on and service status tracking;

[0076] In this embodiment, the unified service portal unit implements SSO based on the OIDC protocol and routes requests through the microservice gateway.

[0077] The personalized data dashboard unit generates a customized list view of access user index intention information based on user historical behavior, such as a high-frequency access mechanism that is displayed at the top. In this embodiment, the personalized data dashboard unit uses a collaborative filtering algorithm to analyze user behavior patterns and implements dynamic dashboard configuration based on D3.js.

[0078] An intelligent semantic search unit is used to support parsing input queries to obtain query intent, combine it with an indexing algorithm, invoke the hash index built into each node, and return cross-system aggregated results. For example, "find records related to X in 2023." Furthermore, the indexing algorithm in this embodiment is preferably a tree index or a variant of a tree index.

[0079] In this embodiment, the intelligent semantic search unit builds a semantic understanding engine based on the BERT model and implements distributed cross-center retrieval through Elasticsearch.

[0080] The process first realizes the secure aggregation and efficient access of cross-center medical data through a multi-dimensional technical architecture. The portrait index module uses a multimodal portrait model and a hierarchical sensitivity assessment model to convert patient basic information and multimodal data such as medical imaging and genetic testing into a hierarchical sensitive feature space, and combines the hash algorithm to construct index nodes with permission identification to achieve accurate classification and structured management of sensitive data, providing underlying support for subsequent cross-center permission control; the global fusion module uses a dynamic feature anchor algorithm to generate cross-center mapping feature codes, and uses a topology algorithm to merge different center index node networks into a global network. The secure index hash table is combined with zero-knowledge proof and other mechanisms to ensure privacy protection when cross-center data is associated, realizing secure aggregation and global semantic mapping of multi-source medical data; the hierarchical permission module constructs hierarchical permission criteria based on access object permissions, data sensitivity levels and transmission security requirements, and simulates cross-center permission indexing through access control algorithms and simulation algorithms. , forming a fine-grained hierarchical access node network to effectively prevent unauthorized access and data leakage risks; the cross-center access module uses a business isolation algorithm to isolate query channels of different centers, and optimizes access paths in combination with historical access frequency analysis to avoid query conflicts in different business scenarios such as cancer prevention and treatment, chronic disease management, and improve cross-center data access efficiency; the patient-side self-service module implements multi-system single sign-on through the OIDC protocol, builds personalized data dashboards based on collaborative filtering algorithms and D3.js, and uses the BERT semantic understanding engine combined with Elasticsearch to implement cross-center semantic retrieval, providing patients with a unified and intelligent self-service entrance; the overall technical architecture forms a closed-loop system of "data classification, cross-center integration, authority stratification, business isolation, and intelligent service", which not only ensures the compliance and security of medical data, but also realizes efficient indexing of cross-center data and intelligent upgrade of patient self-service, thereby improving multi-center medical collaboration efficiency and patient experience.

[0081] Furthermore, the construction process of the cross-center user hash index mapping includes:

[0082] Input the basic information and multimodal monitoring information of each user in a single center into the multimodal feature extraction layer of the modal portrait model to obtain the basic feature space of a single user and a single monitoring state feature space in the current center;

[0083] It should be noted that the specific steps of obtaining a single user basic feature space and a single monitoring state feature space in this embodiment include:

[0084] First, basic information and multimodal monitoring information of patients in a single center are collected. It should be further explained that the basic information in this embodiment includes, but is not limited to, structured data such as the patient's name, ID number, medical record number, and attending department. The multimodal monitoring information in this embodiment includes, but is not limited to: PET-CT images, pathological section images, and gene mutation detection reports corresponding to the cancer prevention and treatment center; dynamic blood sugar and blood pressure monitoring curves and medication order texts corresponding to the chronic disease management center; surgical image sequences and intraoperative physiological indicator data corresponding to the minimally invasive intervention center; and pain score time series records and neuroelectrophysiological detection reports corresponding to the pain diagnosis and treatment center.

[0085] Second, the collected data is input into the multimodal feature extraction layer of the multimodal portrait model, specifically:

[0086] Step 1: Using a 3D convolutional neural network to locate and extract features from medical images of lesions. It should be noted that lesion location and feature extraction include but are not limited to identifying the size, location, and metabolic activity of the tumor;

[0087] Step 2: Use a bidirectional long short-term memory network to analyze the fluctuation patterns and abnormal peaks of time series data such as blood sugar and blood pressure, and use a pre-trained language model in the medical field to parse semantic information such as symptom descriptions, treatment plans, and allergy history in the medical record text; in this embodiment, the pre-trained language model in the medical field is preferably PubMedBERT;

[0088] Step 3: Use graph neural networks to construct a correlation map between gene mutation sites and disease phenotypes, and obtain the patient's basic feature space and a single monitoring state feature space, including but not limited to image feature vectors, time series trend vectors, text semantic vectors, etc.

[0089] Step 4: A feature fusion algorithm is used to map the patient's basic feature space and the single monitoring state feature space into a unified high-dimensional feature vector. Combined with clinical diagnostic criteria, a patient portrait is constructed that includes dimensions such as disease stage, treatment response, and complication risk. Simultaneously, a sensitive feature recognition algorithm uses regular expressions to match identification information such as ID card number and home address. A medical sensitive vocabulary combined with semantic analysis is used to locate sensitive content such as hereditary tumor history and mental illness diagnosis. Based on genetic privacy protection rules, the sensitivity level of pathogenic mutation sites is annotated to form a sensitive feature set.

[0090] Step 5. Based on the health and medical data security guidelines, sensitive features are mapped to a three-dimensional sensitive feature extraction space based on leakage risk: the X-axis represents identity recognizability, for example, ID card number is high risk; the Y-axis represents disease sensitivity, for example, AIDS diagnosis is highly sensitive; and the Z-axis represents data update frequency, for example, real-time physiological indicators are moderately sensitive. Quantify the protection priority of each feature to provide underlying feature support for hierarchical desensitization and permission control for cross-center access.

[0091] The basic features of a single user are set to the highest sensitivity level. At the same time, the single monitoring state feature space is input into the sensitive keyword feature extraction layer with a pre-set sensitive word level library to obtain the single user's graded sensitive word sequence and the number of sensitive words at the corresponding level. It should be noted that the sensitive keyword feature extraction layer in this embodiment preferably uses the Chinese pre-trained Bert model and the comprehensive fuzzy evaluation algorithm.

[0092] Based on the graded sensitive word sequence and the number of sensitive words of the corresponding grade for a single user, the sensitivity level assessment layer is used to obtain the comprehensive sensitivity level and single-level sensitivity of the single user. It should be further explained that the sensitivity level assessment layer in this embodiment has a built-in graded sensitivity assessment model. The steps of constructing and training the graded sensitivity assessment model include:

[0093] First, a sensitive word classification system is defined based on the medical knowledge graph, which divides sensitive words into five levels. For example, ID card numbers are level 5 and disease names are level 3. The medical knowledge graph is optimized as a health and medical data security guide.

[0094] Second, we collected historical medical data to construct a training set, extracted features such as the frequency of sensitive words, word co-occurrence, and semantic similarity, used the XGBoost algorithm to train the classification model, optimized the model parameters through a gradient boosting mechanism, and used 5-fold cross-validation to ensure generalization ability.

[0095] Third, an attention mechanism is introduced to dynamically adjust the weights of each feature, assigning higher weights to high-risk features. The model outputs a single-level sensitivity and a comprehensive sensitivity level, and uses a softmax function to map the probabilities of each level to a continuous value between 0 and 1. Finally, the model performance is evaluated using the ROC curve to ensure that the AUC value is greater than the preset AUC threshold. In this embodiment, AUC is the area under the curve, which usually refers to the area under the receiver operating characteristic curve.

[0096] Based on the comprehensive sensitivity level and single-level sensitivity of a single user, a decision tree structure is used. The comprehensive sensitivity level is used as the trunk, and the basic characteristics of a single user are set as the root node. At the same time, based on the single-level sensitivity, the keyword information corresponding to each sensitivity level is stored in the bifurcation node of the corresponding sensitivity level on the trunk in a manner from near to far. This obtains a single user sensitive information decision tree.

[0097] Based on the information stored in each node of the decision tree for sensitive information of a single user, the label and distinguishing identification information corresponding to each node are obtained through a random interception algorithm;

[0098] It should be further explained that the process of obtaining the label and distinguishing identification information in this embodiment includes:

[0099] First, the feature vector of the node is extracted through the BERT model to generate a 768-dimensional semantic vector; second, a cryptographically secure pseudo-random number generator is used to dynamically generate truncation parameters based on the node sensitivity level. For example, highly sensitive nodes generate anchor points with shorter intervals, and the interval of low-sensitivity nodes is expanded to 15 dimensions; third, a fixed-length subsequence is truncated from the 768-dimensional semantic vector according to the anchor point position, such as 32 dimensions for highly sensitive nodes, and the truncated sequence is Base64 encoded to generate a basic label; at the same time, the node sensitivity level, keyword frequency and other information are hashed using SHA-256 to generate a 128-bit summary, the first 8 bits are taken as the distinguishing identifier prefix, spliced with the basic label and then a CRC-16 check code is added, finally forming a unique identifier that contains semantic features, sensitivity level and verification information, ensuring that the labels of different nodes have semantic distinguishability and data integrity.

[0100] Based on the label and distinguishing identification information corresponding to each node, a hash algorithm is used to obtain a hierarchical hash index corresponding to each node in the single user sensitive information decision tree;

[0101] It should be further explained that this embodiment performs a SHA-256 hash operation on each node of the user sensitive information decision tree, and intercepts the first 16 bytes as the basic hash value; adds a prefix identifier according to the node sensitivity level, for example, high sensitivity = 0x01, medium sensitivity = 0x02, low sensitivity = 0x03, such as the hash of a high-sensitivity node is 0x01-5a3f2b...; uses BloomFilter to deduplicate the hashes of nodes at the same level, and generates an index table containing 128-bit fingerprints for fast retrieval of sensitive features at the same level.

[0102] According to the sensitivity level corresponding to the access rights of historical users in a single center, the access rights-sensitivity mapping function in a single center is obtained through support vector machine;

[0103] It should be further explained that this embodiment first collects access logs from multiple clinical centers in the past Q years, including but not limited to user roles, access data sensitivity levels, and operation time, extracts samples to construct a training set, and then uses the SVM model with RBF kernel function to train the mapping function. The input is the user role, including but not limited to doctors, nurses, administrators and patients, and the access scenarios include but not limited to diagnosis, scientific research, and quality control. The output is the highest sensitivity level allowed for access, for example, chief physicians can access highly sensitive data, and interns are limited to low sensitivity. Finally, the hyperparameters are adjusted through 5-fold cross-validation to ensure that the model accuracy is greater than the preset accuracy threshold.

[0104] The access permission-sensitivity mapping function enables the accessing user to access the current user's sensitive information decision tree in a single center, and at the same time access the user information stored in the nodes of the same sensitivity level and below in the sensitive information decision tree of the associated user;

[0105] Based on the access permission-sensitivity mapping function within a single center and the hierarchical hash index corresponding to each node, the hierarchical hash permission index corresponding to each node is obtained and built into the node corresponding to the single user sensitive information decision tree to obtain a single user hierarchical permission decision tree;

[0106] It should be further explained that this embodiment first embeds the mapping function generated by the SVM into the decision tree nodes, and sets an access permission threshold for each node. For example, a highly sensitive node requires the role of chief physician. Secondly, the LSH (locally sensitive hashing) algorithm is used to calculate the feature similarity between different patient decision trees. When the cosine similarity of the node feature vectors exceeds the preset similarity threshold, a cross-tree connection is established. It should be further explained that the connection weights in this embodiment are calculated based on the co-occurrence frequency of the features using a Bayesian probability model to form a weighted association network. For example, the co-occurrence frequency of "BRCA1 mutation" and "family history of breast cancer" is a%.

[0107] Based on the information stored in the nodes of all single-user hierarchical permission decision trees in a single center and the corresponding hierarchical hash permission indexes, the association similarity between each single-user hierarchical permission decision tree and the node information of the same sensitivity level and the lower-level node information in the remaining single-user hierarchical permission decision trees is obtained through the lower-level association matching algorithm;

[0108] Based on the correlation similarity between each single user hierarchical permission decision tree and the node information of the same sensitivity level and the lower level node information in the remaining single user hierarchical permission decision trees, a correlation similarity connection is constructed, and the hierarchical hash permission index corresponding to the two connected nodes is combined with the public key obtained by the zero-proof algorithm and embedded into the corresponding connection. At the same time, the generated private key is respectively configured into the two connected nodes to obtain a single central user hierarchical permission decision forest;

[0109] See also Figure 2 , assuming that A1 in the figure corresponds to the single user root node of the tumor center, A11 to A14 indicate the bifurcation nodes of different sensitivities corresponding to the single user A1, A2 is the second user root node, A21 to A25 are the bifurcation nodes of different levels of sensitivity corresponding to the second user, A1 and A2 are the corresponding single user hierarchical permission decision trees in the tumor center, and B11 to B14 corresponding to A1 are the single user hierarchical permission decision trees corresponding to the detection information of user A1 in the pain center;

[0110] It should be further explained that this embodiment first uses the EC-Schnorr zero-knowledge proof algorithm for cross-tree connections to generate a 256-bit elliptic curve public key, which is stored in the connection edge and private key and assigned to the corresponding node. When it is necessary to verify the consistency of the features of the two nodes, a non-interactive proof mechanism is used to complete the verification without leaking the original features, thereby reducing access latency while meeting real-time access requirements.

[0111] Based on the hierarchical hash permission indexes built into nodes of different sensitivity levels in the single user hierarchical permission decision tree for the same user in different centers, a dynamic anchor point algorithm is used to randomly intercept index subsequences of length N from the hierarchical hash permission indexes of different levels and then concatenate all intercepted index subsequences to obtain a cross-center fused index sequence.

[0112] It should be further explained that, in this embodiment, a dynamic anchor algorithm is used for the hierarchical authority decision tree of the same patient in different centers to intercept a subsequence of length N=32 from the hash index of each center. It should be explained that the anchor point interval in this embodiment is dynamically adjusted with the sensitivity level, with a high-sensitivity layer interval of 5 bytes and a low-sensitivity layer interval of 15 bytes. For example, the high-sensitivity index 0x01-abc... of the tumor prevention and treatment center intercepts the first 32 bits, and the sensitive index 0x02-def... of the chronic disease management center intercepts the middle 32 bits.

[0113] At the same time, feature index information of length M is extracted from each index subsequence corresponding to the center containing the same user information. Based on the two feature index information extracted from any two centers containing the same user, a verifiable segmentation algorithm is used to obtain the corresponding splicing feature code between any two cross-center fusion index sequences.

[0114] It should be further explained that this embodiment first extracts feature segments from the index subsequences corresponding to any two centers according to a preset offset, encrypts the feature segments using the Paillier homomorphic encryption algorithm, and uses the Groth16 zero-knowledge proof protocol to verify that the hash values of the two encrypted feature segments are equal, ensuring that the verification process does not leak the original feature content;

[0115] Secondly, the hash value of the verified feature segment is concatenated with the cross-center security index factor, the security index factor is embedded in the specified position in the ASN.1 encoding format, and then the concatenated data is operated through the SHA-512 hash algorithm to generate a 64-byte splicing feature code; it should be noted that the cross-center security index factor in this embodiment includes but is not limited to security attributes such as data transmission encryption level and access control policy; further, the feature code in this embodiment integrates encryption verification, security factor embedding and hash operation to ensure the uniqueness, immutability and security traceability of cross-center index splicing, and realize the security association and verification of sensitive data indexes in different centers.

[0116] Based on the corresponding splicing feature codes between any two cross-center fusion index sequences, the cross-center fusion index sequences corresponding to a single user in any two centers are spliced together, and the bidirectional mapping between the access levels of any two centers and the sensitivity levels of non-centers and the cross-center security index factor constructed by the cross-center data security level are built into the splicing feature codes corresponding to the splicing of any two centers to obtain a single user cross-center user hash index mapping;

[0117] Based on the above single user cross-center user hash index mapping construction process, obtain the single user cross-center user hash index mapping corresponding to all users;

[0118] Based on the single-center user hierarchical permission decision forest combined with the single-user cross-center user hash index mapping corresponding to all users, the access control algorithm is combined with the simulation algorithm and the index exception function corresponding to the single-center user hierarchical permission decision forest and the global index exception function corresponding to the cross-center user hierarchical permission decision forest. The single-center sub-authority information index and the cross-center sub-authority information index training are performed to obtain the trained cross-center global permission hierarchical access node network.

[0119] It should be further explained that the training process of the access control algorithm combined with the simulation algorithm in this embodiment includes:

[0120] First, the fusion features of the hierarchical hash permission index of a single-center hierarchical permission decision forest and the cross-center hash index mapping are extracted as training input;

[0121] Secondly, we use the SVM permission mapping function in the access control algorithm to build an initial permission model. Combined with the simulation algorithm, we generate access samples for multi-center scenarios such as cancer prevention and treatment, chronic disease management, etc., covering business scenarios such as normal access, unauthorized access, and cross-center business conflicts.

[0122] Third, during training, the single-center index anomaly function and cross-center global index anomaly function detection model's ability to identify abnormal access are used to verify the accuracy of permission stratification for normal access. For abnormal access, the loss function (including but not limited to fusion conflict probability and access delay) is used to reversely optimize the permission threshold of the decision tree node and the security factor weight of the cross-center mapping feature code;

[0123] Fourth, a Monte Carlo simulation is introduced to dynamically adjust the cross-center security index factor. A gradient descent algorithm is used to iteratively optimize the access control policy. This allows the model to gradually optimize the permission control logic while simulating the access behavior of users with different roles. Finally, adaptive training is used to generate a cross-center global permission-based hierarchical access node network, achieving precise permission interception and path optimization for real access requests. It should be further explained that in this embodiment, the different user roles include doctors, nurses, administrators, and patients.

[0124] This process builds an adaptive cross-center permission control system through the deep integration of access control algorithms and simulation algorithms. Specifically, first, by extracting the fusion features of a single-center hierarchical permission decision forest and a cross-center hash index mapping, the training input is ensured to cover the permission control logic of the hierarchical hash permission index and the security association features of the cross-center fusion index. This provides comprehensive semantic and security dimension feature support for subsequent permission models, enabling the model to understand the sensitivity level and permission mapping relationship of data from different centers. Secondly, the initial model is constructed using the SVM permission mapping function, and the simulation algorithm is combined to generate access samples covering multi-center scenarios such as cancer prevention and treatment, chronic disease management, etc. By simulating the normal and abnormal access behaviors of different roles such as doctors, nurses, administrators, and patients in scenarios such as diagnosis and scientific research, the model can learn the permission boundaries of various business scenarios during the training phase. This multi-scenario simulation training mechanism ensures that the model can accurately identify cross-center business conflicts in actual applications, such as query isolation between chronic disease centers and pain centers, and avoid unauthorized access to data due to ambiguous permission rules.

[0125] Fifth, through the single-center and cross-center dual index abnormal function detection mechanism, combined with the loss function to reversely optimize the decision tree node authority threshold and the cross-center security factor weight, a "detection and optimization" closed loop is formed; this mechanism enables the system to dynamically adjust the access threshold of highly sensitive nodes (such as chief physician authority control) and optimize the cross-tree connection weight according to the frequency of feature co-occurrence (such as the association between BRCA1 mutation and family history of breast cancer), thereby ensuring the security of the most sensitive data such as "ID number" while improving the access efficiency of medium-sensitive data such as "disease name"; Fourth, the Monte Carlo simulation is introduced to dynamically adjust the cross-center security index factor, combined with the gradient descent algorithm to iteratively optimize the access strategy, so that the model It can adapt to differences in data security levels across different centers. For example, a bidirectional mapping between access levels and sensitivity levels is built into the cross-center splicing signature code, enabling real-time compliance with health and medical data security guidelines. It dynamically adjusts data transmission encryption levels and access control policies to ensure privacy protection when cross-center data is associated, such as through Paillier encryption and Groth16 proofs for secure signature verification. Finally, the adaptively trained, cross-center global permission-based hierarchical access node network integrates security mechanisms such as BloomFilter deduplication and EC-Schnorr zero-knowledge proofs to achieve precise permission interception and path optimization for real access requests. While ensuring the non-leakage of highly sensitive data such as genetic test reports, this system, through the collaboration of hierarchical hash indexing and permission mapping, enables authorized users such as doctors to efficiently obtain a complete cross-center patient portrait, improving the efficiency of multi-center medical collaboration. At the same time, through dynamic anchor point algorithms and unique verification of splicing signature codes, it ensures the immutability and secure traceability of cross-center data indexes, forming a technical closed loop of security protection, efficient access, and compliance management.

[0126] Sixth, this embodiment also achieves precise positioning of sensitive features of medical data from structured to semantic through the technical coupling of multimodal feature fusion and hierarchical sensitivity assessment, providing fine-grained semantic support for cross-center permission control. The decision tree architecture and hierarchical hash index work together to organize sensitive features according to risk levels and assign permission identifiers, forming an integrated semantic and permission index structure, which not only ensures the isolated storage of highly sensitive data, but also improves the efficiency of retrieval of features of the same level through Bloom filter deduplication. The dynamic anchor algorithm is combined with Paillier homomorphic encryption to balance security and efficiency by dynamically adjusting the anchor interval when splicing cross-center indexes, and combines Groth16 proof to achieve security association of data availability and invisibility. The integration of access control algorithm and Monte Carlo simulation enables the system to dynamically optimize permission thresholds based on historical access patterns and real-time load, and form flexible permission policies that adapt to the needs of multiple roles such as doctors, nurses, and patients through gradient descent iteration, ultimately achieving the implicit technical effect of "sensitive features can be graded, permission control can be dynamic, and security verification can be traced" for cross-center medical data.

[0127] Example 2

[0128] See also Figure 3 Another embodiment provided by the present invention is an information management method supporting multi-center data fusion, comprising:

[0129] S1, obtaining monitoring information of users at different centers, and obtaining a monitoring node network of different center portrait information by using a graph algorithm based on the monitoring information of users at different centers combined with a preset single user hash index; the graph algorithm of this embodiment is preferably an algorithm such as an adjacency index, a path index, or a subgraph index;

[0130] S2, based on the security index hash table constructed by the portrait information monitoring node network of different centers combined with all user hash indexes, obtain a cross-center global portrait information monitoring node network through an association matching algorithm; the security index hash table includes a single-center user hash index and a cross-center user hash index mapping;

[0131] S3, based on the cross-center global portrait information monitoring node network and the preset hierarchical authority criteria, obtain the cross-center global authority hierarchical access node network through the access control algorithm; the hierarchical authority criteria are constructed based on the access object authority combined with the information sensitivity level of the cross-center global portrait information monitoring node network and the cross-center information security level;

[0132] The single-center user hash index is obtained by combining the user's basic information with the user's corresponding center monitoring information through a hash algorithm; the cross-center user hash index mapping is constructed based on the user hash indexes of the same user in different centers combined with a dynamic feature anchor algorithm and a preset effective time;

[0133] S4, determine the user's access intention and permission level based on the access user information, and obtain the access user index intention information list based on the user's access intention and permission level and the access frequency of historical intention information combined with the cross-center global permission hierarchical access node network.

[0134] Example 3

[0135] An electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements an information management method supporting multi-center data fusion when executing the computer program.

[0136] A computer-readable storage medium stores computer instructions, which, when executed, execute an information management method supporting multi-center data fusion.

[0137] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also change, modify, replace and modify the above-mentioned embodiments without departing from the scope of protection of the purpose of the present invention and the claims, and all of these are protected by the present invention.

[0138] If the technical solution disclosed herein involves personal information, the product using the technical solution disclosed herein has clearly informed the individual of the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the technical solution disclosed herein involves sensitive personal information, the product using the technical solution disclosed herein has obtained the individual's separate consent before processing the sensitive personal information and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, a clear and prominent sign is set to inform that the individual has entered the personal information collection scope and that personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that they agree to the collection of their personal information; or on the personal information processing device, when the personal information processing rules are notified by obvious signs / information, the individual's authorization is obtained through pop-up information or by asking the individual to upload their personal information. The personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.

Claims

1. An information management platform supporting multi-center data fusion, characterized by: include: The portrait index module is used to obtain user monitoring information from different centers, and based on the user monitoring information from different centers combined with the preset single center user hash index, obtain the monitoring node network of different center portrait information through graph algorithms; The global fusion module, based on the security index hash table constructed by the portrait information monitoring node network of different centers and all user hash indexes, obtains a cross-center global portrait information monitoring node network through an association matching algorithm; the security index hash table includes a single-center user hash index and a cross-center user hash index mapping; The hierarchical permission module is based on the cross-center global portrait information monitoring node network combined with the preset hierarchical permission criteria, and obtains the cross-center global permission hierarchical access node network through the access control algorithm; The hierarchical authority criteria are constructed based on the access object authority combined with the cross-center global portrait information monitoring node network information sensitivity level and the cross-center information transmission security level; The cross-center access module determines the user's access intention and permission level based on the access user information, and obtains the access user index intention information list based on the user's access intention and permission level and the access frequency of historical intention information in combination with the cross-center global permission hierarchical access node network through the preset cross-center business isolation algorithm; the single-center user hash index is obtained through the hash algorithm based on the user's basic information combined with the user's corresponding center monitoring information; the cross-center user hash index mapping is constructed based on the user hash index of the same user in different centers combined with the dynamic feature anchor algorithm and the preset effective time.

2. The information management platform supporting multi-center data fusion according to claim 1, characterized in that: The portrait index module includes a portrait unit, a sensitivity evaluation unit, a single index unit and a node construction unit; The profiling unit obtains a single-center user portrait and a corresponding sensitive feature extraction space based on the basic information of a single user in different centers and the multimodal monitoring information combined with a preset multimodal profiling model; The sensitivity evaluation unit obtains a hierarchical sensitivity information space corresponding to the single-center user information based on the single-center user portrait and the corresponding sensitive feature extraction space in combination with a preset hierarchical sensitivity evaluation model; The single indexing unit obtains a hierarchical hash permission index corresponding to the single central user through a hash algorithm based on the hierarchical sensitivity information space corresponding to the single central user information and the preset permission level; The node construction unit constructs edge weights based on the similarity of monitoring information between single portrait nodes constructed based on a single center user portrait and corresponding hierarchical hash authority index, and obtains a single center portrait information monitoring node network.

3. The information management platform supporting multi-center data fusion according to claim 2, characterized in that: The global fusion module includes a global mapping unit and a cross-center mapping unit; The global mapping unit obtains the cross-center mapping feature code of the corresponding user between different centers based on the hierarchical hash authority index corresponding to the same user in the portrait information monitoring node network of different centers combined with the dynamic feature anchor algorithm; The cross-center mapping unit obtains the cross-center user hash index mapping based on the cross-center mapping feature codes of all users in different center portrait information monitoring node networks through a feature splicing algorithm combined with a preset random valid time interval, and obtains the cross-center global portrait information monitoring node network based on the cross-center user hash index mapping combined with the center portrait information monitoring node network through a topology algorithm.

4. The information management platform supporting multi-center data fusion according to claim 3, characterized in that: The hierarchical authority module includes a hierarchical authority unit and a hierarchical authority mapping unit; The hierarchical permission unit obtains the hierarchical access permission level corresponding to the access user based on the basic information of the access user and the preset sensitivity level range of the accessible information; The hierarchical permission mapping unit performs cross-center permission information index simulation through an access control algorithm combined with a simulation algorithm based on the hierarchical access permission level corresponding to the accessing user and the hierarchical sensitivity information space and cross-center mapping feature code corresponding to each user in the cross-center global portrait information monitoring node network, to obtain a cross-center global permission hierarchical access node network.

5. The information management platform supporting multi-center data fusion according to claim 4, characterized in that: The process of constructing the cross-center user hash index mapping includes: Input the basic information and multimodal monitoring information of each user in a single center into the multimodal feature extraction layer of the modal portrait model to obtain the basic feature space of a single user and a single monitoring state feature space of the current center; The basic features of a single user are set to the highest sensitivity level. At the same time, the single monitoring state feature space is input into the sensitive keyword feature extraction layer with a pre-set sensitive word level library to obtain the single user's graded sensitive word sequence and the number of sensitive words at the corresponding level; According to the graded sensitive word sequence of a single user and the number of sensitive words of the corresponding level, the comprehensive sensitivity level and single level sensitivity of the single user are obtained through the sensitivity level assessment layer.

6. The information management platform supporting multi-center data fusion according to claim 5, characterized in that: The process of constructing the cross-center user hash index mapping also includes: Based on the comprehensive sensitivity level and single-level sensitivity of a single user, a decision tree structure is used. The comprehensive sensitivity level is used as the trunk, and the basic characteristics of a single user are set as the root node. At the same time, based on the single-level sensitivity, the keyword information corresponding to each sensitivity level is stored in the bifurcation node of the corresponding sensitivity level on the trunk in a manner from near to far. This obtains a single user sensitive information decision tree. Based on the information stored in each node of the decision tree for sensitive information of a single user, the label and distinguishing identification information corresponding to each node are obtained through a random interception algorithm; Based on the label and distinguishing identification information corresponding to each node, a hierarchical hash index corresponding to each node of the single user sensitive information decision tree is obtained through a hash algorithm.

7. The information management platform supporting multi-center data fusion according to claim 6, characterized in that: The process of constructing the cross-center user hash index mapping also includes: According to the sensitivity level corresponding to the access rights of historical users in a single center, the access rights-sensitivity mapping function in a single center is obtained through support vector machine; The access right-sensitivity mapping function enables the accessing user to access the current user's sensitive information decision tree in a single center, and at the same time access the user information stored in the nodes of the same sensitivity level and below in the sensitive information decision tree of the associated user; Based on the access permission-sensitivity mapping function within a single center and the hierarchical hash index corresponding to each node, the hierarchical hash permission index corresponding to each node is obtained and built into the node corresponding to the single user sensitive information decision tree to obtain a single user hierarchical permission decision tree; Based on the information stored in the nodes of all single-user hierarchical permission decision trees in a single center and the corresponding hierarchical hash permission indexes, the association similarity between each single-user hierarchical permission decision tree and the node information of the same sensitivity level and the lower-level node information in the remaining single-user hierarchical permission decision trees is obtained through the lower-level association matching algorithm.

8. The information management platform supporting multi-center data fusion according to claim 7, characterized in that: The process of constructing the cross-center user hash index mapping also includes: Based on the correlation similarity between each single user hierarchical permission decision tree and the node information of the same sensitivity level and the lower level node information in the remaining single user hierarchical permission decision trees, a correlation similarity connection is constructed, and the hierarchical hash permission index corresponding to the two connected nodes is combined with the public key obtained by the zero-proof algorithm and embedded into the corresponding connection. At the same time, the generated private key is respectively configured into the two connected nodes to obtain a single central user hierarchical permission decision forest; Based on the hierarchical hash permission indexes built into nodes of different sensitivity levels in the single user hierarchical permission decision tree for the same user in different centers, a dynamic anchor point algorithm is used to randomly intercept index subsequences of length N from the hierarchical hash permission indexes of different levels and then concatenate all intercepted index subsequences to obtain a cross-center fused index sequence. At the same time, feature index information of length M is extracted from each index subsequence corresponding to the center where the same user information exists, and based on the two feature index information extracted from any two centers that simultaneously contain the same user, a verifiable segmentation algorithm is used to obtain the corresponding splicing feature code between any two cross-center fusion index sequences, where N>M.

9. The information management platform supporting multi-center data fusion according to claim 8, characterized in that: The process of constructing the cross-center user hash index mapping also includes: Based on the corresponding splicing feature codes between any two cross-center fusion index sequences, the cross-center fusion index sequences corresponding to a single user in any two centers are spliced together, and the bidirectional mapping between the access levels of any two centers and the sensitivity levels of non-centers and the cross-center security index factor constructed by the cross-center data security level are built into the splicing feature codes corresponding to the splicing of any two centers to obtain a single user cross-center user hash index mapping; Based on the above single user cross-center user hash index mapping construction process, obtain the single user cross-center user hash index mapping corresponding to all users; Based on the single-center user hierarchical permission decision forest combined with the single-user cross-center user hash index mapping corresponding to all users, the access control algorithm is combined with the simulation algorithm and the index exception function corresponding to the single-center user hierarchical permission decision forest and the global index exception function corresponding to the cross-center user hierarchical permission decision forest. The single-center sub-authority information index and the cross-center sub-authority information index training are performed to obtain the trained cross-center global permission hierarchical access node network.

10. An information management method supporting multi-center data fusion, which is implemented based on an information management platform supporting multi-center data fusion according to any one of claims 1 to 9, characterized in that: include: Obtain user monitoring information from different centers, and based on this information, combine it with the preset single user hash index and use graph algorithms to obtain a monitoring node network of different center portrait information. Based on the security index hash table constructed by the portrait information monitoring node network of different centers combined with all user hash indexes, a cross-center global portrait information monitoring node network is obtained through an association matching algorithm; the security index hash table includes a single-center user hash index and a cross-center user hash index mapping; Based on the cross-center global portrait information monitoring node network combined with the preset hierarchical authority criteria, through the access control algorithm, a cross-center global authority hierarchical access node network is obtained; The hierarchical authority criteria are constructed based on the access object authority combined with the cross-center global portrait information monitoring node network information sensitivity level and the cross-center information security level; The single-center user hash index is obtained by combining the user's basic information with the user's corresponding center monitoring information through a hash algorithm; the cross-center user hash index mapping is constructed based on the user hash indexes of the same user in different centers combined with a dynamic feature anchor algorithm and a preset effective time; The user's access intention and permission level are determined based on the access user information, and the access user index intention information list is obtained based on the user's access intention and permission level and the access frequency of historical intention information combined with the cross-center global permission hierarchical access node network.

Images