Mobile collaboration platform data management method

Abstract

The invention discloses a mobile collaboration platform data management method, and relates to the technical field of data processing, and the management method comprises the following specific steps: S100, data sensitive grading: calculating sensitivity scores of various types of data in a mobile collaboration platform by using a formula. According to the method, the sensitivity scores of various data are calculated through a formula, and the data are divided according to the sensitivity levels, so that different levels of protection measures are taken for the data with different sensitivity degrees, and meanwhile, through a built-in sensor and a built-in behavior monitoring module, the sensitivity of the data is improved. The operation behavior information, the equipment information and the business process information of the user are collected in real time, the current business scene is intelligently judged, a basis is provided for subsequent privacy protection strategy making, the data processing efficiency is improved, the data security is greatly enhanced, and data leakage and abuse are effectively prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and specifically provides a method for managing data in a mobile collaboration platform. Background Art

[0002] With the rapid development of information technology, mobile collaboration platforms have become an indispensable part of the daily operations of enterprises. These platforms integrate various functional modules to achieve real-time data sharing, collaborative editing, and efficient management, greatly improving the operational efficiency of enterprises. However, with the continuous increase in the amount of data and the increasing diversification of data types, how to ensure the security and privacy of this data has become an urgent problem to be solved. Especially in the case of sensitive data, such as customer information and financial data, once leaked, it will cause immeasurable losses to the enterprise.

[0003] Traditional data management methods have many deficiencies when dealing with the data security challenges of mobile collaboration platforms. On the one hand, traditional methods often adopt a one-size-fits-all data classification method, which cannot perform refined management according to the actual content and potential impact degree of the data. This results in the fact that sensitive data may not be adequately protected, while non-sensitive data may be affected in terms of usage efficiency due to overprotection. On the other hand, traditional methods lack flexibility in privacy protection and cannot be dynamically adjusted according to the privacy preferences of users and changes in business scenarios. This restricts the sharing and circulation of data on the collaboration platform, affecting the collaborative efficiency and innovation ability of enterprises.

[0004] Therefore, developing a method for managing data in a mobile collaboration platform not only improves the level of data security and privacy protection but also ensures the circulation and collaborative efficiency of data, providing strong support for the digital transformation and business development of enterprises. Summary of the Invention

[0005] The purpose of the present invention is to make up for the deficiencies of the prior art and provide a method for managing data in a mobile collaboration platform. This method realizes the full-life cycle management of data from collection, processing, storage to use through seven core steps: data sensitive classification, scenario intelligent discrimination, preference modeling update, policy dynamic adaptation, desensitization precise implementation, measure full-chain control, and feedback-driven optimization. This method not only uses advanced algorithms and technologies to provide personalized protection for user privacy but also effectively responds to data security risks through real-time monitoring and dynamic adjustment of policies.

[0006] To solve the above technical problems, the present invention provides the following technical solution: A method for managing data in a mobile collaboration platform, and the specific steps of this management method are as follows:

[0007] S100, Data Sensitivity Classification: Calculate the sensitivity scores of various types of data in the mobile collaboration platform using formulas, define sensitivity thresholds based on the content characteristics of the data, the business fields to which they belong, and the potential impact factors of data leakage, and classify the data into different sensitive levels according to the sensitivity scores and sensitivity thresholds;

[0008] S200, Scenario Intelligent Discrimination: Real-time collect the user's operation behavior information, device information, and the information of the currently running business process through the sensors and behavior monitoring modules built into the platform, and judge the current business scenario by analyzing and identifying multi-dimensional data;

[0009] S300, Preference Modeling Update: Collect the historical data of the user's data access records, editing records, sharing records, and privacy setting adjustment records on the platform, extract the privacy preference characteristics of the user for data of different sensitive levels in different business scenarios, build a personalized privacy protection model for each user through the user privacy preference modeling formula, and update it according to the real-time data;

[0010] S400, Policy Dynamic Adaptation: Build a policy library containing various privacy protection technologies and policies, generate matching privacy protection policies from the policy library according to the determined data sensitive level, business scenario, and the user's personalized privacy protection model through the dynamic privacy policy matching formula. At the same time, dynamically adjust the generated privacy protection policies according to the platform's performance monitoring data and the user's real-time operation feedback;

[0011] S500, Precise Implementation of Data Masking: When the platform is in a scenario where data needs to be externally displayed or shared, verify the identity and permission information of the data recipient. According to the access level granted to it and the business requirements previously negotiated with the data provider, desensitize the data through the real-time data masking optimization formula. Let the permission level of the recipient be P R , and the business requirement be B R Implement data masking on sensitive data. The masked data is D′, and the original data is D. The formula is: where g() is the masking function, I is the set of sensitive information elements, is the masking weight of each sensitive information element, is the masking operation on the sensitive information element s i in data D. When the data is updated, dynamically adjust the masking weight and masking rules by monitoring the change amount ΔD and update frequency f update of the data. The calculation formula is: where δ is an adjustment coefficient;

[0012] S600, Full-chain control of measures: According to the sensitivity level of the data and the privacy protection policy, store the data in storage areas with different security levels. During the data transmission process, select the network transmission protocol and encryption method according to the privacy protection policy. In the data usage link, through the access control and auditing mechanism at the application programming interface level, verify and record the access requests of the application programs, analyze the relevant data from the data storage, transmission, and usage links in real time, conduct real-time evaluation through the privacy monitoring alarm judgment function to discover potential data security risks, and take corresponding emergency measures according to the alarm type and severity level;

[0013] S700, Feedback-driven optimization: Set up a privacy feedback entry in the user interface of the mobile collaboration platform, which can provide feedback opinions and suggestions to the platform through text input and menu selection methods. The system analyzes the feedback information through the built-in data processing center, and based on the analysis results, conducts targeted optimization and iterative upgrade of the privacy protection model, de-sensitization rules, and privacy policies.

[0014] Furthermore, in the S100, data sensitivity classification, calculate the sensitivity scores of various types of data in the mobile collaboration platform through the data sensitivity evaluation formula. Let the sensitivity score of data D be S D , by calculating the frequency F of the key information in the data w , the potential influence range R of the data, and the importance weight W of the data i to obtain the sensitivity score through comprehensive calculation. The formula is: S D = ∑ w∈W (F w × W w ) + R × α + ∑ i∈I (W i × β), where W w is the weight of each key information word, and α and β are adjustment coefficients.

[0015] Even further, in the S100, data sensitivity classification, for the division of sensitivity levels, when S D < T1 is low-sensitivity data, T1 ≤ S D < T2 is medium-sensitivity data, S D ≥ T2 is high-sensitivity data, and T1 and T2 are preset thresholds.

[0016] Even further, in the S200, scenario intelligent discrimination, identify the business scenarios where users are located on the mobile collaboration platform through the scenario intelligent discrimination formula, and define the business scenario vector which is composed of the user operation behavior feature vector the device environment feature vector and the business process feature vector . The calculation formula is: Among them, ω1, ω2, and ω3 are weight parameters, representing the contribution degree of each feature vector to the business scenario recognition. By calculating the cosine similarity between the business scenario vector and the predefined standard business scenario vector set the business scenarios are classified, and the formula is: When the similarity is greater than the preset threshold θ, it is determined that the current business scenario is the corresponding standard business scenario.

[0017] Furthermore, in the S200, scene intelligent discrimination, the business scenarios include the following scenarios: Office business scenarios: daily office operation scenarios and remote office scenarios; Project collaboration business scenarios: internal project team collaboration scenarios and cross-departmental project collaboration scenarios; Data sharing and external cooperation business scenarios: internal data sharing scenarios and external data cooperation scenarios; High-risk operation business scenarios: sensitive data access scenarios and system maintenance and data backup scenarios.

[0018] Furthermore, in the S300, preference modeling update, a personalized privacy protection model is constructed for each user through the user privacy preference modeling formula. Let the privacy preference value of user U for data D in business scenario BS be P U,D,BS , by analyzing the data access pattern A D in the user's historical operation records, privacy setting behavior S D , and feedback behavior F D for comprehensive quantitative analysis to construct the model, the formula is: P U，D,BS =γ1A D +γ2S D +γ3F D , where γ1, γ2, and γ3 are weight coefficients.

[0019] Furthermore, in the S400, policy dynamic adaptation, the matching privacy protection policy is generated according to the policy library through the dynamic privacy policy matching formula. The formula is: where is the finally generated most matching privacy protection policy vector, is the i-th privacy protection policy vector in the policy library, ω i is the weight coefficient of the i-th policy, S D is the data sensitivity level score, S Di is the data sensitivity level reference score adapted by the i-th policy, σ(x) is a Sigmoid function, is the current business scenario vector, is the business scenario vector adapted by the i-th policy, is the cosine similarity between the two, measuring the matching degree of the business scenario, P U,D,BS is the privacy preference value of the user, is the reference value of the user privacy preference adapted by the i-th policy, and τ is the adjustment coefficient.

[0020] Furthermore, in the S600, for the storage of data with different sensitivity levels in the whole-chain control of measures: highly sensitive data is stored in a dedicated storage device using hardware encryption technology; medium-sensitive data is stored in a software-encrypted partition, and an access control list is used to restrict access only to authorized users and applications; low-sensitive data is stored in a general storage area and encrypted using basic encryption technology.

[0021] Furthermore, in the S600, corresponding measures are taken for data with different sensitivity levels in the whole-chain control of measures: for highly sensitive data, it is encrypted using the two-way authentication SSL / TLS protocol and the AES-256 encryption algorithm, and a traffic monitoring device is deployed to analyze the traffic in real time. In case of a large number of sudden transmissions, abnormal IP-intensive access, and abnormal port connections, the transmission is immediately blocked and a high-level alarm is sent to notify the investigation; for medium-sensitive data, the standard SSL / TLS protocol and the AES-192 encryption algorithm are used, and the traffic is continuously monitored using conventional tools. In case of continuous fluctuations, frequent retransmissions, or a small number of abnormal IP connections, the transmission is suspended for inspection and a medium-level alarm is sent; for low-sensitive data, the basic SSL / TLS protocol and the AES-128 encryption algorithm are selected, and sampling monitoring is performed through the basic monitoring module. In case of continuous transmission failures or a large number of invalid data packet transmissions, the connection is temporarily interrupted and reconnected, and a low-level alarm is sent to notify the maintenance personnel to pay attention to the network stability.

[0022] Furthermore, in the S600, potential data security risks are discovered through real-time evaluation using the privacy monitoring alarm judgment function, and a privacy protection measure monitoring index vector is established. including the data integrity index I d 、the access anomaly index A access 、the system performance index P system , and the formula is: wherein, the weight parameters μ1, μ2, and μ3 are compared with the preset security threshold vector , and the formula is: When the function value is 1, a security alarm is triggered.

[0023] Compared with the prior art, the data management method of a mobile collaboration platform has the following beneficial effects:

[0024] I. Through data sensitivity grading and scenario intelligent discrimination technologies, the present invention realizes refined management and dynamic protection of data. This method calculates the sensitivity scores of various types of data through formula calculation and divides the data according to the sensitivity levels, thereby ensuring different levels of protection measures for data with different sensitivities. At the same time, through built-in sensors and behavior monitoring modules, it collects the operation behavior information, device information, and business process information of users in real time, intelligently discriminates the current business scenario, provides a basis for formulating subsequent privacy protection strategies, not only improves the efficiency of data processing, but also greatly enhances the security of data, effectively preventing data leakage and abuse.

[0025] II. The present invention constructs a policy library containing a variety of privacy protection technologies and strategies, and dynamically generates matching privacy protection strategies according to the determined data sensitivity levels, business scenarios, and users' personalized privacy protection models, making data management more flexible and adaptable to privacy protection requirements in different scenarios. At the same time, when the data needs to be externally displayed or shared, the present invention can also perform precise data desensitization processing according to the identity and permission information of the data recipient and business requirements, not only protecting the security of sensitive data, but also ensuring the availability and business continuity of the data.

[0026] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0028] Figure 1 It is a flowchart of a data management method for a mobile collaboration platform;

[0029] Figure 2 It is a process framework diagram of a data management method for a mobile collaboration platform. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] To further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention objectives, the following will, in conjunction with the accompanying drawings and preferred embodiments, describe in detail the specific embodiments, structures, features, and their effects of the present invention as follows.

[0031] Embodiment 1

[0032] Enterprise internal office scenario

[0033] For various types of data in the enterprise internal office platform with data sensitivity grading, such as employees' basic information (name, employee number, contact information), financial statement data, and project documents. For employees' basic information, the frequency F of its key information w is relatively low, the potential impact range R is mainly within the enterprise, and the importance weight W i is average. After being evaluated by the data sensitivity assessment formula, it is determined as low-sensitivity data. The formula is: S D = ∑ w∈W (F w ×W w ) + R×α + ∑ i∈I (W i ×β); The financial statement data involves the enterprise's financial situation, the frequency F of its key information w is high, the potential impact range R is relatively wide, and the importance weight W i is high, so it is determined as high-sensitivity data; For project documents, according to the importance of the project and the key information involved, some may be medium-sensitivity data. The basis for dividing the sensitivity level is: when S D <T1 is low-sensitivity data, T1 ≤ S D <T2 is medium-sensitivity data, S D ≥ T2 is high-sensitivity data, where T1 and T2 are preset thresholds.

[0034] Scene intelligent discrimination. When an employee is in the daily office operation scenario, and performs document editing and email sending and receiving operations through the platform, the platform's built-in sensors and behavior monitoring modules collect the employee's operation behavior information of using office software, device information (such as device type, IP address), and the information of the current running business process (such as the project process being processed). Define the business scenario vector BS, which consists of the user operation behavior feature vector OB, the device environment feature vector DE, and the business process feature vector BP. The calculation formula is: BS = ω1OB + ω2DE + ω3BP, where ω1, ω2, and ω3 are weight parameters. By calculating the cosine similarity between the business scenario vector BS and the predefined standard business scenario vector set BS i to divide the business scenario. The formula is: When the similarity is greater than the preset threshold θ, it is determined that the current business scenario is the corresponding standard business scenario. For example, it is determined that the current is the daily office operation scenario, belonging to the office type of business scenarios (including the daily office operation scenario and the remote office scenario).

[0035] Preference Modeling Update. Taking an employee as an example, the system collects historical data on their data access records (such as frequently accessing project-related documents), editing records (modification of specific project documents), sharing records (sharing documents with project team members), and privacy setting adjustment records (such as setting access permissions for certain sensitive data) on the platform. Let the privacy preference value of user U for data D in business scenario BS be P U，D,BS , by analyzing the data access pattern A D in the user's historical operation records D , privacy setting behavior S D , and feedback behavior F U，D,BS for comprehensive quantitative analysis to build a model. The formula is: P D = γ1A D + γ2S D + γ3F, where γ1, γ2, and γ3 are weight coefficients. Through analysis, it is found that in the project collaboration scenario, this employee tends to share sensitive-level data in project documents within the project team, while strictly restricting access to highly sensitive-level financial data, thereby building their personalized privacy protection model and updating it in real time according to their subsequent operations.

[0036] Policy Dynamic Adaptation. When an employee accesses project documents (medium-sensitive data) in the daily office operation scenario, let the data sensitivity level score be S D , the current business scenario vector be BS, and the user's privacy preference value be P U，D,BS . According to the data sensitivity level, business scenario, and the employee's personalized privacy protection model, a suitable privacy protection policy is selected from the policy library through the dynamic privacy policy matching formula. The formula is: where PP * is the finally generated most matching privacy protection policy vector, PP i is the i-th privacy protection policy vector in the policy library, ω i is the weight coefficient of the i-th policy, S Di is the data sensitivity level reference score adapted by the i-th policy, σ(x) is a Sigmoid function, BS i is the business scenario vector adapted by the i-th policy, and cos(BS, BS i ) is the cosine similarity between the two, measuring the matching degree of the business scenario. is the reference value of the user privacy preference adapted by the i-th policy, and T is the adjustment coefficient. For example, in terms of data storage, store project documents in a software-encrypted partition and use an access control list to restrict access to only project team members; when transmitting data, use the standard SSL / TLS protocol and the AES-192 encryption algorithm; in the data usage link, through access control and auditing mechanisms at the application programming interface level, record the employee's access operations on project documents. At the same time, according to platform performance monitoring data (such as network bandwidth, server load) and the employee's real-time operation feedback (such as whether the operation is smooth), dynamically adjust the generated privacy protection policy. For example, when the network is congested, appropriately optimize the parameters of the encryption algorithm to improve transmission efficiency.

[0037] Desensitization is precisely implemented. When an enterprise needs to show some project documents to external partners (the permission level P of the external partner has been determined R and business requirement B R ), desensitize the sensitive information in the project documents. Let the sensitive data be D, and the desensitized data be D ′ . Desensitize the data through the real-time data desensitization optimization formula, and the formula is: where g() is the desensitization function, I is the set of sensitive information elements, is the desensitization weight of each sensitive information element, is the desensitization operation on the sensitive information element s i in data D. When the data is updated, by monitoring the change amount ΔD and update frequency f update of the data, dynamically adjust the desensitization weight and desensitization rules, and the calculation formula is: where δ is an adjustment coefficient. For example, for sensitive information elements such as customer names and internal project numbers in the document, according to the desensitization rules and the permission level of external partners, perform partial hiding or replacement processing on them, and generate desensitized data to provide to external partners to ensure the protection of enterprise sensitive information while meeting business requirements.

[0038] Implement full-chain control of measures. Highly sensitive data (such as financial statement data) is stored in dedicated storage devices using hardware encryption technology, encrypted using the two-way authentication SSL / TLS protocol and the AES-256 encryption algorithm. Deploy traffic monitoring devices to analyze traffic in real time. In case of a large number of sudden transmissions, abnormal IP-intensive access, and abnormal port connections, immediately block the transmission and issue a high-level alarm to notify for investigation; Medium-sensitive data (such as project documents) is stored in partitions encrypted by software, using the standard SSL / TLS protocol and the AES-192 encryption algorithm, continuously monitor traffic with conventional tools. If there are continuous fluctuations, frequent retransmissions, or a small number of abnormal IP connections, suspend the transmission for inspection and send a medium-level alarm; Low-sensitive data (such as employee basic information) is stored in ordinary storage areas, encrypted using basic encryption technology, select the basic SSL / TLS protocol and the AES-128 encryption algorithm, sample and monitor through the basic monitoring module. If there are continuous transmission failures or a large number of invalid data packet transmissions, temporarily interrupt and reconnect and issue a low-level alarm to notify the maintenance personnel to pay attention to network stability. At the same time, establish a privacy protection measure monitoring index vector PM, including the data integrity index I D 、access anomaly index A accrss 、system performance index P system ,The formula is: PM = μ1I D +μ2A access +μ3P system where μ1, μ2, and μ3 are weight parameters compared with the preset security threshold vector T: When the function value is 1, trigger a security alarm. In the data usage link, through the access control and auditing mechanism at the application programming interface level, verify and record the access requests of applications to data to ensure the security of data throughout the process.

[0039] Feedback-driven optimization. During the process of employees using the platform, if they find problems or have improvement suggestions in terms of privacy protection, they can provide feedback to the platform through the privacy feedback entry on the user interface (such as entering details of the problem in text or selecting common problem types through a menu). The built-in data processing center of the system analyzes the feedback information. According to the analysis results, specifically optimize and iteratively upgrade the privacy protection model, de-sensitization rules, and privacy policies. For example, if multiple employees feedback that encryption causes slow operations when accessing certain data, the platform will optimize the encryption algorithm or policy, such as adjusting the encryption strength or replacing it with a more efficient encryption method, and at the same time specifically optimize and iteratively upgrade the privacy protection model and de-sensitization rules to improve the user experience and data security.

[0040] In summary, in the internal office scenario of an enterprise, the data management method of the present invention demonstrates good adaptability and effectiveness. Through data sensitivity classification, it accurately evaluates the sensitivity of various office data, providing a basis for subsequent differential processing. The scenario intelligent discrimination accurately identifies the business scenarios where employees are located. Whether it is the scenario of daily office operations or project collaboration, they can be effectively distinguished. The preference modeling updates the personalized privacy protection model constructed for each employee, fully considering the unique needs of employees for data at different sensitivity levels in different scenarios, achieving precise privacy protection. The policy dynamic adaptation generates and adjusts the privacy protection policy in real time according to the data, scenario, and employee preferences. The specific measures in the data storage, transmission, and usage links ensure the security and availability of the data. The desensitization is precisely implemented. When presenting data externally, it protects the sensitive information of the enterprise while meeting the business requirements. The measure full-chain control monitors and manages the data comprehensively from storage, transmission to usage, effectively preventing various data security risks. The feedback-driven optimization further promotes the continuous improvement of the system, enabling the entire data management system to continuously adapt to the changes and developments of enterprise office, improving the efficiency and security of data management, and providing a strong guarantee for the secure and efficient utilization of internal office data of the enterprise.

[0041] Embodiment 2:

[0042] Scientific research project collaboration scenario

[0043] Data sensitivity classification: On the scientific research project collaboration platform, scientific research data (such as experimental data, research results), scientific research personnel information (such as contact information, research directions), and project management data (such as project progress, budget). Scientific research data often contains key research results and experimental data, with a high frequency of key information words, and the potential impact range may cover the entire scientific research field, and the importance weight is extremely high, so it is classified as high-sensitivity data; the importance of scientific research personnel information is relatively low, so it is low-sensitivity data; for project management data, according to the importance of the project and the amount of funds involved, some may be medium-sensitivity data.

[0044] Scenario intelligent discrimination: When scientific research team members conduct experimental data sharing and discussion of research plans in the internal collaboration scenario of the project team, the platform collects the operation behaviors of members using scientific research collaboration software (such as data sharing operations, discussion behaviors in specific research sections), device information (such as the connection status of scientific research equipment used, device types), and the current running scientific research project process information (such as whether it is in the experimental stage or the data analysis stage) through sensors and behavior monitoring modules. After analysis, it is determined that the current business scenario is the internal collaboration scenario of the project team, belonging to the project collaboration type of business scenario.

[0045] Preference modeling update. Taking a certain researcher as an example, historical data on their data access records (such as frequently accessing specific experimental data), editing records (modifying and supplementing experimental data), sharing records (sharing some research results with team members), and privacy setting adjustment records (such as setting the sharing scope of their own research data) on the platform is systematically collected. It is analyzed that in the project collaboration scenario, the researcher is relatively cautious when sharing highly sensitive data of the experimental data they are responsible for within the project team, only allowing specific members to access, while being relatively lenient with medium-sensitive data of project management data. Thus, their personalized privacy protection model is constructed and updated in real-time according to their subsequent operations in the project. For example, when data sharing becomes more strict at a critical stage of the project, the model correspondingly adjusts their privacy preferences.

[0046] Policy dynamic adaptation. When a researcher accesses experimental data (highly sensitive data) in the internal collaboration scenario of the project team, according to the data sensitivity level, business scenario, and the researcher's personalized privacy protection model, a privacy protection policy is matched from the policy library. In terms of data storage, the experimental data is stored in a dedicated storage device using hardware encryption technology; when data is transmitted, a two-way authentication SSL / TLS protocol and AES-256 encryption algorithm are adopted, and the policy is dynamically adjusted according to the platform performance monitoring data (such as server storage capacity, transmission speed) and the researcher's real-time operation feedback (such as whether the data reading speed meets the requirements). For example, the storage structure is optimized to accelerate the data reading speed. In the data usage link, strict access control and auditing mechanisms at the application programming interface level are used to record every access and operation of the researcher on the experimental data, ensuring data security and traceability.

[0047] Precise implementation of data masking. When a scientific research project needs to conduct preliminary result exchanges with external cooperation institutions and show some experimental data (the permission levels and business requirements of the external cooperation institutions have been determined), the experimental data is masked. For example, for sensitive information elements such as exact numerical values and special sample numbers in the experimental data, according to the masking function and the permissions and business requirements of the external cooperation institutions, they are blurred or partially replaced with data to generate masked data and provided to the external cooperation institutions, which not only enables the cooperation institutions to understand the general situation of the project but also protects the core scientific research data.

[0048] Implement full-chain control of measures. Highly sensitive data (such as experimental data) is stored in dedicated hardware-encrypted devices, encrypted for transmission using the two-way authentication SSL / TLS protocol and the AES-256 encryption algorithm. Deploy traffic monitoring devices to analyze traffic in real time. In case of anomalies (such as a large number of abnormal access requests), immediately block the transmission and send a high-level alarm to notify the person in charge of the research team to conduct an investigation. Medium-sensitive data (such as project management data) is stored in software-encrypted partitions, encrypted for transmission using the standard SSL / TLS protocol and the AES-192 encryption algorithm. Continuously monitor traffic with conventional tools. If there are anomalies (such as frequent data modification requests), suspend the transmission for inspection and send a medium-level alarm. Low-sensitive data (such as researchers' information) is stored in ordinary storage areas, encrypted for transmission using the basic SSL / TLS protocol and the AES-128 encryption algorithm. Sampling monitoring is carried out through the basic monitoring module. In case of problems (such as transmission interruption), temporarily interrupt and reconnect and send a low-level alarm to notify the maintenance personnel to check the network. At the same time, in the data usage link, through the access control and audit mechanism of the application programming interface, strictly verify and record the access requests of researchers and external cooperation institutions to the data, ensuring the security of data throughout the process.

[0049] Feedback-driven optimization. During the use of the platform by researchers, if they find that data security protection affects research efficiency (such as slow data processing speed caused by encryption) or there are other privacy protection-related issues, they can provide feedback to the platform through the privacy feedback entry. The system processing center analyzes the feedback. If common problems are found, such as low efficiency of the encryption algorithm in processing large-scale data, the platform will optimize the encryption algorithm or adjust the data storage structure. At the same time, the privacy protection model, de-sensitization rules, and privacy policies will be optimized and iteratively upgraded in a targeted manner to better balance the security of research data and research work efficiency.

[0050] In summary, in the scenario of scientific research project collaboration, the data management method of the present invention plays an important role. The data sensitivity classification reasonably distinguishes the sensitivity levels of scientific research data, personnel information, and project management data, ensuring that key scientific research data is highly valued. The scenario intelligent discrimination accurately judges the operations of scientific research personnel in the internal collaboration scenario of the project team, providing a scenario basis for subsequent processing. The preference modeling updates the constructed personalized privacy protection model to fit the actual needs of scientific research personnel in project collaboration, helping to balance data sharing and privacy protection. The strategy dynamic adaptation selects appropriate strategies based on the data sensitivity level, business scenario, and scientific research personnel preferences, and dynamically optimizes them, ensuring the security and efficiency of scientific research data during storage, transmission, and use. The desensitization is accurately implemented to effectively protect the core scientific research data during external cooperation and communication. The measures for full-chain control target the storage and transmission measures and monitoring systems for scientific research data with different sensitivity levels, comprehensively ensuring the security of scientific research data, preventing data leakage and abnormal situations from interfering with scientific research work. The feedback-driven optimization mechanism prompts the system to continuously improve, better serving the data management needs in scientific research project collaboration, promoting the smooth progress of scientific research work, and enhancing the scientific nature and reliability of scientific research data management.

[0051] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to the above-disclosed technical content to form equivalent embodiments with equivalent changes, as long as they do not depart from the technical solution of the present invention. Any brief modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A mobile collaborative platform data management method, characterized in that: The specific steps of this management method are: S100, Data Sensitivity Classification: Use the formula to calculate the sensitivity scores of various types of data in the mobile collaboration platform, define the sensitivity threshold according to the content characteristics of the data, the business field to which it belongs, and the potential impact of data leakage, and classify the data into different sensitivity levels according to the sensitivity score and sensitivity threshold; S200, intelligent scene identification: Through the platform's built-in sensors and behavior monitoring modules, the platform collects user operation behavior information, device information, and current business process information in real time, and determines the current business scene by analyzing and identifying multi-dimensional data; S300, Preference modeling update: Collect historical data of users’ data access records, editing records, sharing records, and privacy setting adjustment records on the platform, extract users’ privacy preference characteristics for data of different sensitivity levels in different business scenarios, build a personalized privacy protection model for each user through the user privacy preference modeling formula, and update it according to real-time data; S400, dynamic policy adaptation: Build a policy library containing multiple privacy protection technologies and policies. According to the determined data sensitivity level and business scenarios, as well as the user's personalized privacy protection model, generate a matching privacy protection policy based on the policy library through a dynamic privacy policy matching formula. At the same time, dynamically adjust the generated privacy protection policy based on the platform's performance monitoring data and the user's real-time operation feedback; S500, precise implementation of desensitization: When the platform is in a scenario where data needs to be displayed or shared externally, the identity and permission information of the data recipient is verified, and the data is desensitized through the real-time data desensitization optimization formula according to the access level granted to the recipient and the business needs agreed in advance with the data provider. Suppose the permission level of the recipient is P R , business requirement is B R Desensitize sensitive data. The desensitized data is D ′ , the original data is D, and the formula is: Among them, g() is the desensitization function, I is the set of sensitive information elements, is the desensitization weight of each sensitive information element, is the sensitive information element s in data D i Desensitization operation, when the data is updated, by monitoring the data change ΔD and the update frequency f update , dynamically adjust the desensitization weight and desensitization rules, the calculation formula is: Among them, δ is an adjustment coefficient; S600, full-chain control measures: store data in storage areas of different security levels according to the sensitivity level of the data and the privacy protection policy. During the data transmission process, select the network transmission protocol and encryption method according to the privacy protection policy. In the data use link, the access control and audit mechanism at the application interface level is used to verify and record the access request of the application program. The relevant data from the data storage, transmission and use links are analyzed in real time. The privacy monitoring alarm judgment function is used to conduct real-time evaluation to discover potential data security risks, and take corresponding emergency measures according to the alarm type and severity. S700, feedback-driven optimization: A privacy feedback portal is set up in the user interface of the mobile collaboration platform, which can provide feedback and suggestions to the platform through text input and menu selection. The system analyzes the feedback information through the built-in data processing center, and based on the analysis results, conducts targeted optimization and iterative upgrades to the privacy protection model, desensitization rules and privacy policies.

2. A mobile collaborative platform data management method according to claim 1, characterized in that: In S100, the sensitivity scores of various types of data in the mobile collaborative platform are calculated by using a data sensitivity evaluation formula in the data sensitivity classification. Suppose the sensitivity score of data D is S D , by analyzing the key information word frequency F in the data w , the potential impact range R of the data and the importance weight W of the data i The sensitivity score is obtained by comprehensive calculation, and the formula is: S D =∑ w∈W (F w ×W w )+R×α+∑ i∈l (W i ×β), where W w is the weight of each key information word, and α and β are adjustment coefficients.

3. A mobile collaborative platform data management method according to claim 1, characterized in that: The S100, which is the classification of sensitive levels in data sensitivity classification. When S D <T1 is low-sensitive data, T1 ≤ S D <T2 is medium-sensitive data, S D ≥ T2 is high-sensitive data, and T1 and T2 are preset thresholds.

4. A mobile collaborative platform data management method according to claim 1, characterized in that: In the above S200, the business scenario of the user on the mobile collaboration platform is identified by using a scenario intelligent discrimination formula in the scenario intelligent discrimination process, and a business scenario vector is defined. It consists of the user operation behavior feature vector Device environment feature vector and business process feature vector Composition, the calculation formula is: Among them, ω1, ω2 and ω3 are weight parameters, representing the contribution of each feature vector to business scenario recognition. The cosine similarity of is used to divide the business scenarios, and the formula is: When the similarity is greater than a preset threshold θ, the current business scenario is determined to be the corresponding standard business scenario.

5. A mobile collaborative platform data management method according to claim 1, characterized in that: In the above S200, the business scenarios in the scenario intelligent identification include the following scenarios: office business scenarios: daily office operation scenarios and remote office scenarios; project collaboration business scenarios: internal collaboration scenarios of the project team and cross-departmental project collaboration scenarios; Data sharing and external cooperation business scenarios: internal data sharing scenarios and external data cooperation scenarios; High-risk operation business scenarios: sensitive data access scenarios and system maintenance and data backup scenarios.

6. A mobile collaborative platform data management method according to claim 1, characterized in that: In the S300, in the preference modeling update, a personalized privacy protection model is constructed for each user through the user privacy preference modeling formula. Assume that the privacy preference value of user U for data D under the business scenario BS is P U,D,BS , through the data access mode A in the user's historical operation record D , Privacy Setting Behavior D and feedback behavior F D Conduct comprehensive quantitative analysis to build a model, the formula is: P U,D,BS =γ1A D +γ2S D +γ3F D , where γ1, γ2 and γ3 are weight coefficients.

7. A mobile collaborative platform data management method according to claim 1, characterized in that: In the S400, in the dynamic adaptation of the policy, a matching privacy protection policy is generated according to the policy library through a dynamic privacy policy matching formula, and the formula is: in is the best matching privacy protection strategy vector generated in the end. is the i-th privacy protection strategy vector in the strategy library, ω i is the weight coefficient of the i-th strategy, S D is the data sensitivity level score, S Di is the reference score of the data sensitivity level adapted by the i-th strategy, σ(x) is a Sigmoid function, is the current business scenario vector, is the business scenario vector adapted by the i-th strategy, is the cosine similarity between the two, which measures the matching degree of business scenarios. U,D,BS is the user's privacy preference value, is the user privacy preference reference value adapted by the i-th strategy, and τ is the adjustment coefficient.

8. A mobile collaborative platform data management method according to claim 1, characterized in that: The said S600 measures the storage of data of different sensitivity levels in the whole chain control: highly sensitive data is stored in a dedicated storage device using hardware encryption technology; Medium-sensitive data is stored in a partition encrypted by software, and access control lists are used to restrict access to only authorized users and applications; low-sensitivity data is stored in a normal storage area and encrypted using basic encryption technology.

9. A mobile collaborative platform data management method according to claim 1, characterized in that: The S600 measures take corresponding measures for data of different sensitivity levels in the full-chain control: for highly sensitive data, two-way authentication SSL / TLS protocol and AES-256 encryption algorithm are used for encryption, and traffic monitoring equipment is deployed to analyze traffic in real time. In case of large-scale burst transmission, abnormal IP intensive access and abnormal port connection, transmission is immediately blocked and a high-level alarm notification is issued for investigation; for medium-sensitive data, standard SSL / TLS protocol and AES-192 encryption algorithm are used, and conventional tools are used to continuously monitor traffic. In case of continuous fluctuation, frequent retransmission or a small number of abnormal IP connections, transmission is suspended for inspection and a medium-level alarm is issued; For low-sensitivity data, the basic SSL / TLS protocol and AES-128 encryption algorithm are selected, and sampling monitoring is carried out through the basic monitoring module. If continuous transmission failures or a large number of invalid data packets are transmitted, the connection is temporarily interrupted and reconnected, and a low-level alarm is issued to notify maintenance personnel to pay attention to network stability.

10. A mobile collaborative platform data management method according to claim 1, characterized in that: In the S600, measures are used to conduct real-time evaluation in the full-chain control to discover potential data security risks through the privacy monitoring alarm judgment function, and establish a privacy protection measure monitoring indicator vector Including data integrity index I D 、Access abnormality indicator A access 、System performance index P system , the formula is: Among them, the weight parameters μ1, μ2 and μ3 are compared with the preset safety threshold vector For comparison, the formula is: When the function value is 1, a security alert is triggered.

Images