Campus multi-dimensional service resource integration and management method and system

Abstract

The invention discloses a campus multi-dimensional service resource integration and management method and system, and belongs to the technical field of data processing and information management, and the method comprises the steps: collecting a multi-plate interaction record to generate an atomic behavior unit, and storing the atomic behavior unit in a personal data insurance library to construct a dynamic capability portrait; generating an index voucher according to the access rule, and performing semantic feature packaging and resource retrieval on the service request to obtain a preliminarily matched resource response set; a temporary trust integral is generated through identity challenge, and a dynamic trust evaluation result is obtained in combination with a historical interaction mode library, so that priority rearrangement and final service sequence configuration are realized. And finally extracting an execution result and feeding back to the insurance library and updating the mode library. According to the method and the system, an atomic behavior unit is generated from an interaction record, a dynamic capability portrait and an index voucher are constructed, and resources are rearranged and subjected to closed-loop updating in combination with identity challenge verification and dynamic trust evaluation, so that data privacy can be protected, and accurate matching and trusted management of campus service resources can be realized.

Description

Technical Field

[0001] This invention relates to the field of data processing and information management technology, and in particular to a method and system for integrating and managing multi-dimensional service resources on campus. Background Technology

[0002] A campus integrated service platform is an important component of smart campus construction. It aims to integrate multi-dimensional service resources within the campus, including academic research, skills sharing, and daily life services, through information technology, providing faculty and students with a convenient and efficient environment for resource matching and collaboration. These platforms typically involve large-scale user data processing, service supply and demand matching, and trust relationship management among participants, making them typical management and business service systems targeting specific communities.

[0003] Among related technologies, Chinese invention patent CN120410680A discloses a blockchain-based campus secondhand goods trustworthy transaction evaluation system. This system generates user credit scores (Yxf) and classifies credit levels, linking permissions and security deposit mechanisms to effectively prevent fraudulent transactions. It employs convolutional neural networks to detect image tampering traces, combines metadata analysis and multimodal matching mechanisms to calculate image consistency matching scores (Tsv), and classifies trustworthiness to reduce the risk of misjudgment. Furthermore, it uses blockchain to solidify transaction data and construct arbitration evidence packages, automatically assigning responsibility based on preset rules. The system utilizes a consortium blockchain consensus mechanism to ensure data immutability and traceability, improving dispute resolution efficiency and transparency.

[0004] Regarding the aforementioned technologies, the inventors believe they have technical shortcomings in practical applications. Although a blockchain-based credit assessment model and multimodal image verification mechanism have been established, their essence remains a passive and static data processing mode. This results in an inability to accurately reconstruct the evolutionary trajectory of individual capabilities when scheduling high-value, highly interactive service resources. In terms of profiling logic, it only focuses on historical transaction frequency and text sentiment analysis, lacking classification mapping and quality quantification of atomic behaviors such as individual skills and knowledge output. Consequently, it cannot effectively distinguish the true service level under the interference of false evaluations, leading to insufficient accuracy in resource matching. Furthermore, its trust verification mainly relies on external image comparison, lacking an interactive proof mechanism based on identity challenges, resulting in weak robustness. Moreover, its fixed-weight scoring mechanism exhibits linearity shortcomings in service scenarios with multi-objective conflicts. There is significant room for improvement in the semantic matching depth of service requests, the degree of privacy control of personal data, and the self-evolutionary capability based on error feedback. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method and system for integrating and managing multi-dimensional campus service resources. By generating atomic behavior units from interaction records, constructing dynamic capability profiles and index credentials, and combining identity challenge verification and dynamic trust assessment to rearrange and update resources in a closed loop, this method can protect data privacy and achieve accurate matching and reliable management of campus service resources.

[0006] The above objectives can be achieved through the following approach:

[0007] A method for integrating and managing multi-dimensional campus service resources, characterized in that the method includes: acquiring user interaction records generated by multiple independent service modules within a campus integrated service platform, performing multimodal verification processing to generate verifiable atomic behavior units; storing the atomic behavior units in a personal data vault with encrypted isolation function, performing classification mapping and quality quantification using built-in strategy logic to obtain a personal dynamic capability profile, and performing desensitization processing and hash signing on the personal dynamic capability profile and associated data according to user-defined access control rules to generate an index credential associated with a specific service request; receiving the current user's service request, encapsulating the semantic features of the service request and the index credential, sending a retrieval request to the service resource node network, calculating resource matching weights and aggregating feedback information from each node to obtain a preliminary matched resource response set; The resource node features in the resource response set are extracted and verified against the current user's identity challenge capability proof. The verification results are quantified and accumulated to generate a temporary trust score. Successful cases matching the service request and the index credential in the historical interaction pattern library are retrieved. The temporary trust score and historical fit are weighted and calculated using a fusion algorithm to obtain a dynamic trust assessment result. The dynamic trust assessment result is compared with a preset gradient threshold. Based on the comparison result and the resource matching degree, the priority of the initially matched resource response set is rearranged and resource support schemes are configured to generate a final service sequence. The final service sequence is sent to the current user, and the rating and actual result after service execution are extracted, encrypted, and digested to generate a new atomic behavior unit, which is fed back to the personal data vault and the historical interaction pattern library is updated.

[0008] Optionally, generating verifiable atomic behavior units includes: identifying the service module type and interaction content corresponding to the interaction record, and determining the verification method; according to the verification method, extracting key intermediate result information during task execution and recording time consistency for skill demonstration type interactions, extracting confirmation information from both parties to the transaction for performance transaction type interactions, and extracting interaction link association information for knowledge contribution type interactions; and encapsulating the key intermediate result information, the confirmation information, and the interaction link association information to generate verifiable atomic behavior units.

[0009] Optionally, obtaining the personal dynamic capability profile includes: storing the atomic behavior units in a personal data vault; extracting behavior type tags using built-in strategy logic and mapping them to preset capability dimensions to obtain dimension classification results; obtaining behavior quality indicators in the atomic behavior units using the strategy logic; performing decay calculations using a time decay function and combining the weights with the dimension classification results to calculate the real-time weight values ​​of each capability dimension; and vectorizing and combining the real-time weight values ​​corresponding to each capability dimension to obtain the personal dynamic capability profile.

[0010] Optionally, generating an index credential associated with a specific service request includes: performing semantic parsing on the service request to determine the required capability dimensions, and filtering out matching dimension branches from the personal dynamic capability profile to obtain relevant capability data; retrieving the atomic behavior unit supporting the relevant capability data from the personal data vault according to the user-defined invocation strategy, and performing desensitization processing and hash signing to generate an index credential associated with the specific service request.

[0011] Optionally, obtaining the preliminary matching resource response set includes: encapsulating the semantic information of the service request with the capability information contained in the index credential to generate retrieval query information;

[0012] The system obtains the node connection relationships of the service resource node network and filters relevant service resource nodes based on the search query information to obtain a list of relevant service resource nodes. The system sends the search query information to the list of relevant service resource nodes, receives and structures the resource description information and demand matching instructions returned by each service resource node, and generates candidate resource response information. The system summarizes and processes the received candidate resource response information to obtain a preliminary set of matched resource responses.

[0013] Optionally, generating temporary trust points includes: selecting a target node from the initially matched resource response set, and receiving a verification challenge issued by the target node based on the index credential to obtain a challenge question; using the strategy logic to retrieve the corresponding atomic behavior unit in the personal data vault, generating a non-interactive proof for the challenge question to obtain a response proof; returning the response proof to the target node, and accumulating the points based on the verification pass rate fed back by the target node to generate temporary trust points.

[0014] Optionally, obtaining the dynamic trust assessment result includes: calculating the vector distance between the current service request, the index credential, and the historical successful patterns in the historical interaction pattern library to obtain a fit parameter; and using a weighted average algorithm to fuse the temporary trust score and the fit parameter to obtain the dynamic trust assessment result.

[0015] Optionally, generating the final service sequence includes: determining whether the dynamic trust assessment result exceeds a first preset threshold; if it does, extracting the corresponding matching item from the initially matched resource response set and marking it with a priority settlement identifier to obtain a high-reputation service sequence; if the dynamic trust assessment result is between the first preset threshold and the second preset threshold, then generating an intermediate sequence containing a progressive trust scheme by associating the corresponding item with a phased acceptance logic; and rearranging the matching items according to the resource matching weights of the high-reputation service sequence or the intermediate sequence to generate the final service sequence.

[0016] Optionally, the step of generating new atomic behavior units and feeding them back to the personal data vault, and updating the historical interaction pattern library, includes: pushing the final service sequence to the current user, obtaining the response quality rating and actual result after service execution, performing hash operations and encapsulating the response quality rating and actual result with a trusted timestamp to generate new atomic behavior units; storing the new atomic behavior units in the personal data vault, and extracting the quality score and service type label to generate a pattern correction factor; calculating the deviation between the actual result and the dynamic trust assessment result to obtain error feedback data; and using a gradient descent algorithm to correct the matching weights in the historical interaction pattern library based on the error feedback data and the pattern correction factor to obtain updated consensus evaluation parameters.

[0017] Based on the same inventive concept, this invention also provides an integration and management system for multi-dimensional campus service resources. The system includes: a multimodal interaction processing module, used to acquire user interaction records generated by multiple independent service modules within a campus integrated service platform, perform multimodal verification processing, and generate verifiable atomic behavior units; a profile and credential management module, used to store the atomic behavior units in a personal data vault with encrypted isolation function, perform classification mapping and quality quantification using built-in strategy logic to obtain a personal dynamic capability profile, and perform desensitization processing and hash signing on the personal dynamic capability profile and associated data according to user-defined access control rules to generate an index credential associated with a specific service request; and a resource retrieval coordination module, used to receive the current user's service request, encapsulate the semantic features of the service request and the index credential, send a retrieval request to the service resource node network, calculate resource matching weights, and aggregate feedback information from each node to obtain a preliminary matched resource response set. The challenge verification and sorting module is used to extract the resource node features in the resource response set, perform identity challenge-based capability proof verification with the current user, and quantify and accumulate the verification results to generate temporary trust points. The historical consensus evaluation module is used to retrieve successful cases in the historical interaction pattern library that match the service request and the index credential, and perform weighted calculation of the temporary trust points and historical fit through a fusion algorithm to obtain a dynamic trust evaluation result. The decision configuration optimization module is used to compare the dynamic trust evaluation result with a preset gradient threshold, and based on the comparison result and combined with the resource matching degree, prioritize and configure the resource support scheme for the initially matched resource response set to generate a final service sequence. The feedback and closed-loop update module is used to send the final service sequence to the current user, extract the rating and actual result after service execution, perform encrypted digest processing, generate new atomic behavior units to be fed back to the personal data vault, and update the historical interaction pattern library.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] This invention constructs a dynamic profile of an individual's capabilities based on verifiable atomic behavioral units, and combines it with a user-controlled index credential generation mechanism. Under the premise of fully protecting users' personal data sovereignty and privacy, it realizes the quantification and reliable transmission of personal capabilities, and resolves the contradiction between opaque and difficult-to-verify capability information and leakage of personal privacy in traditional service platforms.

[0020] This invention proposes a dynamic trust assessment model that integrates real-time challenge verification and historical consensus evaluation. This model can not only verify the real capabilities of service participants in specific scenarios in real time, but also refer to historical successful interaction patterns, thereby forming a three-dimensional and reliable trust judgment basis, improving the security and success rate of service interaction, and overcoming the shortcomings of traditional single evaluation systems that are easily manipulated and cannot reflect real-time capabilities.

[0021] This invention establishes a complete closed-loop feedback and self-optimization mechanism, which transforms the actual results of each service interaction into new data assets to continuously update individual capability profiles and the global historical interaction pattern library. This enables the system's resource matching accuracy and trust assessment model to continuously learn and evolve, realizing the intelligence and adaptability of the service management system and solving the problem of existing system models being rigid and unable to be continuously improved from practice.

[0022] This invention links dynamic trust assessment results with differentiated resource support schemes, prioritizes and configures schemes for initially matched resource sets, and provides customized support strategies for interactions with different trust levels. For example, it provides priority settlement channels for high-reputation interactions and phased acceptance guarantees for medium-reputation interactions, thereby optimizing resource allocation efficiency and risk management level and improving the operational efficiency and robustness of the entire campus service ecosystem.

[0023] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart illustrating a method for integrating and managing multi-dimensional campus service resources according to an embodiment of the present invention.

[0026] Figure 2 This is a schematic diagram illustrating the evolution of the personal dynamic capability profile according to an embodiment of the present invention.

[0027] Figure 3 This is a schematic diagram of the decision space for dual trust assessment in an embodiment of the present invention.

[0028] Figure 4 This is a schematic diagram of the structure of a campus multi-dimensional service resource integration and management system according to an embodiment of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] Reference Figure 1 One embodiment of the present invention proposes a method for integrating and managing multi-dimensional campus service resources. It adopts the means of generating atomic behavior units from interaction records, constructing dynamic capability profiles and index credentials, and combining identity challenge verification and dynamic trust assessment to rearrange and update resources in a closed loop. This method can protect data privacy and achieve accurate matching and reliable management of campus service resources.

[0031] The method described in this embodiment specifically includes:

[0032] S1. Obtain the interaction records generated by the user in multiple independent service modules within the campus integrated service platform, perform multimodal verification processing, and generate verifiable atomic behavior units;

[0033] Optionally, the generation of verifiable atomic behavior units includes:

[0034] Identify the service module type and interaction content corresponding to the interaction record, and determine the verification method;

[0035] According to the verification method, key intermediate result information during the task execution process is extracted from skill demonstration type interactions and time consistency is recorded; confirmation information of both parties to the transaction is extracted from performance transaction type interactions; and interaction link association information is extracted from knowledge contribution type interactions.

[0036] The key intermediate result information, the confirmation information, and the interaction link association information are encapsulated to generate verifiable atomic behavior units.

[0037] Specifically, upon receiving an interaction record, the system immediately identifies the type of service segment to which it belongs. This identification is based on the service segment identifier contained in the metadata of the interaction record. By maintaining a service segment mapping table, the identifier, such as "SKL-LAB-01" or "TRD-TUTOR-05", is mapped to three preset interaction categories: skill demonstration interaction, contract fulfillment transaction interaction, and knowledge contribution interaction. Based on this classification result, the system determines the verification method to be used subsequently.

[0038] Different information extraction logic is executed for different verification methods. For skill demonstration interactions, such as submitting an online programming task, not only the final result is recorded, but key intermediate result information during task execution is extracted, such as the hash value of a specific version of the code repository, the timestamp of the successful compilation of key functions, and the log summary when unit test coverage reaches 80%. This information is appended with high-precision timestamps for time consistency recording to prove that the task was completed step by step and not plagiarized. For transaction-based interactions, such as completing a paid campus errand service, the confirmation information of both parties on the platform's built-in confirmation interface is extracted. This information is usually a digital signature of the service completion certificate using the private keys of both users, ensuring mutual recognition of the performance. For knowledge contribution interactions, such as answering technical questions on a campus forum, the interaction link information is extracted, including the number of times the answer is cited, the number of likes, and the identifier that the problem has been accepted, to quantify the actual value and influence of the knowledge contribution.

[0039] The extracted key intermediate result information, confirmation information, or interaction link association information are encapsulated to generate atomic behavioral units with a unified data structure. This encapsulation process generates a data object containing multiple fields, the integrity of which is guaranteed by a single hash value. Its structure can be represented by the formula... ,in The final hash signature representing this atomic unit serves as its unique and immutable identifier; The default hash algorithm, such as SHA-256; It is the unique unit ID generated; It is a user identification identifier; It is the original timestamp of the interaction; It is a specific interaction type code, such as 01 representing a skill demonstration; It refers to the evidence data itself, which has been structured, such as a JSON string containing code submission hashes and timestamp sequences.

[0040] For example, when User B submits their final project, "ReadShare App Interface Design," in the "Digital Media Design" section of the campus platform, the multimodal interaction processing module is activated. The system recognizes that the metadata of this record contains the identifier "SKL-LAB-01" and categorizes it as "Skill Demonstration Interaction" through an internal mapping table. The system then executes differentiated extraction logic, recording not only the final score but also extracting key intermediate results: the specific version submission hash value of the Figma project file (Hash_v1.0), the export timestamp of key design drafts, and the system log summary of the design specification document. The system encapsulates this information to generate atomic behavior units. Known For “U20251201”, For "User-B", interaction timestamp The type code is "1733020800". It is "01". Evidence data The format is JSON: {"figma_hash":"a7b8c9","log_summary":"pass_standard"}. The system uses the SHA-256 algorithm to process the concatenated string, obtaining a unique identifier e3b0c442... By performing differentiated feature extraction and hash encapsulation on skill demonstrations, contract fulfillment transactions, and knowledge contributions, the unstructured process records are transformed into immutable atomic behavioral units, thereby ensuring the authenticity and integrity of the underlying evidence chain at the technical level and providing data support for subsequent evaluation stages.

[0041] S2. Store the atomic behavior unit in a personal data vault with encryption and isolation functions, perform classification mapping and quality quantification using built-in policy logic to obtain a personal dynamic capability profile, and perform desensitization processing and hash signing on the personal dynamic capability profile and related data according to the access control rules set by the user to generate an index credential associated with a specific service request.

[0042] Optionally, obtaining the personal dynamic capability profile includes:

[0043] The atomic behavior units are stored in a personal data vault, and behavior type tags are extracted using built-in strategy logic and mapped to preset capability dimensions to obtain dimension classification results.

[0044] The behavior quality index in the atomic behavior unit is obtained by using the strategy logic, the decay is calculated by using the time decay function, and the weight is accumulated by combining the dimensional classification results to calculate the real-time weight value of each capability dimension.

[0045] The real-time weight values ​​corresponding to each capability dimension are vectorized and combined to obtain a dynamic personal capability profile.

[0046] Specifically, newly added atomic behavior units are read from the personal data vault, and the built-in strategy logic is immediately used to parse them. This strategy logic is a predefined set of rules that, based on the behavior type label in the atomic behavior unit, such as "SKL-LAB-01", queries the internal capability dimension mapping table and maps it to one or more preset capability dimensions, such as "programming practice ability" or "data analysis ability". The output of this step is the dimension classification result, which labels each behavior unit with capability attributes.

[0047] Quantifiable behavioral quality indicators are extracted from atomic behavioral units using strategic logic. These indicators vary depending on the behavior type and may include a normalized value of a project's final score (ranging from 0 to 1), the converted score of a 5-star rating earned for a service, or the number of times a knowledge contribution is adopted. To reflect the real-time nature of the capability, a time decay function is used to calculate the decay of these quality indicators. The calculation formula is as follows:

[0048] ,

[0049] In this formula, This represents the effective weight value of the atomic behavior unit at the current moment. These are the raw behavioral quality indicators extracted and normalized from behavioral units. It is the base of the natural logarithm. This is a preset time decay coefficient, typically ranging from 0.001 to 0.01, used to adjust the rate at which ability value is forgotten over time. It is the current timestamp. This is the timestamp when the atomic behavior unit was generated. After calculation, combined with the obtained dimensional classification results, all valid weight values ​​belonging to the same capability dimension are... The values ​​are accumulated to calculate the real-time weights for each capability dimension. The total weight for that dimension is then calculated. The summation iterates through all elements categorized into dimensions. atomic behavior unit .like Figure 2 As shown, this reflects the non-linear decrease in the weights of different capability dimensions as the time difference increases.

[0050] The real-time weight values ​​corresponding to all ability dimensions are vectorized and combined to generate the final dynamic personal ability profile. For example, if N ability dimensions are preset, such as academic research, teamwork, and innovative thinking, then the dynamic personal ability profile is an N-dimensional vector. , where each component of the vector Each corresponds to a real-time weight value for a specific capability dimension.

[0051] For example, the profile and credential management module retrieves atomic behavior units from user B's vault. The strategy logic identifies the "SKL-LAB-01" tag and queries the mapping table to categorize it under the "UI / UX design capability" dimension (j=1). The system then extracts normalized raw behavior quality indicators from the unit. (Converted from a course grade of A+). To reflect timeliness, the system applies a time decay function. Preset attenuation coefficient The time difference is 180 days. If user B had a competition record with a quality metric of 0.85 30 days ago, then its weight... The real-time total weight of this dimension. This method utilizes time decay functions and multidimensional vectorization modeling to scientifically simulate the evolution of human skills over time, enabling personal competency profiles to reflect a user's current professional level in real time and accurately, thus avoiding the credit lag or inflated competency issues caused by static evaluation models.

[0052] Optionally, generating the index credential associated with a specific service request includes:

[0053] The service request is semantically parsed to determine the required capability dimensions, and the matching dimension branches are selected from the personal dynamic capability profile to obtain relevant capability data.

[0054] According to the user-defined calling strategy, the atomic behavior unit supporting the relevant capability data is retrieved from the personal data vault, and then de-identified and hash-signed to generate an index credential associated with a specific service request.

[0055] Specifically, the semantic parsing engine is activated to process the text content of the service request. This engine is typically based on a pre-trained natural language processing model, such as a BERT model fine-tuned for a campus scenario, to identify core capability keywords in the request, such as "data analysis," "poster design," or "event organization." These keywords are then mapped to preset capability dimension labels to determine the set of capability dimensions required for this request. Next, using these labels, matching dimension branches are selected from the current user's dynamic capability profile vector, and the corresponding real-time weight values ​​are extracted to form a set of relevant capability data containing specific capability items and their quantified scores.

[0056] Subsequent operations are performed according to the retrieval policy pre-defined by the user in their personal data vault. This retrieval policy is a set of access control rules, such as "for research collaboration requests initiated by certified teachers, access to all relevant project completion certificates is allowed" or "for skills inquiries from anonymous users, only competency scores are provided, without displaying any original evidence." Based on this policy, an authorized data retrieval request is initiated to the personal data vault to obtain the original atomic behavioral units supporting the aforementioned competency data. Upon data return, a rigorous anonymization process is immediately performed, removing or replacing personally identifiable information in the atomic behavioral units. For example, real names are replaced with hash values ​​of user IDs, specific project or course names are hidden, and only the category and summary of achievements are retained. After anonymization, this data packet containing the relevant competency data summary and the list of anonymized atomic behavioral unit evidence is hash-signed as a whole to generate the final index credential. The structure of this credential can be represented by the following formula: .in, It is the generated index credential; This represents a digital signature made using the user's private key, ensuring the non-repudiation of the credentials; It uses a hash algorithm, such as SHA-256; It is a unique identifier for the current service request, ensuring the contextual uniqueness of the credentials; It is the timestamp when the voucher was generated; It is a subset of relevant capability data extracted from an individual's dynamic capability profile; and It is a set of hash values ​​of atomic behavioral units that support these capability data after they have been anonymized.

[0057] For example, in response to a request from User A to "design the UI / UX for a social app, requiring familiarity with Figma, complete mobile application design experience, and a 3-week project timeline," the profile and credential management module activates the semantic parsing engine. This engine, based on a fine-tuned BERT model, identifies the core competency keywords "UI / UX design" and "Figma," and extracts the corresponding real-time weight value of 1.115 from User B's profile. User B has preset a calling strategy in the vault: "For ordinary service requests initiated by certified students, only anonymized evidence will be displayed." The system performs anonymization processing, removing the specific scores and full course names from the original atomic behavior units, and generates index credentials. .in For the service request ID "REQ-2025-001", The timestamp for when the voucher was generated is 1735441200. For a subset of relevant capability data , This is the set of de-identified evidence hashes. This method combines semantic parsing and access control strategies to achieve on-demand authorization and privacy de-identification for specific service requests. While ensuring users possess the evidentiary power to complete tasks, it reduces the exposure of sensitive personal information, balancing data flow and individual privacy rights.

[0058] S3. Receive the service request from the current user, encapsulate the semantic features of the service request with the index credential, send a retrieval request to the service resource node network, calculate the resource matching weight and aggregate the feedback information of each node to obtain a preliminary matching resource response set.

[0059] Optionally, the set of resource responses that have been initially matched includes:

[0060] The semantic information of the service request is encapsulated with the capability information contained in the index credential to generate retrieval query information;

[0061] Obtain the node connection relationship of the service resource node network, and filter relevant service resource nodes based on the search query information to obtain a list of relevant service resource nodes;

[0062] Send the retrieval query information to the list of relevant service resource nodes, receive and structure the resource description information and demand matching instructions returned by each service resource node, and generate candidate resource response information;

[0063] The received candidate resource response information is aggregated and processed to obtain a preliminary set of matched resource responses.

[0064] Specifically, the semantic information of the service request—that is, the core demand vector extracted through natural language processing—is encapsulated in a unified data format along with the capability information contained in the index credentials—that is, the user's capability dimension summary and the corresponding evidence hash set—to generate a structured retrieval query. The structure of this information can be represented as follows: ,in This represents the information requested in the search query; It is a unique identifier for the service request; It is the semantic feature vector of the service request; It is the overall hash value of the index credential, used to ensure its integrity; This is the valid timestamp of the query information. Set a short lifespan, such as 180 seconds, to avoid the proliferation of expired requests on the network.

[0065] Next, the resource retrieval coordination module accesses the topology map of the service resource node network it maintains. This map records the addresses, service category labels, and connections between all online nodes. The module then uses the semantic feature vectors from the retrieval query information to... The system performs initial matching and filtering with the service category tags of the nodes. For example, a request vector related to "software development" will prioritize filtering service resource nodes tagged with "information technology" and "programming outsourcing" to obtain a list of related service resource nodes with a controllable scale.

[0066] Retrieve query information The query is sent in parallel to each relevant service resource node in this list. Upon receiving the query, each node parses it based on its own currently available resources. and The system generates a resource description and a demand matching specification, and returns them. The resource description is a standardized statement of the node's capabilities, while the demand matching specification is a response to this specific request, which may include a preliminary self-assessment score of matching degree, ranging from 0 to 1. The resource retrieval coordination module asynchronously receives feedback from all nodes, structures and organizes it, and uniformly parses the returned heterogeneous data into standardized candidate resource response information containing fields such as node ID, resource description, and matching specification.

[0067] All in The candidate resource response information received within the validity period is aggregated and processed to form a set. This set is the preliminary matching resource response set, which is an unsorted and unverified raw list containing information on all service resource nodes that have shown preliminary interest and capability matching in response to the current service request.

[0068] For example, the resource retrieval coordination module uniformly encapsulates user A's request semantic information regarding "social app design" with the index credentials of users B and C. The module then generates retrieval query information. .in The value is “REQ-2025-001”. It is the extracted core requirement vector (This represents a high-dimensional representation of proficiency in UI design and Figma.) The overall hash value of user B's index credentials; Set to the current timestamp plus 180 seconds. The module accesses the service resource node topology graph and identifies that the nodes where users B and C reside both maintain the service category label "UI / UX Design". The system will... Send in parallel to the relevant nodes. User B node parses. Then, based on its current load and capabilities, it generates a demand matching description with a self-scoring score of 0.92. The resource retrieval and coordination module then... This method asynchronously receives and parses heterogeneous data within a given period, standardizing the information of users B and C into a response list containing node IDs, resource descriptions, and matching instructions, thus forming a preliminary set of matched resource responses. By semanticizing requests and utilizing node topology graphs for targeted broadcasting, this method reduces redundant communication overhead in the distributed network. While improving resource retrieval response speed, it ensures that the initially selected resource set has high business relevance, laying the foundation for subsequent deep validation.

[0069] S4. Extract the resource node features from the resource response set, perform capability verification based on identity challenge with the current user, and quantify and accumulate the verification results to generate temporary trust points.

[0070] Optionally, generating temporary trust points includes:

[0071] Select a target node from the initially matched resource response set, and receive a verification challenge issued by the target node based on the index credential to obtain the challenge question;

[0072] Using the strategy logic, the corresponding atomic behavior unit in the personal data vault is retrieved to generate a non-interactive proof for the challenge question, thus obtaining a response proof;

[0073] The response proof is returned to the target node, and the verification pass rate fed back by the target node is obtained, and the points are accumulated to generate a temporary trust score.

[0074] Specifically, one or more target nodes are selected sequentially or in parallel from the set, and the system actively receives verification challenges issued by these target nodes. This verification challenge is a verification request, or challenge question, generated by the target node after parsing the user-submitted index credentials, targeting a specific capability declared within them. For example, if the index credentials contain a hash of an atomic behavior unit completed in a project, the challenge question might require the user to provide the source code digest hash of a specific module within that project to verify their actual participation.

[0075] Upon receiving a challenge, built-in policy logic is triggered, securely requesting access to the atomic unit of action directly related to the challenge from the personal data vault. This access is strictly authorized, limited to providing only the minimum information required to generate the proof. This information is then used to generate a non-interactive proof as the response proof. A non-interactive proof is a cryptographic credential that can be verified without continuous communication between the parties; for example, a zero-knowledge proof can prove that the user indeed possesses the data and meets the challenge requirements without revealing the specific content of the atomic unit of action. In a simpler implementation, it could be the hash value of the requested data, accompanied by the user's digital signature.

[0076] After generating the response proof, it is returned to the target node that issued the challenge. The target node verifies the proof locally and feeds back the verification result, typically a pass rate between 0 and 1, to the challenge verification and sorting module. This module sums the verification results received from each target node to generate a final provisional trust score. The calculation formula can be expressed as follows: ,in It is the generated temporary trust integral; summation symbol This indicates that all target nodes that have initiated and responded to the challenge are traversed and accumulated. It is the verification pass rate fed back from the i-th target node, which is calculated by the target node based on the correctness and completeness of the response proof; It is the weight coefficient for the i-th challenge. This coefficient is set according to the relevance between the verified capability and the core requirements of the current service request. The higher the relevance, the weight is usually set between 0.8 and 1.0, and vice versa.

[0077] For example, after obtaining the initial matching set of resource responses, the challenge verification and sorting module initiates an evidence-based capability verification procedure for users B and C. The module generates a specific verification challenge for the "Yuexiang App Interface Design" project mentioned in user B's credentials, requiring them to provide the SHA-256 hash digest of a specific layer structure of the "Personal Homepage" canvas in the design draft. User B's system, authorized through the personal data vault, retrieves the corresponding atomic action unit, generates a non-interactive zero-knowledge proof as the response proof, and returns it. The target node verifies the proof, confirming its consistency with the original evidence chain, and provides feedback on the verification pass rate. UI design skills are known to be the core requirement of this request, and the system sets a weighting factor for this challenge. According to the formula The temporary trust score of user B is calculated. In contrast, User C, unable to provide corresponding wireframe evidence, had a lower pass rate in the verification process. With a value of only 0.4, its temporary trust score is calculated to be 0.4. This method introduces a challenge-response mechanism based on zero-knowledge proofs, transforming the user's passive ability statement into an active cryptographic verification, thereby generating an instant trust score without disclosing the original data. This enhances the system's ability to identify false claims and constructs a high-strength instant trust barrier.

[0078] S5. Retrieve successful cases in the historical interaction pattern library that match the service request and the index credential, and use a fusion algorithm to weight and calculate the temporary trust score and the historical fit to obtain a dynamic trust evaluation result;

[0079] Optionally, obtaining the dynamic trust assessment result includes:

[0080] Calculate the vector distance between the current service request, the index credential, and the historical success pattern in the historical interaction pattern library to obtain the fit parameter;

[0081] The temporary trust score and the fit parameter are fused using a weighted average algorithm to obtain the dynamic trust assessment result.

[0082] Specifically, upon receiving the temporary trust points, a search of the historical interaction pattern library is immediately activated. The historical interaction pattern library stores a large number of successful service interaction cases. Each case is abstracted into a feature vector, which is composed of the semantic features of the service request and a summary of the service provider's index credentials capabilities at that time.

[0083] The semantic feature vector of the current service request is concatenated with the capability summary vector of the index credential to form a query vector. Then, the vector distance between this query vector and all historical successful pattern vectors is calculated in the database. This calculation typically uses a cosine similarity algorithm to assess the pattern similarity between the current context and historical successful cases, rather than the absolute numerical difference. A series of similarity scores are obtained after calculation, and the highest score is selected as the fit parameter. The formula for this parameter is as follows: ,in This represents the final selected fit parameter, and its value ranges from 0 to 1. Represents the cosine similarity between vector A and vector B; It is a query vector consisting of the current service request and index credentials; It is the first in the historical interaction mode library Feature vectors of successful cases.

[0084] A weighted average algorithm is used to fuse the temporary trust score and the fit parameter to obtain the final dynamic trust assessment result. This fusion process aims to balance immediate performance and long-term reputation, and its calculation formula is as follows:

[0085] ,

[0086] in, It is the output of the dynamic trust assessment result, which is a normalized trust score; It is a temporary trust score calculated by the challenge verification and sorting module; It is the parameter with the highest matching degree obtained by retrieving and matching from the historical interaction pattern library; These are preset fusion weight coefficients, typically ranging from 0.4 to 0.7. For example... Figure 3 As shown in the figure, the diagram uses a three-dimensional surface to illustrate how the real-time ability challenge score and the fit of historical interactions jointly determine the final dynamic trust assessment result.

[0087] For example, upon receiving a temporary trust score, a search of the historical interaction pattern library is initiated to enhance the robustness of the evaluation. The system concatenates the semantic vector of the current social app design request with the index credential summary of user B to generate a feature vector, and calculates the cosine similarity between this feature vector and the feature vectors of historical success cases stored in the library. The calculation results show that user B has previously completed a highly similar social application design task, and the highest similarity score is selected as the fit parameter. The system then employs a weighted average algorithm to fuse real-time performance and historical reputation, setting the fusion weights. The dynamic trust assessment result for user B was calculated. User C, lacking relevant historical success records, has a lower fit parameter. The initial score was only 0.35, while the fused evaluation result was 0.380. This method utilizes cosine similarity to deeply integrate current interaction performance with historical success patterns. It can capture users' immediate highlights of ability and also refer to their long-term reputation accumulation, thereby deriving a robust, multi-dimensional, and fault-tolerant dynamic trust score, improving the scientific rigor and comprehensiveness of trust assessment.

[0088] S6. Compare the dynamic trust assessment result with the preset gradient threshold, and based on the comparison result and in combination with the resource matching degree, prioritize and configure the resource support scheme for the initially matched resource response set to generate the final service sequence.

[0089] Optionally, generating the final service sequence includes:

[0090] Determine whether the dynamic trust assessment result exceeds a first preset threshold. If it does, extract the corresponding matching item from the preliminary matching resource response set and mark it with a priority settlement identifier to obtain a high-reputation service sequence.

[0091] If the dynamic trust assessment result is between the first preset threshold and the second preset threshold, then the corresponding item is associated with the phased acceptance logic, generating an intermediate sequence containing a progressive trust scheme.

[0092] The final service sequence is generated by rearranging the matching items in the high-reputation service sequence or the intermediate sequence according to their resource matching weights.

[0093] Specifically, upon receiving the dynamic trust assessment results, the following steps are executed: [The dynamic trust assessment results are then processed / executed]. It is compared with two preset gradient thresholds, namely the first preset threshold. Second preset threshold ,in It is usually set to a high value, such as 0.85, to represent high confidence. Set it to a moderate value, such as 0.60, to represent basic confidence.

[0094] Execute conditional branch logic based on the comparison results. If the dynamic trust assessment result... Exceeding the first preset threshold The system extracts corresponding matches from the initially matched resource response set and adds a priority settlement flag to their metadata. This flag is a boolean value that triggers fast-track processes in subsequent transactions, such as automatic authorization of prepayments or simplified settlement approval, thus forming a high-reputation service sequence. If the dynamic trust assessment result... At the first preset threshold With the second preset threshold Between these points, a pre-defined phased acceptance logic is associated with the corresponding matching item. This logic is a standardized service delivery framework, such as a smart contract template with three milestone nodes, requiring the service provider to deliver results in stages, and unlocking the corresponding portion of the reward only after obtaining user confirmation at each stage, thus generating an intermediate sequence containing a progressive trust scheme. Below If the match is deemed high-risk, it will typically be removed directly from the candidate list.

[0095] Regardless of whether a high-reputation service sequence or an intermediate sequence is generated, the final priority is reordered based on the resource matching weights of each matching item in the sequence. The resource matching weight is a normalized value, calculated by the service node itself during the resource retrieval phase, representing the semantic similarity between its service content and the user's request. All items in the sequence are sorted in descending order according to this weight value to generate the final service sequence. The structure of this sequence can be represented as an ordered list. ,in It is the final service sequence; Representatives Perform a descending sort operation; It is the first One resource response item; It is its corresponding resource matching weight; This refers to the service support scheme configured based on the threshold judgment result, namely the reference to the priority settlement identifier or the phased acceptance logic.

[0096] For example, a differentiated configuration strategy is implemented for candidate nodes based on the dynamic trust assessment results. The system presets a first gradient threshold. With the second gradient threshold The comparison results show that user B's... If the first threshold is exceeded, the system adds a "priority settlement identifier" to the matching item's metadata and configures a fast transaction channel that allows for partial prepayment, forming a high-reputation sequence member. User C's... If the match is below the second threshold, the system classifies it as a high-risk match and removes it from the recommendation list. Finally, the module uses the resource matching weight of user B recorded in the resource response set. The service sequence is reordered in descending order to generate a final service sequence containing user B and their optimized service plan, which is then pushed to user A. This method uses gradient threshold judgment to transform trust scores into differentiated service configuration logic, opening a green settlement channel for high-credit users and providing phased risk management solutions for ordinary users, thereby maximizing the efficiency of campus resource allocation and minimizing the risk of interactive default.

[0097] S7. Send the final service sequence to the current user, extract the rating and actual result after service execution, perform encrypted digest processing, generate a new atomic behavior unit and feed it back to the personal data vault, and update the historical interaction pattern library.

[0098] Optionally, the step of generating new atomic behavior units and feeding them back to the personal data vault, and updating the historical interaction pattern library, includes:

[0099] The final service sequence is pushed to the current user, the response quality rating and actual result after the service execution are obtained, and the response quality rating and actual result are hashed and encapsulated with a trusted timestamp to generate a new atomic behavior unit.

[0100] The new atomic behavior unit is stored in the personal data vault, and the quality score and service type label are extracted to generate a pattern correction factor;

[0101] Calculate the deviation between the actual result and the dynamic trust assessment result to obtain error feedback data;

[0102] The gradient descent algorithm is used to correct the matching weights in the historical interaction pattern library based on the error feedback data and the pattern correction factor, so as to obtain the updated consensus evaluation parameters.

[0103] Specifically, after pushing the final service sequence and receiving the service completion signal, the system proactively requests or automatically extracts the response quality rating and actual results after service execution from the user. The response quality rating is typically a user-submitted 1-5 star rating, which is normalized to a value between 0 and 1; the actual results may be a Boolean flag indicating project acceptance or a specific performance indicator of a task. These two data points are hashed and encapsulated with a trusted timestamp to generate new atomic behavioral units, the structure of which can be represented as follows: ,in It is the hash identifier of the new atomic behavior unit, where H is the hash algorithm. This is the unique ID for this service interaction. It is a normalized quality rating. It is structured, actual result data. It is a trusted timestamp after the service is completed.

[0104] Once generated, this new atomic behavior unit is immediately sent and stored in the user's personal data vault. The quality score contained within it will directly participate in the next iteration of the user's dynamic personal capability profile calculation, enabling real-time updates to personal capabilities. Simultaneously, the quality score and service type tag are extracted from this unit and combined to form a pattern correction factor. This factor is used to adjust the weight of the feedback in the model update; for example, the weight of successful feedback for core professional services will be higher than that for general services. Next, the deviation between the actual result and the dynamic trust assessment result generated before this interaction is calculated to obtain the error feedback data. ,in It is the previously calculated dynamic trust assessment score. This is the actual result after normalization.

[0105] Using the gradient descent algorithm, the weight parameters that best match the current interaction pattern in the historical interaction pattern library are adjusted based on error feedback data and a pattern correction factor. This update process follows the formula:

[0106] ,

[0107] in and These are the values ​​of the target historical pattern weights before and after the update. It is a preset learning rate, typically ranging from 0.01 to 0.1. This is the input vector representing the characteristics of this interaction. Through this iterative process, the consensus evaluation parameters are continuously fine-tuned, so that the predictive ability of the historical interaction pattern library gradually improves with the accumulation of actual interaction data.

[0108] For example, after User A and User B complete the UI / UX design project for the social app, the feedback and closed-loop update module is activated to enable the continuous evolution of system capabilities. The system receives a 5-star rating submitted by User A and normalizes the response quality rating. and the actual results of project acceptance. This feedback data, combined with a trusted timestamp, is encapsulated into a new atomic behavior unit and stored in user B's personal data vault to enhance their profile weight. Simultaneously, the system calculates the prediction deviation value. The system uses the gradient descent algorithm to correct the weight parameters in the historical interaction pattern library that best match the current "social app design" pattern. The weights for the current pattern are then set. Learning rate Pattern correction factor The magnitude or simplified influencing factor of the feature vector of this interaction . This method feeds back service effectiveness to individual profiles and a global pattern library, forming an adaptive learning mechanism based on gradient descent. This ensures that the system can continuously correct evaluation parameters based on historical errors, driving the entire campus service ecosystem to continuously evolve towards higher accuracy and greater reliability.

[0109] Based on the same inventive concept, such as Figure 4As shown, the present invention also provides an integration and management system for multi-dimensional campus service resources, including:

[0110] The multimodal interaction processing module is used to acquire the interaction records generated by users in multiple independent service modules within the campus integrated service platform, perform multimodal verification processing, and generate verifiable atomic behavior units.

[0111] The profile and credential management module is used to store the atomic behavior units in a personal data vault with encrypted isolation function, perform classification mapping and quality quantification using built-in policy logic to obtain a personal dynamic capability profile, and perform desensitization processing and hash signing on the personal dynamic capability profile and related data according to the access control rules set by the user to generate an index credential associated with a specific service request.

[0112] The resource retrieval coordination module is used to receive the service request from the current user, encapsulate the semantic features of the service request with the index credentials, send the retrieval request to the service resource node network, calculate the resource matching weight and aggregate the feedback information of each node to obtain a preliminary set of matched resource responses.

[0113] The challenge verification and sorting module is used to extract the resource node features in the resource response set, perform identity challenge-based capability verification with the current user, and quantify and accumulate the verification results to generate temporary trust points.

[0114] The historical consensus assessment module is used to retrieve successful cases in the historical interaction pattern library that match the service request and the index credential, and to calculate the dynamic trust assessment result by weighting the temporary trust score and the historical fit through a fusion algorithm.

[0115] The decision configuration optimization module is used to compare the dynamic trust assessment results with preset gradient thresholds, and based on the comparison results and combined with the resource matching degree, to prioritize and configure resource support schemes for the initially matched resource response set, and generate the final service sequence.

[0116] The feedback and closed-loop update module is used to send the final service sequence to the current user, extract the rating and actual result after service execution, perform encrypted digest processing, generate new atomic behavior units to be fed back to the personal data vault, and update the historical interaction pattern library.

[0117] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0118] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A method for integrating and managing multi-dimensional campus service resources, characterized in that, The method includes: The system acquires user interaction records from multiple independent service modules within the campus integrated service platform, performs multimodal verification processing, and generates verifiable atomic behavior units. The atomic behavior units are stored in a personal data vault with encrypted isolation function. The built-in policy logic is used to classify, map and quantify the quality to obtain a personal dynamic capability profile. The personal dynamic capability profile and related data are desensitized and hashed according to the access control rules set by the user to generate an index credential associated with a specific service request. Receive the service request from the current user, encapsulate the semantic features of the service request with the index credential, send a retrieval request to the service resource node network, calculate the resource matching weight and aggregate the feedback information of each node to obtain a preliminary matching resource response set. Extract the resource node features from the resource response set, perform an identity challenge-based capability verification with the current user, and quantify and accumulate the verification results to generate a temporary trust score. Successful cases matching the service request and the index credential are retrieved from the historical interaction pattern library. The temporary trust score and the historical fit are weighted and calculated using a fusion algorithm to obtain a dynamic trust evaluation result. The dynamic trust assessment results are compared with preset gradient thresholds. Based on the comparison results and combined with the resource matching degree, the priority of the initially matched resource response set is rearranged and the resource support scheme is configured to generate the final service sequence. The final service sequence is sent to the current user, and the rating and actual result after service execution are extracted, encrypted and digested to generate a new atomic behavior unit, which is fed back to the personal data vault and the historical interaction pattern library is updated.

2. The method for integrating and managing multi-dimensional campus service resources according to claim 1, characterized in that, The generation of verifiable atomic behavior units includes: Identify the service module type and interaction content corresponding to the interaction record, and determine the verification method; According to the verification method, key intermediate result information during the task execution process is extracted from skill demonstration type interactions and time consistency is recorded; confirmation information of both parties to the transaction is extracted from performance transaction type interactions; and interaction link association information is extracted from knowledge contribution type interactions. The key intermediate result information, the confirmation information, and the interaction link association information are encapsulated to generate verifiable atomic behavior units.

3. The method for integrating and managing multi-dimensional campus service resources according to claim 1, characterized in that, The obtained personal dynamic capability profile includes: The atomic behavior units are stored in a personal data vault, and behavior type tags are extracted using built-in strategy logic and mapped to preset capability dimensions to obtain dimension classification results. The behavior quality index in the atomic behavior unit is obtained by using the strategy logic, the decay is calculated by using the time decay function, and the weight is accumulated by combining the dimensional classification results to calculate the real-time weight value of each capability dimension. The real-time weight values ​​corresponding to each capability dimension are vectorized and combined to obtain a dynamic personal capability profile.

4. The method for integrating and managing multi-dimensional campus service resources according to claim 1, characterized in that, The generation of index credentials associated with a specific service request includes: The service request is semantically parsed to determine the required capability dimensions, and the matching dimension branches are selected from the personal dynamic capability profile to obtain relevant capability data. According to the user-defined calling strategy, the atomic behavior unit supporting the relevant capability data is retrieved from the personal data vault, and then de-identified and hash-signed to generate an index credential associated with a specific service request.

5. The method for integrating and managing multi-dimensional campus service resources according to claim 1, characterized in that, The set of resource responses that have been initially matched includes: The semantic information of the service request is encapsulated with the capability information contained in the index credential to generate retrieval query information; Obtain the node connection relationship of the service resource node network, and filter relevant service resource nodes based on the search query information to obtain a list of relevant service resource nodes; Send the retrieval query information to the list of relevant service resource nodes, receive and structure the resource description information and demand matching instructions returned by each service resource node, and generate candidate resource response information; The received candidate resource response information is aggregated and processed to obtain a preliminary set of matched resource responses.

6. The method for integrating and managing multi-dimensional campus service resources according to claim 1, characterized in that, The generation of temporary trust points includes: Select a target node from the initially matched resource response set, and receive a verification challenge issued by the target node based on the index credential to obtain the challenge question; Using the strategy logic, the corresponding atomic behavior unit in the personal data vault is retrieved to generate a non-interactive proof for the challenge question, thus obtaining a response proof; The response proof is returned to the target node, and the verification pass rate fed back by the target node is obtained, and the points are accumulated to generate a temporary trust score.

7. The method for integrating and managing multi-dimensional campus service resources according to claim 1, characterized in that, The obtained dynamic trust assessment results include: Calculate the vector distance between the current service request, the index credential, and the historical success pattern in the historical interaction pattern library to obtain the fit parameter; The temporary trust score and the fit parameter are fused using a weighted average algorithm to obtain the dynamic trust assessment result.

8. The method for integrating and managing multi-dimensional campus service resources according to claim 1, characterized in that, The generation of the final service sequence includes: Determine whether the dynamic trust assessment result exceeds a first preset threshold. If it does, extract the corresponding matching item from the preliminary matching resource response set and mark it with a priority settlement identifier to obtain a high-reputation service sequence. If the dynamic trust assessment result is between the first preset threshold and the second preset threshold, then the corresponding item is associated with the phased acceptance logic, generating an intermediate sequence containing a progressive trust scheme. The final service sequence is generated by rearranging the matching items in the high-reputation service sequence or the intermediate sequence according to their resource matching weights.

9. The method for integrating and managing multi-dimensional campus service resources according to claim 1, characterized in that, The process of generating new atomic behavior units and feeding them back to the personal data vault, and updating the historical interaction pattern library, includes: The final service sequence is pushed to the current user, the response quality rating and actual result after the service execution are obtained, and the response quality rating and actual result are hashed and encapsulated with a trusted timestamp to generate a new atomic behavior unit. The new atomic behavior unit is stored in the personal data vault, and the quality score and service type label are extracted to generate a pattern correction factor; Calculate the deviation between the actual result and the dynamic trust assessment result to obtain error feedback data; The gradient descent algorithm is used to correct the matching weights in the historical interaction pattern library based on the error feedback data and the pattern correction factor, so as to obtain the updated consensus evaluation parameters.

10. A system for integrating and managing multi-dimensional campus service resources, characterized in that, The system includes: The multimodal interaction processing module is used to acquire the interaction records generated by users in multiple independent service modules within the campus integrated service platform, perform multimodal verification processing, and generate verifiable atomic behavior units. The profile and credential management module is used to store the atomic behavior units in a personal data vault with encrypted isolation function, perform classification mapping and quality quantification using built-in policy logic to obtain a personal dynamic capability profile, and perform desensitization processing and hash signing on the personal dynamic capability profile and related data according to the access control rules set by the user to generate an index credential associated with a specific service request. The resource retrieval coordination module is used to receive the service request from the current user, encapsulate the semantic features of the service request with the index credentials, send the retrieval request to the service resource node network, calculate the resource matching weight and aggregate the feedback information of each node to obtain a preliminary set of matched resource responses. The challenge verification and sorting module is used to extract the resource node features in the resource response set, perform identity challenge-based capability verification with the current user, and quantify and accumulate the verification results to generate temporary trust points. The historical consensus assessment module is used to retrieve successful cases in the historical interaction pattern library that match the service request and the index credential, and to calculate the dynamic trust assessment result by weighting the temporary trust score and the historical fit through a fusion algorithm. The decision configuration optimization module is used to compare the dynamic trust assessment results with preset gradient thresholds, and based on the comparison results and combined with the resource matching degree, to prioritize and configure resource support schemes for the initially matched resource response set, and generate the final service sequence. The feedback and closed-loop update module is used to send the final service sequence to the current user, extract the rating and actual result after service execution, perform encrypted digest processing, generate new atomic behavior units to be fed back to the personal data vault, and update the historical interaction pattern library.

Images