Room ticket management information system construction method based on edge calculation

Abstract

The invention relates to the technical field of room ticket information management, and discloses a room ticket management information system construction method based on edge computing, and the method comprises the steps: deploying an edge computing node to collect room ticket management original data in real time, including user qualification verification information, real estate resource information and the like; performing multi-dimensional policy matching degree optimization on the original data by adopting a genetic algorithm to generate a multi-dimensional feature matrix; constructing a room ticket transaction risk prediction model based on a capsule network, and outputting a risk level and a potential illegal behavior identifier; constructing a dynamic room ticket circulation graph according to the risk level and the identifier, and presenting a topological relation between related nodes and transaction paths; generating a transaction strategy matching rule including qualification verification conditions and the like based on the topological relation; and issuing the rule to an edge computing node to execute real-time transaction control. According to the method, the precision, dynamic and efficient room ticket management can be realized, and the standardization and controllability of room ticket transaction are improved.

Description

Technical Field

[0001] This invention relates to the field of housing ticket information management technology, specifically a method for constructing a housing ticket management information system based on edge computing. Background Technology

[0002] In the current field of housing voucher management, insufficient real-time data collection is a common problem. Traditional centralized data collection models often require various types of information to be aggregated and processed at a central server. This makes it easy for user qualification verification information, housing resource information, etc., to be delayed during transmission, making it difficult to meet the timeliness requirements of housing voucher transactions.

[0003] The management of housing vouchers involves numerous and multi-dimensional policy provisions, with policy constraints varying across different regions and time periods. Current management methods rely heavily on manual operations or simple rule engines for policy matching, failing to comprehensively and accurately match multi-dimensional policy information. This results in problems with some housing voucher transactions due to insufficient policy compatibility.

[0004] Predicting risks during the housing voucher transaction process is also a major challenge. Traditional risk assessment methods are mostly based on single characteristics or simple combinations of characteristics, making it difficult to capture the complex potential risk factors in housing voucher transactions. They also have a weak ability to identify potential violations, which can easily lead to transaction disputes or illegal operations.

[0005] The circulation of housing vouchers lacks a dynamic and intuitive graphical representation. The topological relationships between voucher holders, property resource nodes, and transaction paths are unclear, making it difficult for administrators to fully grasp the circulation status of vouchers and hindering effective monitoring and management of the transaction process. Regarding transaction strategy formulation, existing rules are often relatively fixed and cannot be adjusted in real time according to the dynamic changes in voucher circulation. This leads to problems such as unreasonable resource allocation and suboptimal transaction paths, affecting the overall efficiency and fairness of housing voucher management. Summary of the Invention

[0006] The purpose of this invention is to provide a method for constructing a housing ticket management information system based on edge computing, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides a method for constructing a housing ticket management information system based on edge computing, the method comprising:

[0008] Edge computing nodes are deployed to collect raw data on housing ticket management in real time. The raw data on housing ticket management includes user qualification verification information, real estate resource information, transaction dynamic information, and regional policy constraint information.

[0009] A genetic algorithm is used to optimize the original data of housing ticket management in multiple dimensions to generate a multi-dimensional feature matrix containing policy adaptability weights.

[0010] A housing voucher transaction risk prediction model based on capsule network is constructed. The multidimensional feature matrix is ​​input into the housing voucher transaction risk prediction model, and the housing voucher transaction risk level and potential violation behavior identifier are output.

[0011] Based on the risk level of the housing voucher transaction and the identification of potential violations, a dynamic housing voucher circulation graph is constructed. The dynamic housing voucher circulation graph includes the topological relationship of housing voucher holding nodes, real estate resource nodes and transaction path edges.

[0012] Based on the topological relationship of the dynamic housing ticket circulation graph, housing ticket transaction strategy matching rules are generated. The housing ticket transaction strategy matching rules include qualification verification conditions, resource allocation priority and transaction path constraints.

[0013] The matching rules for the housing voucher transaction strategy are distributed to edge computing nodes to execute real-time transaction control.

[0014] Preferably, the specific method for optimizing the multi-dimensional policy matching degree of the original data for housing ticket management using a genetic algorithm includes:

[0015] Extract the social security payment duration, household registration status, and family size from the user qualification verification information as individual gene codes;

[0016] The regional policy constraint information is converted into a fitness function, which includes policy clause coverage and constraint conflict penalty terms.

[0017] A new generation of policy adaptability weight set is generated through crossover and mutation operations, and the individual with the highest fitness function value is selected and decoded into the multidimensional feature matrix.

[0018] Preferably, the specific method for constructing the capsule network-based housing ticket transaction risk prediction model includes:

[0019] The multidimensional feature matrix is ​​split into a spatial correlation feature subset and a temporal evolution feature subset;

[0020] In the primary capsule layer, convolutional compression operations are performed on the spatial correlation feature subsets, and gated cyclic encoding operations are performed on the temporal evolution feature subsets.

[0021] The coupling coefficient between the primary capsule output vector and the risk category capsule is calculated at the routing protocol layer, and the coupling coefficient is iteratively updated through a dynamic routing algorithm.

[0022] The high-dimensional risk feature vector is aggregated at the output layer to generate the risk level of the housing voucher transaction and the identifier of potential violations.

[0023] Preferably, the specific method for constructing the dynamic housing ticket circulation map includes:

[0024] The high-risk transaction path identifiers in the risk level of the housing voucher transactions are analyzed to generate a topology blacklist;

[0025] Update the credibility attribute of the property resource node based on the potential violation identifier;

[0026] By integrating the policy adaptability weights and the priority attributes of housing voucher holder nodes, the connection weights of the transaction path edges are reconstructed.

[0027] The adjacency matrix of the dynamic housing ticket circulation graph is established based on the updated node attributes and connection weights.

[0028] Preferably, the specific method for generating the matching rules for the housing voucher transaction strategy includes:

[0029] Traverse the adjacency matrix of the dynamic housing ticket circulation graph and extract the set of housing ticket holding nodes that meet the eligibility verification conditions;

[0030] Resource allocation priority is sorted in descending order based on the credibility attribute of real estate resource nodes;

[0031] Calculate the optimal transaction path constraints by combining the connection weights of the transaction path edges;

[0032] The qualification verification conditions, resource allocation priority, and transaction path constraints are compiled into executable strategy code.

[0033] Preferably, the specific method for performing real-time transaction control includes:

[0034] The executable strategy code is loaded on the edge computing node to verify the user qualification information in the transaction request in real time;

[0035] Compare the target property resources in the transaction request with the resource allocation priority list;

[0036] When a transaction path is detected to deviate from the optimal transaction path constraints, a transaction termination instruction is triggered and the violation is recorded.

[0037] The transaction execution results and strategy deviation data are encapsulated into the transaction feedback data.

[0038] Preferably, the method further includes:

[0039] Record transaction feedback data and strategy execution logs;

[0040] The model update mechanism, which uses the transaction feedback data and strategy execution logs to trigger a dual-time-dimensional mechanism, is used to iteratively optimize the parameters of the housing ticket transaction risk prediction model.

[0041] Preferably, the specific method of the dual-time-dimensional model update mechanism includes:

[0042] Collect the transaction success rate and violation interception rate from the strategy execution logs at an hourly granular level;

[0043] Aggregate the resource matching deviation value and strategy deviation data from the transaction feedback data at a daily granularity.

[0044] Hourly granular data is input into the capsule routing protocol layer of the housing ticket transaction risk prediction model for weight fine-tuning.

[0045] The daily granular data is input into the primary capsule layer of the housing ticket transaction risk prediction model for feature encoder reconstruction.

[0046] Preferably, the specific method for the iterative optimization of parameters includes:

[0047] The fine-tuned capsule routing protocol layer weights are extracted as the initial population for the genetic algorithm;

[0048] A new generation of spatially correlated feature subsets is generated based on the reconstructed feature encoder;

[0049] The policy matching degree of the next-generation spatially associated feature subset is evaluated using a fitness function;

[0050] When the strategy matching degree improves beyond a preset threshold, the parameter set of the housing ticket transaction risk prediction model is updated.

[0051] Preferably, it also includes a trusted evidence storage mechanism for transaction data, specifically: during the transaction control process executed by the edge computing node, the transaction request, strategy matching result and transaction execution result are transmitted through an encrypted channel.

[0052] Compared with the prior art, the beneficial effects of the present invention are:

[0053] This edge computing-based method for building a housing ticket management information system allows for the real-time collection of raw housing ticket management data by deploying edge computing nodes. This enables the data collection process to be closer to the source of information, reduces intermediate data transmission links, and allows various types of information to be quickly acquired and processed. It ensures the timeliness of information such as user qualification verification and housing resources, facilitating timely subsequent management operations.

[0054] By employing a genetic algorithm to optimize the policy matching degree of the original data in multiple dimensions, a multi-dimensional feature matrix containing policy adaptability weights is generated. This fully considers the correlation and differences between different policy provisions, integrates and matches policy information from multiple dimensions, making the policy matching process more comprehensive and accurate, and making the adaptation between housing voucher transactions and policy constraints more in line with the actual situation.

[0055] A risk prediction model for housing voucher transactions based on capsule networks is constructed. By inputting a multi-dimensional feature matrix into the model, the risk level and potential violation identifiers are output. This model can leverage the capsule network's ability to extract and process complex features to deeply explore potential risk factors in housing voucher transactions, accurately identify possible violations, and make risk predictions more closely match the actual risk situation of the transaction. This helps to avoid potential problems in the transaction in advance.

[0056] A dynamic housing ticket circulation map is constructed based on risk levels and potential violation indicators, clearly presenting the topological relationships of housing ticket holding nodes, real estate resource nodes, and transaction path edges. This allows managers to intuitively and comprehensively understand the circulation dynamics of housing tickets, grasp the relationship between each node, and facilitate effective monitoring and sorting of the housing ticket circulation process.

[0057] Based on the topological relationship of the dynamic housing voucher circulation graph, the rules for matching housing voucher transaction strategies are generated. These rules include eligibility verification conditions, resource allocation priorities, and transaction path constraints. This allows the formulation of transaction strategies to be closely integrated with the actual situation of housing voucher circulation, making eligibility verification more in line with actual needs, resource allocation more reasonable, and transaction paths more optimized, thereby improving the overall standardization and efficiency of housing voucher transactions.

[0058] By distributing transaction strategy matching rules to edge computing nodes for real-time transaction control, transaction control instructions can be quickly applied to the transaction process, responding promptly to various changes in housing voucher transactions, ensuring that the transaction process proceeds in an orderly manner under the constraints of the rules, and guaranteeing the fairness and stability of housing voucher management. Attached Figure Description

[0059] Figure 1 This is a schematic diagram illustrating the working principle of the edge computing-based housing ticket management information system construction method described in this invention.

[0060] Figure 2 This is a flowchart illustrating the multi-dimensional policy matching optimization using a genetic algorithm.

[0061] Figure 3 A flowchart for constructing a dynamic housing ticket circulation map;

[0062] Figure 4 A flowchart for implementing real-time transaction control. Detailed Implementation

[0063] 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, and 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.

[0064] Please see Figure 1 This invention provides a method for constructing a housing ticket management information system based on edge computing, the method comprising:

[0065] During system construction, edge computing nodes were first deployed to collect raw data for housing voucher management in real time. These edge computing nodes are distributed across housing voucher management terminals in various regions, acquiring user qualification verification information, property resource information, transaction dynamics information, and regional policy constraints information through distributed data collection interfaces. User qualification verification information includes social security payment records, household registration documents, and family population registration information; property resource information includes static attributes such as property location, unit size, and ownership status, as well as dynamic attributes such as listing time and price fluctuations; transaction dynamics information records time-series data such as changes in housing voucher holdings, transaction initiation and confirmation status, and fund transfer records; and regional policy constraints information includes textual and structured data such as restrictions on home purchase qualifications, regulations on the transfer period of housing vouchers, and detailed rules for resource allocation policies.

[0066] After data collection, a genetic algorithm was used to optimize the policy matching degree of the original housing voucher management data in multiple dimensions. During the optimization process, key parameters were first extracted from user qualification verification information as individual gene codes. Then, regional policy constraint information was converted into a fitness function. A policy fit weight set was generated through crossover and mutation operations of the genetic algorithm, and the optimal individual was selected and decoded into a multi-dimensional feature matrix. Subsequently, a housing voucher transaction risk prediction model based on capsule networks was constructed. The multi-dimensional feature matrix was split into spatial correlation feature subsets and temporal evolution feature subsets. After processing by the primary capsule layer and iterative updating of the coupling coefficients by the routing protocol layer, the output layer aggregated the features to generate risk levels and potential violation identifiers.

[0067] Based on risk prediction results, a dynamic housing voucher circulation graph is constructed. This graph establishes an adjacency matrix by analyzing high-risk transaction paths, updating node credibility, and reconstructing path edge weights, clearly presenting the topological relationships between housing voucher holding nodes, real estate resource nodes, and transaction path edges. Based on this topological relationship, housing voucher transaction strategy matching rules are generated, including eligibility verification conditions, resource allocation priorities, and transaction path constraints. These rules are then compiled into executable code and distributed to edge computing nodes to achieve real-time transaction control.

[0068] Example 1: See Figure 2When optimizing the original data for housing voucher management for multi-dimensional policy matching, the first step is to extract key elements from user eligibility verification information. Among these, social security contribution duration is a crucial indicator reflecting a user's participation and stability in local economic activities. In practice, the system accurately obtains the start and current dates of a user's social security contributions, calculating the contribution duration in months. Household registration status determines whether a user meets local housing purchase preferential policies or restrictions. The system records this in binary code, with 1 representing local household registration and 0 representing non-local household registration. This concise and efficient coding method facilitates subsequent algorithm processing. Family size also significantly impacts housing purchase eligibility and housing resource allocation. The system accurately enters the number of people registered on the family household registration book and categorizes them according to preset population size ranges, such as 1-3 people as small families and 4-6 people as medium-sized families. These three key pieces of information combine to form a gene sequence of a specific length, becoming the foundational individual gene code for subsequent genetic algorithm operations.

[0069] The system transforms regional policy constraints into fitness functions. These constraints contain a large amount of text and structured data, such as minimum social security contribution years required for home purchases, restrictions on the number of homes that can be purchased for different household registrations, and recommended apartment sizes for different family sizes. The system parses these policy clauses and converts them into quantifiable calculation rules. When calculating policy clause coverage, the system iterates through all information in an individual's genetic code and compares each clause with the policy clause. For example, if a policy requires 36 months of social security contributions, and an individual's genetic code shows a contribution period that meets or exceeds this value, then the policy clause is considered satisfied. The proportion of satisfied clauses to the total number of key policy clauses is the policy clause coverage rate. A constraint conflict penalty term handles potential policy conflicts that may arise in an individual's genetic code. For example, if a policy stipulates that only local residents with 24 months of social security contributions are eligible to purchase a second home, and an individual's genetic code indicates local residency but their social security contribution period is only 12 months, yet they show a tendency to purchase a second home, the system will deduct points from the individual's fitness function value according to preset weights. This reflects the degree of conflict between the individual and the policy. The fitness function is ultimately calculated by the difference between the policy coverage rate and the constraint conflict penalty term; the higher this value, the better the match between the individual and the regional policy.

[0070] After constructing the fitness function, the core operation of the genetic algorithm—crossover and mutation—takes place. When selecting parent individuals, the system sorts the individuals in the current population based on their fitness function values, prioritizing individuals with high fitness as parents. For example, individuals in the top 30% of fitness are selected for subsequent operations. Crossover uses a single-point crossover method, randomly selecting a non-critical feature position in the parent individual's gene sequence as the crossover point. Taking a gene sequence composed of social security payment duration, household registration status, and family size as an example, if the randomly selected crossover point is located between the household registration status and family size codes, then the gene segments before and after that point in the two parent individuals will be exchanged, generating two new offspring gene sequences. Mutation introduces new genetic diversity into the population; the mutation probability is usually set within a reasonable range, such as 0.05. For mutation operations, the system randomly selects a certain proportion of offspring individuals for processing. During processing, the system may fine-tune the social security contribution duration value, such as adding or subtracting a certain number of months from the original value; it may also flip the household registration status code, changing 1 to 0 or 0 to 1; or adjust the family population code, such as moving it up or down by one interval from the original size range. Through repeated cross-mutation operations, a new generation of policy adaptability weight sets is continuously generated. After multiple iterations, generally set to 50-100 generations, the system will select the individual with the highest fitness function value from the last generation population and decode its gene sequence into a multidimensional feature matrix. During the decoding process, the social security contribution duration, household registration status, and family population code in the gene sequence will be converted back into corresponding actual values ​​and statuses, and associated with the calculated policy adaptability weights to form a multidimensional feature matrix with user qualification features as rows and policy adaptability weights as columns.

[0071] In the stage of constructing a capsule network-based model for predicting housing voucher transaction risks, the first step is to rationally decompose the multidimensional feature matrix. The spatial correlation feature subset mainly focuses on the static spatial relationship features between users and properties. For example, the matching between the user's registered residence and the area where the property is located; if the user is a local resident, the probability and risk characteristics of choosing a property in a popular local area; the fit between family size and property size; for example, large families choosing small properties may face the risk of overcrowding, thus affecting transaction stability. These features are integrated into the spatial correlation feature subset. The temporal evolution feature subset focuses on dynamic features that change over time. The cumulative change in social security payment duration reflects the trend of user economic stability. If the social security payment duration continues to increase, it indicates that the user's economic stability is enhanced, and the transaction risk is relatively reduced. The length of time the housing voucher is held also affects transaction risk; too short a holding period may indicate speculative transaction risk. The fluctuation of transaction price over a period of time is also an important temporal feature; excessive price fluctuations may indicate market instability or abnormal transactions.

[0072] The primary capsule layer employs different processing methods for these two feature subsets. For the spatially correlated feature subset, convolutional compression is performed. The system uses 3×3 convolutional kernels to perform convolution operations on the spatially correlated feature data across multiple channels. Each convolutional kernel is designed to capture spatial feature patterns; for example, a specific kernel may be designed to identify a particular combination pattern of matching a user's household registration and property area. The result of the convolution operation undergoes a nonlinear transformation using an activation function, such as the ReLU function, to enhance feature representation and form spatial feature capsule vectors. These vectors contain compressed and refined spatially correlated feature information. For the temporally evolved feature subset, gated recurrent encoding is used, employing gated recurrent units (GRUs) to process the temporal sequence data. GRUs can effectively capture the changing patterns of features over time. When processing temporal data such as social security payment duration, housing voucher holding time, and transaction price fluctuations, the GRU uses a gating mechanism based on the current input and the hidden state of the previous time step to determine which information needs to be retained and which needs to be updated, thus outputting capsule vectors that reflect the temporally evolved features.

[0073] The routing protocol layer is responsible for the crucial task of calculating and updating the coupling coefficients. In the initial stage, the relationship between the primary capsule output vector and the risk category capsule is normalized using the softmax function to obtain the initial coupling coefficients. Subsequently, based on the dynamic routing algorithm, the system continuously calculates the similarity between the primary capsule vector and the category capsule vector. If a primary capsule vector has a high similarity to a certain type of risk capsule vector, their coupling coefficient is increased in subsequent iterations; otherwise, it is decreased. Typically, after 3-5 iterations, the coupling coefficient matrix stabilizes, accurately reflecting the degree of correlation between different feature vectors and risk categories. In the output layer, the high-dimensional risk feature vector processed by the routing protocol layer is mapped through a fully connected layer. The neurons in the fully connected layer are connected to each element of the input vector. Through linear transformation of the weight matrix and nonlinear transformation of the activation function, the high-dimensional risk feature vector is mapped to a preset risk level space. For example, if the preset risk levels are low, medium, and high, the output layer determines the risk level of the housing ticket transaction based on the calculated results and a threshold judgment. Meanwhile, a multi-label classifier is used to identify potential violations. The multi-label classifier learns the correspondence between different risk features and violations through training. After inputting a high-dimensional risk feature vector, it outputs a potential violation identifier, such as a fake listing risk identifier or an abnormal transaction price identifier.

[0074] Example 2: See Figure 3The process of constructing a dynamic housing voucher circulation graph begins with an in-depth analysis of the risk levels of housing voucher transactions. The risk level of housing voucher transactions is output by a risk prediction model, which includes risk assessment results for various transaction paths. The system extracts identifiers for transaction paths marked as high-risk. These identifiers record detailed information such as the unique ID of the housing voucher-holding node participating in the transaction, the unique ID of the corresponding real estate resource node, the specific timestamp of the transaction initiation, and transaction status change records. The system organizes this information in chronological order of the transactions, removes duplicate records, and forms a set containing key information for all high-risk transaction paths—a topology blacklist.

[0075] After acquiring potential violation identifiers, the system updates the credibility attribute of property resource nodes accordingly. Potential violation identifiers cover various possible violation types, including but not limited to: fake listings (inconsistent information), unclear property rights (incomplete or disputed ownership documents), and duplicate listings (the same property listed on multiple platforms). For different types of violations, the system pre-sets corresponding credibility deduction rules. For example, for a property resource node marked as a fake listing, its initial credibility value will be reduced by a preset percentage; if there is an unclear property rights identifier, the percentage reduction is determined based on the severity of the ownership dispute; the credibility deduction percentage for duplicate listings is relatively low, but accumulates with the number of repetitions. The credibility attribute of property resource nodes uses a quantitative score of 0-100. The initial value is set based on the completeness and compliance of the information at the time of the property's initial registration. Each time a new potential violation identifier is detected, the system updates the corresponding node's credibility score in real time to ensure that the score accurately reflects the current compliance status of the property resource.

[0076] The system integrates policy adaptability weights and the priority attributes of housing voucher holders to reconstruct the connection weights of transaction path edges. The policy adaptability weights are derived from a multi-dimensional feature matrix generated after optimization using a genetic algorithm. This matrix quantifies the degree of matching between the eligibility information of housing voucher holders and regional policy provisions; a higher value indicates that the holder is more aligned with policy guidelines. The priority attributes of housing voucher holders incorporate multiple factors, including the duration of the housing voucher holding (longer holding time may increase priority), the level of housing need within the holder's family (e.g., families without housing have higher priority than families with housing), and the compliance status of past transaction records. These factors are weighted to form a quantified priority value. Connection weights are calculated using a weighted summation method, with policy adaptability weights accounting for 60% and housing voucher holder priority attributes accounting for 40%. The weighted result is normalized and mapped to a 0-1 range. Values ​​within this range directly reflect the compliance and rationality of the transaction path; higher values ​​indicate better overall performance in policy adaptability and resource matching.

[0077] Based on the updated node attributes and connection weights, the system establishes an adjacency matrix for the dynamic housing voucher circulation graph. The number of rows in the adjacency matrix corresponds to the number of housing voucher-holding nodes, and the number of columns corresponds to the number of real estate resource nodes. Each element in the matrix represents the connection weight of the transaction path edge between the corresponding housing voucher-holding node and the real estate resource node. If a transaction path is blacklisted in the topology relationship, its corresponding matrix element value is forcibly set to 0 to explicitly mark the high-risk attribute of that path. Simultaneously, the adjacency matrix also has additional attribute columns to record key information such as the priority attribute of each housing voucher-holding node and the credibility attribute of each real estate resource node. This information, together with the matrix element values, constitutes a complete dynamic housing voucher circulation graph data structure. It not only clearly displays the topological relationships between housing voucher-holding nodes, real estate resource nodes, and transaction path edges, but also dynamically reflects the compliance status and risk changes during the housing voucher circulation process through real-time updates of node attributes and connection weights.

[0078] Throughout the construction process, the system continuously monitors changes in node attributes and connection weights. When the eligibility information of a housing voucher holder node changes (e.g., an increase in social security payment duration), the violation flag of a real estate resource node is removed or added, or regional policy constraint information is updated leading to adjustments in policy adaptability weights, the system immediately triggers a dynamic update mechanism. This mechanism recalculates the attribute values ​​of the relevant nodes and the connection weights of path edges, and synchronously updates the corresponding elements in the adjacency matrix. This real-time update mechanism ensures that the dynamic housing voucher circulation graph accurately captures the latest changes in the housing voucher circulation process, providing a dynamic and accurate topological view for housing voucher management.

[0079] Example 3: See Figure 4 When generating matching rules for housing voucher transaction strategies, the system comprehensively traverses the adjacency matrix of the dynamic housing voucher circulation graph, using preset eligibility verification thresholds to filter eligible housing voucher-holding nodes. These thresholds include several specific indicators. The minimum weight for policy adaptability is determined based on the strictness of regional housing purchase policies; for example, a higher value may be set in core urban areas to ensure resources are prioritized for groups with high policy adaptability. The minimum holding period for housing vouchers references local regulations on housing voucher circulation cycles to prevent resource waste caused by frequent short-term transactions. The minimum standard for family essential housing needs is determined based on family housing situation, member structure, and other information; more lenient standards apply to essential housing need groups such as families without their own housing or families with elderly members. During the traversal, the system extracts the association attributes of housing voucher-holding nodes row by row in the adjacency matrix, checking each node against all threshold conditions. Only node IDs that meet all criteria are included in the qualified node set; nodes that do not meet the conditions are marked as pending review, and their subsequent transaction requests require additional manual review.

[0080] Based on the credibility attributes of real estate resource nodes, the system prioritizes resource allocation by ranking them from highest to lowest score. Credibility attributes comprehensively reflect the compliance status of properties, encompassing the completeness of property ownership certificates, consistency between property information and actual condition, and records of disputes in past transactions. During sorting, all real estate resource nodes are first arranged in descending order of credibility score to form a preliminary sequence. If multiple nodes have the same credibility score, the system will further refine the ranking by considering the static attributes of the properties. For example, it will prioritize properties with a higher suitability between unit size and family size, or properties with more complete surrounding amenities. Through multi-dimensional consideration, a final resource allocation priority list is formed. Properties ranked higher in the list will be prioritized for recommendation to qualified property holders during the transaction matching stage.

[0081] When calculating the constraints of the optimal transaction path, the system first calculates the connection weights of all valid transaction path edges in the adjacency matrix, and then determines the upper and lower limits of the path constraints through statistical analysis. The connection weights reflect the overall compliance level of the transaction path, and their thresholds are calculated using the following formula:

[0082] T high =α+2β

[0083] T low =α-2β

[0084] Where α represents the arithmetic mean of the connection weights of all valid transaction paths in the current adjacency matrix, β represents the standard deviation of the connection weights, and T high T is the upper limit threshold for path weights. low This represents the lower threshold for path weights. Connection weights exceeding the upper threshold may indicate anomalies caused by artificially inflating path priority, while connection weights below the lower threshold suggest deficiencies in policy adaptation or resource matching. Simultaneously, the system sets path taboo rules based on a topology blacklist, including high-risk node pairs recorded in the blacklist within the prohibited transaction scope, explicitly prohibiting direct transaction paths between these nodes, thus forming a complete constraint boundary encompassing weight thresholds and taboo nodes.

[0085] After completing the constraint calculations, the system converts the eligibility verification conditions, resource allocation priorities, and transaction path constraints into executable strategy code. The eligibility verification conditions are compiled into the rule engine's condition judgment logic, such as "eligibility verification passes when the policy adaptability weight of the housing voucher holder node is ≥0.7 and the holding period is ≥6 months." The resource allocation priority is converted into sorting algorithm code, which updates housing priority in real time by calling a dynamic sorting function. The transaction path constraints are compiled into path filtering logic, which filters out transaction paths that meet the constraints by iteratively comparing path weights with thresholds and checking whether nodes are on the prohibition list. The generated code includes a standardized input interface for receiving real-time transaction data uploaded by edge computing nodes; the logic processing module is responsible for executing various rule judgments and calculations; and the output interface returns transaction permission or interception instructions, ensuring that the code can run directly in the hardware environment of the edge computing nodes.

[0086] After the edge computing node loads the executable policy code, it enters the real-time transaction control process. When a transaction request is received, the system first extracts the user's eligibility information from the request, including social security payment details, household registration certificate number, and family population registration information, and compares it item by item with the eligibility verification conditions in the policy rules. During the comparison process, if it is found that the user's social security payment duration has not reached the threshold or the household registration status does not meet the policy requirements, the system will immediately mark it as unqualified and suspend the transaction process. At the same time, the system will match the target property resource ID in the transaction request with the resource allocation priority list, query the specific ranking of the property in the list, and if the ranking is low, it will push higher-ranked alternative property information to the user.

[0087] During the transaction path detection phase, the system calculates the connection weight corresponding to the current transaction path and compares it with the upper and lower thresholds in the optimal transaction path constraints. If the connection weight is greater than the upper threshold or less than the lower threshold, or if it involves node pairs in the topology blacklist, the system immediately triggers a transaction termination command, suspends the transaction operation, and records detailed information about the violation, including the violation type (e.g., abnormal path weight, involvement of high-risk nodes), the ID of the involved housing voucher holder node, the ID of the property resource node, and the precise time of the violation. After the transaction is completed or terminated, the edge computing node encapsulates the transaction result data. The result data includes the final transaction status (success or termination), the ranking of the matched property in the priority list, and detailed information about the actual transaction path. The strategy deviation data includes the difference between the actual path weight and the threshold, and the matching deviation between the qualification information and the verification conditions. This data will be stored as historical records for subsequent system operation analysis. The entire process achieves rapid response through localized processing by the edge computing node, ensuring the timeliness and accuracy of transaction control.

[0088] Example 4: During the operation of the edge computing-based housing ticket management information system, edge computing nodes continuously record transaction feedback data and strategy execution logs. Transaction feedback data covers the final execution status of each transaction, including identifiers for successful transaction completion, identifiers for transactions terminated due to risk interception, and information on the actual matched property resources and the node IDs of both parties involved in the transaction. The strategy execution logs record in detail the operation trajectory of the rule engine, such as the eligibility verification conditions triggered by each transaction request, the resource priority comparison process, the path constraint check results, and the trigger time and execution duration of various instructions. This data is categorized and stored according to the timestamp of each transaction. Each record contains complete field information, such as transaction ID, request time, user ID, property ID, execution result code, and rule matching details, forming a structured historical dataset that provides raw data support for subsequent model updates.

[0089] The system achieves dynamic optimization through a dual-time-dimensional model update mechanism. It extracts transaction success rate and violation interception rate from the strategy execution logs at an hourly granularity. The transaction success rate is the ratio of the number of successfully executed transactions to the total number of transaction requests within a specific hour. The violation interception rate is the ratio of the number of transactions triggering termination commands to the total number of transaction requests within the same hour. For example, during the 9:00-10:00 time slot on a weekday, if the system receives 120 transaction requests, of which 95 pass the qualification verification and are completed, and 25 are intercepted due to excessively high risk levels, the transaction success rate for this time slot is 95 / 120, and the violation interception rate is 25 / 120. After each hour, the system automatically calculates these metrics and generates hourly granular data, including fields such as the start time of the time slot, total number of transactions, number of successful transactions, number of intercepted transactions, success rate, and interception rate. This data is transmitted to the model optimization module in real time via a dedicated interface.

[0090] Table: Example of data update for a two-time-dimensional model

[0091]

[0092] When aggregating transaction feedback data on a daily granular basis, the system calculates the daily average values ​​of resource matching deviation and strategy deviation. The resource matching deviation reflects the difference between the actually allocated property resources and the optimal recommended resources. It is calculated by summing the differences between the ranking of the actually selected property in the priority list and the ranking of the optimal property recommended by the system across all transactions on that day, and dividing by the total number of transactions that day. For example, if there are 1200 transactions on a workday, and the total difference between the actual property ranking and the optimal ranking is 1440, then the resource matching deviation is 1.2. The strategy deviation comprehensively considers path weight deviation and qualification matching deviation. Path weight deviation is the average absolute value of the difference between the connection weight of the actual transaction path and the weight of the optimal path. Qualification matching deviation is the average difference between the matching degree of user qualification information and verification conditions. The two are weighted and summed according to a preset ratio to obtain the daily strategy deviation. For example, if the path weight deviation is 0.4 and the qualification matching deviation is 0.3 on a certain day, the weighted strategy deviation is 0.35. After standardization, the daily granular data eliminates the dimensional differences between different indicators, forming optimized data with a unified scale.

[0093] After inputting hourly granular data into the capsule routing protocol layer of the housing ticket transaction risk prediction model, the system analyzes the correlation between transaction success rate, violation interception rate, and the coupling coefficient of the routing protocol layer. If the violation interception rate increases for two consecutive hours, it indicates that the model's sensitivity to risk identification is insufficient. The system will then increase the update step size of the coupling coefficient, allowing the routing protocol layer to focus on high-risk characteristics more quickly during the iteration process. If the transaction success rate continues to decline, it may be that the model is over-intercepting normal transactions. In this case, the update step size will be reduced, and the number of iterations will be extended, allowing the coupling coefficient to more accurately reflect the actual risk distribution. Through this dynamic adjustment, the model's identification of recent transaction risks is kept consistent with actual interception needs.

[0094] After daily granular data is input into the primary capsule layer, the system adjusts the convolutional kernel parameters of spatial correlation features based on the resource matching deviation value. If the resource matching deviation value is high, it indicates that the model is not sufficiently extracting spatial adaptation features between users and housing resources. The system will optimize the size and number of convolutional kernels to enhance the ability to capture key spatial features such as unit area and regional location. For strategy deviation data, the system will adjust the parameters of the gated recurrent unit (GRU), such as the weights of the reset gate and update gate, so that dynamic features such as social security payment duration and housing voucher holding time in the time-series evolution feature subset are encoded more accurately. For example, when the strategy deviation shows a large qualification matching deviation, the update gate weight of the GRU will be tilted towards core qualification features such as social security payment records, improving the model's ability to capture long-term qualification evolution patterns. Through the collaborative optimization of the hourly and daily time dimensions, the model can simultaneously adapt to short-term transaction fluctuations and long-term trend changes, continuously optimizing the accuracy of risk prediction and strategy matching.

[0095] Example 5: During the parameter iterative optimization process, the system first extracts the weight parameters, fine-tuned with hourly granular data, from the capsule routing protocol layer. These weight parameters are key parameters formed by the model during recent transaction risk identification, reflecting the correlation strength between different primary capsules and risk category capsules. The system converts these weight parameters into an initial population that can be processed by the genetic algorithm, with each individual in the population corresponding to a complete set of routing weight parameters. The gene encoding length of the individual is consistent with the number of weight parameters, and the encoded value directly corresponds to the quantified value of the weight parameter. Through this conversion method, the model parameters and the optimization algorithm are effectively connected, enabling the genetic algorithm to perform optimization operations based on the parameter distribution formed by actual transaction data. The size of the initial population is determined according to the dimension of the weight parameters to ensure that the population has sufficient diversity.

[0096] After reconstructing the feature encoder at the primary capsule layer, the system uses the updated encoder to extract features from the original housing ticket management data, generating a new generation of spatial association feature subsets. The reconstruction of the feature encoder involves adjusting the convolution kernel parameters and gated recurrent unit parameters, based on daily aggregated resource matching deviation values ​​and policy deviation data. The updated encoder can more accurately capture the spatial association patterns between user qualifications and housing resources, such as more detailed encoding of matching features between user's registered residence and housing area, and adaptation features between family size and apartment size. Compared to the original version, the new generation of spatial association feature subsets contains more feature patterns reflecting actual transaction patterns, reducing policy matching errors caused by feature extraction biases.

[0097] The system evaluates the strategy matching degree of a new generation of spatial association feature subsets using a fitness function. The fitness function is calculated using daily aggregated resource matching deviation and strategy deviation data as core inputs. The resource matching deviation reflects the difference between the actual allocated housing and the optimal recommended housing, while the strategy deviation reflects the deviation of the transaction path from the rule constraints. Both are weighted and summed according to a preset weight ratio, with the resource matching deviation having a higher weighting to ensure that the optimization process prioritizes the accuracy of resource allocation. A lower fitness function value indicates a better strategy matching effect for the feature subset. The system evaluates the fitness of all individuals in the genetic algorithm population, retaining individuals with lower fitness function values ​​for the next generation iteration while eliminating those with poor fitness. Through multiple iterations, the strategy matching capability of the feature subset is gradually improved.

[0098] When the improvement in strategy matching degree exceeds a preset threshold after three consecutive iterations, the system initiates the model parameter update process. The threshold is set by comprehensively considering typical improvement rates in historical optimization data and the stability requirements of actual business operations, ensuring that only significant optimization effects trigger parameter updates and avoiding model fluctuations caused by frequent updates. The system selects the individual with the lowest fitness function value from the current population and decodes its genetic code into specific model parameters, including the coupling coefficient matrix of the routing protocol layer, the convolution kernel parameters of the primary capsule layer, and the weight parameters of the gated recurrent unit. These parameters are synchronously updated to the corresponding modules of the housing ticket transaction risk prediction model, enabling the model to apply the optimized parameter combinations to conduct subsequent risk prediction work and achieve continuous improvement in model performance.

[0099] Throughout the entire transaction control process executed at the edge computing nodes, the system employs a trusted evidence storage mechanism to ensure the security and integrity of transaction data. Transaction request data undergoes symmetric encryption before transmission. The edge node and the central system generate a session key through a pre-defined key negotiation mechanism. This key is valid only within a single transaction session and automatically expires after the session ends. Sensitive data such as user qualification information, property resource IDs, and transaction amounts are all encrypted using this session key. The encryption process is completed locally at the edge node, avoiding the risk of transmitting sensitive data in an unencrypted state.

[0100] After the policy matching result is generated, the system signs the result data using a digital signature algorithm. The signature information includes the hash value of the result data and a unique identifier for the edge node. The hash value is generated by performing a one-way hash calculation on the result data, and it can uniquely identify the data content. Once the data is tampered with, the hash value will change. The digital signature is transmitted along with the policy matching result, and the recipient can verify the authenticity of the data's source and the integrity of its content by verifying the signature.

[0101] The transmission of transaction execution results uses a TLS protocol to establish an encrypted channel, ensuring data confidentiality during transmission. The TLS protocol negotiates encryption algorithms and session keys through a handshake process, encrypting and decrypting transmitted data in real time to prevent eavesdropping or tampering during transmission. The transaction execution result data includes transaction status, information about involved nodes, timestamps, etc., and is encapsulated before being transmitted to the central system through the encrypted channel, forming a complete transaction record.

[0102] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0103] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for constructing a housing ticket management information system based on edge computing, characterized in that, include: Edge computing nodes are deployed to collect raw data on housing ticket management in real time. The raw data on housing ticket management includes user qualification verification information, real estate resource information, transaction dynamic information, and regional policy constraint information. A genetic algorithm is used to optimize the original data of housing ticket management in multiple dimensions to generate a multi-dimensional feature matrix containing policy adaptability weights. A housing voucher transaction risk prediction model based on capsule network is constructed. The multidimensional feature matrix is ​​input into the housing voucher transaction risk prediction model, and the housing voucher transaction risk level and potential violation behavior identifier are output. Based on the risk level of the housing voucher transaction and the identification of potential violations, a dynamic housing voucher circulation graph is constructed. The dynamic housing voucher circulation graph includes the topological relationship of housing voucher holding nodes, real estate resource nodes and transaction path edges. Based on the topological relationship of the dynamic housing ticket circulation graph, housing ticket transaction strategy matching rules are generated. The housing ticket transaction strategy matching rules include qualification verification conditions, resource allocation priority and transaction path constraints. The matching rules for the housing voucher transaction strategy are distributed to edge computing nodes to execute real-time transaction control.

2. The method for constructing a housing ticket management information system based on edge computing according to claim 1, characterized in that, The specific method for optimizing the multi-dimensional policy matching degree of the original data for housing ticket management using a genetic algorithm includes: Extract the social security payment duration, household registration status, and family size from the user qualification verification information as individual gene codes; The regional policy constraint information is converted into a fitness function, which includes policy clause coverage and constraint conflict penalty terms. A new generation of policy adaptability weight set is generated through crossover and mutation operations, and the individual with the highest fitness function value is selected and decoded into the multidimensional feature matrix.

3. The method for constructing a housing ticket management information system based on edge computing according to claim 2, characterized in that, The specific method for constructing a housing ticket transaction risk prediction model based on capsule networks includes: The multidimensional feature matrix is ​​split into a spatial correlation feature subset and a temporal evolution feature subset; In the primary capsule layer, convolutional compression operations are performed on the spatial correlation feature subsets, and gated cyclic encoding operations are performed on the temporal evolution feature subsets. The coupling coefficient between the primary capsule output vector and the risk category capsule is calculated at the routing protocol layer, and the coupling coefficient is iteratively updated through a dynamic routing algorithm. The high-dimensional risk feature vector is aggregated at the output layer to generate the risk level of the housing voucher transaction and the identifier of potential violations.

4. The method for constructing a housing ticket management information system based on edge computing according to claim 3, characterized in that, The specific methods for constructing a dynamic housing ticket circulation map include: The high-risk transaction path identifiers in the risk level of the housing voucher transactions are analyzed to generate a topology blacklist; Update the credibility attribute of the property resource node based on the potential violation identifier; By integrating the policy adaptability weights and the priority attributes of housing voucher holder nodes, the connection weights of the transaction path edges are reconstructed. The adjacency matrix of the dynamic housing ticket circulation graph is established based on the updated node attributes and connection weights.

5. The method for constructing a housing ticket management information system based on edge computing according to claim 4, characterized in that, The specific methods for generating matching rules for housing voucher transaction strategies include: Traverse the adjacency matrix of the dynamic housing ticket circulation graph and extract the set of housing ticket holding nodes that meet the eligibility verification conditions; Resource allocation priority is sorted in descending order based on the credibility attribute of real estate resource nodes; Calculate the optimal transaction path constraints by combining the connection weights of the transaction path edges; The qualification verification conditions, resource allocation priority, and transaction path constraints are compiled into executable strategy code.

6. The method for constructing a housing ticket management information system based on edge computing according to claim 5, characterized in that, The specific methods for implementing real-time transaction control include: The executable strategy code is loaded on the edge computing node to verify the user qualification information in the transaction request in real time; Compare the target property resources in the transaction request with the resource allocation priority list; When a transaction path is detected to deviate from the optimal transaction path constraints, a transaction termination instruction is triggered and the violation is recorded. The transaction execution results and strategy deviation data are encapsulated into the transaction feedback data.

7. The method for constructing a housing ticket management information system based on edge computing according to claim 1, characterized in that, The method further includes: Record transaction feedback data and strategy execution logs; The model update mechanism, which uses the transaction feedback data and strategy execution logs to trigger a dual-time-dimensional mechanism, is used to iteratively optimize the parameters of the housing ticket transaction risk prediction model.

8. The method for constructing a housing ticket management information system based on edge computing according to claim 7, characterized in that, The specific methods of the dual-time-dimensional model update mechanism include: Collect the transaction success rate and violation interception rate from the strategy execution logs at an hourly granular level; Aggregate the resource matching deviation value and strategy deviation data from the transaction feedback data at a daily granularity. Hourly granular data is input into the capsule routing protocol layer of the housing ticket transaction risk prediction model for weight fine-tuning. The daily granular data is input into the primary capsule layer of the housing ticket transaction risk prediction model for feature encoder reconstruction.

9. The method for constructing a housing ticket management information system based on edge computing according to claim 8, characterized in that, The specific methods for parameter iterative optimization include: The fine-tuned capsule routing protocol layer weights are extracted as the initial population for the genetic algorithm; A new generation of spatially correlated feature subsets is generated based on the reconstructed feature encoder; The policy matching degree of the next-generation spatially associated feature subset is evaluated using a fitness function; When the strategy matching degree improves beyond a preset threshold, the parameter set of the housing ticket transaction risk prediction model is updated.

10. The method for constructing a housing ticket management information system based on edge computing according to claim 1, characterized in that, It also includes a trusted evidence storage mechanism for transaction data, specifically: during the transaction control process at the edge computing node, transaction requests, strategy matching results, and transaction execution results are transmitted via an encrypted channel.

Images