A data analysis-based security house management method and system

Abstract

This invention belongs to the field of smart city governance technology. It discloses a data analysis-based method and system for managing affordable housing. The system includes: collection and standardized preprocessing of multi-source heterogeneous data; anomaly detection of residential behavior based on multimodal feature fusion; in-depth analysis of the root causes of abnormal behavior based on causal inference; automatic generation and classification of intelligent intervention strategies; execution and feedback collection of multi-channel intervention strategies; quantitative evaluation and iterative optimization of intervention effects; and full-process data analysis and verification of the housing security effect evaluation. By automatically discovering residential anomalies through multimodal feature fusion and accurately locating the root causes of problems using causal inference technology, it achieves an intelligent leap from anomaly monitoring to root cause diagnosis. This enables deep integration and dynamic perception of data resources, significantly improving the initiative of dynamic supervision, avoiding information fragmentation and lag, and laying a solid foundation for precise management.

Description

Technical Field

[0001] This invention relates to the field of smart city governance technology, and more specifically, to a data analysis-based method and system for managing affordable housing. Background Technology

[0002] In the field of smart city governance, affordable housing is an important project for people's livelihood. Its construction scale and management complexity are increasing day by day. The traditional affordable housing management model mainly relies on manual operation and static information registration, which has prominent problems such as information update lag, low resource allocation efficiency, difficulty in qualification review and insufficient dynamic supervision. Meanwhile, because affordable housing involves multiple stages such as application, review, allocation, occupancy, and withdrawal, the data sources are scattered and the standards are inconsistent, making it difficult for management departments to fully and accurately grasp the dynamics of housing resources, residents, and needs, which affects the fair allocation and effective use of affordable housing resources.

[0003] In view of this, the present invention proposes a data analysis-based method and system for managing affordable housing to solve the above problems. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a data analysis-based method and system for managing affordable housing, comprising: Step 1: Acquisition and Standardization Preprocessing of Multi-Source Heterogeneous Data Through the data acquisition module, multi-source heterogeneous raw data from affordable housing databases, government platforms, IoT devices and rental platforms are collected. The data is then cleaned, de-identified, standardized and integrated to generate standardized data records with unified identifiers and timestamps. Step 2: Anomaly Detection of Residential Behavior Based on Multimodal Feature Fusion Based on the generated standardized data records, multidimensional features such as electricity consumption, water consumption, door switch, density of anonymous equipment, and market rent ratio are extracted. A residential consistency index is generated through a calculation model. When the index is continuously lower than the dynamic threshold, a preliminary warning report is generated and the resident is marked as pending verification. Step 3: In-depth analysis of the root causes of abnormal behavior based on causal inference The early warning information is input into the causal inference engine. Using similar normal households as the control group, the difference-in-differences analysis method is used to calculate the contribution of each factor. A causal graph model is constructed to quantify the weights and generate a root cause analysis report that clearly points out the main abnormal root causes and their credibility. Step 4: Automatic generation and classification of intelligent intervention strategies Based on the factor combinations and weights in the root cause analysis report, the system automatically matches a preset intervention strategy template and generates an intelligent intervention strategy package with a root cause summary and evidence chain index. Step 5: Implementation and Feedback Collection of Multi-channel Intervention Strategies Based on the type of intelligent intervention strategy package, instructions are sent through a multi-channel execution system via mobile application push or manual task distribution, and the recipient is forced to upload the execution feedback results of the preset type, and the feedback information is associated, stored and a complete timeline is recorded. Step 6: Quantitative evaluation of intervention effects and iterative optimization of the model The feedback data is statistically analyzed for success rate and response rate according to strategy type. Counterfactual analysis is used to compare the differences in subsequent behavior between the intervention group and the control group to quantify the effect. An evaluation report is generated and the calculation model parameters and strategy matching rules are automatically adjusted accordingly. Step 7: Full-process data analysis and verification to ensure effectiveness evaluation By repeatedly executing the above steps, a complete data analysis closed loop is formed, from data collection to effect evaluation. This closed loop ensures that all conclusions have a traceable data chain and logical path, and clarifies the role of data analysis in risk identification, root cause diagnosis, precise intervention, and effect quantification by outputting a comprehensive verification report.

[0005] Furthermore, the acquisition and standardization preprocessing of multi-source heterogeneous data includes: The system obtains application information, resident identity information, and housing information from the affordable housing management database; it also obtains authorized household registration information, social security records, marriage information, and real estate registration information from the government's data sharing platform; it acquires anonymized energy consumption data and door magnetic switch events from IoT devices; and it obtains market rent reference information from the housing rental information platform. All raw data acquired is cleaned to remove invalid records, desensitized to handle sensitive fields, standardized to unify data format and units, and finally linked and integrated to merge multi-source data of the same entity and assign a unique serial number and timestamp to generate standardized data records.

[0006] Furthermore, the detection of residential behavior anomalies through multimodal feature fusion includes: Based on the standardized data records, electricity consumption characteristics are extracted, the total daily electricity consumption is calculated, and the deviation from the household's historical average electricity consumption is obtained. Extract water usage characteristics and calculate the standard deviation of water usage during the early morning period each day as the nighttime water usage volatility. Extract portal switch characteristics, record and count the number of daily switches, and calculate the multi-day coefficient of variation as a regularity indicator; Extract the density characteristics of anonymous devices and calculate the seven-day moving average of the number of daily active anonymous devices as the density trend. Extract the characteristics of market rent ratios and calculate the ratio of the latest listed rent to the rent of affordable housing; Input the electricity consumption characteristics, water consumption characteristics, gateway switch characteristics, anonymous equipment density characteristics, and market rent ratio characteristics into the calculation model.

[0007] Furthermore, the in-depth analysis of the root causes of abnormal behavior based on causal inference includes: Based on the resident information and warning trigger time in the preliminary warning report, historical characteristic data of abnormal residents were retrieved from standardized data records; similar normal residents were selected as a control group based on housing resources, family structure and occupancy duration; with the warning trigger time as the intervention point, the difference-in-differences method was used to calculate the contribution of each factor, such as changes in commuting distance, changes in work unit, changes in family income and fluctuations in market rent, to changes in residential behavior. A causal graph model is constructed based on the contribution of each factor, and the standardized weight of each factor is calculated. A root cause analysis report is generated, which lists the influencing factors in order of standardized weight and indicates that the one with the highest weight is the main root cause of the anomaly. At the same time, the confidence interval of the contribution is attached as a confidence assessment.

[0008] Furthermore, the automatic generation and classification of intelligent intervention strategies include: Based on the standardized weights of each influencing factor in the root cause analysis report, key and secondary factors are identified, and the corresponding intervention strategy templates in the pre-set intervention strategy template library are matched by the type combination of key and secondary factors. The specific data from the root cause analysis report is used to fill the specified fields of the intervention strategy template, while generating a root cause summary containing descriptions of the main abnormal root causes and an evidence chain index related to the underlying data. This results in the final filled intervention strategy template, root cause summary, and evidence chain index being packaged into an intelligent intervention strategy package with a unified format and type identifier.

[0009] Furthermore, the implementation and feedback collection of multi-channel intervention strategies include: The execution channel is determined by parsing the strategy type identifier of the intelligent intervention strategy package; if the identifier corresponds to the proactive suggestion strategy, a notification message is pushed to the designated resident through the application interface; if the identifier corresponds to the strategy requiring manual intervention, a task order is created and assigned to the designated administrator in the background; after execution in any channel, the recipient is forced to select the preset type of execution feedback result on the designated interface and submit it. The execution feedback results are associated with and stored with the corresponding intelligent intervention strategy package, and the complete processing timeline from strategy generation to feedback reception is recorded.

[0010] Furthermore, the quantitative evaluation of intervention effects and iterative optimization of the model include: The execution feedback results were categorized and summarized according to the associated strategy type, and the execution success rate and resident response rate of each strategy type were calculated. Using counterfactual analysis, residents who had received the smart intervention strategy package were set as the intervention group, and residents with similar conditions but who had not received the intervention were matched as the control group. The average change value of the residential consistency index of the two groups of residents within a specific time period after the intervention was calculated, and the difference was used as the average treatment effect to quantify the intervention effect. Generate an intervention effect evaluation report that includes the success rate, response rate, and average treatment effect; based on the quantitative results in the intervention effect evaluation report, automatically adjust the weight coefficients of each feature in the calculation model, and update the matching priority of factor combinations and strategy types in the intervention strategy template library.

[0011] Furthermore, the full-process data analysis and verification to ensure effectiveness evaluation includes: Steps 1 through 6 are executed cyclically to form a data analysis closed loop, and intermediate data and reports for each cycle are associated and stored by assigning a unified cycle number. After completing a preset number of cycles, the preliminary early warning reports, root cause analysis reports, intelligent intervention strategy package processing results, and intervention effect evaluation reports for each cycle are summarized; quantitative analysis is performed based on the summarized data to evaluate and clarify the role of data analysis in ensuring the accuracy of risk identification, the accuracy of root cause diagnosis, the effectiveness of precise intervention, and the sustainability of quantitative effects; a comprehensive verification report is generated and published.

[0012] The technical effects and advantages of the data analysis-based affordable housing management method and system of this invention are as follows: 1. Through multi-source heterogeneous data collection and standardized preprocessing, real-time or near-real-time data from government affairs, the Internet of Things, and the market are integrated to achieve deep data fusion and dynamic perception. This enables management departments to comprehensively, accurately, and dynamically grasp the status of housing resources, the actual living conditions of residents, and changes in market demand, laying a solid data foundation for accurate decision-making and effectively overcoming the slow updates and limited perspectives caused by traditional reliance on static and manual registration information. 2. By automatically generating and tracking the execution of intervention strategies, personalized strategies can be intelligently matched and pushed based on the root cause analysis results. A forced feedback mechanism is then implemented to form an online closed-loop management process, which greatly optimizes resource allocation and execution efficiency. Furthermore, by establishing a quantifiable and iterative self-optimization mechanism, a sustainable evaluation of intervention effects is formed, which is used to automatically optimize models and strategies, so that management effectiveness can continuously evolve with the accumulation of data.

[0013] 3. By establishing a smart governance mechanism that is traceable throughout the entire process and driven by data, the regulatory challenges of the traditional model will be effectively solved. This will form a complete closed loop, from front-end data perception to intelligent analysis, precise intervention conclusions, and final effect evaluation, ensuring the transparency, scientific nature, and adaptability of the management process. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating a data analysis-based affordable housing management method and system according to the present invention. Figure 2 A schematic diagram of the intelligent encapsulation strategy package for the root cause analysis report of this invention; Figure 3 This is a schematic diagram illustrating the execution and feedback collection of the multi-channel intervention strategy of the present invention.

[0015] Figure 4 This diagram illustrates the quantitative evaluation and iterative optimization of the intervention effect of this invention. Detailed Implementation

[0016] 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.

[0017] Example 1 Please see Figures 1-4 As shown in the figure, the main design contents of the data analysis-based affordable housing management method and system described in this embodiment are as follows: A data analysis-based affordable housing management method and system is designed, including: Step S1: Acquisition and Standardization Preprocessing of Multi-Source Heterogeneous Data Through the data acquisition module, multi-source heterogeneous raw data from affordable housing databases, government platforms, IoT devices and rental platforms are collected. The data is then cleaned, de-identified, standardized and integrated to generate standardized data records with unified identifiers and timestamps. Step S2: Anomaly detection of residential behavior based on multimodal feature fusion Based on the generated standardized data records, multidimensional features such as electricity consumption, water consumption, door switch, density of anonymous equipment, and market rent ratio are extracted. A residential consistency index is generated through a calculation model. When the index is continuously lower than the dynamic threshold, a preliminary warning report is generated and the resident is marked as pending verification. Step S3: In-depth analysis of the root causes of abnormal behavior based on causal inference The early warning information is input into the causal inference engine. Using similar normal households as the control group, the difference-in-differences analysis method is used to calculate the contribution of each factor. A causal graph model is constructed to quantify the weights and generate a root cause analysis report that clearly points out the main abnormal root causes and their credibility. Step S4: Automatic generation and classification of intelligent intervention strategies Based on the factor combinations and weights in the root cause analysis report, the system automatically matches a preset intervention strategy template and generates an intelligent intervention strategy package with a root cause summary and evidence chain index. Step S5: Implementation and Feedback Collection of Multi-channel Intervention Strategies Based on the type of intelligent intervention strategy package, instructions are sent through a multi-channel execution system via mobile application push or manual task distribution, and the recipient is forced to upload the execution feedback results of the preset type, and the feedback information is associated, stored and a complete timeline is recorded. Step S6: Quantitative evaluation of intervention effects and iterative optimization of the model The feedback data is statistically analyzed for success rate and response rate according to strategy type. Counterfactual analysis is used to compare the differences in subsequent behavior between the intervention group and the control group to quantify the effect. An evaluation report is generated and the calculation model parameters and strategy matching rules are automatically adjusted accordingly. Step S7: Full-process data analysis and verification to ensure effectiveness evaluation By repeatedly executing the above steps, a complete data analysis closed loop is formed, from data collection to effect evaluation. This closed loop ensures that all conclusions have a traceable data chain and logical path, and clarifies the role of data analysis in risk identification, root cause diagnosis, precise intervention, and effect quantification by outputting a comprehensive verification report.

[0018] In this embodiment, a complete intelligent closed loop from data perception to decision optimization is constructed based on the collection of multi-source data. This closed loop uses standardized data records as a unified data foundation. First, a residential consistency index is generated through feature extraction and fusion calculation to achieve quantitative perception of abnormal risks. When an anomaly is detected, a causal inference engine is used to perform deep root cause diagnosis, generate a root cause analysis report to clarify the problem, and automatically match the diagnosis results to generate an intelligent intervention strategy package. A dual-channel execution system promotes the implementation of the strategy and forcibly collects the execution feedback results. Subsequently, the intervention effect is scientifically quantified and evaluated through counterfactual analysis, generating an intervention effect evaluation report and back-optimizing the feature calculation model and strategy matching rules. Finally, a self-iteratio verification closed loop is formed through cyclical execution, and a comprehensive verification report is output to fully trace and prove the guarantee role of data analysis in the whole process, ensuring that management decisions are always based on a traceable data evidence chain.

[0019] Example 2 Please see Figures 1 to 4 As shown in this embodiment, a data analysis-based affordable housing management method and system are described. The acquisition and standardization preprocessing of multi-source heterogeneous data includes: The system obtains application information, resident identity information, and housing information from the affordable housing management database; it also obtains authorized household registration information, social security records, marriage information, and real estate registration information from the government's data sharing platform; it acquires anonymized energy consumption data and door magnetic switch events from IoT devices; and it obtains market rent reference information from the housing rental information platform. All raw data acquired is cleaned to remove invalid records, desensitized to handle sensitive fields, standardized to unify data format and units, and finally linked and integrated to merge multi-source data of the same entity and assign a unique serial number and timestamp to generate standardized data records. Multimodal feature fusion-based residential behavior anomaly detection includes: Based on the standardized data records, electricity consumption characteristics are extracted, the total daily electricity consumption is calculated, and the deviation from the household's historical average electricity consumption is obtained. Extract water usage characteristics and calculate the standard deviation of water usage during the early morning period each day as the nighttime water usage volatility. Extract portal switch characteristics, record and count the number of daily switches, and calculate the multi-day coefficient of variation as a regularity indicator; Extract the density characteristics of anonymous devices and calculate the seven-day moving average of the number of daily active anonymous devices as the density trend. Extract the characteristics of market rent ratios and calculate the ratio of the latest listed rent to the rent of affordable housing; Input the electricity consumption characteristics, water consumption characteristics, gateway switch characteristics, anonymous device density characteristics, and market rent ratio characteristics into the calculation model; In-depth analysis of the root causes of abnormal behavior based on causal inference includes: Based on the resident information and warning trigger time in the preliminary warning report, historical characteristic data of abnormal residents were retrieved from standardized data records; similar normal residents were selected as a control group based on housing resources, family structure and occupancy duration; with the warning trigger time as the intervention point, the difference-in-differences method was used to calculate the contribution of each factor, such as changes in commuting distance, changes in work unit, changes in family income and fluctuations in market rent, to changes in residential behavior. A causal graph model is constructed based on the contribution of each factor, and the standardized weight of each factor is calculated. A root cause analysis report is generated, which lists the influencing factors in order of standardized weight and indicates that the one with the highest weight is the main source of the anomaly. The confidence interval of the contribution is also attached as a confidence assessment. The automatic generation and classification of intelligent intervention strategies include: Based on the standardized weights of each influencing factor in the root cause analysis report, key and secondary factors are identified, and the corresponding intervention strategy templates in the pre-set intervention strategy template library are matched by the type combination of key and secondary factors. The specific data from the root cause analysis report is used to fill the specified fields of the intervention strategy template, while generating a root cause summary containing descriptions of the main abnormal root causes and an evidence chain index related to the underlying data; so that the final filled intervention strategy template, root cause summary and evidence chain index will be packaged into an intelligent intervention strategy package with a unified format and type identifier; The implementation and feedback collection of multi-channel intervention strategies include: The execution channel is determined by parsing the strategy type identifier of the intelligent intervention strategy package; if the identifier corresponds to the proactive suggestion strategy, a notification message is pushed to the designated resident through the application interface; if the identifier corresponds to the strategy requiring manual intervention, a task order is created and assigned to the designated administrator in the background; after execution in any channel, the recipient is forced to select the preset type of execution feedback result on the designated interface and submit it. The execution feedback results are associated with and stored with the corresponding intelligent intervention strategy package, and the complete processing timeline from strategy generation to feedback reception is recorded. Quantitative evaluation of intervention effects and iterative optimization of the model include: The execution feedback results were categorized and summarized according to the associated strategy type, and the execution success rate and resident response rate of each strategy type were calculated. Using counterfactual analysis, residents who had received the smart intervention strategy package were set as the intervention group, and residents with similar conditions but who had not received the intervention were matched as the control group. The average change value of the residential consistency index of the two groups of residents within a specific time period after the intervention was calculated, and the difference was used as the average treatment effect to quantify the intervention effect. Generate an intervention effectiveness evaluation report that includes the success rate, response rate, and average treatment effect; based on the quantitative results in the intervention effectiveness evaluation report, automatically adjust the weight coefficients of each feature in the calculation model, and update the matching priority of factor combinations and strategy types in the intervention strategy template library; The full-process data analysis and verification ensures the effectiveness evaluation, including: Steps 1 through 6 are executed cyclically to form a data analysis closed loop, and intermediate data and reports for each cycle are associated and stored by assigning a unified cycle number. After completing a preset number of cycles, the preliminary early warning reports, root cause analysis reports, intelligent intervention strategy package processing results, and intervention effect evaluation reports for each cycle are summarized; quantitative analysis is performed based on the summarized data to evaluate and clarify the role of data analysis in ensuring the accuracy of risk identification, the accuracy of root cause diagnosis, the effectiveness of precise intervention, and the sustainability of quantitative effects; a comprehensive verification report is generated and published.

[0020] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0021] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only one method, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0022] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

[0023] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A data analysis-based method and system for managing affordable housing, characterized in that, The data analysis-based affordable housing management method and system includes: Step 1: Acquisition and Standardization Preprocessing of Multi-Source Heterogeneous Data Through the data acquisition module, multi-source heterogeneous raw data from affordable housing databases, government platforms, IoT devices and rental platforms are collected. The data is then cleaned, de-identified, standardized and integrated to generate standardized data records with unified identifiers and timestamps. Step 2: Anomaly Detection of Residential Behavior Based on Multimodal Feature Fusion Based on the generated standardized data records, multidimensional features such as electricity consumption, water consumption, door switch, density of anonymous equipment, and market rent ratio are extracted. A residential consistency index is generated through a calculation model. When the index is continuously lower than the dynamic threshold, a preliminary warning report is generated and the resident is marked as pending verification. Step 3: In-depth analysis of the root causes of abnormal behavior based on causal inference The early warning information is input into the causal inference engine. Using similar normal households as the control group, the difference-in-differences analysis method is used to calculate the contribution of each factor. A causal graph model is constructed to quantify the weights and generate a root cause analysis report that clearly points out the main abnormal root causes and their credibility. Step 4: Automatic generation and classification of intelligent intervention strategies Based on the factor combinations and weights in the root cause analysis report, the system automatically matches a preset intervention strategy template and generates an intelligent intervention strategy package with a root cause summary and evidence chain index. Step 5: Implementation and Feedback Collection of Multi-channel Intervention Strategies Based on the type of intelligent intervention strategy package, instructions are sent through a multi-channel execution system via mobile application push or manual task distribution, and the recipient is forced to upload the execution feedback results of the preset type, and the feedback information is associated, stored and a complete timeline is recorded. Step 6: Quantitative evaluation of intervention effects and iterative optimization of the model The feedback data is statistically analyzed for success rate and response rate according to strategy type. Counterfactual analysis is used to compare the differences in subsequent behavior between the intervention group and the control group to quantify the effect. An evaluation report is generated and the calculation model parameters and strategy matching rules are automatically adjusted accordingly. Step 7: Full-process data analysis and verification to ensure effectiveness evaluation By repeatedly executing the above steps, a complete data analysis closed loop is formed, from data collection to effect evaluation. This closed loop ensures that all conclusions have a traceable data chain and logical path, and clarifies the role of data analysis in risk identification, root cause diagnosis, precise intervention, and effect quantification by outputting a comprehensive verification report.

2. The data analysis-based affordable housing management method and system according to claim 1, characterized in that, The acquisition and standardization preprocessing of the multi-source heterogeneous data includes: The system obtains application information, resident identity information, and housing information from the affordable housing management database; it also obtains authorized household registration information, social security records, marriage information, and real estate registration information from the government's data sharing platform; it acquires anonymized energy consumption data and door magnetic switch events from IoT devices; and it obtains market rent reference information from the housing rental information platform. All raw data acquired is cleaned to remove invalid records, desensitized to handle sensitive fields, standardized to unify data format and units, and finally linked and integrated to merge multi-source data of the same entity and assign a unique serial number and timestamp to generate standardized data records.

3. The data analysis-based affordable housing management method and system according to claim 2, characterized in that, The multimodal feature fusion-based residential behavior anomaly detection includes: Based on the standardized data records, electricity consumption characteristics are extracted, the total daily electricity consumption is calculated, and the deviation from the household's historical average electricity consumption is obtained. Extract water usage characteristics and calculate the standard deviation of water usage during the early morning period each day as the nighttime water usage volatility. Extract portal switch characteristics, record and count the number of daily switches, and calculate the multi-day coefficient of variation as a regularity indicator; Extract the density characteristics of anonymous devices and calculate the seven-day moving average of the number of daily active anonymous devices as the density trend. Extract the characteristics of market rent ratios and calculate the ratio of the latest listed rent to the rent of affordable housing; Input the electricity consumption characteristics, water consumption characteristics, gateway switch characteristics, anonymous equipment density characteristics, and market rent ratio characteristics into the calculation model.

4. The data analysis-based affordable housing management method and system according to claim 3, characterized in that, The in-depth analysis of the root causes of abnormal behavior in the causal inference includes: Based on the resident information and warning trigger time in the preliminary warning report, historical characteristic data of abnormal residents were retrieved from standardized data records; similar normal residents were selected as a control group based on housing resources, family structure and occupancy duration; with the warning trigger time as the intervention point, the difference-in-differences method was used to calculate the contribution of each factor, such as changes in commuting distance, changes in work unit, changes in family income and fluctuations in market rent, to changes in residential behavior. A causal graph model is constructed based on the contribution of each factor, and the standardized weight of each factor is calculated. A root cause analysis report is generated, which lists the influencing factors in order of standardized weight and indicates that the one with the highest weight is the main root cause of the anomaly. At the same time, the confidence interval of the contribution is attached as a confidence assessment.

5. The data analysis-based affordable housing management method and system according to claim 4, characterized in that, The automatic generation and classification of the intelligent intervention strategies include: Based on the standardized weights of each influencing factor in the root cause analysis report, key and secondary factors are identified, and the corresponding intervention strategy templates in the pre-set intervention strategy template library are matched by the type combination of key and secondary factors. The specific data from the root cause analysis report is used to fill the specified fields of the intervention strategy template, while generating a root cause summary containing descriptions of the main abnormal root causes and an evidence chain index related to the underlying data. This results in the final filled intervention strategy template, root cause summary, and evidence chain index being packaged into an intelligent intervention strategy package with a unified format and type identifier.

6. The data analysis-based affordable housing management method and system according to claim 5, characterized in that, The implementation and feedback collection of the multi-channel intervention strategy include: The execution channel is determined by parsing the strategy type identifier of the intelligent intervention strategy package; if the identifier corresponds to the proactive suggestion strategy, a notification message is pushed to the designated resident through the application interface; if the identifier corresponds to the strategy requiring manual intervention, a task order is created and assigned to the designated administrator in the background; after execution in any channel, the recipient is forced to select the preset type of execution feedback result on the designated interface and submit it. The execution feedback results are associated with and stored with the corresponding intelligent intervention strategy package, and the complete processing timeline from strategy generation to feedback reception is recorded.

7. The data analysis-based affordable housing management method and system according to claim 6, characterized in that, The quantitative evaluation and iterative optimization of the intervention effect include: The execution feedback results were categorized and summarized according to the associated strategy type, and the execution success rate and resident response rate of each strategy type were calculated. Using counterfactual analysis, residents who had received the smart intervention strategy package were set as the intervention group, and residents with similar conditions but who had not received the intervention were matched as the control group. The average change value of the residential consistency index of the two groups of residents within a specific time period after the intervention was calculated, and the difference was used as the average treatment effect to quantify the intervention effect. Generate an intervention effect evaluation report that includes the success rate, response rate, and average treatment effect; based on the quantitative results in the intervention effect evaluation report, automatically adjust the weight coefficients of each feature in the calculation model, and update the matching priority of factor combinations and strategy types in the intervention strategy template library.

8. The data analysis-based affordable housing management method and system according to claim 7, characterized in that, The full-process data analysis and verification to ensure effectiveness evaluation includes: Steps 1 through 6 are executed cyclically to form a data analysis closed loop, and intermediate data and reports for each cycle are associated and stored by assigning a unified cycle number. After completing a preset number of cycles, the preliminary early warning reports, root cause analysis reports, intelligent intervention strategy package processing results, and intervention effect evaluation reports for each cycle are summarized; quantitative analysis is performed based on the summarized data to evaluate and clarify the role of data analysis in ensuring the accuracy of risk identification, the accuracy of root cause diagnosis, the effectiveness of precise intervention, and the sustainability of quantitative effects; a comprehensive verification report is generated and published.

Images