Search

CN-122022054-A - Accurate service touch method of interactive community range based on ROMS simulation result calculation

CN122022054ACN 122022054 ACN122022054 ACN 122022054ACN-122022054-A

Abstract

The invention provides an accurate service touch method of an interactive community range calculated based on a ROMS simulation result, which comprises the steps of obtaining multisource community data of a target user, inputting the multisource community data into a user behavior dynamic simulation model, predicting future service demands of the target user based on historical accumulated features, current behavior features and community environment data of the user behavior data by the model, determining target service strategies from candidate service strategies based on the predicted service demands through an optimization function, determining evaluation values of the candidate service strategies by the optimization function, determining the evaluation values based on matching degree of the candidate service strategies and the service demands and matching degree of the evaluation values with community environment data, executing the target service strategies, touching corresponding service contents to the target user through touch channels, collecting user feedback data, and carrying out iterative optimization on parameters of the user behavior dynamic simulation model based on differences between the user feedback data and the predicted service demands.

Inventors

  • CHEN CHEN
  • ZHANG JIANCHAO
  • WU XIANLI
  • YE JIANWEI
  • WENG YONGSHENG

Assignees

  • 中邮科通信技术股份有限公司

Dates

Publication Date
20260512
Application Date
20260205

Claims (10)

  1. 1. The community service touch method based on the behavior dynamic simulation is characterized by comprising the following steps of: The method comprises the steps of obtaining multi-source community data of a target user, wherein the multi-source community data comprise user behavior data and community environment data; inputting the multi-source community data into a user behavior dynamic simulation model, and predicting future service demands of a target user by the model based on historical accumulated characteristics, current behavior characteristics and the community environment data of the user behavior data; Determining a target service policy from candidate service policies by an optimization function based on the predicted service demand, wherein the optimization function is used for calculating an evaluation value of each candidate service policy, and the evaluation value is determined based on the matching degree of the candidate service policy and the service demand and the matching degree of the candidate service policy and community environment data; Executing the target service strategy, touching the corresponding service content to a target user through a touch channel, and collecting user feedback data; and carrying out iterative optimization on parameters of the dynamic simulation model of the user behavior based on the difference between the user feedback data and the predicted service demand.
  2. 2. The community service touch method based on dynamic behavior simulation of claim 1, wherein the method comprises the following steps: The user behavior data comprises basic data and service interaction data; The basic data comprise access control data, one-key door opening data and property payment data; The service interaction data comprise community convenience activity participation records, endowment or support service consultation records, third party service reservation records and service evaluation data; the community environmental data includes community street planning activities, seasons, holidays, and community development stages.
  3. 3. The community service touch method based on dynamic behavior simulation of claim 1, wherein the method comprises the following steps: The history accumulation feature is obtained by processing behavior data of a target user in a set history time period through a time sequence integration or sliding window statistical method; The user behavior dynamic simulation model is configured to receive as input a current behavior feature of a target user, community environment variables, and the historical accumulated features, and to output future service demands.
  4. 4. The community service touch method based on behavior dynamic simulation of claim 3, wherein the user behavior dynamic simulation model is constructed based on a recurrent neural network.
  5. 5. The community service touch method based on dynamic behavior simulation of claim 1, wherein the method comprises the following steps: the optimization function obtains an evaluation value by calculating the sum of the first weighting value and the second weighting value; The first weighted value is the product of the matching degree and a first weighted coefficient, and the second weighted value is the product of the matching degree and a second weighted coefficient; wherein the sum of the first weight coefficient and the second weight coefficient is 1.
  6. 6. The community service touch method based on dynamic behavior simulation of claim 1, wherein the method comprises the following steps: The reach channel comprises voice call and short message pushing; The triggering conditions for executing the target service strategy comprise that a target user triggers a preset behavior event; the preset behavior event comprises the steps of home returning of the gate inhibition data display, one-key gate opening data display and use, and the property payment data display is about to expire.
  7. 7. The community service touch method based on the behavior dynamic simulation of claim 1, wherein in the process of performing iterative optimization on parameters of a user behavior dynamic simulation model, model parameters are updated through a gradient optimization algorithm based on differences between user feedback data and predicted service requirements.
  8. 8. The community service touch method based on dynamic behavior simulation of claim 1, wherein the method comprises the following steps: The method also comprises the step of dynamically updating: When a target user triggers a preset updating condition, updating a target service strategy based on updated multi-source community data and the latest prediction result of the user behavior dynamic simulation model; the preset updating condition comprises that a target user generates new community behaviors, and the service demand change amplitude predicted by the model reaches a preset threshold value.
  9. 9. The method for community service reach based on dynamic behavior simulation of claim 1, wherein the user feedback data comprises service consultation record, activity participation record, service reservation record and satisfaction evaluation data.
  10. 10. A community service touch system, comprising: the data acquisition module is used for acquiring multi-source community data of a target user, wherein the multi-source community data comprises user behavior data and community environment data; The behavior dynamic simulation module is used for predicting future service demands of a target user based on the historical accumulated characteristics of the user behavior data, the current behavior characteristics and the community environment data through a user behavior dynamic simulation model; The policy decision module is used for determining a target service policy from candidate service policies through an optimization function based on the predicted service demand, wherein the optimization function is used for calculating an evaluation value of each candidate service policy, and the evaluation value is determined based on the matching degree of the candidate service policy and the service demand and the matching degree of the candidate service policy and community environment data; the touch feedback module is used for executing the target service strategy to touch the service content and collecting user feedback data; And the iterative optimization module is used for carrying out iterative optimization on the parameters of the dynamic user behavior simulation model based on the difference between the user feedback data and the predicted service demand.

Description

Accurate service touch method of interactive community range based on ROMS simulation result calculation Technical Field The invention belongs to the technical field of community digital service synergy, and particularly relates to an interactive community range accurate service touch method based on ROMS simulation result calculation. Background With the penetration of digital transformation of community management, community streets and operation subjects increasingly seek to improve service accuracy and sense of acquisition of residents through data driving. At present, although a technical support system of community service is provided with a primitive model, a series of technical bottlenecks still face in the aspects of data utilization, service adaptation and system evolution, and the continuous improvement of service efficiency is restricted. At the data level, the prior art schemes mostly stay in shallow applications for single-dimension or isolated traffic data. For example, it is common practice to make a prompt reminder based on property payment records, or to issue an activity notification based on simple community registration data. Although the data can reflect a certain side of resident behaviors, effective time sequence association and fusion analysis cannot be carried out on gate inhibition data reflecting resident physical access dynamics, high-frequency used 'one-key door opening' convenience service data and behavior track data participating in various online and offline activities. Because of the lack of integrated modeling of multidimensional and continuous behavior patterns of residents in communities, the system is difficult to construct a dynamic portrait capable of reflecting the real life state and potential demand change of individuals, so that service judgment stays in a static and one-sided level. In the service policy generation and touch link, the prior art generally has the problems of policy staticization and touch rough. Most community service platforms employ push logic based on fixed rules or simple tags, such as sending activity announcements to population at fixed times, or coarse group push based solely on demographic attributes (e.g., age). This model fails to dynamically correlate the generation of service policies with real-time behavioral contexts of the resident (e.g., post-home period, post-specific service usage), and lacks an intelligent matching mechanism between the reach channel (e.g., phone, SMS, APP notification) and specific service content, resident reach preferences. The result is malposition of touch time, channel discomfort, lack of individuation of content, low resident participation, and non-ideal input-output ratio of service resources. More importantly, the existing community service technology framework generally lacks a closed-loop optimization mechanism based on data feedback. The feedback of residents to services, whether explicit evaluation and complaint or implicit participation and consultation actions, are only archived as records in most systems, and cannot be systematically collected, quantized and converted into optimized signals for driving the iteration of the service model. This means that once deployed, the service policy cannot be automatically evaluated and revised by the actual operation data, and the system does not have self-evolution capability. Community operators thus have difficulty accumulating "data assets" and achieving a sustained "service synergy", often putting into periodic repeated investments without breaking through the effective ceilings. In summary, although the prior art locally realizes the online community service, there are significant technical breakpoints in three key links of deep fusion and dynamic characterization of multi-source heterogeneous behavior data, real-time service policy generation and accurate touch based on behavior context, and continuous self-optimization of service model driven by feedback data. The community operation main body needs a new technical scheme to solve the core pain points of scattered data, static service, low touch efficiency and lack of intelligent iteration, and truly realizes the upgrading of the data-driven community service operation mode. Disclosure of Invention Aiming at the defects and shortcomings in the prior art, the invention provides a community service touch method and system based on behavior dynamic simulation. The method constructs a dynamic user behavior simulation model by deeply fusing multisource community data such as access control, one-key door opening, property payment, service interaction and the like of a target user. The model is particularly focused on the historical accumulated characteristics of the user behaviors, and integrates the current behaviors and community environment data to predict future service demands. On the basis, a target service strategy is decided from candidate service strategies through an optimization function, and t