Geo-fence based real-time coexistence user social system and method

Abstract

The present application belongs to the technical field of geographic fence and real-time social, and discloses a real-time coexisting user social system and method based on geographic fence, which comprises the following steps: collecting multi-dimensional multi-modal original data of users through a dynamic sampling strategy and an energy efficiency balance mechanism, and generating a four-dimensional data set with information desensitization; constructing four layers of fences, evaluating the value attributes of each fence, dynamically adjusting the fence state through life cycle management, and forming a four-dimensional fence value network; hierarchically identifying basic events, composite events and collaborative events, constructing a real-time value flow through event value quantification, and generating a dynamic rule set adapted to the scene; distributing user permissions through a three-dimensional value permission matrix, completing score generation and transaction through a multi-dimensional value credit system, and forming a value circulation record containing permission state, score changes and transaction records; analyzing the value health degree through a three-level optimization mechanism of real-time, short-term and long-term, generating optimization parameters, and iteratively returning to the corresponding previous link.

Description

Technical Field

[0001] This invention relates to the fields of geofencing and real-time social technology, and more specifically, to a geofencing-based real-time coexistence user social system and method. Background Technology

[0002] With the development of mobile internet and location technology, the demand for real-time coexistence social networking based on geolocation has become prominent. Users expect accurate social matching in the same scene, scenario-based services and value monetization. However, existing systems generally have problems such as static design that is difficult to adapt to dynamic scenes, insufficient accuracy of social matching, and limited commercial conversion efficiency, which cannot fully meet the deep needs of users and the demands of scenario development.

[0003] Existing technologies still have some shortcomings: fence construction mostly adopts a static design with fixed boundaries, which is difficult to adapt to changes in scenarios; data collection strategies are rigid, resulting in an imbalance between power consumption and data accuracy, and privacy desensitization is difficult to balance usability; event recognition only focuses on single independent behaviors and has not formed a systematic hierarchical recognition system, making it impossible to capture user behavior links and group synergy effects; permission allocation lacks a multi-dimensional dynamic adaptation mechanism, the points system is disconnected from permissions, there is no value anchor, and the system parameters are fixed, which can easily lead to problems such as efficiency decay and experience decline in long-term use. 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 real-time coexistence user social system based on geofencing, comprising:

[0005] Data acquisition module: Collects multi-dimensional, multi-modal raw data from users through dynamic sampling strategies and energy efficiency balance mechanisms, and generates a four-dimensional data set with information anonymization;

[0006] Fence Construction and Value Assessment Module: Based on a four-dimensional dataset, a four-layer fence is constructed, consisting of physical, logical, social, and commercial layers. The value attributes of each fence are evaluated through an ecological niche value assessment mechanism, and the fence status is dynamically adjusted through lifecycle management to form a four-dimensional fence value network.

[0007] Event recognition and rule generation module: Based on a four-dimensional fence value network, it hierarchically identifies basic events, composite events and collaborative events, constructs a real-time value stream through event value quantification, and generates a dynamic rule set adapted to the scenario by combining a three-level rule generation system;

[0008] Permission allocation and points circulation module: Based on dynamic rule set, user permissions are allocated through three-dimensional value permission matrix, and a multi-dimensional value points system is linked to complete points generation and transactions, forming a value circulation record that includes permission status, points changes and transaction records.

[0009] Value Optimization and Reverse Iteration Module: Based on value circulation records, the module analyzes value health through a three-level optimization mechanism of real-time, short-term and long-term, generates optimization parameters, and iterates back to the corresponding preceding stage.

[0010] Furthermore, the generation method of the four-dimensional data set includes:

[0011] In the process of collecting multimodal raw data of users in four dimensions of space, behavior, time and value, the system combines the heat map data of people flow in the spatial dimension, executes a dynamic sampling strategy according to people flow density, and identifies the mobile and stationary states according to the changes in user location, and adapts the positioning method through an energy efficiency balance mechanism.

[0012] Sensitive information is desensitized from the collected multimodal raw data, and the data is associated with user ID and timestamp to generate a four-dimensional data set containing precise location, behavioral characteristics, time stamps, and value attributes.

[0013] Furthermore, the four-layer fence is constructed in the following ways:

[0014] Based on a four-dimensional dataset, and following a hierarchical structure of physical, logical, social, and commercial elements, a four-layered fence is constructed covering physical space, scenario logic, social relationships, and commercial links:

[0015] Physical fences are defined based on the spatial dimension of building structures and POI category labels, and the boundaries are dynamically calibrated by combining real-time location data.

[0016] Logical fences are generated and bound to lifecycles based on behavioral event logs and short-term trends at the behavioral dimension, triggered by events or predicted by demand.

[0017] Social fences are adapted to the intimacy and user relationship types of the value dimension, and are constructed in combination with the interactive scenarios of the behavioral dimension;

[0018] Commercial fences are stratified according to consumption potential in the value dimension, linking upstream and downstream of the value chain, and dynamically adjusting their scope based on user consumption behavior in the behavioral dimension.

[0019] Furthermore, the formation of the four-dimensional fence value network includes:

[0020] Based on a four-layer fence and a four-dimensional dataset, the value attributes of each fence are quantitatively evaluated according to hierarchical indicators through an ecological niche value assessment mechanism.

[0021] Furthermore, based on the value assessment results of each fence, the fence status is dynamically adjusted through the entire lifecycle management of creation, activation, optimization, and destruction;

[0022] Furthermore, by binding physical fences and commercial fences according to spatial overlap and logical fences and social fences according to value complementarity, a cross-fence collaborative relationship is constructed, forming a four-dimensional fence value network that includes fence boundaries, fence value attributes, and collaborative relationships.

[0023] Furthermore, the methods for identifying basic events, composite events, and collaborative events include:

[0024] Based on a four-dimensional fence value network and a four-dimensional dataset, events are identified in three levels:

[0025] Based on the value attributes of the fence, a scenario-based threshold is set to capture a single behavior and associate it with the fence tag to form a basic event;

[0026] By linking basic events through temporal continuity and logical coherence, and matching typical behavioral chains to form composite events;

[0027] By identifying collaborative behaviors among multiple individuals, classifying them according to group behavioral characteristics, and assessing their value levels, collaborative events are formed.

[0028] The three types of events are integrated and associated with fence attributes and value levels to form an event set.

[0029] Furthermore, the generation method of the dynamic rule set includes:

[0030] Based on event sets, combined with a four-dimensional fence value network and a four-dimensional data set, the value of a single event is quantified according to event type and fence attributes. Using user behavior trajectory as the axis, the three levels of event values ​​are connected according to time sequence and fence collaboration relationship to construct a real-time value stream and monitor health.

[0031] The rule base system is based on a three-tier architecture of basic rules, dynamic rules, and evolutionary rules. It matches scenarios and rules according to scenario tags, value stream health, and fence type, and sorts them by priority to form a dynamic rule set adapted to the scenario.

[0032] Furthermore, the method of allocating user permissions through a three-dimensional value permission matrix includes:

[0033] Based on dynamic rule sets, a three-dimensional value permission matrix is ​​constructed with relational, spatial, and temporal dimensions as the three axes, defining dimensional hierarchy, permission level, and operational scope.

[0034] Extract the user's real-time status in three dimensions, match initial permissions from the three-dimensional value permission matrix, resolve permission conflicts in the order of scenario priority > relationship strength > time value, and generate three-dimensional value permission credentials.

[0035] Furthermore, the formation of the value circulation record includes:

[0036] Based on three-dimensional value authorization credentials, a multi-dimensional value points system is constructed by linking dynamic rule sets, event sets, and value stream data:

[0037] Basic points are generated based on permission levels, and reward points are distributed based on event value and scenario weight; scenario-based transactions are executed within the scope of permissions, and risk control is carried out simultaneously.

[0038] Integrate permission status, points changes, and transaction details to form a value circulation record.

[0039] Furthermore, the method of generating optimized parameters and iterating backwards to the corresponding preceding steps includes:

[0040] Based on value circulation records, a three-level optimization mechanism is constructed:

[0041] The real-time mechanism monitors sudden anomalies in value circulation and corrects them immediately by temporarily optimizing parameters; the short-term mechanism analyzes the trend changes in value circulation and generates periodic optimization parameters to adapt to different scenarios.

[0042] The long-term mechanism is based on mining structural problems in value circulation from long-term historical data, generating systematic optimization parameters to upgrade the adaptability of the entire chain;

[0043] The optimization parameters are iterated backwards to the preceding steps according to the problem attribution, forming a closed loop for the entire process.

[0044] Furthermore, the geofencing-based real-time coexistence user social method is characterized by including:

[0045] S1: Collect multi-dimensional, multimodal raw data from users through dynamic sampling strategies and energy efficiency balance mechanisms to generate a four-dimensional data set with desensitized information;

[0046] S2: Based on a four-dimensional data set, construct four layers of fences: physical, logical, social, and commercial. Evaluate the value attributes of each fence through an ecological niche value assessment mechanism, and dynamically adjust the fence status through life cycle management to form a four-dimensional fence value network.

[0047] S3: Based on a four-dimensional fence value network, it hierarchically identifies basic events, composite events, and collaborative events. It constructs a real-time value stream through event value quantification and generates a dynamic rule set adapted to the scenario by combining a three-level rule generation system.

[0048] S4: Based on a dynamic rule set, user permissions are allocated through a three-dimensional value permission matrix, and a multi-dimensional value points system is linked to complete the generation and transaction of points, forming a value circulation record that includes permission status, points changes and transaction records.

[0049] S5: Based on value circulation records, analyze value health through a three-level optimization mechanism of real-time, short-term and long-term, generate optimization parameters, and iterate backward to the corresponding preceding stage.

[0050] The technical effects and advantages of the geofencing-based real-time coexistence user social system and method of this invention are as follows:

[0051] This invention focuses on the core scenario of real-time coexistence social interaction. Through a full-link modular closed-loop design, it connects data, fences, events, permissions, and points, significantly improving the accuracy of social matching, scenario adaptability, and commercial conversion efficiency while ensuring privacy and security.

[0052] First, by combining dynamic sampling strategies with energy efficiency balancing mechanisms, power consumption is reduced while ensuring data accuracy. Through scientific desensitization, data can be used but not seen, balancing privacy and value.

[0053] Secondly, we construct a four-dimensional dynamic fence system encompassing physical, logical, social, and commercial dimensions, combining value assessment and full lifecycle management to adapt to real-time changes in scenarios and solve the problem of rigid static fences.

[0054] Then, by capturing individual single behaviors, behavioral links, and group collaborative behaviors in a hierarchical manner, and combining real-time value stream analysis with a three-level rule generation system, the service can be upgraded from passive response to proactive adaptation.

[0055] Next, a dynamic permission matrix is ​​constructed using the three dimensions of relationship, space, and time, which deeply binds permission control with a multi-dimensional points system to ensure accurate and secure permissions, and incentivizes user interaction and consumption through the points mechanism;

[0056] Finally, based on the value circulation record, real-time anomaly correction, short-term trend adaptation and long-term system upgrade are achieved, so that the system parameters can be continuously optimized and adjusted through data feedback, ensuring the stability and long-term adaptability of the whole process.

[0057] This invention, through a modular closed-loop design, comprehensively covers the core needs of real-time coexisting social interaction, and is applicable to diverse scenarios such as local life, business social interaction, and group activities, significantly improving the accuracy, security, efficiency, and user stickiness of social systems. Attached Figure Description

[0058] Figure 1 This is a schematic diagram of the real-time coexistence user social system based on geofencing according to the present invention.

[0059] Figure 2 This is a schematic diagram illustrating the process of event set formation in the geofence-based real-time coexistence user social system of the present invention;

[0060] Figure 3 This is a schematic diagram of the real-time coexistence user social method based on geofencing according to the present invention. Detailed Implementation

[0061] 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.

[0062] Example 1

[0063] Please see Figure 1 and Figure 2 As shown in this embodiment, the real-time coexistence user social system based on geofencing includes:

[0064] Data acquisition module: Collects multi-dimensional, multi-modal raw data from users through dynamic sampling strategies and energy efficiency balance mechanisms, and generates a four-dimensional data set with information anonymization;

[0065] Fence Construction and Value Assessment Module: Based on a four-dimensional dataset, a four-layer fence is constructed, consisting of physical, logical, social, and commercial layers. The value attributes of each fence are evaluated through an ecological niche value assessment mechanism, and the fence status is dynamically adjusted through lifecycle management to form a four-dimensional fence value network.

[0066] Event recognition and rule generation module: Based on a four-dimensional fence value network, it hierarchically identifies basic events, composite events and collaborative events, constructs a real-time value stream through event value quantification, and generates a dynamic rule set adapted to the scenario by combining a three-level rule generation system;

[0067] Permission allocation and points circulation module: Based on dynamic rule set, user permissions are allocated through three-dimensional value permission matrix, and a multi-dimensional value points system is linked to complete points generation and transactions, forming a value circulation record that includes permission status, points changes and transaction records.

[0068] Value Optimization and Reverse Iteration Module: Based on value circulation records, the module analyzes value health through a three-level optimization mechanism of real-time, short-term and long-term, generates optimization parameters, and iterates back to the corresponding preceding stage.

[0069] The methods for generating four-dimensional datasets include:

[0070] The system collects multimodal raw data across four dimensions: user space, behavior, time, and value. During the collection process, it combines spatial dimension heatmap data to execute a dynamic sampling strategy. Based on changes in user location, it adapts the positioning method through an energy efficiency balancing mechanism. Specifically:

[0071] Spatial Dimension Data Acquisition: User location data is acquired through fusion positioning technologies such as UWB (Ultra-Wideband), Bluetooth beacons, and WiFi fingerprints; building CAD structural drawings, POI category tags (such as "dining area" and "office area"), and real-time pedestrian flow heat maps are collected simultaneously to form spatial semantic data; environmental perception data is collected through light, temperature, humidity, and sound sensors to achieve multi-dimensional characterization of the spatial environment.

[0072] Dynamic sampling strategy: Real-time reading of spatial heatmap data on pedestrian flow. When the pedestrian density in a region exceeds a preset density threshold (e.g., >200 people / 100㎡), the sampling frequency is reduced (e.g., from 1Hz to 0.33Hz (3 seconds / time)), and the behavioral data is compressed using the LZ4 compression algorithm (compression rate 60%, retaining core fields). When the pedestrian density is ≤ density threshold, full sampling is restored to ensure data accuracy.

[0073] Energy efficiency balance mechanism: The user terminal's built-in status monitoring module identifies the user's moving and stationary states. When moving, GPS+UWB positioning is prioritized to ensure accuracy; when stationary, GPS is automatically turned off, and only Bluetooth and WiFi positioning are retained to reduce power consumption; when the terminal's battery is low, the data sampling frequency is uniformly reduced to the minimum regardless of the crowd density.

[0074] Behavioral data collection: Individual behavioral data such as movement trajectory, dwell time, interaction operations, and consumption actions (e.g., scanning QR codes, payment) are recorded through user terminal motion sensors; group clustering is identified using the DBSCAN clustering algorithm, and cluster density and duration are extracted; the direction of group flow is analyzed using a directed graph model; and interactive network topology and social graph (nodes are users, edge weight = interaction frequency) are constructed by combining value-dimensional social relationship chain data as group behavioral data; preset event triggering rules are used to capture event information in real time and generate behavioral event logs when the terminal crosses the fence boundary, switches POI type labels, or enters a sensitive area as real-time event data;

[0075] Time-based data collection: User location, behavioral status (moving and stationary), and current fence ID (ID refers to a unique identifier) ​​are updated at a second-level frequency to form real-time data; user dwell time (≥30 seconds), interaction frequency, and consumption amount are counted in a minute-level sliding window to generate trend curves as short-term trend data; at fixed times each day, short-term data is aggregated into historical behavioral data (including high-frequency POI category tags (≥5 occurrences), interaction preferences (text and voice ratio), consumption habits (average order value, repurchase cycle), etc.) through ETL tools to build a time-series database as long-term trend data;

[0076] Value Dimension Data Collection: User social relationship chains are acquired through in-app social modules, and relationship types are categorized by interaction frequency (≥10 monthly interactions for friends, 1-9 for weak relationships, and 0 for strangers). Intimacy is calculated (the weighted sum of interaction frequency, single interaction duration, and shared dwell time within a statistical period, compared to the weighted sum of the theoretical maximum values ​​of these three indicators within that period). Relationship type and intimacy are used as social relationship data. Merchant APIs are integrated to obtain user consumption intent (e.g., browsing history, favorited items), payment ability (historical average order value), and conversion probability (coupon redemption rate = redemption count ÷ coupon redemption count). Collaborative filtering algorithms are used to predict user preferences for different categories of businesses (range 0-1, e.g., "restaurant preference 0.8, retail preference 0.3"), which are then integrated into commercial value data.

[0077] Sensitive information anonymization: For location data of sensitive areas (such as residences and hospitals), K-anonymization (K=10) is used to blur the data to areas shared by more than 10 people, removing specific coordinates and retaining only area labels; User identity information (such as mobile phone numbers and real names) is pseudonymous (replaced with "user ID + random string"), and the association relationship is stored only through an internal mapping table; All anonymization operations follow the principle of data minimization, retaining only the feature fields necessary for subsequent analysis (such as "duration of stay" is retained, while "specific coordinates of stay" are blurred).

[0078] The anonymized data in the four dimensions of space, behavior, time, and value are linked by user ID and timestamp, and integrated to form a four-dimensional data set that includes precise location (precise coordinates + POI category label), behavioral characteristics (individual behavioral labels (such as "high-frequency interaction", "dining consumption") + group interaction probability), time stamp (time stamp + time period label)), and value attributes (intimacy + business preference score + social influence score (i.e., centrality in interaction network analysis theory)).

[0079] The construction methods for a four-layer fence include:

[0080] Based on a four-dimensional dataset, and following a hierarchical structure of physical, logical, social, and commercial elements, a four-layered fence is constructed covering physical space, scenario logic, social relationships, and commercial links:

[0081] Physical fences are defined based on the spatial dimension of the building structure (building CAD structural drawings (including wall and passageway coordinate boundaries)) and POI category labels, and the boundaries are dynamically calibrated using real-time location data. Specific construction rules are as follows:

[0082] Basic Boundaries: Based on the architectural CAD structural drawings, the initial boundaries are defined according to the physical space functions (e.g., the boundary of the "dining area on the east side of the 3rd floor" is the X1-Y1 to X2-Y2 coordinate range), and the functional attributes are clarified by associating POI category labels; combined with the user's real-time accurate coordinate positioning data, the boundaries are dynamically calibrated (e.g., the boundary of the passage area is extended outward by 0.5m due to frequent pedestrian traffic).

[0083] Logical fences are generated and bound to lifecycles based on behavioral event logs and short-term trends, triggered by events or predicted by demand. Specific construction rules are as follows:

[0084] Event-triggered type: When "≥5 cluster events in the same area per hour" occur, a temporary logical fence is automatically generated, with the boundary being the smallest bounding rectangle of the cluster area (the coordinate range is fitted according to the positioning data).

[0085] Demand-forecasting: Based on individual behavioral tags (such as "User A has stayed in the supermarket on the 1st floor 3 times in the past 7 days and then entered the restaurant on the 2nd floor"), cross-floor logical fences (such as "supermarket-restaurant linkage area") are generated to associate scenario needs.

[0086] Lifecycle binding: The existence duration of logical fences is bound to the triggering event and predicted demand (e.g., the fence will automatically disappear 30 minutes after the temporary gathering event ends, while the fence for long-term demand scenarios will be retained until the demand disappears).

[0087] Social fences are adapted to the value dimension of intimacy and user relationship type, and are constructed in conjunction with the behavioral dimension of interactive scenarios. Specific construction rules are as follows:

[0088] Relationship strength adaptation: For "friends" with intimacy ≥ 0.6, the social fence boundary is the union of the three POI areas where the two people have the highest frequency of co-occurrence; for "weak ties" with intimacy 0.3-0.6, the fence only includes the area of ​​the most recent co-interaction.

[0089] Interactive scenario binding: Based on the interactive network topology, high-frequency interactive users (such as group chat members) are included in the same social fence, and the boundary is the area where the group participates in the event (such as "the shopping mall area where 5 people successfully formed a group purchase"); the boundary range of the social fence is linked to the relationship type (the friend fence is fully visible, while the weak relationship fence only shows the basic location).

[0090] Commercial fences are stratified according to consumption potential in the value dimension, linking upstream and downstream of the value chain, and dynamically adjusted in scope based on user consumption behavior in the behavioral dimension. Specific construction rules are as follows:

[0091] Consumer potential segmentation: users are matched to corresponding business fences (e.g., "high preference dining area" includes restaurant POIs with a rating of ≥4.5) based on business type preference (≥0.7 is high preference) and conversion probability (≥0.6 is high conversion).

[0092] Value chain linkage: Automatically linking upstream and downstream business scenarios (e.g., binding "fresh food supermarket" fence to "kitchenware store" and "takeout pickup point" to form a consumer chain fence);

[0093] Dynamic adjustment: Based on user consumption behavior (e.g., a user's redemption rate in a shopping mall is less than 20% for 3 consecutive times), the commercial fence corresponding to the user is automatically narrowed (low-conversion POIs are removed).

[0094] The formation of a four-dimensional fence value network includes:

[0095] Based on a four-layer fence and a four-dimensional dataset, the value attributes of each fence are quantitatively evaluated according to hierarchical indicators through an ecological niche value assessment mechanism.

[0096] The specific method for quantifying the value of the four-layer fence using the niche value assessment mechanism is as follows:

[0097] Physical fences: Layered indicators include space utilization efficiency (average daily number of people staying within the fence ÷ fence area), pedestrian flow carrying capacity (peak pedestrian flow ÷ design capacity), and boundary adaptability (the overlap rate between the user's actual trajectory and the fence boundary). The layered indicators are weighted and fused to map a physical fence value score (the weighted fusion value is in the range of 0-1, which is mapped to a score range of 0-100 points, and ≥60 points are defined as high-value physical fences).

[0098] Logical fence value assessment: Layered indicators include scenario adaptability (matching rate between triggering events and fence functions), user engagement (user interaction within the fence ÷ consumption frequency), and lifecycle adaptability (overlap rate between fence duration and demand cycle); similarly, after weighted fusion, it is mapped to a logical fence value score (defined as ≥50 points for high-value logical fences).

[0099] Social fence value assessment: The tiered indicators include relationship activity (average daily number of user interactions within the fence ÷ number of users), network density (ratio of the number of interactive edges in the interactive network topology to the square of the total number of users), and social association rate (number of commercial behaviors after social interaction ÷ total number of interactions); similarly, after weighted fusion, it is mapped to the social fence value score (defined as ≥55 points as high-value social fence).

[0100] Commercial fence value assessment: The tiered indicators include conversion efficiency (repurchase rate within the fence ÷ coupon redemption rate), average order value contribution (average average order value within the fence ÷ overall average), and user repurchase rate (number of repurchases by users within the fence in 30 days); similarly, after weighted integration, they are mapped to a commercial fence value score (defined as ≥65 points for high-value commercial fences).

[0101] Furthermore, based on the value assessment results of each fence, the fence status is dynamically adjusted through the entire lifecycle management of creation, activation, optimization, and destruction;

[0102] The creation phase involves formally creating a four-layer fence based on the fence construction results.

[0103] Activation phase: Fences with a value score of ≥60 will automatically activate all functions; fences with a score <60 will only activate basic functions (such as only displaying the list of POIs within the fence).

[0104] Optimization phase: If the efficiency of physical fence usage decreases (e.g., <2 people / ㎡), the boundary is reduced by a preset ratio (e.g., 20%); if the conversion efficiency of commercial fences is <preset conversion threshold (e.g., 30%), low-conversion POI category tags are removed; if the relationship activity of social fences decreases (e.g., <0.3), they are merged into nearby high-value social fences.

[0105] Extinction Phase: Logical fences automatically disappear 30 minutes after the event is triggered; commercial fences are automatically deactivated if their score is less than 40 points for 7 consecutive days; social fences are automatically disbanded if users have no interaction for 30 days.

[0106] Construct cross-fence collaborative relationships according to collaborative rules: preset spatial overlap ≥ 50% and value complementarity ≥ 0.7 as collaborative conditions, bind physical fences and commercial fences according to spatial overlap (e.g., "3rd floor physical fence" is associated with "3rd floor catering commercial fence"), and bind logical fences and social fences according to scenario complementarity (e.g., "group buying logical fence" is associated with "friends social fence").

[0107] Among them, the spatial overlap degree = the area of ​​the overlapping area of ​​the two fences ÷ the total area of ​​the smaller fence of the two fences, and the ratio should be ≤1 to avoid evaluation deviation due to the difference in fence size.

[0108] Value complementarity: Based on the hierarchical indicators of fence value assessment, the functional difference score (the lower the overlap of indicator types, the higher the score) and synergistic gain score (the degree of positive impact of one party's indicator improvement on the other party) of the two fences are quantified according to the preset assessment logic. After weighted fusion, they are normalized to 0-1 (each with a weight of 0.5); Example: Group buying logic fence and friend social fence, functional difference score = 0.9 (no overlap of indicators), synergistic gain score = 0.6 (social relationship can increase group buying participation), then value complementarity = (0.9 + 0.6) × 0.5 = 0.75;

[0109] Based on the value of each fence and the collaborative relationships across fences, a four-dimensional fence value network is formed, which includes fence boundaries, fence value attributes, and collaborative relationships.

[0110] The methods for identifying basic events, composite events, and collaborative events include:

[0111] Based on a four-dimensional fence value network and a four-dimensional dataset, events are identified in three levels:

[0112] Basic event identification: Basic events are independent, quantifiable single behaviors (such as dwell time, interaction, and consumption) triggered by users within a single fence. Valid single behaviors are defined by setting scenario-based thresholds based on the fence's value attributes (e.g., dwelling for ≥20 seconds within a high-value commercial fence, initiating one text or voice interaction within a social fence, or crossing boundaries and dwelling for ≥5 seconds within a physical fence). Basic events are formed by binding timestamps and precise location data to the corresponding fence and adding multi-dimensional tags (behavior type (dwell time, interaction, consumption), fence type, and value level (high, medium, low, determined by threshold segmentation based on the associated fence value score)).

[0113] Composite event identification: Connecting basic events that satisfy temporal continuity (interval between adjacent basic events ≤ 10 minutes) and logical correlation (events related by collaborative fence relationships or scene links); matching preset typical behavioral links using sequence pattern mining algorithms (such as PrefixSpan), including social (e.g., "matching event → interaction event → friend adding event"), commercial, and scene switching events; it should be noted that the symbol → indicates a logical relationship; scoring based on the completeness of basic events (e.g., a complete sequence of "receiving coupon → staying → redeeming" gets 100 points, while the missing "redeeming" gets 60 points), a score ≥ 80 is considered a valid composite event;

[0114] Collaborative event identification: Groups are identified using the DBSCAN clustering algorithm (cluster radius ≤ 10 meters, minimum cluster size ≥ 3 people), requiring ≥ 50% of users within the group to trigger similar basic events; group behavior characteristics are categorized into clustering, collaborative, and mobility (multiple people moving synchronously from fence A to fence B, with mobility percentage ≥ 30%); the value level of collaborative events is assessed (high, medium, low) by combining group size (based on expert experience and pre-set evaluation rules, such as ≥ 10 people receiving a high score) and associated fence value (similarly pre-set evaluation rules, such as collaborative events within high-value fences receiving a high score), with high-value events triggering subsequent collaborative responses first;

[0115] The three types of events are integrated and associated with fence attributes and value levels to form an event set.

[0116] The methods for generating dynamic rule sets include:

[0117] Based on the event type and associated fence value attributes of the event set, each event is assigned an immediate value score (ranging from 0 to 10 points) as its individual event value. Specifically:

[0118] Basic events are calculated as the value of a single action multiplied by the fence value coefficient (e.g., a single "consumption event" in a commercial fence has a value of 8 points, a fence value score of 74 points corresponds to a coefficient of 0.9, and the immediate value score = 8 × 0.9 = 7.2 points); Composite events are calculated as the link integrity multiplied by the synergy gain coefficient; Synergistic events are calculated as the scale of participation multiplied by the network effect coefficient.

[0119] Among them, the initial base values ​​of the indicators used to calculate the real-time value score are preset by expert experience to avoid the inability to implement the system when there is no data support in the early stage. The data of the previous 24 hours are statistically analyzed by ETL tools every day and dynamically adjusted according to the value relevance.

[0120] Example of setting initial base values ​​for indicators:

[0121] Value of a single behavior: Preset according to the priority of commercial value (consumption behavior > interaction behavior > dwell behavior), with the initial value set as "consumption = 8 points, interaction = 4 points, dwell = 2 points";

[0122] Link integrity: Preset according to the degree of closure of basic events in the composite event link (complete link = 10 points, missing one basic event = 6 points, only one basic event = 3 points), such as the complete link of "receive coupon → redeem" gets 10 points;

[0123] Participation scale: Preset according to the group effect threshold (10 or more people = 9 points, 5-10 people = 6 points, 3-5 people = 3 points);

[0124] The adjustment logic in the example is as follows: If the redemption rate after a dwell time is found to be higher than expected (e.g., >30%), the single value of "dwell time behavior" will be increased from 2 points to 3 points; if the conversion efficiency of the complete link "coupon redemption → dwell time → redemption" (60%) is much higher than "coupon redemption → redemption" (40%), the integrity score of the complete link will be increased from 10 points to 12 points (still normalized, without exceeding the scoring range); if the redemption rate of points for collaborative events with more than 10 people (50%) is lower than that of groups of 5-10 people (70%), the participation scale score for groups of more than 10 people will be reduced to 8 points to avoid misjudging events with large scale but low value.

[0125] The initial base values ​​of each coefficient used to calculate the instant value score are preset by humans (bound to the previous evaluation results, not assigned out of thin air); the purpose is to associate with existing fence value network, collaborative relationship and other data to ensure that the coefficients are consistent with the previous process; based on real-time data feedback iteration, when the rule execution effect (such as redemption rate, user activity) corresponding to a certain type of coefficient is lower than the threshold (such as <30%) for 7 consecutive days, coefficient optimization is triggered.

[0126] Fence value coefficient: Based on the fence value score (≥65 points = 0.9, 50-64 points = 0.7, <50 points = 0.5), initially mapped directly according to the score range (no additional subjective judgment required);

[0127] Synergy gain coefficient: Based on the existing synergy judgment results, the coefficient is assigned by binding the cross-fence synergy relationship (spatial overlap ≥ 50% and value complementarity ≥ 0.7 = 0.8, single condition compliance = 0.5, no synergy = 0.3).

[0128] Network effect coefficient: The density of people in the interactive network topology (≥0.8=0.9, 0.5-0.7=0.7, <0.5=0.5), mapped according to prior social relationship data;

[0129] Example of coefficient optimization logic:

[0130] If the event redemption rate (25%) within the high-value commercial fence (coefficient 0.9) is lower than that within the medium-value fence (coefficient 0.7, redemption rate 40%), then the coefficient of the high-value fence will be lowered to 0.8, and the coefficient of the medium-value fence will be raised to 0.85; if the composite event conversion efficiency (55%) corresponding to the gain coefficient (0.8) of the social-commercial collaboration fence is higher than that of the logical-social collaboration fence (coefficient 0.8, conversion efficiency 35%), then the gain coefficient of the logical-social collaboration fence will be raised to 0.9 to strengthen the high-value collaboration link;

[0131] Using user ID and timeline as the core axis, and linking the value of continuous events of the same user in chronological order and in coordination with fences, a real-time value stream is constructed to achieve value superposition. Specifically:

[0132] The value of consecutive events within 10 minutes is accumulated (e.g., 3 points for a stay event + 7.2 points for a consumption event, totaling 10.2 points); the value of events across collaborative fences is additionally weighted (e.g., if a social fence event triggers a related commercial fence event, the total value is increased by an additional 20%); forming a real-time value stream ledger of user ID-time-event chain-cumulative value;

[0133] The method for monitoring the health of the value stream is as follows: preset the value stream health threshold (if a user's cumulative value in 1 hour is ≥10 points, the health is judged as high activity; 3-10 points is medium activity; <3 points is low activity), and the health status is marked in real time. High-activity users trigger priority response, and low-activity users trigger wake-up strategy.

[0134] A rule base system with a three-tiered architecture of pre-defined basic rules, dynamic rules, and evolutionary rules;

[0135] Among them, the basic rule library covers standardized scenarios. The operation logic is as follows: the built-in basic rule template (such as "high-value consumption event triggers double points" in business scenarios and "intimacy ≥ 0.6 triggers friend permission opening" in social scenarios) matches the corresponding rules according to the event tags (such as "business-high value" and "social-high activity"). The rule parameters are fixed (such as the points doubling ratio is 2 times).

[0136] Dynamic rule base: Adjust rule parameters or add scenario-based rules based on real-time value stream and high-frequency events; Operation logic: The business rule points doubling ratio for highly active users (value stream health is high activity) is increased from 2x to 2.5x, and the wake-up rule of "give a small amount of points for staying for 5 minutes" is triggered for low-activity users; When there are ≥5 group purchase events in a certain area within 1 hour, a temporary rule of "add 50% extra points for group purchases with ≥10 people" is automatically added, which will automatically expire after the event frequency decreases;

[0137] Evolutionary rule base: Daily analysis of rule execution performance over the previous 24 hours using machine learning algorithms (such as decision trees), solidifying rules with a conversion rate of ≥60% into the evolutionary rule base, and replacing inefficient rules in the basic rule base;

[0138] Based on a three-level rule base, candidate rules are first selected from the rule base according to the scene tags associated with the event (such as "catering business scene"). Then, the parameters are adjusted in combination with the user value stream health. Finally, cross-fence rules are added according to the fence collaboration relationship of the corresponding fence type.

[0139] And prioritize them: use the value weight method to sort (high value event rules priority > low value events, real-time rules > basic rules);

[0140] Integrate and generate a structured dynamic rule set adapted to the scenario. Each rule contains "rule ID, adapted scenario, execution conditions, operation instructions, priority, and validity period". For example: R2024111301-catering business scenario-consumption ≥ 100 yuan and high value flow health-points doubled by 2.5 times-priority 10-validity period 24 hours.

[0141] The methods for allocating user permissions through a three-dimensional value-based permission matrix include:

[0142] Based on dynamic rule sets, a three-dimensional value permission matrix is ​​constructed with relational, spatial, and temporal dimensions as the three axes, defining dimensional hierarchy, permission level, and operational scope.

[0143] Specifically, dimensional hierarchy and permission level mapping (a standardized framework that can be directly implemented):

[0144] Relationship Dimension Grading Standard: Based on social relationship data, Level 4 (highest level, intimacy ≥ 0.8, friends), Level 3 (intimacy 0.5-0.8, friends or high-frequency weak ties), Level 2 (intimacy 0.2-0.5, weak ties), and Level 1 (lowest, intimacy < 0.2, strangers or sensitive relationships); Permission Mapping: Level 4 grants full data viewing and collaborative operations (e.g., friends viewing real-time locations, jointly initiating group purchases), while Level 1 only grants basic POI information viewing (e.g., strangers can only see public facility tags within the fenced area);

[0145] Spatial Dimension Classification Standards: Based on fence type and value score, Level 4 (high-value commercial fence or non-sensitive area, score ≥ 65 points), Level 3 (medium-value collaborative fence, score 50-64 points), Level 2 (low-value physical fence, score < 50 points), and Level 1 (sensitive area, such as private offices or hospital clinics); Permission Mapping: Level 4 grants both consumption and points redemption permissions (e.g., direct coupon redemption within high-value restaurant fences), while Level 1 only grants location reporting permissions (e.g., submitting a reason for visit is required to enter a hospital clinic).

[0146] Time-based hierarchical standards: Binding time tags to event value, Level 3 (peak hours or high-value event periods, such as 11:30-13:30 peak dining hours, group buying events), Level 2 (off-peak hours or medium-value event periods), Level 1 (low-peak hours or low-value event periods); Permission mapping: Level 3 grants priority response permissions (such as priority redemption of points and priority matching of service resources during peak hours), while Level 1 only grants basic operation permissions (such as points redemption limits during low-peak hours).

[0147] Matrix integration: The three-dimensional hierarchical cross mapping forms a three-dimensional value permission matrix of "relationship × space × time". Each intersection point corresponds to a clear permission level (level 1-5, level 5 is the highest) and operation scope. For example, "level 4 + level 4 + level 3" corresponds to level 5 permission (full commercial + social operation), and "level 1 + level 1 + level 1" corresponds to level 1 permission (basic reporting only).

[0148] Using user ID as the core, and associating it with their real-time status data across three dimensions, the corresponding permissions are extracted from the three-dimensional value-permission matrix as initial permissions. Specifically:

[0149] Extract the user's real-time relationship status (e.g., intimacy with the current interactive object is 0.7, corresponding to relationship level 3), the fence status corresponding to the real-time location (e.g., in the high-value restaurant fence, corresponding to space level 4), and the real-time time tag (e.g., 12:00 restaurant peak, corresponding to time level 3) from the four-dimensional dataset.

[0150] Locate the intersection points of the corresponding levels in the three-dimensional value permission matrix, match the corresponding level permissions, and use them as initial permissions; compare with the scenario priority of the dynamic rule set (such as "dining scenario has higher priority than ordinary social scenario"), supplement permission details (such as unlimited points redemption in the dining scenario), and generate the user's initial permission list.

[0151] A permission conflict is defined as a conflict arising from the overlap of multiple states or multiple scenarios of the same user at the same time (e.g., a user is simultaneously in a Level 4 high-value commercial fence and a Level 1 sensitive area, resulting in conflicting permission requirements).

[0152] Conflicts are resolved first based on scenario priority > relationship strength > time value: Sorted by scenario priority in dynamic rule set (commercial scenario > social scenario > public scenario > sensitive scenario). For example, if a user initiates commercial verification (space level 4 scenario) in a hospital clinic (space level 1 sensitive scenario), sensitive scenario permissions will be applied first (only the reporting permission will be retained, and the verification will be suspended).

[0153] Then, conflicts are resolved based on the strength of the relationship: when multiple users interact in the same space, the decision is made according to the relationship level. For example, if a stranger (relationship level 1) and a friend (relationship level 4) apply to view the data within the fence at the same time, only the friend's permission is granted.

[0154] Finally, conflicts are resolved based on time value: when peak and off-peak periods overlap, the higher timeliness is used to override the lower timeliness. For example, if a user initiates a points redemption at the end of a peak period (transitioning from time level 3 to time level 2), it will still be executed according to time level 3 permissions.

[0155] The system integrates and generates a structured three-dimensional value-based authorization certificate, which includes user ID, authorization level, three-dimensional status, scope of operation, and conflict resolution records.

[0156] The ways in which value circulation records are formed include:

[0157] Based on three-dimensional value authorization credentials, a multi-dimensional value points system is constructed by linking dynamic rule sets, event sets, and value stream data, including points generation and points trading:

[0158] Points generation: Basic points are generated according to permission level, and reward points are issued in combination with event value and scenario weight. User behavior value and permission rights are converted into tradable points, ensuring that points generation is strongly correlated with the value assessment of previous steps.

[0159] Basic points (used to ensure basic access rights): A basic points pool is set according to the access level (Level 1 = 100 points / day...Level 5 = 500 points / day), and points are regularly distributed to the points account every day (for circulation only); points are deducted in real time when the access level changes.

[0160] Reward points (used for behavioral value incentives): Issued based on the value level of three-tiered events. Basic events are redeemed based on immediate value points (e.g., 1 immediate value point can be redeemed for 10 reward points), composite events are redeemed based on completeness scores, and collaborative events are redeemed based on the scale of participation. High-value scenarios receive additional weighting (e.g., peak-hour restaurant redemption, group buying with friends), and reward points are additionally added (e.g., peak-hour consumption event points × 1.5). Reward points are issued after the event is completed.

[0161] For the generated basic points and reward points, check that "the total number of points is less than or equal to the daily limit of the scenario points pool" (e.g., the daily points pool for commercial scenarios is 100,000 points). If the limit is exceeded, the points will be distributed according to the priority of the events (high-value events will be given priority).

[0162] Points trading (the core guarantee of circulation efficiency) is to execute scenario-based transactions within the scope of authority and simultaneously control risks. The core logic is to define transaction scenarios based on the scope of authorized operations, set boundaries according to rules, ensure security through risk control, and improve efficiency through collaboration in executing transactions. It needs to cover three core scenarios: exchange, transfer, and collaboration.

[0163] Specifically, the scenarios and operations for points trading are as follows:

[0164] Commercial redemption transactions (core scenario): Users can only redeem goods and services within their authorized operation scope (e.g., Level 4 users can redeem high-value restaurant vouchers, while Level 1 users can only redeem public facility usage vouchers); the redemption rules are set according to a dynamic rule set to determine the points redemption ratio (e.g., 100 points = 1 yuan, and the redemption ratio for high-value scenarios is increased to 80 points = 1 yuan). Points are automatically deducted and a verification code is generated during redemption.

[0165] Social transfer transactions (relationship linkage scenario): Only supports transfers between users with relationship levels 2 to 4 (level 1 strangers are prohibited), and the transfer limit is graded according to the intimacy level (e.g.: level 4 has no limit, level 3 ≤ 5000 points per day).

[0166] Group Collaborative Transactions (Event Linkage Scenarios): Applicable scenarios are for 3 or more users participating in collaborative events (such as group buying or collaborative consumption). Users can initiate point-based order redemption (points can be used cumulatively). The rule is that when the total points of the combined order are greater than or equal to 1.2 times the total points rewarded by the collaborative event, an additional 10% points will be awarded. After redemption, the points will be deducted from each user's account proportionally.

[0167] Transaction risk control: Real-time monitoring of abnormal transactions (such as ≥10 exchanges in a single day, or a sudden increase of more than 5 times in transfer amount), triggering risk control verification (such as SMS verification, secondary authorization confirmation), and abnormal records are synchronously stored in the risk control log;

[0168] By integrating three types of information—permission status, points changes, and transaction details—a structured and traceable record of value circulation is formed, providing data support for subsequent value analysis and rule optimization.

[0169] The core fields of the value circulation record include circulation ID, user ID, permission status (level + operation scope), points change type (generation, transaction), change amount, associated event ID or fence ID, transaction object, and timestamp.

[0170] Methods for generating optimization parameters and iterating backwards to the corresponding preceding steps include:

[0171] Based on value circulation records, a three-level optimization mechanism is constructed:

[0172] The real-time mechanism monitors all data in the value circulation process and compares it with corresponding preset thresholds. If the threshold is not met (e.g., the transaction success rate drops sharply to 85%, failing to meet the >95% threshold), it is judged as a sudden anomaly. The root cause of the anomaly is located (e.g., weak signal from a commercial fence causing positioning deviation → permission matching error → transaction failure), and temporary optimization parameters are generated (e.g., temporarily increasing the positioning sampling frequency of the fence from 1Hz to 2Hz, relaxing the permission verification conditions of the area, etc.). The optimization parameters are pushed to the corresponding links in real time. After the anomaly is repaired, the monitoring is re-monitored, and if the target is met, the temporary parameters are fixed (to avoid frequent adjustments).

[0173] The short-term mechanism regularly aggregates data from the previous 24 hours daily and uses common trend analysis algorithms (such as time series ARIMA) to identify trends in value circulation (e.g., the redemption rate of the catering category during weekday lunch hours is only 30%). It delves into the root causes of these trends (e.g., the points redemption ratio during this period is 100 points = 1 yuan, which is lower than user expectations) and generates periodic optimization parameters (e.g., adjusting the redemption ratio of the catering category during this period to 80 points = 1 yuan, or increasing the priority of the rules for this scenario). These parameters are then pushed to the dynamic rule set and the category value network. If the parameters meet the criteria, they are solidified as weekly parameters; otherwise, they are optimized a second time.

[0174] The long-term mechanism uses long-term historical data (such as data from each quarter) and machine learning algorithms (such as decision trees and cluster analysis) to perform structural analysis (e.g., low-activity user retention rate is only 15%, because long-term holding of points is not attractive); it identifies the root cause of the system and explores structural problems in value circulation (e.g., no appreciation rate for long-term holding, no tiered appreciation strategy); it generates systematic optimization parameters (e.g., adjusting the appreciation rate to 5%, adding an additional 1% tiered interest rate for highly active users; optimizing the accuracy of the data collection and anonymization process for sensitive areas).

[0175] The optimization parameters are iterated backwards according to the problem attribution to the preceding stages (data collection, fence construction, rule generation, etc.), forming a closed loop of data collection, value circulation and optimization iteration, so as to enable the system to continuously adapt to the needs of the scenario.

[0176] Example 2

[0177] Please see Figure 3As shown, parts not described in detail in this embodiment are described in Embodiment 1. A real-time coexistence user social method based on geofencing is provided, including:

[0178] S1: Collect multi-dimensional, multimodal raw data from users through dynamic sampling strategies and energy efficiency balance mechanisms to generate a four-dimensional data set with desensitized information;

[0179] S2: Based on a four-dimensional data set, construct four layers of fences: physical, logical, social, and commercial. Evaluate the value attributes of each fence through an ecological niche value assessment mechanism, and dynamically adjust the fence status through life cycle management to form a four-dimensional fence value network.

[0180] S3: Based on a four-dimensional fence value network, it hierarchically identifies basic events, composite events, and collaborative events. It constructs a real-time value stream through event value quantification and generates a dynamic rule set adapted to the scenario by combining a three-level rule generation system.

[0181] S4: Based on a dynamic rule set, user permissions are allocated through a three-dimensional value permission matrix, and a multi-dimensional value points system is linked to complete the generation and transaction of points, forming a value circulation record that includes permission status, points changes and transaction records.

[0182] S5: Based on value circulation records, analyze value health through a three-level optimization mechanism of real-time, short-term and long-term, generate optimization parameters, and iterate backward to the corresponding preceding stage.

[0183] Example 3

[0184] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the geofence-based real-time coexistence user social system described above.

[0185] Since the electronic device described in this embodiment is the one used to implement the geofencing-based real-time coexistence user social networking method in this application embodiment, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the geofencing-based real-time coexistence user social networking method described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any electronic device used by those skilled in the art to implement the geofencing-based real-time coexistence user social networking method in this application embodiment falls within the scope of protection of this application.

[0186] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0187] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A real-time coexistence user social system based on geofencing, characterized in that, include: Data acquisition module: Collects multi-dimensional, multi-modal raw data from users through dynamic sampling strategies and energy efficiency balance mechanisms, and generates a four-dimensional data set with information anonymization; The four-dimensional data set is generated in the following ways: In the process of collecting multimodal raw data of users in four dimensions of space, behavior, time and value, the system combines the heat map data of people flow in the spatial dimension, executes a dynamic sampling strategy according to people flow density, and identifies the mobile and stationary states according to the changes in user location, and adapts the positioning method through an energy efficiency balance mechanism. Sensitive information is desensitized from the collected multimodal raw data, and the data is associated with user ID and timestamp to generate a four-dimensional data set containing precise location, behavioral characteristics, time stamps and value attributes. Fence Construction and Value Assessment Module: Based on a four-dimensional dataset, a four-layer fence is constructed, consisting of physical, logical, social, and commercial layers. The value attributes of each fence are evaluated through an ecological niche value assessment mechanism, and the fence status is dynamically adjusted through lifecycle management to form a four-dimensional fence value network. The four-layer fence is constructed in the following ways: Based on a four-dimensional dataset, and following a hierarchical structure of physical, logical, social, and commercial elements, a four-layered fence is constructed covering physical space, scenario logic, social relationships, and commercial links: Physical fences are defined based on the spatial dimension of building structures and POI category labels, and the boundaries are dynamically calibrated by combining real-time location data. Logical fences are generated and bound to lifecycles based on behavioral event logs and short-term trends at the behavioral dimension, triggered by events or predicted by demand. Social fences are adapted to the intimacy and user relationship types of the value dimension, and are constructed in combination with the interactive scenarios of the behavioral dimension; Commercial fences are stratified according to consumption potential in the value dimension, linking upstream and downstream of the value chain, and dynamically adjusting their scope based on user consumption behavior in the behavioral dimension. Event recognition and rule generation module: Based on a four-dimensional fence value network, it hierarchically identifies basic events, composite events and collaborative events, constructs a real-time value stream through event value quantification, and generates a dynamic rule set adapted to the scenario by combining a three-level rule generation system; The methods for identifying basic events, composite events, and collaborative events include: Based on a four-dimensional fence value network and a four-dimensional dataset, events are identified in three levels: Based on the value attributes of the fence, a scenario-based threshold is set to capture a single behavior and associate it with the fence tag to form a basic event; By linking basic events through temporal continuity and logical coherence, and matching typical behavioral chains to form composite events; By identifying collaborative behaviors among multiple individuals, classifying them according to group behavioral characteristics, and assessing their value levels, collaborative events are formed. Integrate the three types of events and associate them with fence attributes and value levels to form an event set; The dynamic rule set is generated in the following ways: Based on event sets, combined with a four-dimensional fence value network and a four-dimensional data set, the value of a single event is quantified according to event type and fence attributes. Using user behavior trajectory as the axis, the three levels of event values ​​are connected according to time sequence and fence collaboration relationship to construct a real-time value stream and monitor health. The rule base system is based on a three-tier architecture of basic rules, dynamic rules, and evolutionary rules. It matches scenarios and rules according to scenario tags, value stream health, and fence type, and sorts them by priority to form a dynamic rule set adapted to the scenario. Permission allocation and points circulation module: Based on dynamic rule set, user permissions are allocated through three-dimensional value permission matrix, and a multi-dimensional value points system is linked to complete points generation and transactions, forming a value circulation record that includes permission status, points changes and transaction records. The method of allocating user permissions through a three-dimensional value permission matrix includes: Based on dynamic rule sets, a three-dimensional value permission matrix is ​​constructed with relational, spatial, and temporal dimensions as the three axes, defining dimensional hierarchy, permission level, and operational scope. Extract the user's real-time status in three dimensions, match the initial permissions from the three-dimensional value permission matrix, resolve permission conflicts in the order of scenario priority > relationship strength > time value, and generate three-dimensional value permission credentials. The methods for forming the value circulation record include: Based on three-dimensional value authorization credentials, a multi-dimensional value points system is constructed by linking dynamic rule sets, event sets, and value stream data: Basic points are generated based on permission levels, and reward points are distributed based on event value and scenario weight; scenario-based transactions are executed within the scope of permissions, and risk control is carried out simultaneously. Integrate permission status, points changes, and transaction details to form a value circulation record; Value Optimization and Reverse Iteration Module: Based on value circulation records, it analyzes value health through a three-level optimization mechanism of real-time, short-term and long-term, generates optimization parameters, and iterates back to the corresponding preceding stage; The method of generating optimization parameters and iterating backwards to the corresponding preceding steps includes: Based on value circulation records, a three-level optimization mechanism is constructed: The real-time mechanism monitors sudden anomalies in value circulation and corrects them immediately by temporarily optimizing parameters; the short-term mechanism analyzes the trend changes in value circulation and generates periodic optimization parameters to adapt to different scenarios. The long-term mechanism is based on mining structural problems in value circulation from long-term historical data, generating systematic optimization parameters to upgrade the adaptability of the entire chain; The optimization parameters are iterated backwards to the preceding steps according to the problem attribution, forming a closed loop for the entire process.

2. The real-time coexistence user social system based on geofencing according to claim 1, characterized in that, The formation methods of the four-dimensional fence value network include: Based on a four-layer fence and a four-dimensional dataset, the value attributes of each fence are quantitatively evaluated according to hierarchical indicators through an ecological niche value assessment mechanism. Furthermore, based on the value assessment results of each fence, the fence status is dynamically adjusted through the entire lifecycle management of creation, activation, optimization, and destruction; Furthermore, by binding physical fences and commercial fences according to spatial overlap and logical fences and social fences according to value complementarity, a cross-fence collaborative relationship is constructed, forming a four-dimensional fence value network that includes fence boundaries, fence value attributes, and collaborative relationships.

3. A geofencing-based real-time coexistence user social networking method, implemented based on the geofencing-based real-time coexistence user social networking system according to any one of claims 1 to 2, characterized in that, include: S1: Collect multi-dimensional, multimodal raw data from users through dynamic sampling strategies and energy efficiency balance mechanisms to generate a four-dimensional data set with desensitized information; S2: Based on a four-dimensional data set, construct four layers of fences: physical, logical, social, and commercial. Evaluate the value attributes of each fence through an ecological niche value assessment mechanism, and dynamically adjust the fence status through life cycle management to form a four-dimensional fence value network. S3: Based on a four-dimensional fence value network, it hierarchically identifies basic events, composite events, and collaborative events. It constructs a real-time value stream through event value quantification and generates a dynamic rule set adapted to the scenario by combining a three-level rule generation system. S4: Based on a dynamic rule set, user permissions are allocated through a three-dimensional value permission matrix, and a multi-dimensional value points system is linked to complete the generation and transaction of points, forming a value circulation record that includes permission status, points changes and transaction records. S5: Based on value circulation records, analyze value health through a three-level optimization mechanism of real-time, short-term and long-term, generate optimization parameters, and iterate backward to the corresponding preceding stage.

Images