Bus and subway ticket association risk control method based on hierarchical scanning, medium and equipment

Abstract

The invention discloses a hierarchical scanning-based public transportation and subway associated ticket risk control method, a medium and equipment. The method comprises the following steps of: performing offline analysis on total transaction data in a day end period of a T day to generate a risk list record; checking whether the users in the real-time transaction data exist in the risk list record or not in the T + 1 day real-time period, and performing multi-dimensional abnormal behavior rule matching on the users with risks; according to a matching result, updating a blacklist library and triggering real-time risk control disposal; performing long-period comprehensive analysis on the accumulated transaction data in the T + n-day analysis period to generate a comprehensive judgment result; generating statistical analysis data according to a comprehensive judgment result, updating a blacklist library and triggering periodic risk control disposal; and finally, implementing a linked ticket use limiting operation on the blacklist library user and pushing a risk control warning message. According to the invention, multi-level cooperation of off-line, real-time and periodic risk control is realized, security of use of associated tickets and risk control response efficiency are effectively improved, and system resource allocation is optimized through grading processing.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of public transportation ticketing risk control, and particularly relates to a bus and subway joint ticket risk control method based on hierarchical scanning, a medium and equipment. BACKGROUND

[0002] As a convenient form of public transportation ticketing, the bus and subway joint ticket aims to improve passenger travel experience through preferential mechanisms, but its free or preferential qualifications are easily misused, and there are risks such as fraud and impersonation. Therefore, such a ticketing system usually involves a large amount of transaction data processing and user behavior analysis to ensure the reasonable use of preferential policies. Existing risk control technologies mostly use single-dimensional detection methods based on fixed rules, such as relying on single-day transaction frequency or simple qualification verification, which are difficult to cope with complex and variable abnormal behavior patterns. Due to the lack of collaborative analysis capabilities for real-time transactions and historical data, the system often has problems such as response lag, high misjudgment rate, and cannot effectively identify potential misuse behaviors. At the same time, traditional methods lack the ability to mine long-period behavior characteristics, resulting in limited adaptability and accuracy of risk control strategies, making it difficult to balance user experience and risk control needs. How to realize multi-level, multi-period risk dynamic perception and collaborative disposal has become a technical problem to be solved in the field of joint ticket risk control. SUMMARY

[0003] In view of the above problems, the present application provides a bus and subway joint ticket risk control method based on hierarchical scanning, a medium and equipment, which realizes dynamic identification and accurate control of abnormal behaviors through a multi-level risk control mechanism combining offline analysis, real-time judgment and long-period analysis, and solves the problems of real-time blocking of joint ticket preferential fraud behaviors and long-period risk lag disposal.

[0004] To achieve the above purpose, in a first aspect, the present application provides a bus and subway joint ticket risk control method based on hierarchical scanning, comprising:

[0005] At the end of the day T, offline analysis and processing are performed on the full amount of transaction data of day T, and a risk list record is generated according to the preset free qualification judgment rule and the single-day ride frequency threshold;

[0006] In the real-time period of T+1 day, real-time transaction data is subjected to risk control judgment, and it is verified whether the user exists in the risk list record;

[0007] If yes, the user existing in the risk list record is subjected to multi-dimensional abnormal behavior rule matching, and the multi-dimensional abnormal behavior rule includes parallel ride behavior rule, short-time high-frequency ride behavior rule and super-long duration ride behavior rule;

[0008] According to the matching result, the black list library is updated and real-time risk control disposal is triggered;

[0009] In the analysis period of T+n day, the long-period comprehensive analysis is performed on the cumulative transaction data from T day to T+n day, and the comprehensive judgment result is generated according to the free qualification judgment rule, the cumulative ride frequency threshold and the blacklist screening condition;

[0010] According to the comprehensive judgment result, statistical analysis data is generated, and the blacklist library is updated and the periodic risk control treatment is triggered;

[0011] And, the user in the blacklist library is implemented to limit the use of the joint ticket, and the risk control warning message is pushed to the user end.

[0012] Further, the T day full transaction data is analyzed and processed offline, and the risk list record is generated according to the preset free qualification judgment rule and the single-day ride frequency threshold, including:

[0013] Extracting user identification information, ride time information and ride type information in the T day full transaction data;

[0014] According to the preset free qualification judgment rule, the user type with free ride qualification is identified, including joint ticket user, talent card user and old card user;

[0015] The number of free rides of the user with free qualification in T day is counted;

[0016] The user identification information with free ride frequency exceeding the single-day ride frequency threshold is written into the risk list record;

[0017] The user identification information in the risk list record exceeding the preset time limit is automatically removed.

[0018] Further, the real-time transaction data is subjected to risk control judgment, and whether the user exists in the risk list record is checked, including:

[0019] Extracting user identification information and transaction time information in real-time transaction data;

[0020] Real-time matching query is performed between the user identification information in the real-time transaction data and the user identification information in the risk list record;

[0021] If the user identification information in the real-time transaction data does not exist in the risk list record, the real-time transaction data is implemented to release;

[0022] If the user identification information in the real-time transaction data exists in the risk list record, the multi-dimensional abnormal behavior rule matching process is triggered;

[0023] Recording the matching query result and updating the real-time risk control log.

[0024] Further, the user existing in the risk list record is subjected to multi-dimensional abnormal behavior rule matching, including parallel riding behavior rule, short-time high-frequency riding behavior rule and super-long duration riding behavior rule, including:

[0025] Obtaining real-time transaction data and historical travel trajectory data of the user in the risk list record;

[0026] Performing parallel riding behavior rule matching, including:

[0027] Detecting whether the current user has parallel riding behavior of bus and subway or multiple valid riding behaviors at the same time in the same time period to obtain a first matching result;

[0028] Performing short-time high-frequency riding behavior rule matching, including:

[0029] Statistically analyzing the time interval and occurrence frequency of adjacent two bus rides of the current user in a preset time window, detecting whether there is an abnormal high-frequency riding mode, and obtaining a second matching result;

[0030] Performing super-long duration riding behavior rule matching, including:

[0031] Calculating the duration of a single subway trip of the current user or the cumulative riding duration in a continuous time window, detecting whether there is behavior exceeding the reasonable use duration, and obtaining a third matching result;

[0032] Generating a final matching result according to the first matching result, the second matching result and the third matching result.

[0033] Further, updating the blacklist library and triggering real-time risk control treatment according to the matching result, including:

[0034] Generating a real-time risk control treatment instruction according to the matching result;

[0035] Updating the user identification information corresponding to the user triggering any of the parallel riding behavior rule, the short-time high-frequency riding behavior rule or the super-long duration riding behavior rule to the blacklist library;

[0036] Setting the initial risk level and treatment time limit of the user identification information in the blacklist library, and recording the user as a blacklist user;

[0037] Triggering real-time risk control treatment operations, including:

[0038] Implementing real-time transaction interception, ticket association function limitation or riding qualification suspension processing on the blacklist user;

[0039] And generating a real-time risk control treatment record and synchronizing it to a risk control log database;

[0040] Send the risk control state update notification to the business system through the message queue.

[0041] Further, long-period comprehensive analysis is performed on the cumulative transaction data from T day to T+n day, and a comprehensive determination result is generated according to the free qualification determination rule, the cumulative ride frequency threshold, and the blacklist screening condition, including:

[0042] All joint ticket ride transaction data in the T day to T+n day period are obtained, denoted as cumulative transaction data;

[0043] The free qualification determination rule matching is performed, including:

[0044] Verify whether the user has special free ride qualification, and exclude abnormal ride records of users with legal free qualification, to generate a first analysis result;

[0045] The cumulative ride frequency threshold detection is performed, including:

[0046] The total number of rides of the user in the period is counted, and it is detected whether it exceeds the preset reasonable use frequency threshold, to generate a second analysis result;

[0047] The blacklist screening condition matching is performed, including:

[0048] Combined with real-time blacklist library data, the user with high-risk behavior in the period is screened again to generate a third analysis result;

[0049] The comprehensive determination result is generated according to the first analysis result, the second analysis result, and the third analysis result;

[0050] In addition, the user identification information of the user with joint ticket abuse behavior in the period from T day to T+n day is marked, denoted as a period abnormal user.

[0051] Further, statistical analysis data is generated according to the comprehensive determination result, and the blacklist library is updated and the period risk control disposal is triggered, including:

[0052] The period risk control statistical analysis report is generated according to the comprehensive determination result, and the period risk control statistical analysis report includes the number of rule violation users, the type of high-frequency abnormal behavior, and the distribution characteristics of risk period;

[0053] The user identification information of the period abnormal user is updated to the blacklist library, and the corresponding risk level and disposal validity are adjusted according to the severity of the violation;

[0054] The period risk control disposal operation is triggered, including:

[0055] The joint ticket function of the user in the blacklist library whose risk level reaches the risk threshold is frozen, the ride qualification is suspended, or the credit points are deducted;

[0056] Generate a cycle risk control handling record and synchronize to the risk control log database;

[0057] Send a risk control state update notification to the business system through the message queue;

[0058] And, synchronize the black list state and user credit change information to the joint ticket management system and the third party credit platform through the data interface.

[0059] Further, the method further comprises:

[0060] Obtain risk control handling feedback data;

[0061] According to the statistical analysis data and the risk control handling feedback data, adaptively adjust the parameters in the free qualification judgment rule, the single-day ride frequency threshold, the parallel ride behavior rule, the short-time high-frequency ride behavior rule, the ultra-long duration ride behavior rule, and the cumulative ride frequency threshold.

[0062] In a second aspect, the present application also provides a computer readable storage medium, which stores computer program instructions, the computer program instructions being executed by a processor to implement the method of the first aspect.

[0063] In a third aspect, the present application also provides an electronic device, comprising a memory and a processor, the memory being used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method of the first aspect.

[0064] Unlike the prior art, the above technical solution provides a bus and subway joint ticket risk control method, medium and equipment based on hierarchical scanning, which generates a risk list record by offline analysis of full transaction data at the end of T day; in the real-time period of T+1 day, it is checked whether the user exists in the risk list record in the real-time transaction data, and the user with risk is matched with multi-dimensional abnormal behavior rules; according to the matching result, update the black list library and trigger real-time risk control handling; in the analysis period of T+n day, the cumulative transaction data is analyzed to generate a comprehensive judgment result; according to the comprehensive judgment result, generate statistical analysis data, update the black list library and trigger cycle risk control handling; finally, implement joint ticket use restriction operation on the black list library user and push the risk control warning message. The present application realizes multi-level cooperation of offline, real-time and cycle risk control, effectively improves the security and risk control response efficiency of joint ticket use, and optimizes the system resource allocation through hierarchical processing.

[0065] The above summary related to the invention is only a summary of the technical solutions of the present application. In order to enable those skilled in the art to more clearly understand the technical solutions of the present application, and then can be implemented according to the content of the description and the drawings, and in order to let the above-mentioned purposes and other purposes, characteristics and advantages of the present application can be more easily understood, the following is described in combination with the specific embodiments of the present application and the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0066] The accompanying drawings are only used to show the principles, implementation manners, applications, characteristics and effects of the specific embodiments of the present application and other related contents, and cannot be considered as limitations of the present application.

[0067] In the drawings of the specification:

[0068] Figure 1 The specific flowchart of the risk control method described in the specific embodiment is shown in the figure.

[0069] Figure 2 The step diagram of steps S101 to S105 of the risk control method described in the specific embodiment is shown in the figure.

[0070] Figure 3 The step diagram of steps S201 to S205 of the risk control method described in the specific embodiment is shown in the figure.

[0071] Figure 4 The step diagram of steps S301 to S303 of the risk control method described in the specific embodiment is shown in the figure.

[0072] Figure 5 The structure diagram of the electronic device described in the specific embodiment is shown in the figure.

[0073] The reference signs involved in the above drawings are explained as follows:

[0074] 1. Electronic device;

[0075] 11. Memory;

[0076] 12. Processor. DETAILED DESCRIPTION

[0077] In order to explain the possible application scenarios, technical principles, specific schemes that can be implemented, and the purposes and effects that can be achieved of the present application in detail, the following is described in combination with the specific examples listed and the accompanying drawings. The examples described in this paper are only used to more clearly explain the technical solutions of the present application, therefore only as an example, and cannot limit the protection scope of the present application.

[0078] The term "embodiment" is mentioned herein means that the specific features, structures, or characteristics described in connection with the embodiment can be included in at least one embodiment of the present application. The term "embodiment" appearing in various places in the specification does not necessarily refer to the same embodiment, nor does it particularly limit the independence or association between other embodiments. In principle, in the present application, as long as there is no technical contradiction or conflict, each technical feature mentioned in each embodiment can be combined in any way to form a corresponding implementable technical solution.

[0079] Unless otherwise defined, the meanings of the technical terms used herein are the same as those commonly understood by those skilled in the art to which the present application belongs; the use of related terms herein is only for the purpose of describing specific embodiments, and is not intended to limit the present application.

[0080] In the description of the present application, the phrase "and / or" is a description of the logical relationship between the objects, which means that there can be three relationships, for example, A and / or B, which means that there are three cases: A exists, B exists, and A and B exist at the same time. In addition, the character " / " herein generally represents that the associated objects before and after are a "or" logical relationship.

[0081] In the present application, the terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual quantity, primary and secondary or order relationship between the entities or operations.

[0082] In the present application, without more limitation, the "includes", "contains", "has" or other similar open expressions used in the sentence are intended to cover non-exclusive inclusion, and these expressions do not exclude the presence of other elements in the process, method or product including the described elements, so that the process, method or product including a series of elements can not only include those limited elements, but also include other elements not explicitly listed, or also include the elements inherent to such process, method or product.

[0083] In the present application, "greater than", "less than", "exceed" and other expressions are understood as not including the number; "above", "below", "within" and other expressions are understood as including the number. In addition, in the description of the embodiments of the present application, the meaning of "multiple" is more than two (including two), and similar expressions related to "multiple" are also understood in this way, for example, "multiple groups", "multiple times" and the like, unless otherwise explicitly limited.

[0084] In the description of the embodiments of the present application, the spatially relative terms used, such as "central", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "vertical", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. The indicated orientation or positional relationship is based on the orientation or positional relationship shown in the specific embodiment or the drawing, and is only for the convenience of describing the specific embodiment of the present application or for the reader to understand, and does not indicate or imply that the indicated device or component must have a particular position, a particular orientation, or be constructed or operated in a particular orientation, and therefore cannot be understood as a limitation on the embodiments of the present application.

[0085] Please refer to Figures 1 to 2 In the first aspect, the embodiment provides a bus and subway joint ticket risk control method based on hierarchical scanning, comprising:

[0086] S101, at the end of the day of T, the full amount of transaction data of T is analyzed and processed offline, and the risk list record is generated according to the preset free qualification judgment rule and the single-day riding frequency threshold;

[0087] S102, in the real-time period of T+1 day, the real-time transaction data is judged, and it is verified whether the user exists in the risk list record;

[0088] If yes, the user existing in the risk list record is matched with multi-dimensional abnormal behavior rules, and the multi-dimensional abnormal behavior rules include parallel riding behavior rules, short-time high-frequency riding behavior rules and super-long time riding behavior rules;

[0089] S103, updating the blacklist library and triggering real-time risk control disposal according to the matching result;

[0090] S104, in the analysis period of T+n day, the cumulative transaction data from T day to T+n day is analyzed, and the comprehensive judgment result is generated according to the free qualification judgment rule, the cumulative riding frequency threshold and the blacklist screening condition;

[0091] S105, generating statistical analysis data according to the comprehensive judgment result, and updating the blacklist library and triggering periodic risk control disposal;

[0092] And, the user in the blacklist library is implemented with joint ticket use restriction operation, and the risk control warning message is pushed to the user end.

[0093] In step S101, the end-of-day period of T day refers to a data processing time window preset by the system after the operation of buses and subways ends, which is used to perform offline computing tasks. The full-amount transaction data of T day contains all the ride transaction records of the day, which are extracted from the transaction database through a data collection interface. The preset free qualification determination rule is a qualification verification logic based on user types, which is implemented through a configuration rule engine. The single-day ride frequency threshold is a reasonable upper limit of ride frequency set according to historical travel rules, which is used to identify abnormal high-frequency behaviors. The risk list records are generated through batch data processing and stored in a risk control database, and the preset time limit is automatically cleaned up by a timing task.

[0094] In step S102, the real-time period of T+1 day refers to an online risk control period during the business peak period. Real-time transaction data is collected through a message queue. Risk list record matching uses distributed cache query to reduce delay. The multi-dimensional abnormal behavior rule contains multiple detection logics based on behavior characteristics, which are used to analyze the user travel data through a real-time computing engine to identify abnormal behaviors that meet the preset risk characteristics.

[0095] In step S103, the matching result is output through a rule engine to trigger real-time risk control disposal instructions. Preferably, the black list library update uses a database transaction mechanism to ensure data consistency, and sets an initial risk level and disposal time limit for new user identifiers; real-time risk control disposal includes immediate restriction operation on high-risk users, records disposal details through a risk control log database, and pushes state updates to the business system through a message mechanism.

[0096] In step S104, the analysis period of T+n day refers to a periodic data analysis window. Cumulative transaction data is aggregated through a data warehouse. Preferably, the long-period comprehensive analysis reuses the free qualification determination rule, the cumulative ride frequency threshold is set based on periodic travel characteristics, the black list screening conditions are comprehensively judged in combination with risk control data, and the comprehensive determination result is generated through multi-rule comprehensive analysis and marked as a periodic abnormal user.

[0097] In step S105, statistical analysis data generates a report through a data visualization tool, and the risk level of the black list library is adjusted according to the degree of violation. Periodic risk control disposal includes subsequent processing measures for violating users, and synchronizes the state with external systems through a data interface. The joint ticket usage restriction operation is performed through a user service system, and the risk warning message is sent through a message push channel.

[0098] The embodiment realizes the risk control closed loop through hierarchical time sequence processing: the risk list record is generated through offline analysis at the end of the day, the risk list matching and multi-dimensional abnormal behavior rule detection are performed in real time, and the accumulated data is comprehensively analyzed in the cycle stage. The processing efficiency is improved by using the rule engine and distributed calculation, the complex scene is adapted through dynamic threshold and multi-level rules, the whole-process risk control from real-time interception to cycle control is realized, and the safety of joint ticket use and the accuracy of system response are effectively improved.

[0099] Please refer to Figure 3 In some embodiments, the T-day full transaction data is analyzed and processed offline, and the risk list record is generated according to the preset free qualification judgment rule and the single-day ride frequency threshold, including:

[0100] S201, extracting user identification information, ride time information and ride type information in T-day full transaction data;

[0101] S202, identifying the user type with free ride qualification according to the preset free ride qualification judgment rule, the user type with free ride qualification including joint ticket users, talent card users and old card users;

[0102] S203, counting the number of free rides of the user type with free qualification within T-day;

[0103] S204, writing the user identification information with free ride frequency exceeding the single-day ride frequency threshold into the risk list record;

[0104] S205, automatically removing the user identification information exceeding the preset time limit in the risk list record.

[0105] In step S201, the user identification information, ride time information and ride type information in the T-day full transaction data are batched from the transaction database through the data interface, the user identification information is used to uniquely identify the user identity, the ride time information records the time point of the trip, and the ride type information distinguishes ordinary tickets and discount tickets. This step provides a data basis for subsequent qualification judgment and frequency statistics.

[0106] In step S202, the preset free qualification judgment rule is realized by configuring the rule engine, the joint ticket users, talent card users and old card users are the user types with free ride qualification, and are automatically identified by matching the user type field and the rule library conditions. This step ensures that only users with special qualifications are subjected to subsequent risk control analysis.

[0107] In step S203, the extracted data is grouped by user identification, and the number of daily ride records is counted to count the number of free rides of the user type with free qualification within T-day. This statistical focus is on free ride behavior, providing data input for threshold comparison.

[0108] In step S204, by comparing the statistical result with the preset threshold, the database insertion operation is performed on the user exceeding the standard, and the user identification information whose free ride frequency exceeds the single-day ride frequency threshold is written into the risk list record. Preferably, the preset threshold is set based on the historical travel mode. This step realizes the preliminary screening of the abnormally high-frequency behavior.

[0109] In step S205, by scanning the list library through a timing task, the records with timestamps earlier than the set threshold are deleted, and the user identification information exceeding the preset time limit in the risk list record is automatically removed. This mechanism ensures the timeliness of the list data and avoids the continuous influence of historical misjudgment on users.

[0110] The embodiment forms a closed-loop process through data extraction, qualification filtering, frequency statistics, threshold comparison, and time management, accurately identifies the abnormal free ride behavior within a single day, and provides reliable data support for real-time risk control. For example, by quickly matching the user type through the rule engine, combining time series data statistics and dynamic threshold determination, the normal user rights are protected while the potential fraud risk is efficiently captured.

[0111] Please refer to Figure 4 In some embodiments, real-time transaction data is subjected to risk control judgment to verify whether the user exists in the risk list record, including:

[0112] S301, extracting user identification information and transaction time information in real-time transaction data;

[0113] S302, real-time matching query of the user identification information in the real-time transaction data with the user identification information in the risk list record;

[0114] If the user identification information in the real-time transaction data does not exist in the risk list record, the real-time transaction data is subjected to release processing;

[0115] If the user identification information in the real-time transaction data exists in the risk list record, a multi-dimensional abnormal behavior rule matching process is triggered;

[0116] S303, recording the matching query result and updating the real-time risk control log.

[0117] In step S301, preferably, the real-time transaction data is collected in real time through a high-throughput message middleware, the user identification information adopts a standardized coding format to ensure global uniqueness, and the transaction time information is accurate to the millisecond level and is synchronized with the coordinated universal time. This step provides a timely data basis for low-latency risk control judgment.

[0118] In step S302, the real-time matching query adopts an in-memory database to achieve a microsecond-level response, and the query path can be optimized through a Bloom filter pre-screening mechanism. For a matching hit scenario, a rule engine dynamic loading process is triggered; for a miss scenario, a lightweight state verification mechanism is adopted to ensure release efficiency. Preferably, a soft state marking mechanism is introduced in the matching process to handle critical concurrent requests.

[0119] In step S303, the matching result log is recorded in a structured log paradigm, including a session identifier, a rule trigger identifier, and a disposition serial number. The real-time risk control log is persisted through an event trace mode, supporting real-time stream analysis and offline backtracking verification. This mechanism optimizes storage efficiency through log compression technology while ensuring data trace integrity.

[0120] The embodiment collects transaction data in real time, performs high-speed matching and verification with a risk list, and shunts disposition according to the matching result and records a risk control log. Microsecond-level response is achieved through in-memory computing and pre-screening mechanism, and operation auditability is ensured through structured log and event trace, which not only ensures smooth business flow but also accurately intercepts high-risk transactions, thereby improving the real-time risk control efficiency and data trace integrity of the system.

[0121] In some embodiments, a user existing in a risk list record is subjected to multi-dimensional abnormal behavior rule matching, including parallel riding behavior rules, short-time high-frequency riding behavior rules, and super-long duration riding behavior rules, including:

[0122] Obtaining real-time transaction data and historical travel trajectory data of the user in the risk list record;

[0123] Performing parallel riding behavior rule matching, including:

[0124] Detecting whether the current user has a parallel riding behavior of bus and subway in the same time period or a behavior of simultaneously existing multiple valid riding trips, to obtain a first matching result;

[0125] Performing short-time high-frequency riding behavior rule matching, including:

[0126] Statistically analyzing the time interval and occurrence frequency of adjacent two bus rides of the current user within a preset time window, detecting whether there is an abnormally high-frequency riding mode, and obtaining a second matching result;

[0127] Performing super-long duration riding behavior rule matching, including:

[0128] Calculating the duration of a single subway trip of the current user or the cumulative riding duration within a continuous time window, detecting whether there is a behavior of exceeding a reasonable use duration, and obtaining a third matching result;

[0129] The final matching result is generated according to the first matching result, the second matching result and the third matching result.

[0130] In the embodiment, real-time transaction data and historical trajectory data of the user in the risk list record are acquired to provide multi-source data support for rule matching, and the historical trajectory data is stored in a time series database to realize efficient range query.

[0131] When performing parallel riding behavior rule matching, preferably, the parallel riding behavior of buses and subways in the same time period is identified through a time overlap detection algorithm, and the concurrency of multiple valid riding trips is judged based on a trip state machine model. The rule realizes millisecond-level overlap window calculation through a distributed stream processing engine, effectively identifying the illegal ticket sharing behavior.

[0132] When performing short-time high-frequency riding behavior rule matching, the preset time window is dynamically adjusted according to the line operation characteristics, the time interval and frequency of adjacent two bus rides are counted through a sliding window, and an abnormal high-frequency mode is detected in combination with a Poisson distribution model. The rule adopts incremental calculation to optimize resource consumption and accurately identifies short-term malicious ticket fraud behavior.

[0133] When performing super-long duration riding behavior rule matching, the duration of a single subway trip is calculated by the difference between the time stamps of entering and leaving the station, and the cumulative duration of a continuous time window is realized by an accumulator mechanism. The reasonable use of time length threshold is dynamically configured according to the line mileage and average running speed, effectively identifying long-term abnormal occupation behaviors such as retention.

[0134] When generating the final matching result according to the first matching result, the second matching result and the third matching result, a weighted decision fusion mechanism is adopted, in which the parallel riding behavior rule has the highest priority. The final matching result is identified by state coding to identify the specific violation type, providing a basis for subsequent hierarchical disposal.

[0135] The embodiment reduces the false positive rate while improving the detection capability of complex abnormal patterns through parallel execution of multiple rules and decision fusion. Through parallel detection of abnormal riding patterns (parallel riding, short-time high-frequency, super-long duration), in combination with real-time transaction and historical trajectory data, a dynamic threshold and a multi-rule weighted decision mechanism are adopted. Its beneficial effects are: accurately identifying complex fraud behaviors such as ticket sharing, malicious ticket fraud and illegal occupation, optimizing system resource consumption through parallel calculation and incremental processing, and improving the accuracy and real-time performance of multi-dimensional anomaly detection.

[0136] In some embodiments, the blacklist library is updated and real-time risk control disposal is triggered according to the matching result, including:

[0137] Generating a real-time risk control disposal instruction according to the matching result;

[0138] updating the user identification information of the user triggering any of the parallel riding behavior rule, the short-time high-frequency riding behavior rule, or the super-long-time riding behavior rule to the blacklist library;

[0139] setting the initial risk level and the disposal validity of the user identification information in the blacklist library, and recording the user as a blacklisted user;

[0140] triggering a real-time risk control disposal operation, including:

[0141] implementing real-time transaction interception, joint ticket function restriction, or riding qualification suspension processing on the blacklisted user;

[0142] generating a real-time risk control disposal record and synchronizing it to a risk control log database;

[0143] sending a risk control state update notification to a business system through a message queue.

[0144] In the embodiment, the execution instruction is dynamically generated according to the result of multi-dimensional abnormal behavior rule matching, the instruction is dynamically configured through a rule engine to realize accurate mapping of the instruction and the violation type. When the user triggering any rule is updated to the blacklist library, the user identification information is stored by using hash encryption to ensure data security, and the atomicity and consistency of the update operation are guaranteed through a database transaction mechanism.

[0145] When setting the initial risk level and the disposal validity of the user identification information in the blacklist library, the initial risk level is hierarchically set according to the severity of the triggered rule, and the disposal validity is dynamically adjusted based on historical behavior pattern analysis, wherein the high-risk level is associated with a shorter disposal validity by default to realize rapid response. Preferably, the process of recording the user as a blacklisted user is realized through data marking bit update and version number mechanism to realize state tracing.

[0146] When triggering a real-time risk control disposal operation, preferably, real-time transaction interception is realized by API gateway fuse mechanism to block in milliseconds, joint ticket function restriction adopts permission identification bit dynamic disable strategy, and riding qualification suspension processing is realized through distributed lock to guarantee multi-node state synchronization; when generating a real-time risk control disposal record, the record content includes disposal time, operation type and associated rule code, and is batched and persisted to the risk control log database through asynchronous writing mechanism, the database adopts time series data model to optimize high-frequency writing performance; when sending a risk control state update notification to a business system through a message queue, a publish-subscribe mode is adopted to realize decoupling communication, the message body includes user identification, risk level and effective time stamp, and the business system realizes load balancing processing through a consumer group.

[0147] The embodiment realizes end-to-end automation of risk control response while guaranteeing system throughput through closed-loop management of the chain. Through dynamic risk grading and real-time interception strategy, the disposal efficiency is improved; asynchronous log and message queue are used to guarantee system performance; and the risk control response is ensured to be accurate, traceable and not to affect the business throughput.

[0148] In some embodiments, long-period comprehensive analysis is performed on the cumulative transaction data of T day to T+n day, and a comprehensive judgment result is generated according to the free qualification judgment rule, the cumulative ride frequency threshold and the blacklist screening condition, including:

[0149] All joint ticket ride transaction data in the T day to T+n day period are obtained, denoted as cumulative transaction data;

[0150] The free qualification judgment rule matching is performed, including:

[0151] Verify whether the user has special free ride qualification, and exclude abnormal ride records of users with legal free qualification, to generate a first analysis result;

[0152] The cumulative ride frequency threshold detection is performed, including:

[0153] The total number of rides of the user in the period is counted, and whether it exceeds the preset reasonable use frequency threshold is detected, to generate a second analysis result;

[0154] The blacklist screening condition matching is performed, including:

[0155] Combined with real-time blacklist library data, the user with high-risk behavior in the period is screened again to generate a third analysis result;

[0156] The comprehensive judgment result is generated according to the first analysis result, the second analysis result and the third analysis result;

[0157] And, the user identification information of the user with irregular use of joint ticket behavior in the period from T day to T+n day is marked, denoted as period abnormal user.

[0158] In the embodiment, the period from T day to T+n day refers to a continuous time interval, and the length n is dynamically configured through business scenarios, usually combined with ticket settlement period or operation statistical period. The cumulative transaction data refers to the set of all joint ticket ride records in the period, which is collected in real time from each business node by a distributed data collection component and is stored by merging according to the user identification.

[0159] The free qualification judgment rule is used to distinguish between legal free ride behavior and irregular behavior, which verifies the special identity (such as the elderly, the disabled, etc.) of the user by accessing an external qualification certification system, and excludes compliant records based on a pre-defined whitelist mechanism. In the first analysis result generation process, data labeling technology is used to identify abnormal records to ensure that only the abnormal ride behavior of unqualified users is processed subsequently.

[0160] The cumulative ride frequency threshold refers to the maximum reasonable ride number allowed within a unit period, which is obtained through historical travel big data analysis and dynamically adjusted in combination with urban travel characteristics. The second analysis result is generated by comparing the actual ride number of the user with the threshold in real time through a distributed counting service.

[0161] The blacklist screening condition refers to the associated matching rule in combination with the real-time blacklist library, which performs secondary filtering on the user behavior within a period through multi-dimensional association query (including user identification, device fingerprint, behavior characteristics, etc.). The third analysis result identifies high-risk users through risk weight calculation to ensure linkage with real-time risk control status.

[0162] The comprehensive judgment result is the logical combination output based on the first analysis result, the second analysis result, and the third analysis result, and a rule engine is used to realize multi-condition decision-making, in which any analysis result triggering an exception will lead to a final judgment of irregularity. The periodic abnormal user marking updates the user state identification through the data persistence layer and provides data basis for subsequent disposal processes.

[0163] This embodiment effectively improves the identification accuracy of irregular use of joint tickets through long-period data analysis and multi-rule hierarchical judgment, while avoiding false positives for legitimate users. By aggregating joint ticket data within a period, the free qualification verification, frequency threshold detection, and blacklist secondary screening are performed in turn to generate a comprehensive judgment result. Multi-rule hierarchical verification improves the accuracy of irregular identification, avoids false positives by combining dynamic thresholds and real-time blacklists, and provides precise tracing capabilities for long-period fraudulent behavior.

[0164] In some embodiments, statistical analysis data is generated according to the comprehensive judgment result, and the blacklist library is updated and periodic risk control disposal is triggered, including:

[0165] A periodic risk control statistical analysis report is generated according to the comprehensive judgment result, including the number distribution of irregular users, the type statistics of high-frequency abnormal behavior, and the distribution characteristics of risk periods;

[0166] The user identification information of the periodic abnormal user is updated to the blacklist library, and the corresponding risk level and disposal validity are adjusted according to the severity of the irregularity;

[0167] The periodic risk control disposal operation is triggered, including:

[0168] Implementing joint ticket function freezing, ride eligibility suspension, or credit score deduction processing on users in the blacklist library whose risk level reaches a risk threshold;

[0169] Generating a periodic risk control disposition record and synchronizing it to a risk control log database;

[0170] Sending a risk control state update notification to a business system through a message queue;

[0171] And synchronously updating blacklist state and user credit change information to a joint ticket management system and a third-party credit platform through a data interface.

[0172] In the embodiment, the periodic risk control statistical analysis report refers to a quantitative analysis document generated based on a comprehensive judgment result, which visually presents multi-dimensional characteristics of a number of rule violation users, high-frequency abnormal behavior type statistics, and risk period distribution characteristics through a data aggregation engine. The number of rule violation users is displayed by combining a geographic information system with time series analysis to show regional rule violation density, the high-frequency abnormal behavior type statistics uses a clustering algorithm to identify behavior pattern characteristics, and the risk period distribution characteristics are extracted by a time series analysis model to provide data support for operational strategy adjustment.

[0173] Writing user identification information of periodic abnormal users into a distributed memory database to update the blacklist library, wherein a risk level is dynamically calculated and generated according to historical rule violation times, behavior harm degrees, and other dimensions, and a disposition time limit is automatically matched with preset rules (such as short-term monitoring, long-term ban, etc.) according to the risk level. Preferably, the process ensures data consistency through an atomic transaction and establishes a version management mechanism to support rollback operations.

[0174] In the periodic risk control disposition operation, a risk threshold is dynamically calibrated through a machine learning model after analyzing historical disposition effects, joint ticket function freezing is implemented by calling a real-time interface to realize business system linkage, ride eligibility suspension is implemented by a permission management center for hierarchical management and control, and credit score deduction is calculated by a rule engine to determine a deduction range and trigger credit system update. The risk control log database uses a columnar storage structure to record complete disposition procedures and supports traceability queries.

[0175] The message queue uses a publish-subscribe mode to push a risk control state update notification to a business system, and through a retry mechanism and a dead letter queue, ensures reliable message delivery. The data interface is designed based on RESTful specifications, uses an incremental synchronization strategy to reduce bandwidth pressure when synchronizing blacklist state to a joint ticket management system, and uses asymmetric encryption and digital signature to ensure data security and compliance when transmitting user credit change information to a third-party credit platform. Through disposition closed loop and multi-party collaboration mechanism, the embodiment realizes systematic landing of risk control measures and cross-platform data consistency guarantee.

[0176] In some embodiments, the method further comprises:

[0177] obtain risk control treatment feedback data;

[0178] According to the statistical analysis data and the risk control treatment feedback data, parameters in the free qualification determination rule, the single-day ride frequency threshold, the parallel ride behavior rule, the short-time high-frequency ride behavior rule, the ultra-long-time ride behavior rule, and the cumulative ride frequency threshold are adaptively adjusted.

[0179] In the embodiment, the risk control treatment feedback data refers to treatment execution result data collected through a business system interface, including operation feedback information such as joint ticket function freezing success rate, ride qualification suspension execution state, and credit score deduction completion degree. The data is obtained through a real-time monitoring interface and an asynchronous callback mechanism, and is associated and matched with the treatment records in the risk control log database to form a complete treatment effect tracking chain.

[0180] Adaptive adjustment refers to a two-way evaluation mechanism based on statistical analysis data and risk control treatment feedback data, and dynamic optimization of parameters is realized through the cooperation of a rule engine and a machine learning model. The adjustment of the free qualification determination rule is to calibrate the whitelist update frequency and the verification strictness by analyzing the change of the misjudgment rate and the missed judgment rate; the single-day ride frequency threshold is dynamically adjusted by using a time series prediction model according to the misinterception rate in the peak period ride behavior distribution and the treatment feedback.

[0181] The parallel ride behavior rule optimizes the time window parameters according to the actual concurrent behavior characteristics through device fingerprint association analysis and geographic location collision detection; the short-time high-frequency ride behavior rule elastically adjusts the time granularity and the frequency threshold in combination with the time period distribution characteristics and the treatment feedback data; the ultra-long-time ride behavior rule optimizes the maximum allowed time length parameter in segments according to the line operation time characteristics and the abnormal time length distribution. The cumulative ride frequency threshold is gradually adjusted by using a gradient descent algorithm based on long-period behavior pattern analysis and in combination with the credit score change feedback data. The embodiment realizes the self-evolution of the risk control parameters through a closed-loop feedback mechanism, which reduces the misjudgment rate while ensuring the interception accuracy.

[0182] In a second aspect, the embodiment also provides a computer-readable storage medium having computer program instructions stored thereon, the computer program instructions being executed by a processor to implement the method in the first aspect.

[0183] The computer program involved in the embodiment can be stored in a computer device readable storage medium, including but not limited to magnetic disk, magnetic tape, magnetic card, floppy disk, flash memory, optical disc, optical card, read-only memory (ROM), random access memory (RAM), erasable programmable ROM (EPROM) and electrically erasable programmable ROM (EEPROM) and the like, and also includes other biological, physical or chemical structures that can realize similar or equivalent functions as the above-mentioned storage media, such as DNA, RNA, protein and the like units with information storage ability. In specific embodiments, the storage medium involved can be one of the above-mentioned medium types, or a combination of the above-mentioned medium types. In different embodiments, the computer program involved in the embodiment can be centrally stored in a single medium, or can be distributedly stored in multiple media. The storage medium containing the computer device readable storage medium can be a non-volatile memory or a random access memory. These computer device readable storage media can be built-in in the device, or connected with the device as an external device or part of the external device. In some embodiments, the storage medium with the computer device readable storage medium is deployed locally; in other embodiments, the storage medium can also be deployed remotely from the processor, for example, network attached storage accessed via RF circuit or external port and communication network, wherein the communication network can be Internet, one or more intranets, local area network (LAN), wide area network (WAN), storage area network (SAN) and the like, or appropriate combination thereof, as long as the access of the computer device to the storage medium can be realized. In addition, the computer program involved in the embodiment can be stored in plaintext / ciphertext form, or can be designed as training data, and integrated and reorganized by model training to be implicitly saved in the parameter state of the deep neural network or other machine learning model.

[0184] Please refer to Figure 5 In a third aspect, the embodiment also provides an electronic device 1 comprising a memory 11 and a processor 12, the memory 11 being configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor 12 to implement the method of the first aspect.

[0185] The processor described in the embodiment can be implemented by hardware, firmware, software or a combination thereof, and can use at least one of a circuit, a single or multiple application specific integrated circuits (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, a microprocessor, and other physical, biological or chemical structures that can realize the same or equivalent functions as the above-mentioned processors, such as biological neurons, quantum computing units, DNA computing units, etc., so that the processor can execute part or all of the steps or any combination of the steps mentioned in the computer programs or methods of various embodiments of the present application.

[0186] Further, the following examples can be developed in combination with the above technical solutions:

[0187] 1. At the end of T day, the T-day full transaction data is analyzed and processed by the T-day risk control computing component. The analysis rules include having free qualifications, including purchasing joint tickets, talent cards, and old age cards, and taking more than 10 free rides on T day. The free ride qualifications and the number of rides can be automatically adjusted through dynamic rule updates. Users who meet the rules are written into the risk list record; the list in the risk list that exceeds a certain time is removed, and the specific time can be configured.

[0188] 2. On T+1 day, the transaction data generated on T+1 day is analyzed by the T+1 day real-time risk control computing component. First, it is determined whether the user is in the risk list. If not, it is directly released. If so, the rule is judged. The rule includes parallel rides, taking the subway during the bus ride, and matching immediately writing into the blacklist library and performing risk disposal. The time difference between two bus rides is less than 5 minutes, and more than 3 times a day, and matching immediately writing into the blacklist library and performing risk disposal. The time difference and the number of occurrences are automatically updated and adjusted through dynamic rules. The ride time within 8 hours exceeds 6 hours, and the bus is calculated at 10 minutes. Matching immediately writes into the blacklist library and performs risk disposal. The transaction window period is 8 hours, the threshold is 6 hours, and the bus default time is automatically updated and adjusted through dynamic rules.

[0189] 3. On T+n day, the transaction information received from T day to T+n day is comprehensively analyzed, and the analysis rules include three rules, 1. Free qualification, including the purchase of joint tickets, talent cards, and old age cards, 2. Total number of rides exceeds 200 times, 3. Not added to the blacklist, for users who meet all three rules, add them to the blacklist and perform risk disposal, the free qualification and total number of rides involved in the rules can be updated through dynamic rules. The T+n period risk control calculation component will generate dynamic rule data based on statistical analysis data and disposal feedback data, and update the rule parameters.

[0190] 4. Risk control disposal, immediately push a message to the user, and limit the use of joint tickets according to the multi-person use mode in the joint ticket agreement.

[0191] By adopting the above technical scheme, the application is different from the prior art and has the following beneficial effects: the risk control closed-loop management is realized through the hierarchical time sequence processing mechanism, the risk list record is generated based on the free qualification judgment rule and the single-day ride frequency threshold at the end of T day, the high-risk transaction interception is realized through multi-dimensional abnormal behavior rule matching in the real-time period of T+1 day, and the comprehensive judgment result is generated through long-period comprehensive analysis in the analysis period and triggers the periodic disposal. Dynamic threshold adjustment and multi-rule collaborative judgment are adopted to effectively identify abnormal behavior patterns such as parallel riding, short-time high-frequency, and super-long duration. Through the combination of real-time risk control disposal and periodic statistical analysis, the detection accuracy and response timeliness of joint ticket fraud behavior are significantly improved. At the same time, the self-adaptive optimization of rule parameters is driven by risk control disposal feedback data, forming a continuously improved risk control closed loop, which greatly reduces the misjudgment rate and the omission rate on the premise of protecting the rights and interests of normal users, and realizes the coordinated improvement of joint ticket use safety and system control accuracy.

[0192] Finally, it should be noted that although the above embodiments have been described in the specification and drawings of the present application, they do not limit the patent protection scope of the present application. Any equivalent structure or equivalent flow replacement or modification based on the essential concept of the present application, using the content described in the specification and drawings, and directly or indirectly implementing the technical solutions of the above embodiments in other related technical fields, etc. are all included in the patent protection scope of the present application.

Claims

1. A bus and subway ticket risk control method based on hierarchical scanning, characterized in that: include: At the end of day T, all transaction data for day T is analyzed offline and processed, and a risk list is generated based on the preset free qualification determination rules and the daily ride frequency threshold. During the real-time period of T+1 day, risk control judgment is performed on real-time transaction data to verify whether the user is on the risk list record; If so, the user in the risk list record is matched with multi-dimensional abnormal behavior rules, including parallel ride behavior rules, short-term high-frequency ride behavior rules, and excessively long ride behavior rules; Update the blacklist database and trigger real-time risk control measures based on the matching results; During the analysis period of day T+n, a long-term comprehensive analysis is conducted on the accumulated transaction data from day T to day T+n, and a comprehensive judgment result is generated based on the free qualification judgment rules, the cumulative ride frequency threshold, and the blacklist screening conditions; Generate statistical analysis data based on the comprehensive judgment results, update the blacklist database and trigger periodic risk control measures; In addition, restrictions on the use of joint tickets will be implemented for users in the blacklist library, and risk control warning messages will be pushed to the user end.

2. The bus and subway combined ticket risk control method based on hierarchical scanning according to claim 1 is characterized in that: Perform offline analysis and processing of all transaction data on day T, and generate risk list records based on preset free qualification determination rules and single-day ride frequency thresholds, including: Extract user identification information, ride time information, and ride type information from the full transaction data on day T; Identify the types of users who are eligible for free rides based on the preset free qualification determination rules. The types of users who are eligible for free rides include combined ticket users, talent card users, and senior citizen card users. Count the number of free rides for users with free eligibility within T days; The identification information of users whose free rides exceed the daily ride frequency threshold is recorded in the risk list; Automatically remove user identification information from risk list records that exceeds the preset validity period.

3. The bus and subway combined ticket risk control method based on hierarchical scanning according to claim 1 is characterized in that: Real-time transaction data is used to conduct risk control judgments and verify whether the user exists in the risk list records, including: Extracting user identification information and transaction time information from real-time transaction data; Perform real-time matching query on user identification information in real-time transaction data and user identification information in risk list records; If the user identification information in the real-time transaction data does not exist in the risk list record, the real-time transaction data will be released; If the user identification information in the real-time transaction data exists in the risk list record, the multi-dimensional abnormal behavior rule matching process is triggered; Record matching query results and update real-time risk control logs.

4. The bus and subway combined ticket risk control method based on hierarchical scanning according to claim 1 is characterized in that: Multi-dimensional abnormal behavior rules are matched against users in the risk list. The multi-dimensional abnormal behavior rules include parallel ride behavior rules, short-term high-frequency ride behavior rules, and extremely long ride behavior rules, including: Obtain real-time transaction data and historical travel trajectory data of users on the risk list; Perform parallel ride behavior rule matching, including: Detect whether the current user has taken both the bus and the subway in the same time period, or has multiple valid travel itineraries at the same time, and obtain the first matching result; Implement short-term high-frequency ride behavior rule matching, including: Counting the time interval and frequency between two consecutive bus rides of the current user within a preset time window, detecting whether there is an abnormally high-frequency riding pattern, and obtaining a second matching result; Execute excessively long ride behavior rule matching, including: Calculate the duration of the current user's single subway trip or the cumulative riding time within a continuous time window, detect whether there is any behavior that exceeds the reasonable usage time, and obtain a third matching result; A final matching result is generated according to the first matching result, the second matching result, and the third matching result.

5. The bus and subway combined ticket risk control method based on hierarchical scanning according to claim 1 is characterized in that: Update the blacklist database based on the matching results and trigger real-time risk control measures, including: Generate real-time risk control disposal instructions based on the matching results; For users who trigger any of the parallel ride behavior rules, short-term high-frequency ride behavior rules, or excessively long ride behavior rules, their corresponding user identification information is updated to the blacklist database; Set the initial risk level and handling time limit for user identification information in the blacklist database, and record them as blacklist users; Trigger real-time risk control and disposal operations, including: Implement real-time transaction blocking, combined ticket function restrictions, or suspension of travel eligibility for blacklisted users; Also, generate real-time risk control disposal records and synchronize them to the risk control log database; Send risk control status update notifications to the business system through the message queue.

6. The bus and subway combined ticket risk control method based on hierarchical scanning according to claim 1 is characterized in that: Conduct a long-term comprehensive analysis of the accumulated transaction data from day T to day T+n, and generate a comprehensive judgment result based on the free qualification judgment rules, the cumulative ride frequency threshold, and the blacklist screening conditions, including: Obtain all combined ticket transaction data from day T to day T+n, and record it as cumulative transaction data; Execute free qualification determination rule matching, including: Verify whether the user has special free ride qualifications, and exclude abnormal ride records of users with legal free ride qualifications to generate the first analysis result; Perform cumulative ride frequency threshold detection, including: Counting the total number of rides taken by the user during the period, detecting whether the total number of rides exceeds a preset reasonable frequency threshold, and generating a second analysis result; Perform blacklist screening matching, including: Combined with real-time blacklist database data, users with high-risk behaviors within the period are screened twice to generate third-party analysis results; generating a comprehensive determination result based on the first analysis result, the second analysis result, and the third analysis result; Also, the identification information of users who have violated the rules in using the coupon during the period from T to T+n is marked as abnormal users of the period.

7. The bus and subway combined ticket risk control method based on hierarchical scanning according to claim 6 is characterized in that: Generate statistical analysis data based on the comprehensive judgment results, update the blacklist database and trigger periodic risk control measures, including: Generate a periodic risk control statistical analysis report based on the comprehensive judgment results. The periodic risk control statistical analysis report includes the number of illegal users, statistics on high-frequency abnormal behavior types, and distribution characteristics of risk periods; Update the user identification information of users with abnormal cycles to the blacklist database, and adjust the corresponding risk level and handling time limit according to the severity of the violation; Trigger periodic risk control disposal operations, including: For users whose risk level in the blacklist reaches the risk threshold, the combined ticket function will be frozen, their ride eligibility will be suspended, or their credit points will be deducted; Generate periodic risk control disposal records and synchronize them to the risk control log database; Send risk control status update notifications to the business system through the message queue; In addition, the blacklist status and user credit change information are synchronously updated to the joint ticket management system and the third-party credit platform through the data interface.

8. The bus and subway combined ticket risk control method based on hierarchical scanning according to claim 7 is characterized in that: The method further comprises: Obtain risk control and disposal feedback data; Based on statistical analysis data and risk control feedback data, the parameters in the free qualification determination rules, single-day ride frequency threshold, parallel ride behavior rules, short-time and high-frequency ride behavior rules, extra-long ride behavior rules, and cumulative ride frequency threshold are adaptively adjusted.

9. A computer-readable storage medium storing computer program instructions, characterized in that: The computer program instructions implement the method according to any one of claims 1 to 8 when executed by a processor.

10. An electronic device comprising a memory and a processor, characterized in that: The memory is configured to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method according to any one of claims 1 to 8.

Images