Behavior data analysis method, device, equipment and storage medium

Abstract

This application relates to the field of artificial intelligence technology, disclosing a method, apparatus, device, and storage medium for behavioral data analysis. The method includes: acquiring target behavioral data; parsing behavioral element data from the target behavioral data; generating atomic events from the behavioral element data as target atomic events; inputting the target atomic events into a preset complex event state library for complex event state matching to obtain state matching results; and determining behavioral data analysis results based on each of the state matching results. By performing state matching of complex events based on atomic events generated from behavioral element data parsed in real time from the behavioral data, behavioral data is integrated into a unified processing system, thereby enabling behavioral data analysis of user behavioral data and application behavioral data, allowing for the full utilization of behavioral data and fully exploring its data value.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a behavior data analysis method and device, equipment and a storage medium. BACKGROUND

[0002] Behavior data is divided into user behavior data and application behavior data. The user behavior data is used to describe the behavior of the user, and the application behavior data is used to describe the behavior of the application. The traditional behavior data analysis is mainly for the point-in-time data in the user behavior data. The point-in-time data has a unified specification and standard, but the analysis of other types of behavior data is less. For example, in the financial industry, there are a large amount of transaction behavior data, financial behavior data and behavior data other than the point-in-time data related to finance. These behavior data are inconsistent with the specification of the point-in-time data, and there are differences in the frequency of use and application scenarios, which makes it difficult to fully apply the behavior data other than the point-in-time data and to mine the data value of the behavior data other than the point-in-time data. SUMMARY

[0003] The main purpose of the present application is to provide a behavior data analysis method, device, equipment and storage medium, which aims to solve the technical problem that the behavior data analysis of the prior art is mainly for the point-in-time data in the user behavior data. Because the behavior data other than the point-in-time data is inconsistent with the specification of the point-in-time data, and there are differences in the frequency of use and application scenarios, it is difficult to fully apply the behavior data other than the point-in-time data and to mine the data value of the behavior data other than the point-in-time data.

[0004] In order to achieve the above-mentioned purpose of the application, the present application provides a behavior data analysis method, which comprises:

[0005] acquiring target behavior data;

[0006] parsing behavior element data from the target behavior data;

[0007] generating the behavior element data as an atomic event as a target atomic event;

[0008] inputting the target atomic event into a preset complex event state library for state matching of complex events to obtain a state matching result;

[0009] determining a behavior data analysis result according to each state matching result.

[0010] Further, the step of parsing behavior element data from the target behavior data comprises:

[0011] Resolving a parameter initial value corresponding to each behavior element parameter from the target behavior data, the behavior element parameter including: subject identification, behavior time, behavior location, behavior mode and behavior content;

[0012] Obtaining one of the behavior element parameters as a to-be-processed parameter, and obtaining the parameter initial value corresponding to the to-be-processed parameter as a to-be-analyzed initial value;

[0013] Obtaining a conversion type corresponding to the to-be-processed parameter from a preset conversion type table as a target conversion type;

[0014] When the target conversion type is format conversion, obtaining a conversion format corresponding to the to-be-processed parameter from the conversion type table as a target conversion format, and converting the to-be-analyzed initial value according to the target conversion format to obtain a parameter target value corresponding to the to-be-processed parameter;

[0015] When the target conversion type is data mapping, obtaining a mapping table corresponding to the to-be-processed parameter as a target mapping table, performing original value lookup on the to-be-analyzed initial value in the target mapping table, and taking a mapping value corresponding to the original value found in the target mapping table as the parameter target value corresponding to the to-be-processed parameter;

[0016] When the target conversion type is data query, obtaining query configuration data corresponding to the to-be-processed parameter, and obtaining data corresponding to the to-be-analyzed initial value from a preset target database through the query configuration data as the parameter target value corresponding to the to-be-processed parameter;

[0017] Repeating the step of obtaining one of the behavior element parameters as a to-be-processed parameter until the parameter target value corresponding to each behavior element parameter is determined;

[0018] Taking each parameter target value as the behavior element data.

[0019] Further, the target behavior data is in JSON message format, and the step of resolving a parameter initial value corresponding to each behavior element parameter from the target behavior data includes:

[0020] Resolving a JSON original variable corresponding to each behavior element parameter from the target behavior data;

[0021] Performing preset format conversion on each JSON original variable to obtain the parameter initial value corresponding to each behavior element parameter.

[0022] Furthermore, the step of inputting the target atomic event into a preset complex event state library for complex event state matching to obtain the state matching result includes:

[0023] Search a preset list of complex events for the complex event identifier that corresponds to the target atomic event, and use it as a set of complex event identifiers;

[0024] Based on the target atomic event and each complex event identifier in the complex event identifier set, keywords are generated to obtain each retrieval keyword corresponding to the complex event identifier set;

[0025] Each of the search keywords is searched in the complex event state database to obtain the search results;

[0026] When there is a failed search result, the complex event and the search keyword corresponding to each failed search result are added as associated data to the complex event state library. The target atomic event is input into each complex event corresponding to the failed search result in the complex event state library for state matching to obtain the state matching result.

[0027] When a successful search result exists, each complex event in the complex event state library corresponding to the successful search result is taken as a complex event to be matched. The target atomic event is input into each complex event to be matched for state matching to obtain the state matching result.

[0028] Furthermore, the step of determining the behavioral data analysis results based on each of the state matching results includes:

[0029] Each complex event in the complex event state library that corresponds to a successful state matching result is taken as a hit complex event;

[0030] The output of the subject identifier and event attribute of each of the aforementioned complex events is used as the behavioral data analysis result.

[0031] Furthermore, after the step of determining the behavioral data analysis result based on each of the state matching results, the method further includes:

[0032] Retrieve subscription configuration dataset;

[0033] The filtering conditions for each subscription configuration data in the subscription configuration dataset are used to filter data from the behavioral data analysis results to obtain the analysis results to be sent for each subscription configuration data.

[0034] The analysis results to be sent are sent to the downstream application corresponding to the subscription configuration data.

[0035] Furthermore, after the step of determining the behavioral data analysis result based on each of the state matching results, the method further includes:

[0036] Retrieve the historical subscription configuration dataset corresponding to the preset list of complex events;

[0037] Using the historical subscription configuration dataset, determine whether each complex event corresponding to the preset complex event list has been subscribed to within a preset duration.

[0038] If not, the complex events that have not been subscribed to within the preset duration will be removed from the complex event list.

[0039] This application also proposes a behavioral data analysis device, the device comprising:

[0040] The data acquisition module is used to acquire target behavior data;

[0041] A behavior element data determination module is used to parse behavior element data from the target behavior data;

[0042] The target atomic event determination module is used to generate atomic events from the behavioral element data, which are then used as target atomic events.

[0043] The state matching result determination module is used to input the target atomic event into a preset complex event state library to perform state matching of complex events and obtain the state matching result;

[0044] The behavioral data analysis result determination module is used to determine the behavioral data analysis result based on the matching results of each state.

[0045] This application also proposes a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods described above.

[0046] This application also proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0047] This application discloses a behavioral data analysis method, apparatus, device, and storage medium. The method involves acquiring target behavioral data; parsing behavioral element data from the target behavioral data; generating atomic events from the behavioral element data as target atomic events; inputting the target atomic events into a preset complex event state library for complex event state matching to obtain state matching results; and determining behavioral data analysis results based on each state matching result. By performing complex event state matching based on atomic events generated from behavioral element data parsed in real-time from the behavioral data, behavioral data is integrated into a unified processing system. This enables behavioral data analysis of user behavioral data and application behavioral data, allowing for the full utilization and exploitation of the data value of user behavioral data and application behavioral data. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating a behavioral data analysis method according to an embodiment of this application;

[0049] Figure 2 This is a schematic block diagram of the structure of a behavioral data analysis device according to an embodiment of this application;

[0050] Figure 3 This is a schematic block diagram of the structure of a computer device according to an embodiment of this application.

[0051] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0053] Reference Figure 1 This application provides a behavioral data analysis method, the method comprising:

[0054] S1: Obtain target behavior data;

[0055] S2: Extract behavioral element data from the target behavioral data;

[0056] S3: Generate atomic events from the behavioral element data, and use them as target atomic events;

[0057] S4: Input the target atomic event into a preset complex event state library to perform complex event state matching and obtain the state matching result;

[0058] S5: Determine the behavioral data analysis results based on the matching results of each state.

[0059] This embodiment performs state matching of complex events based on atomic events generated from behavioral element data parsed in real time from behavioral data, integrating behavioral data into a unified processing system. This enables behavioral data analysis of user behavior data and application behavior data, allowing for full utilization of behavioral data and fully exploring its data value.

[0060] For S1, behavioral data can be obtained from the message queue as target behavioral data, or behavioral data sent directly by the target application can be obtained as target behavioral data.

[0061] The target applications include: clients, application plugins, and one or more application systems.

[0062] Message queues include one or more of the following: RocketMQ (a distributed message middleware), RabbitMQ (a message-oriented middleware), and Kafka (a high-throughput distributed publish-subscribe messaging system).

[0063] Understandably, this application can be applied to the financial sector.

[0064] Behavioral data includes user behavior data and / or application behavior data. User behavior data includes both event-tracked and non-event-tracked behavioral data. When this application is applied to the financial field, non-event-tracked behavioral data includes, but is not limited to, transaction behavior data, wealth management behavior data, and other finance-related behavioral data.

[0065] Application behavioral data includes, but is not limited to: data synchronization behavioral data.

[0066] Optionally, behavioral data in JSON (Lightweight Data Interchange Format) message format can be retrieved from the message queue as the target behavioral data.

[0067] For S2, the behavioral element data is represented as 4W1H data.

[0068] The subject identifier, behavior time, behavior location, behavior method, and behavior content are parsed from the target behavior data. The parsed subject identifier (WHO in 4W1H), behavior time (WHEN in 4W1H), behavior location (WERE in 4W1H), behavior method (HOW in 4W1H), and behavior content (WHAT in 4W1H) are used as behavior element data. The subject identifier can be data that uniquely identifies a subject, such as subject name or subject ID. The subject can be a user or an application.

[0069] For example, when the target behavior data is transfer behavior data, the subject identifier (i.e., the account identifier of the transferor), behavior time (i.e., the time when the transfer was successful), behavior location (i.e., the longitude and latitude of the device used for the transfer), behavior method (i.e., self-service transfer machine A11), and behavior content (i.e., the transfer transaction) are parsed from the target behavior data as behavior element data. This example is not specifically limited.

[0070] For S3, a preset atomic event generation rule is used to generate atomic events from the behavioral element data, and the generated atomic events are used as target atomic events.

[0071] An atomic event, also known as a basic event or a simple event, is an event that has only a single result in the sample space.

[0072] An atomic event consists of an atomic event identifier and atomic event description data. The atomic event description data is essentially descriptive data about the behavioral element data.

[0073] For S4, the target atomic event is input into the complex event state library, and each complex event related to the target atomic event is matched for state matching to obtain the state matching result corresponding to each complex event related to the target atomic event. When the logical operation relationship and window calculation relationship between the atomic events in the event description information of the complex event are satisfied, the state matching result is determined to be successful; otherwise, the state matching result is determined to be unsuccessful.

[0074] Complex events involve summarizing and exploring behavioral patterns. This involves comparing and providing feedback on multiple behavioral patterns to identify typical patterns, applying discovered patterns, or specifying particular patterns for application. Finally, a matching identifier is issued. A complex event is a set of atomic events, containing multiple atomic events. Processing complex events involves identifying user-defined meaningful events within this set of atomic events. Atomic events in the set do not contain specific parameter values.

[0075] Complex events include: a complex event identifier, an event description, and attribute output configuration data. The complex event identifier can be a unique identifier for a complex event, such as a complex event name or a complex event ID. The event description describes the logical operations (AND, OR, NOT) between the atomic events, as well as the window calculation relationships. The attribute output configuration data is the configuration data for the attributes to be output, which originate from the atomic events.

[0076] The complex event state library includes event keywords and complex event state data. Complex event state data is data on the state status of a complex event.

[0077] For S5, the matched complex events are determined based on the matching results of each state, and the behavioral data analysis results are determined based on the hit subject identifier and event attribute output results of each matched complex event.

[0078] A hit complex event is a complex event in which the logical operation relationship and window calculation relationship between the atomic events in the event description information are satisfied.

[0079] In one embodiment, the step of parsing behavioral element data from the target behavioral data includes:

[0080] S21: Parse the initial value of each behavior element parameter from the target behavior data. The behavior element parameters include: subject identifier, behavior time, behavior location, behavior method, and behavior content.

[0081] S22: Obtain one of the behavioral element parameters as a parameter to be processed, and use the initial value of the parameter corresponding to the parameter to be processed as the initial value to be analyzed;

[0082] S23: Obtain the conversion type corresponding to the parameter to be processed from the preset conversion type table, and use it as the target conversion type;

[0083] S24: When the target conversion type is format conversion, obtain the conversion format corresponding to the parameter to be processed from the conversion type table as the target conversion format, and convert the initial value to be analyzed according to the target conversion format to obtain the parameter target value corresponding to the parameter to be processed;

[0084] S25: When the target transformation type is data mapping, obtain the mapping table corresponding to the parameter to be processed as the target mapping table, search for the original value of the initial value to be analyzed in the target mapping table, and use the mapping value corresponding to the original value found in the target mapping table as the parameter target value corresponding to the parameter to be processed.

[0085] S26: When the target conversion type is data query, obtain the query configuration data corresponding to the parameter to be processed, and obtain the data corresponding to the initial value to be analyzed from the preset target database through the query configuration data, as the parameter target value corresponding to the parameter to be processed;

[0086] S27: Repeat the step of obtaining one of the behavioral element parameters as a parameter to be processed until the target value of the parameter corresponding to each of the behavioral element parameters is determined;

[0087] S28: Use the target values ​​of each parameter as the behavioral element data.

[0088] This embodiment first parses the initial value of each behavioral element parameter from the target behavioral data, and then transforms the initial value of the parameter according to the transformation type of the behavioral element parameter. This provides a foundation for generating behavioral element data with unified processing specifications and for performing behavioral data analysis on user behavior data and application behavior data.

[0089] For S21, the keyword identifier corresponding to each behavioral element parameter is searched from the target behavioral data, and the value corresponding to each keyword identifier found is used as the initial value of the parameter.

[0090] The initial value of the parameter can be any one of the following: the initial value of the subject identifier, the initial value of the behavior time, the initial value of the behavior location, the initial value of the behavior method, and the initial value of the behavior content.

[0091] For S22, any one of the behavioral element parameters is obtained as a parameter to be processed. The initial value of the parameter to be processed is used as the initial value to be analyzed, thus providing a basis for subsequent format conversion of the initial value of the parameter to be processed.

[0092] For S23, the conversion type corresponding to the parameter to be processed is obtained from the preset conversion type table, and the obtained conversion type is used as the target conversion type.

[0093] The conversion type table includes: behavioral element parameters, conversion type, and conversion format. The possible values ​​for conversion types include: format conversion, data mapping, and data query. The possible values ​​for conversion formats include, but are not limited to: time format and currency format.

[0094] For S24, when the target conversion type is format conversion, it means that the initial value to be analyzed needs to be format converted. Therefore, the conversion format corresponding to the parameter to be processed is obtained from the conversion type table, the obtained conversion format is used as the target conversion format, the initial value to be analyzed is converted according to the target conversion format, and the converted data is used as the parameter target value corresponding to the parameter to be processed.

[0095] For example, when the parameter to be processed is behavior time, the conversion format corresponding to the parameter to be processed is time format. It is necessary to convert the initial value to be analyzed into time format data, and use the converted time format data as the parameter target value corresponding to the parameter to be processed. This example does not make specific limitations.

[0096] For S25, when the target transformation type is data mapping, it means that the initial value to be analyzed needs to find a mapping value. Therefore, a mapping table corresponding to the parameter to be processed is obtained from the database, and the obtained mapping table is used as the target mapping table. The initial value to be analyzed is used to find the original value in the target mapping table, and the mapping value corresponding to the original value found in the target mapping table is used as the parameter target value corresponding to the parameter to be processed, thereby realizing the mapping value of the initial value to be analyzed as the parameter target value corresponding to the parameter to be processed.

[0097] The mapping table includes: original values ​​and mapped values.

[0098] For S26, when the target transformation type is data query, it means that data needs to be queried based on the initial value to be analyzed. The queried data is used as the target value of the parameter to be processed. Therefore, configuration data corresponding to the parameter to be processed is obtained from the configuration database as query configuration data. Through the query configuration data, data corresponding to the initial value to be analyzed is obtained from the target database corresponding to the database link address of the query configuration data, and the obtained data is used as the target value of the parameter to be processed.

[0099] The query configuration data includes: database link address, table identifier, query field identifier, and output field identifier set. A table identifier can be a unique identifier for a table, such as a table name or table ID. A query field identifier is the identifier for the field to be queried. A field identifier can be a unique identifier for a field, such as a field name or field ID. The output field identifier set includes a set of field identifiers and field assembly rules, used to describe the assembly rules for the fields and field values ​​to be output. The field identifier set includes one or more field identifiers.

[0100] For S27, the step of obtaining one of the behavioral element parameters as a parameter to be processed is repeated, that is, steps S22 to S27 are repeated until the target value of the parameter corresponding to each behavioral element parameter is determined.

[0101] In one embodiment, the target behavior data is in JSON message format, and the step of parsing the initial value of each behavior element parameter from the target behavior data includes:

[0102] S211: Parse the JSON raw variables corresponding to each of the behavioral element parameters from the target behavioral data;

[0103] S212: Convert each of the original JSON variables into a preset format to obtain the initial value of the parameter corresponding to each behavioral element parameter.

[0104] In this embodiment, the target behavior data is in JSON message format. JSON message format has a smaller data volume, which helps to improve the transmission speed of behavior data.

[0105] For S211, the JSON parameter value corresponding to each behavior element parameter is parsed from the target behavior data using a JSON parser, and an original variable is created for each JSON parameter value as the original JSON variable.

[0106] For S212, each of the original JSON variables is converted from JSON message format to character format, and the data obtained after format conversion is used as the initial value of the parameter corresponding to each of the behavioral element parameters.

[0107] In one embodiment, the step of inputting the target atomic event into a preset complex event state library for complex event state matching to obtain a state matching result includes:

[0108] S41: Search for the complex event identifier corresponding to the target atomic event from the preset complex event list, and use it as a complex event identifier set;

[0109] S42: Generate keywords based on the target atomic event and each complex event identifier in the complex event identifier set to obtain each search keyword corresponding to the complex event identifier set;

[0110] S43: Search for each of the search keywords in the complex event state database to obtain the search results;

[0111] S44: When there is a failed search result, the complex event and the search keyword corresponding to each failed search result are added as associated data to the complex event state library. The target atomic event is input into each complex event corresponding to the failed search result in the complex event state library for state matching to obtain the state matching result.

[0112] S45: When there is a successful search result, each complex event in the complex event state library corresponding to the successful search result is taken as a complex event to be matched, and the target atomic event is input into each complex event to be matched for state matching to obtain the state matching result.

[0113] In this embodiment, the complex event identifier corresponding to the target atomic event is searched from the complex event list. Then, search keywords are generated based on the found complex event identifier. Finally, the target atomic event is input into the complex event state database for complex event state matching based on the search keywords, which provides a basis for determining the matched complex event.

[0114] For S41, the complex event list includes: complex event identifiers and a set of atomic event identifiers, wherein the set of atomic event identifiers includes one or more atomic event identifiers.

[0115] Specifically, the atomic event identifier corresponding to the target atomic event is searched from each set of atomic event identifiers in a preset list of complex events, and the complex event identifiers corresponding to each set of atomic event identifiers containing the atomic event identifier corresponding to the target atomic event are taken as complex event identifier sets.

[0116] For S42, keywords are generated based on the subject identifier of the target atomic event and each complex event identifier in the complex event identifier set, and each generated keyword is used as a search keyword. That is, the number of search keywords is the same as the number of complex event identifiers in the complex event identifier set.

[0117] For S43, each of the search keywords is searched in the various event keywords of the complex event state database. The search result corresponding to the search keyword that is found is determined to be successful, and the search result corresponding to the search keyword that is not found is determined to be unsuccessful.

[0118] For S44, when there is a failed search result, it means that the complex event corresponding to each failed search result has not been added to the complex event state library. Therefore, firstly, the complex event corresponding to each failed search result and the search keyword are added to the complex event state library as associated data. Then, the target atomic event is input into the complex event state data corresponding to each failed search result in the complex event state library for state matching to obtain the state matching result, so as to realize the state matching of the newly added complex event.

[0119] For S45, when there is a successful search result, it means that each complex event corresponding to the successful search result already exists in the complex event state library. Therefore, each complex event in the complex event state library corresponding to the successful search result is taken as a complex event to be matched, so as to find each complex event in the complex event state library corresponding to the target atomic event. The target atomic event is input into the complex event state data corresponding to each complex event to be matched for state matching, and the state matching result is obtained, which provides a basis for determining the hit complex event.

[0120] In one embodiment, the step of determining the behavioral data analysis result based on each of the state matching results includes:

[0121] S51: Each complex event in the complex event state library that corresponds to a successful state matching result is taken as a hit complex event;

[0122] S52: The output results of the hit subject identifier and event attribute of each of the aforementioned hit complex events are used as the behavioral data analysis results.

[0123] In this embodiment, each complex event corresponding to a successful state matching result in the complex event state library is first taken as a hit complex event. Then, the output of each hit complex event is taken as the behavioral data analysis result. This realizes the state matching of complex events and the determination of hit complex events based on atomic events generated from behavioral element data parsed in real time from behavioral data. This allows behavioral data to be fully utilized and fully explores the data value of user behavioral data and application behavioral data.

[0124] For S52, a preset result combination rule is adopted to combine the hit subject identifier and event attribute output results of each hit complex event, and the combined data is used as the behavioral data analysis result.

[0125] The event attribute output result is obtained by combining the data obtained from the complex event status data corresponding to the complex event, based on the attribute output configuration data of the complex event.

[0126] Optionally, the result combination rule is to generate a list based on the hit order.

[0127] Optionally, after the step of using the output results of the hit subject identifier and event attributes of each of the hit complex events as the behavioral data analysis results, the method further includes: deleting each of the hit complex events from the complex event state database.

[0128] In one embodiment, after the step of determining the behavioral data analysis result based on each of the state matching results, the method further includes:

[0129] S61: Obtain the subscription configuration dataset;

[0130] S62: Filter the data from the behavioral data analysis results using the filtering conditions of each subscription configuration data in the subscription configuration dataset to obtain the analysis results to be sent for each subscription configuration data;

[0131] S63: Send the analysis results to be distributed to the downstream application corresponding to the subscription configuration data.

[0132] This embodiment enables the filtering of data from the behavioral data analysis results based on filtering criteria for distribution, thereby improving the accuracy of distribution and increasing user satisfaction.

[0133] For S61, the subscription configuration dataset can be obtained from the database.

[0134] The subscription configuration dataset includes one or more subscription configuration data entries.

[0135] Subscription configuration data includes: application identifier and filter criteria. The application identifier can be data that uniquely identifies an application, such as the application name or application ID.

[0136] For S62, the filtering conditions for each subscription configuration data in the subscription configuration dataset are used to filter data from the behavioral data analysis results, and the data obtained for each subscription configuration data is used as an analysis result to be sent out.

[0137] For S63, the analysis result to be sent is sent to the downstream application corresponding to the application identifier of the subscription configuration data corresponding to the analysis result to be sent.

[0138] In one embodiment, after the step of determining the behavioral data analysis result based on each of the state matching results, the method further includes:

[0139] S71: Obtain the historical subscription configuration dataset corresponding to the preset list of complex events;

[0140] S72: Using the historical subscription configuration dataset, determine whether each complex event corresponding to the preset complex event list has been subscribed to within a preset duration;

[0141] S73: If not, the complex events that have not been subscribed to within the preset duration will be removed from the complex event list.

[0142] This embodiment enables the automatic deletion of complex events that have not been subscribed to for a long time, saving computing resources.

[0143] For S71, the historical subscription configuration dataset corresponding to the preset complex event list can be obtained from the database.

[0144] Historical subscription configuration dataset.

[0145] For S72, the complex event identifier corresponding to each complex event in the preset complex event list is searched for in the historical subscription configuration dataset. The expiration time of the subscription configuration data corresponding to the event identifier found in the historical subscription configuration dataset is taken as the single configuration expiration time. The latest time of each single configuration expiration time corresponding to each complex event in the preset complex event list is taken as the latest subscription time corresponding to each complex event in the preset complex event list. Complex events whose interval between the latest subscription time and the current time is within a preset time are determined to be subscribed within the preset time. Complex events whose interval between the latest subscription time and the current time is outside the preset time are determined to be not subscribed within the preset time.

[0146] For S73, if not, that is, if it is not subscribed within a preset time period, then the complex event that is not subscribed within the preset time period will be deleted from the complex event list, so that the complex event list always retains active complex events.

[0147] Reference Figure 2 This application also proposes a behavioral data analysis device, the device comprising:

[0148] Data acquisition module 100 is used to acquire target behavior data;

[0149] The behavior element data determination module 200 is used to parse behavior element data from the target behavior data;

[0150] The target atomic event determination module 300 is used to generate atomic events from the behavioral element data as target atomic events.

[0151] The state matching result determination module 400 is used to input the target atomic event into a preset complex event state library to perform state matching of complex events and obtain a state matching result.

[0152] The behavior data analysis result determination module 500 is used to determine the behavior data analysis result based on the matching results of each state.

[0153] This embodiment performs state matching of complex events based on atomic events generated from behavioral element data parsed in real time from behavioral data, integrating behavioral data into a unified processing system. This enables behavioral data analysis of user behavior data and application behavior data, allowing for full utilization of behavioral data and fully exploring its data value.

[0154] In one embodiment, the behavioral element data determination module 200 includes: a parameter initial value determination submodule and

[0155] The initial values ​​of each behavior element parameter are parsed from the target behavior data. The behavior element parameters include: subject identifier, behavior time, behavior location, behavior method, and behavior content.

[0156] Obtain one of the behavioral element parameters as a parameter to be processed, and use the initial value of the parameter corresponding to the parameter to be processed as the initial value to be analyzed.

[0157] Obtain the conversion type corresponding to the parameter to be processed from the preset conversion type table, and use it as the target conversion type;

[0158] When the target conversion type is format conversion, the conversion format corresponding to the parameter to be processed is obtained from the conversion type table and used as the target conversion format. The initial value to be analyzed is converted according to the target conversion format to obtain the parameter target value corresponding to the parameter to be processed.

[0159] When the target transformation type is data mapping, obtain the mapping table corresponding to the parameter to be processed as the target mapping table, search for the original value of the initial value to be analyzed in the target mapping table, and use the mapping value corresponding to the original value found in the target mapping table as the parameter target value corresponding to the parameter to be processed.

[0160] When the target conversion type is data query, obtain the query configuration data corresponding to the parameter to be processed, and obtain the data corresponding to the initial value to be analyzed from the preset target database through the query configuration data, as the parameter target value corresponding to the parameter to be processed;

[0161] Repeat the step of obtaining one of the behavioral element parameters as a parameter to be processed until the target value of the parameter corresponding to each of the behavioral element parameters is determined.

[0162] The target values ​​of each parameter are used as the behavioral element data.

[0163] In one embodiment, the target behavior data is in JSON message format, and the step of parsing the initial value of each behavior element parameter from the target behavior data includes:

[0164] Parse the JSON raw variables corresponding to each of the behavioral element parameters from the target behavioral data;

[0165] Each of the original JSON variables is converted to a preset format to obtain the initial value of the parameter corresponding to each behavioral element parameter.

[0166] In one embodiment, the step of inputting the target atomic event into a preset complex event state library for complex event state matching to obtain a state matching result includes:

[0167] Search a preset list of complex events for the complex event identifier that corresponds to the target atomic event, and use it as a set of complex event identifiers;

[0168] Based on the target atomic event and each complex event identifier in the complex event identifier set, keywords are generated to obtain each retrieval keyword corresponding to the complex event identifier set;

[0169] Each of the search keywords is searched in the complex event state database to obtain the search results;

[0170] When there is a failed search result, the complex event and the search keyword corresponding to each failed search result are added as associated data to the complex event state library. The target atomic event is input into each complex event corresponding to the failed search result in the complex event state library for state matching to obtain the state matching result.

[0171] When a successful search result exists, each complex event in the complex event state library corresponding to the successful search result is taken as a complex event to be matched. The target atomic event is input into each complex event to be matched for state matching to obtain the state matching result.

[0172] In one embodiment, the step of determining the behavioral data analysis result based on each of the state matching results includes:

[0173] Each complex event in the complex event state library that corresponds to a successful state matching result is taken as a hit complex event;

[0174] The output of the subject identifier and event attribute of each of the aforementioned complex events is used as the behavioral data analysis result.

[0175] In one embodiment, after the step of determining the behavioral data analysis result based on each of the state matching results, the method further includes:

[0176] Retrieve subscription configuration dataset;

[0177] The filtering conditions for each subscription configuration data in the subscription configuration dataset are used to filter data from the behavioral data analysis results to obtain the analysis results to be sent for each subscription configuration data.

[0178] The analysis results to be sent are sent to the downstream application corresponding to the subscription configuration data.

[0179] In one embodiment, after the step of determining the behavioral data analysis result based on each of the state matching results, the method further includes:

[0180] Retrieve the historical subscription configuration dataset corresponding to the preset list of complex events;

[0181] Using the historical subscription configuration dataset, determine whether each complex event corresponding to the preset complex event list has been subscribed to within a preset duration.

[0182] If not, the complex events that have not been subscribed to within the preset duration will be removed from the complex event list.

[0183] Reference Figure 3 This application also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 3 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data such as behavioral data analysis methods. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a behavioral data analysis method. The behavioral data analysis method includes: acquiring target behavioral data; parsing behavioral element data from the target behavioral data; generating atomic events from the behavioral element data as target atomic events; inputting the target atomic events into a preset complex event state library for complex event state matching to obtain state matching results; and determining behavioral data analysis results based on each state matching result.

[0184] This embodiment performs state matching of complex events based on atomic events generated from behavioral element data parsed in real time from behavioral data, integrating behavioral data into a unified processing system. This enables behavioral data analysis of user behavior data and application behavior data, allowing for full utilization of behavioral data and fully exploring its data value.

[0185] One embodiment of this application also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements a behavior data analysis method, including the steps of: acquiring target behavior data; parsing behavior element data from the target behavior data; generating atomic events from the behavior element data as target atomic events; inputting the target atomic events into a preset complex event state library for complex event state matching to obtain state matching results; and determining behavior data analysis results based on each of the state matching results.

[0186] The aforementioned behavioral data analysis method performs state matching of complex events based on atomic events generated from behavioral element data parsed in real time from behavioral data. This integrates behavioral data into a unified processing system, enabling behavioral data analysis of user behavior data and application behavior data. This allows for the full utilization and exploitation of the data value of user behavior data and application behavior data.

[0187] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media provided in this application and in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0188] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0189] The above description is only a preferred embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural changes made based on the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A behavioral data analysis method, characterized in that, The method includes: Obtain target behavior data; Extract behavioral element data from the target behavioral data; The behavioral element data is used to generate atomic events, which are then used as target atomic events. The target atomic event is input into a preset complex event state library for complex event state matching to obtain the state matching result; The behavioral data analysis results are determined based on the matching results of each state. The step of parsing behavioral element data from the target behavioral data includes: The initial values ​​of each behavior element parameter are parsed from the target behavior data. The behavior element parameters include: subject identifier, behavior time, behavior location, behavior method, and behavior content. Obtain one of the behavioral element parameters as a parameter to be processed, and use the initial value of the parameter corresponding to the parameter to be processed as the initial value to be analyzed. Obtain the conversion type corresponding to the parameter to be processed from the preset conversion type table, and use it as the target conversion type; When the target conversion type is format conversion, the conversion format corresponding to the parameter to be processed is obtained from the conversion type table and used as the target conversion format. The initial value to be analyzed is converted according to the target conversion format to obtain the parameter target value corresponding to the parameter to be processed. When the target transformation type is data mapping, obtain the mapping table corresponding to the parameter to be processed as the target mapping table, search for the original value of the initial value to be analyzed in the target mapping table, and use the mapping value corresponding to the original value found in the target mapping table as the parameter target value corresponding to the parameter to be processed. When the target conversion type is data query, obtain the query configuration data corresponding to the parameter to be processed, and obtain the data corresponding to the initial value to be analyzed from the preset target database through the query configuration data, as the parameter target value corresponding to the parameter to be processed; Repeat the step of obtaining one of the behavioral element parameters as a parameter to be processed until the target value of the parameter corresponding to each of the behavioral element parameters is determined. The target values ​​of each parameter are used as the behavioral element data; After the step of determining the behavioral data analysis results based on each of the state matching results, the method further includes: Retrieve the historical subscription configuration dataset corresponding to the preset list of complex events; Using the historical subscription configuration dataset, determine whether each complex event corresponding to the preset complex event list has been subscribed to within a preset time period. If not, the complex events that have not been subscribed to within the preset duration will be removed from the complex event list.

2. The behavioral data analysis method according to claim 1, characterized in that, The target behavior data is in JSON message format. The step of parsing the initial parameter values ​​corresponding to each behavior element parameter from the target behavior data includes: Parse the JSON raw variables corresponding to each of the behavioral element parameters from the target behavioral data; Each of the original JSON variables is converted to a preset format to obtain the initial value of the parameter corresponding to each behavioral element parameter.

3. The behavioral data analysis method according to claim 1, characterized in that, The step of inputting the target atomic event into a preset complex event state library for complex event state matching to obtain the state matching result includes: Search a preset list of complex events for the complex event identifier that corresponds to the target atomic event, and use it as a set of complex event identifiers; Based on the target atomic event and each complex event identifier in the complex event identifier set, keywords are generated to obtain each retrieval keyword corresponding to the complex event identifier set; Each of the search keywords is searched in the complex event state database to obtain the search results; When there is a failed search result, the complex event and the search keyword corresponding to each failed search result are added as associated data to the complex event state library. The target atomic event is input into each complex event corresponding to the failed search result in the complex event state library for state matching to obtain the state matching result. When a successful search result exists, each complex event in the complex event state library corresponding to the successful search result is taken as a complex event to be matched, and the target atomic event is input into each complex event to be matched for state matching to obtain the state matching result.

4. The behavioral data analysis method according to claim 1, characterized in that, The step of determining the behavioral data analysis results based on each of the state matching results includes: Each complex event in the complex event state library that corresponds to a successful state matching result is taken as a hit complex event; The output of the subject identifier and event attribute of each of the aforementioned complex events is used as the behavioral data analysis result.

5. The behavioral data analysis method according to claim 1, characterized in that, After the step of determining the behavioral data analysis results based on each of the state matching results, the method further includes: Retrieve subscription configuration dataset; The filtering conditions for each subscription configuration data in the subscription configuration dataset are used to filter data from the behavioral data analysis results to obtain the analysis results to be sent for each subscription configuration data. The analysis results to be sent are sent to the downstream application corresponding to the subscription configuration data.

6. A behavioral data analysis apparatus for performing the method according to any one of claims 1-5, characterized in that, The device includes: The data acquisition module is used to acquire target behavior data; A behavior element data determination module is used to parse behavior element data from the target behavior data; The target atomic event determination module is used to generate atomic events from the behavioral element data, which are then used as target atomic events. The state matching result determination module is used to input the target atomic event into a preset complex event state library to perform state matching of complex events and obtain the state matching result; The behavioral data analysis result determination module is used to determine the behavioral data analysis result based on the matching results of each state.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images