CN113486983A

CN113486983A - Big data office information analysis method and system for anti-fraud processing

Info

Publication number: CN113486983A
Application number: CN202110878011.XA
Authority: CN
Inventors: 冯春梅; 赵琦
Original assignee: Dongguan Daojiao Hongnuo Computer Technology Development Service Center
Current assignee: Dongguan Daojiao Hongnuo Computer Technology Development Service Center
Priority date: 2021-08-02
Filing date: 2021-08-02
Publication date: 2021-10-08

Abstract

The application relates to a big data office information analysis method and a system for anti-fraud processing, wherein scene office behavior data covering migration office scene data in cloud office behavior information with anti-fraud correlation analysis requirements are obtained, and then anti-fraud correlation analysis of the scene office behavior data is carried out through the scene office behavior data covering the migration office scene data, so that the information of the office scene level of the cloud office behavior information can be taken into consideration in the anti-fraud correlation analysis process of the scene office behavior data, the integrity of the anti-fraud correlation analysis condition can be ensured as much as possible in the process of mining the first anti-fraud correlation analysis condition, omission of the anti-fraud correlation analysis caused by omission of the office scene is avoided, and thus, when anti-fraud detection is carried out on the cloud office behavior information subsequently, the method can be based on the complete and accurate anti-fraud correlation analysis condition so as to ensure the accuracy and credibility of anti-fraud detection.

Description

Big data office information analysis method and system for anti-fraud processing

Technical Field

The embodiment of the application relates to the technical field of anti-fraud and big data office, in particular to a big data office information analysis method and system for anti-fraud processing.

Background

Data information security issues have always been a major concern in the big data era. With the continuous expansion of the fields related to big data intelligence, the data information security problem faced by these fields is also not negligible. Taking the office field as an example, the online office has the operation advantages of higher efficiency and lower cost due to the addition of big data, cloud computing and artificial intelligence. However, security of office data information and the associated anti-fraud measures are also challenges that the online office needs face.

In general, anti-fraud processing of office information is to avoid loss of office data information or to prevent important software programs from being tampered, and a prerequisite for execution of anti-fraud processing is to detect a risk. However, the inventor finds that, in the research process, when the related technology carries out information risk detection, the detection basis used has the technical problems of incomplete content and low richness, so that the accuracy and the reliability of anti-fraud detection are difficult to ensure.

Disclosure of Invention

In view of this, the present application provides a big data office information analysis method and system for anti-fraud processing.

The embodiment of the application provides a big data office information analysis method for anti-fraud processing, which is applied to a big data office information analysis system, and the method comprises the following steps:

collecting a set number of groups of cloud office behavior information with anti-fraud correlation analysis requirements;

performing collaborative behavior item analysis on each group of the cloud office behavior information one by one to obtain an office collaborative item record of each group of the cloud office behavior information, wherein the office collaborative item record comprises a plurality of first active collaborative items;

mapping the first active cooperation item to obtain first scene office behavior data of the migration office scene data carrying the cloud office behavior information;

and performing anti-fraud correlation analysis on the first scene office behavior data in the set quantity group of cloud office behavior information with anti-fraud correlation analysis requirements to obtain a first anti-fraud correlation analysis condition of the set quantity group of cloud office behavior information with anti-fraud correlation analysis requirements.

Under some design ideas which can be independently implemented, the office collaboration item records comprise a fuzzy record subset and a clear record subset, quantitative significance evaluation data of the fuzzy record subset is smaller than that of the clear record subset, the office collaboration item corresponding to the fuzzy record subset is the first active collaboration item, and the office collaboration item corresponding to the clear record subset is the second active collaboration item;

after the anti-fraud association analysis is performed on the first scene-based office behavior data in the set number of sets of cloud office behavior information with the anti-fraud association analysis requirement, and a first anti-fraud association analysis condition of the set number of sets of cloud office behavior information with the anti-fraud association analysis requirement is obtained, the method further includes:

extracting a related cooperation item cluster from the set number of clear record subsets of the cloud office behavior information with anti-fraud related analysis requirements based on the first anti-fraud related analysis condition, wherein the related cooperation item cluster comprises a set number of target item sub-clusters, and each target item sub-cluster comprises a plurality of second active cooperation items extracted from a set of clear record subsets of the cloud office behavior information;

and performing anti-fraud association analysis on second scenario-based office behavior data corresponding to the relevance cooperation item cluster to obtain a second anti-fraud association analysis condition of the set number of groups of cloud office behavior information with anti-fraud association analysis requirements, wherein the second scenario-based office behavior data is a second activity cooperation item in the relevance cooperation item cluster or generated by mapping the second activity cooperation item in the relevance cooperation item cluster.

Under some design ideas which can be independently implemented, before performing anti-fraud association analysis on the second scenario-based office behavior data corresponding to the relevance cooperation item cluster to obtain a second anti-fraud association analysis condition of the set number of groups of cloud office behavior information having anti-fraud association analysis requirements, the method further includes:

and mapping the second active cooperation item in the target item sub-cluster to obtain second scene office behavior data covering the migration office scene data of the target item sub-cluster.

Under some design ideas which can be independently implemented, the mapping processing is performed on the first active collaboration item to obtain first scenized office behavior data of migration office scene data carrying the cloud office behavior information, or the mapping processing is performed on the second active collaboration item in the target item sub-cluster to obtain second scenized office behavior data of the migration office scene data covering the target item sub-cluster, including:

setting the first active cooperation item as first office business operation data, setting the first scene office behavior data as second office business operation data, and setting each group of cloud office behavior information as a specified constraint condition; or, the second active cooperation item is set as first office business operation data, the second scene office behavior data is set as second office business operation data, and each target item sub-cluster is a specified constraint condition;

and obtaining the second office business operation data based on the dynamic arrangement of the first office business operation data in the same specified constraint condition and/or the dynamic arrangement of the first office business operation data in the specified constraint condition with difference.

Under some design ideas which can be independently implemented, the obtaining the second office business operation data based on the dynamic arrangement of the first office business operation data in the same specified constraint and/or the dynamic arrangement of the first office business operation data in the specified constraint with a difference comprises:

and taking each specified constraint condition as a current specified constraint condition one by one, and performing at least one round of content mapping processing on the current specified constraint condition as follows:

taking each first office business operation data in the current specified constraint condition as current office business operation data;

dynamically sorting the current office business operation data in the current specified constraint condition and other first office business operation data to obtain third office business operation data corresponding to the current office business operation data;

dynamically sorting the third office business operation data of the current specified constraint condition and the third office business operation data of other specified constraint conditions to obtain fourth office business operation data corresponding to the current office business operation data;

on the premise that the current round of content mapping processing is not the last round of content mapping processing, the fourth office business operation data is used as the first office business operation data in the next content mapping processing;

and on the premise that the content mapping processing of the current round is the last content mapping processing of the last round, taking the fourth office business operation data as the second office business operation data.

Under some design ideas which can be independently implemented, the step of dynamically arranging the current office business operation data in the current specified constraint condition and the other first office business operation data is realized by an independent hierarchical processing unit in a set thread;

and the step of dynamically arranging the third office business operation data of the current specified constraint condition and the third office business operation data of other specified constraint conditions is realized by an interactive hierarchical processing unit in the set thread.

Under some design ideas which can be independently implemented, the strategy executed by the independent hierarchical processing unit and/or the interactive hierarchical processing unit is a standard hierarchical processing strategy.

Under some design ideas which can be independently implemented, the first scenario-based office behavior data with the anti-fraud association relationship in the set number of sets of cloud office behavior information with the anti-fraud association analysis requirement is an association behavior data set, the distribution descriptions of the association behavior data set in the set number of sets of cloud office behavior information with the anti-fraud association analysis requirement are first distribution descriptions, the first anti-fraud association analysis condition includes distribution description information expressing the first distribution descriptions, and the range covered by the target item sub-cluster in the cloud office behavior information includes the first distribution description;

and/or performing anti-fraud association analysis on second scenario-based office behavior data corresponding to the relevance collaboration item cluster to obtain a second anti-fraud association analysis condition of the set number of groups of cloud office behavior information with anti-fraud association analysis requirements, wherein the second anti-fraud association analysis condition comprises:

taking any one target item sub-cluster of the relevance cooperation item cluster as a calibration sub-cluster, and taking the second scene office behavior data set and distributed and described in the calibration sub-cluster as template scene behavior data;

searching for the second scene office behavior data which has an anti-fraud association relation with the template scene behavior data in other target matter sub-clusters of the relevance cooperation matter cluster;

and obtaining the second anti-fraud association analysis condition based on the template scene behavior data and the second scene office behavior data with the anti-fraud association relationship.

Under some design ideas which can be independently implemented, the extracting, based on the first anti-fraud correlation analysis condition, a correlation and collaboration item cluster from the set number of clear record subsets of the cloud office behavior information having anti-fraud correlation analysis requirements includes:

determining a second distribution description corresponding to the first distribution description in the clear record subset;

and extracting the target item sub-cluster which is based on the second distribution description and has a set size from the clear record subset to obtain the relevance cooperation item cluster.

Under some independently implementable design considerations, the set distribution is described as a reference for the calibration sub-cluster;

and/or searching for the second scenario-based office behavior data having an anti-fraud association relationship with the template scenario behavior data in other target item sub-clusters of the relevance collaboration item cluster, including:

acquiring visual association conditions between the template scene behavior data and each second scene office behavior data in the other target matter sub-clusters;

and searching the second scene office behavior data which has an anti-fraud association relation with the template scene behavior data from the other target item sub-clusters based on the visual association condition.

Under some design ideas which can be independently implemented, the obtaining of the visual association between the template scenario behavior data and each second scenario office behavior data in the other target item sub-cluster includes:

performing visual optimization processing on the template scene behavior data and the second scenized office behavior data in the other target item sub-clusters to obtain a relevance quantization matrix, wherein quantization data described in different distributions in the relevance quantization matrix represent anti-fraud relevance between the template scene behavior data and the second scenized office behavior data with differences;

searching for the second scene office behavior data which has an anti-fraud association relation with the template scene behavior data from the other target item sub-clusters based on the visualization association condition, wherein the searching comprises:

and processing the relevance quantization matrix by setting a data conversion strategy to obtain the second scene office behavior data which has an anti-fraud incidence relation with the template scene behavior data.

Under some design ideas which can be independently implemented, before the mapping processing is performed on the first active collaboration item to obtain first scenarized office behavior data of the migratory office scene data carrying the cloud office behavior information, the method further includes one or more than one of the following steps:

binding distribution description information corresponding to the cloud office behavior information of the first active cooperation item to the first active cooperation item;

and updating the plurality of first active cooperation matters into a second statistical mode from a first statistical mode.

Under some design ideas which can be independently implemented, performing fraud prevention association analysis on the first scenario-based office behavior data in the set number of groups of cloud office behavior information with fraud prevention association analysis requirements to obtain a first fraud prevention association analysis condition of the set number of groups of cloud office behavior information with fraud prevention association analysis requirements, including:

acquiring data association evaluation indexes among the first scene office behavior data with differences in the set number of groups of cloud office behavior information with anti-fraud association analysis requirements;

determining a set number of associated behavior data sets in the set number of groups of cloud office behavior information with anti-fraud associated analysis requirements based on the data associated evaluation index, wherein the associated behavior data sets comprise one first scenized office behavior data in each group of cloud office behavior information;

and obtaining the first anti-fraud correlation analysis condition based on the correlation behavior data set.

Under some design ideas which can be independently implemented, the obtaining of the data association evaluation index between the first scenario-based office behavior data with the difference in the set number of sets of cloud office behavior information with the anti-fraud association analysis requirement includes:

obtaining quantitative common analysis results among different first scene office behavior data in the set quantity groups of cloud office behavior information with anti-fraud correlation analysis requirements;

processing the quantitative commonality analysis result through a set processing instruction to obtain a data association evaluation index between the first scene office behavior data with difference in the set quantity group of cloud office behavior information with anti-fraud association analysis requirements;

and/or determining the associated behavior data set in the set number of groups of cloud office behavior information with anti-fraud associated analysis requirements based on the data associated evaluation index, wherein the determining comprises the following steps:

and capturing the first scene office behavior data with the data association evaluation index meeting the association analysis index from the set number of groups of cloud office behavior information with anti-fraud association analysis requirements to form the association behavior data set.

The embodiment of the application also provides a big data office information analysis system, which comprises a processor, a communication bus and a memory; the processor and the memory communicate via the communication bus, and the processor reads the computer program from the memory and runs the computer program to perform the method described above.

An embodiment of the present application further provides a computer storage medium, where a computer program is stored, and the computer program implements the method when running.

Compared with the prior art, based on the technical scheme, the scene-based office behavior data covering the migration office scene data in the cloud office behavior information with the anti-fraud correlation analysis requirement is obtained, and the anti-fraud correlation analysis of the scene-based office behavior data is performed through the scene-based office behavior data covering the migration office scene data, so that the information of the office scene level of the cloud office behavior information with the fraud anti-fraud correlation analysis requirement can be considered in the anti-fraud correlation analysis process of the scene-based office behavior data, the integrity of the anti-fraud correlation analysis condition can be ensured as much as possible in the process of mining the first anti-fraud correlation analysis condition of the cloud office behavior information, the omission of the anti-fraud correlation analysis caused by the omission of the office scene is avoided, and therefore, when the anti-fraud detection is performed on the subsequent cloud office behavior information, the method can be based on the complete and accurate anti-fraud correlation analysis condition so as to ensure the accuracy and credibility of anti-fraud detection.

In the description that follows, additional features will be set forth, in part, in the description. These features will be in part apparent to those skilled in the art upon examination of the following and the accompanying drawings, or may be learned by production or use. The features of the present application may be realized and attained by practice or use of various aspects of the methodologies, instrumentalities and combinations particularly pointed out in the detailed examples that follow.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.

Fig. 1 is a block diagram illustrating a big data office information analysis system according to an embodiment of the present disclosure.

Fig. 2 is a flowchart of a big data office information analysis method for anti-fraud processing according to an embodiment of the present application.

Fig. 3 is a block diagram of a big data office information analysis device for anti-fraud processing according to an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.

Fig. 1 is a block diagram illustrating a big data office information analysis system 10 according to an embodiment of the present disclosure. The big data office information analysis system 10 in the embodiment of the present application may be a server with data storage, transmission, and processing functions, as shown in fig. 1, the big data office information analysis system 10 includes: memory 11, processor 12, communication bus 13 and big data office information analysis device 20 for anti-fraud processing.

The memory 11, processor 12 and communication bus 13 are electrically connected, directly or indirectly, to enable the transfer or interaction of data. For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The memory 11 stores big data office information analysis device 20 for anti-fraud processing, the big data office information analysis device 20 for anti-fraud processing comprises at least one software functional module which can be stored in the memory 11 in the form of software or firmware (firmware), and the processor 12 executes various functional applications and data processing by running the software programs and modules stored in the memory 11, such as the big data office information analysis device 20 for anti-fraud processing in the embodiment of the present application, so as to realize the big data office information analysis method for anti-fraud processing in the embodiment of the present application.

The Memory 11 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like. The memory 11 is used for storing a program, and the processor 12 executes the program after receiving an execution instruction.

The processor 12 may be an integrated circuit chip having data processing capabilities. The Processor 12 may be a general-purpose Processor including a Central Processing Unit (CPU), a Network Processor (NP), and the like. The various methods, steps and logic blocks disclosed in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The communication bus 13 is used for establishing communication connection between the big data office information analysis system 10 and other communication terminal devices through a network, and realizing the transceiving operation of network signals and data. The network signal may include a wireless signal or a wired signal.

It will be appreciated that the configuration shown in FIG. 1 is merely illustrative and that the big data office information analysis system 10 may include more or fewer components than shown in FIG. 1 or may have a different configuration than shown in FIG. 1. The components shown in fig. 1 may be implemented in hardware, software, or a combination thereof.

Fig. 2 shows a flow chart of big data office information analysis for anti-fraud processing provided by an embodiment of the present application. The method steps defined by the flow related to the method are applied to the big data office information analysis system 10 and can be realized by the processor 12, and the method comprises the following contents.

Step S11: and collecting the cloud office behavior information with the anti-fraud correlation analysis requirement in a set number of groups.

In the embodiment of the present application, the set number of groups may be two or more groups, that is, at least two groups. The form of obtaining the cloud office behavior information with the anti-fraud correlation analysis requirement may be obtained by starting a crawler thread (obtaining authorization of an office service terminal corresponding to the cloud office behavior information before obtaining the cloud office behavior information) in a big data office information analysis system of the big data office information analysis method for anti-fraud processing, or may be transmitted to the big data office information analysis system executing the big data office information analysis method for anti-fraud processing through other service terminals or service terminals in other network communication manners.

Further, the big data office information analysis system can be understood as a cloud server, a big data server or a computer device. Other network communication methods include, but are not limited to, local area network communication, wireless network communication, hot spot communication, wired optical fiber communication, and the like.

In addition, the cloud office behavior information with the fraud-resistant association analysis requirement in the embodiment of the application can be understood as the cloud office behavior information after various information optimizations (such as information denoising and information missing item repairing), and can be the cloud office behavior information without the information optimizations. The cloud office behavior information may reflect operation behavior habits, operation behavior intentions, or operation behavior contents of the office user. In addition, the information recording forms of the cloud office behavior information with the anti-fraud correlation analysis requirement may be the same, or there may be a difference, for example, one set of the cloud office behavior information is the cloud office behavior information in a text form, and the other set of the cloud office behavior information is the cloud office behavior information in a voice form.

The scale of the cloud office behavior information (such as the size of the information amount) and quantitative significance evaluation data (feature recognition degree) which are set in the quantity group and have anti-fraud correlation analysis requirements can be the same or different. Namely, any two groups of cloud office behavior information can be used as the cloud office behavior information with the anti-fraud correlation analysis requirement. In the embodiment of the application, two sets of cloud office behavior information with anti-fraud correlation analysis requirements are taken as an example, and of course, in other embodiments, three or more sets of cloud office behavior information with anti-fraud correlation analysis requirements may be present, and the number of the cloud office behavior information with anti-fraud correlation analysis requirements is not further limited here.

Step S12: and performing collaborative behavior item analysis on each cloud office behavior information with the anti-fraud correlation analysis requirement one by one to obtain office collaborative item records of each cloud office behavior information with the anti-fraud correlation analysis requirement, wherein the office collaborative item records comprise a plurality of first active collaborative items.

In the embodiment of the application, the collaborative behavior event analysis may be performed in various manners, for example, the collaborative behavior event analysis may be performed through various intelligent threads (such as a machine learning algorithm), and the collaborative behavior event may be understood as one of the features corresponding to the cloud office behavior information, so the collaborative behavior event analysis may also be understood as feature extraction. The office collaboration item record includes a plurality of first active collaboration items, where the office collaboration item record may be presented in the form of an information table. The active collaboration item refers to the scenario office behavior data of the migration office scenario data which does not include the cloud office behavior information, in other words, the active collaboration item can be understood as an item including the fragmented content of the cloud office behavior information which has the anti-fraud association analysis requirement. Viewed from another perspective, the active collaboration item may be a feature corresponding to a behavior information segment with a higher interaction heat.

Step S13: and mapping the first active cooperation item to obtain first scene office behavior data of the migration office scene data carrying the cloud office behavior information.

By mapping the first active cooperation item, the first scene office behavior data after mapping can carry the migration office scene data of the cloud office behavior information with anti-fraud correlation analysis requirements. In other words, the first scenized office behavior data covers office scene information of the cloud office behavior information with the anti-fraud correlation analysis requirement. The mapping process can be implemented according to a preset mapping rule or a mapping list, for example, the first scenarized office behavior data corresponding to the first active collaboration item is queried or matched according to a mapping relationship, so that the overall consideration of the office scene can be increased on the office item level, and the richness of the first scenarized office behavior data is improved. In a sense, the first scenized office behavior data can be understood as another type of information characteristic of the cloud office behavior information.

Step S14: and performing anti-fraud correlation analysis on the first scene office behavior data in the set quantity group of cloud office behavior information with anti-fraud correlation analysis requirements to obtain a first anti-fraud correlation analysis condition of the set quantity group of cloud office behavior information with anti-fraud correlation analysis requirements.

In this embodiment of the application, there may be multiple manners of performing anti-fraud association analysis on the scenized office behavior data, for example, performing anti-fraud association analysis on the scenized office behavior data by setting a processing instruction (such as an optimal mode algorithm), which is, of course, only an example, and in other embodiments, other anti-fraud association analysis manners may be adopted.

In addition, the anti-fraud association analysis can be understood as matching association analysis of different first scene office behavior data, so that different cloud office behavior information can be associated from different angles, a first anti-fraud association analysis condition (anti-fraud association analysis result) with richer information content and larger information amount is obtained, and the first anti-fraud association analysis condition can be used as subsequent anti-fraud analysis, so that accuracy and reliability of anti-fraud detection in the office interaction process are realized.

Based on the above technical solution, by obtaining the scenario-based office behavior data covering the migration office scenario data in the cloud office behavior information with the anti-fraud correlation analysis requirement, and further performing the anti-fraud correlation analysis on the scenario-based office behavior data covering the migration office scenario data, the information at the office scenario level of the cloud office behavior information with the fraud anti-fraud correlation analysis requirement can be taken into account in the anti-fraud correlation analysis process of the scenario-based office behavior data, so that in the process of mining the first anti-fraud correlation analysis condition of the cloud office behavior information, the integrity of the anti-fraud correlation analysis condition can be ensured as much as possible, and the omission of the anti-fraud correlation analysis due to the omission of the office scenario is avoided, so that when performing the anti-fraud detection on the subsequent cloud office behavior information, the method can be based on the complete and accurate anti-fraud correlation analysis condition so as to ensure the accuracy and credibility of anti-fraud detection.

In an embodiment of the present application, the office collaboration matter record includes a fuzzy record subset and a clear record subset, and the quantitative significance assessment data of the fuzzy record subset is smaller than the clear record subset. Further, the fuzzy record subset may be understood as a less recognizable transaction record subset, and the clear record subset may be understood as a more recognizable transaction record subset. The office cooperation item in the fuzzy record subset is a first active cooperation item, and the office cooperation item in the clear record subset is a second active cooperation item.

On the basis of the above contents, the cooperative behavior item analysis is performed on each piece of cloud office behavior information with the anti-fraud correlation analysis requirement one by one, and the office cooperative item record of each piece of cloud office behavior information with the anti-fraud correlation analysis requirement can be obtained through a multidimensional intelligent thread (for example, a related neural network model). The method includes the steps that diversified visual descriptions of cloud office behavior information with anti-fraud associated analysis requirements can be obtained through a multi-dimensional intelligent thread, for example, visual descriptions of a first proportion and a second proportion of quantized significance evaluation data of the cloud office behavior information with anti-fraud associated analysis requirements are extracted, or visual descriptions of a third proportion and a fourth proportion of quantized significance evaluation data of the cloud office behavior information with anti-fraud associated analysis requirements are extracted.

In some possible embodiments, the quantized significance evaluation data of the subset of blurred records is a fourth scale of the quantized significance evaluation data of the subset of sharp records. Quantitative significance assessment data between the fuzzy record subset and the clear record subset can be determined according to requirements of collaborative behavior event parsing performance (timeliness or accuracy), for example, the former is inferior in timeliness but better in accuracy compared with the visual description of the significance assessment data which extracts the cloud office behavior information quantification which has anti-fraud correlation analysis requirements in the first proportion and the second proportion, and the latter is superior in timeliness but poorer in accuracy compared with the visual description of the significance assessment data which extracts the cloud office behavior information quantification which has anti-fraud correlation analysis requirements in the third proportion and the fourth proportion. In the embodiment of the application, migration office scene data of cloud office behavior information without anti-fraud association analysis requirements is obtained according to the first activity collaboration items included in the fuzzy record subset and the second activity collaboration items included in the clear record subset, which are obtained by the multi-dimensional intelligent thread.

In some independently implementable technical solutions, before mapping the first active collaboration item to obtain the first scenarized office behavior data of the migration office scene data carrying the cloud office behavior information, one or more than one of the following steps is further included.

The first aspect is that the distribution description information corresponding to the cloud office behavior information with the anti-fraud association analysis requirement of the first active cooperation item is bound to the first active cooperation item. Illustratively, the distribution description mining process causes the respective first active collaboration items to pair non-repeating distribution description keywords.

In this embodiment of the application, the distribution description information corresponding to the cloud office behavior information of the first activity collaboration item may be understood as paragraph position information or distribution area information of the first activity collaboration item in the cloud office behavior information. In this way, accurate positioning of the first active collaboration item can be achieved in the cloud office behavior information.

The second aspect is to update the plurality of first active cooperation items from the first statistical manner to the second statistical manner. Illustratively, the first statistical manner may be two-dimensional statistics, such as that each first active collaboration event forms a fuzzy record subset in the form of a two-dimensional list, and the second statistical manner may be a manner of updating the two-dimensional list into a linear queue according to a certain sorting order.

It can be understood that, by adding the distribution description information of the first active cooperation item corresponding to the cloud office behavior information having the anti-fraud association analysis requirement to the first active cooperation item, the first scenario office behavior data after the content mapping processing can carry the distribution description information of the first scenario office behavior data in the cloud office behavior information having the anti-fraud association analysis requirement. In addition, the plurality of first active cooperation items are updated to the second statistical mode from the first statistical mode, so that the set thread can perform content mapping processing on the first active cooperation items quickly and orderly, and resource overhead caused by the content mapping processing is reduced.

Compared with the method that the cloud office behavior information with the anti-fraud correlation analysis requirement is directly led into the set thread, the fuzzy record subset can reduce the information amount led into the set thread, and therefore operation resources can be saved.

In some embodiments, step S13 may include the following steps: and taking the first active cooperation matters as first office business operation data, setting the first scene office behavior data as second office business operation data, and taking the cloud office behavior information with anti-fraud correlation analysis requirements as a specified constraint condition.

And obtaining second office business operation data based on the dynamic arrangement of the first office business operation data in the same appointed constraint condition and/or the dynamic arrangement of the first office business operation data in different appointed constraint conditions.

Illustratively, each specified constraint is taken as a current specified constraint one by one, and at least one round of content mapping processing as follows is performed on the current specified constraint.

The first aspect is to take each first office business operation data in the currently specified constraint as the current office business operation data.

And the second aspect is that the current office business operation data in the current specified constraint condition and other first office business operation data are dynamically arranged to obtain third office business operation data corresponding to the current office business operation data.

The step of dynamically sorting the current office business operation data in the current specified constraint condition and other first office business operation data is realized by an independent hierarchical processing unit in a set thread, wherein the mode that the independent hierarchical processing unit (a first local focusing unit) and an interactive hierarchical processing unit (a second local focusing unit) dynamically sort the relevant office business operation data can be understood as carrying out aggregation processing on the office business operation data (carrying out content selection, analysis and classification on the relevant data). Illustratively, an independent hierarchical processing unit includes a plurality of independent hierarchical processing nodes deployed in parallel, and all the first office business operation data of each specified constraint condition is input into the independent hierarchical processing node to perform dynamic arrangement of the first office business operation data within the specified constraint condition, that is, each independent hierarchical processing node only inputs the first office business operation data of one specified constraint condition, and the first office business operation data of a plurality of specified constraint conditions cannot be simultaneously input into the same independent hierarchical processing node.

Further, office business operation data in the form of a second statistical mode is input into the independent hierarchical processing node. The first office business operation data are dynamically arranged through the independent hierarchical processing unit, so that the obtained third office business operation data comprise migration office scene data with cloud office behavior information.

And the third aspect is that the third office business operation data of the current specified constraint condition and the third office business operation data of other specified constraint conditions are dynamically arranged to obtain the fourth office business operation data corresponding to the current office business operation data.

The step of dynamically arranging the third office business operation data of the current specified constraint condition and the third office business operation data of other specified constraint conditions is realized by an interactive hierarchical processing unit in a set thread. Because the interactive hierarchical processing unit has a non-limiting performance, that is, the output content of the interactive hierarchical processing unit only includes the output corresponding to any one input. Therefore, the interactive hierarchical processing unit also includes a set number of groups of interactive hierarchical processing nodes deployed in parallel, and the third office business operation data in the currently specified constraint condition and the third office business operation data of other specified constraint conditions are simultaneously input into the synchronous interactive hierarchical processing nodes, and in the process, it is necessary to adjust the precedence relationship of the currently specified constraint condition and the third office business operation data of other specified constraint conditions input into the interactive hierarchical processing nodes, for example, in the first interactive hierarchical processing node, the third office business operation data of the currently specified constraint condition is taken as input _1, the third office business operation data of other specified constraint conditions is taken as input _2, and in the second interactive hierarchical processing node, the third office business operation data of the currently specified constraint condition is taken as input _2, and the other third office business operation data with specified constraint conditions is used as input _ 1.

And acquiring fourth office business operation data through two synchronous interactive hierarchical processing nodes, so that the third office business operation data corresponding to each specified constraint condition has the corresponding fourth office business operation data, and ensuring one-to-one accurate pairing of the third office business operation data corresponding to each specified constraint condition and the fourth office business operation data.

Optionally, the set of independent hierarchical processing units and the set of interactive hierarchical processing units are used as a primary mapping adjustment process, the set thread includes a plurality of primary mapping adjustment processes, and process importance coefficients (weights) included in each primary mapping adjustment process are independent of each other. In addition, the number of the preliminary mapping adjustment processes can be determined according to the accuracy of the content mapping process and the timeliness of the content mapping process.

For example, if the accuracy of the content mapping process is required to be high, the number of the initial mapping adjustments can be increased appropriately, and if the time-based requirement for the content mapping process is high, the number of the initial mapping adjustments can be reduced appropriately. And on the premise that the current round of content mapping processing is not the last round of content mapping processing, taking the fourth office business operation data as the first office business operation data in the next content mapping processing. Of course, on the premise that the current round of content mapping processing is the last round of content mapping processing, the fourth office operation data is set as the second office operation data. In other words, the output content of the previous primary mapping adjustment will be used as the input of the next primary mapping adjustment, and the result of the last primary mapping adjustment will be set as the second office operation data.

The content of the visual description of the high-quantization significance evaluation data is extracted and processed into the scene office behavior data with the target item sub-cluster migration office scene data, and the scene office behavior data is subjected to anti-fraud association analysis through the scene office behavior number, so that office scene information can be comprehensively considered in the anti-fraud association analysis process, and the anti-fraud association analysis condition is more accurate and complete.

Illustratively, the third office business operation data can have office scene information of the current specified constraint condition by dynamically sorting the first office business operation data of the current specified constraint condition, the fourth office business operation data can have office scene information of other specified constraint conditions by dynamically sorting the third office business operation data of different specified constraint conditions, and in addition, the finally obtained second office business operation data is more complete and accurate through at least one round of content mapping processing, so that when the anti-fraud correlation analysis of the scene-based office behavior data is carried out through the second office business operation data, the anti-fraud correlation analysis condition of more accurate office behavior information can be obtained.

In some other examples, the policy executed by the independent hierarchical processing unit and/or the interactive hierarchical processing unit is a canonical hierarchical processing policy. Illustratively, the computation kernel (kernel) used in the independent hierarchical processing unit and/or the interactive hierarchical processing unit may be any computation kernel, and the computation kernel is modified into the weighted results of the two migration algorithms by code debugging according to a preset path, so that the operation precedence relationship of the hierarchical processing unit is changed by the matching index (fusion rate) weighted by the list, and the operation load is optimized from the traditional multidimensional operation load to the one-dimensional operation load. By the method, the operation load in the content mapping processing process can be simplified into one dimension by using the standardized hierarchical processing strategy, and compared with a non-standardized hierarchical processing strategy, the content mapping processing method has the advantages of less time consumption and lower operation load.

In some independently implementable technical solutions, the method for performing anti-fraud association analysis on the first scene-based office behavior data in the set number of sets of cloud office behavior information having anti-fraud association analysis requirements to obtain the first anti-fraud association analysis condition of the set number of sets of cloud office behavior information having anti-fraud association analysis requirements includes the following steps.

The first aspect is to collect data correlation evaluation indexes of different first scene office behavior data in a set number of groups of cloud office behavior information with anti-fraud correlation analysis requirements.

In some independently implementable technical solutions, the manner of collecting the data association evaluation indexes of different first scenario office behavior data in a set number of groups of cloud office behavior information with anti-fraud association analysis requirements includes the following steps.

Firstly, quantitative commonality analysis results among different first scenario office behavior data in a set number of groups of cloud office behavior information with anti-fraud correlation analysis requirements are collected.

For example, the quantitative commonality analysis result (similarity) may be obtained by determining the similarity between every two first scenario office behavior data in two cloud office behavior information with anti-fraud association analysis requirements, and forming a similarity distribution (such as a similarity matrix). The method for determining the similarity may be numerical similarity with state change, vector similarity or other similarity calculation methods.

Secondly, the quantitative commonality analysis result is processed through the set processing instruction, and a data association evaluation index between different first scene office behavior data in the set number of groups of cloud office behavior information with anti-fraud association analysis requirements is obtained. Illustratively, the similarity distribution is transformed to serve as a reference distribution, and the reference distribution is iterated for a set number of times by an optimal solution method to obtain a data association evaluation index. In other words, the calculation result of the data association evaluation index between different first scene-based office behavior data in the cloud office behavior information with the anti-fraud association analysis requirement is updated to be the calculation result with the standardized separated optimal solution operation. The set number capture determines the discrete condition of the data association evaluation index, and can be captured according to specific requirements so as to realize the coordination between accuracy and timeliness.

The sum of the first quantized positioning mark and the second quantized positioning mark of the distribution formed by the obtained data correlation evaluation index is 1 respectively.

In the embodiment of the application, the cloud office behavior information with the anti-fraud association analysis requirement is regarded as the first cloud office behavior information with the anti-fraud association analysis requirement and the second cloud office behavior information with the anti-fraud association analysis requirement respectively, wherein a data association evaluation index corresponding to a certain first quantitative positioning identifier in data association evaluation index distribution represents a data association evaluation index between certain first scenarized office behavior data in the first cloud office behavior information with the anti-fraud association analysis requirement and all first scenarized office behavior data in the second cloud office behavior information with the anti-fraud association analysis requirement. And the data association evaluation index corresponding to a certain second quantitative positioning identifier in the data association evaluation index distribution represents the data association evaluation index between certain first scene office behavior data in the second cloud office behavior information with the anti-fraud association analysis requirement and all first scene office behavior data in the first cloud office behavior information with the anti-fraud association analysis requirement.

The second aspect is that the associated behavior data set in the set number of groups of cloud office behavior information with anti-fraud associated analysis requirements is determined based on the data association evaluation index.

The method comprises the steps that a set number of sets of first scene office behavior data with anti-fraud association relation in cloud office behavior information with anti-fraud association analysis requirements are set as association behavior data sets. The association behavior data set comprises first scene office behavior data in each piece of cloud office behavior information with anti-fraud association analysis requirements. Namely, the associated behavior data set is composed of a first scene office behavior data in a plurality of pieces of cloud office behavior information with anti-fraud associated analysis requirements.

The method for determining the associated behavior data set in the set quantity groups of cloud office behavior information with the anti-fraud associated analysis requirement based on the data association evaluation index can be that the first scene office behavior data with the data association evaluation index meeting the associated analysis index is captured to form the associated behavior data set in the set quantity groups of cloud office behavior information with the anti-fraud associated analysis requirement.

Alternatively, the correlation analysis index may be a maximum first quantized location identifier and a maximum second quantized location identifier corresponding to the same time in the captured data correlation evaluation index distribution. For example, if the credibility index of the first quantized positioning identifier horizontal _1 and the second quantized positioning identifier vertical _2 in the data association evaluation index distribution is both at the maximum first quantized positioning identifier and at the maximum second quantized positioning identifier, it indicates that the maximum evaluation index of the data association between the first scenario office behavior data in the cloud office behavior information with the first anti-fraud association analysis requirement and the second data association evaluation index of the data association between the first scenario office behavior data in the cloud office behavior information with the first anti-fraud association analysis requirement and the second scenario office behavior data in the cloud office behavior information with the second anti-fraud association analysis requirement in the cloud office behavior information with the first anti-fraud association analysis requirement is the first active collaboration event.

The data association evaluation indexes among different first scenarized office behavior data are obtained through setting the processing instructions, and then the first scenarized office behavior data meeting the association analysis indexes are captured from the data association evaluation indexes, so that the anti-fraud association of the finally obtained associated behavior data set can meet the requirements.

A third aspect is to derive a first anti-fraud association analysis case based on the association behavior dataset.

Illustratively, the first anti-fraud correlation analysis condition is obtained based on distribution descriptions of the correlation behavior data sets in the set number of groups of cloud office behavior information with anti-fraud correlation analysis requirements.

The distribution description of the associated behavior data sets in the set number of groups of cloud office behavior information with anti-fraud associated analysis requirements is a first distribution description, and the first anti-fraud associated analysis condition comprises distribution description information expressing the first distribution description. The distribution description information may be a relative position of the behavior data in the associated behavior data set in the cloud office behavior information with the anti-fraud associated analysis requirement, and of course, may also be a distribution description area of the behavior data in the fuzzy record subset, and the distribution description area may map the first distribution description. By acquiring the data association evaluation indexes among different first scenized office behavior data and acquiring the association behavior data set based on the data association evaluation indexes, the finally acquired credibility index of the association behavior data set can meet the requirement.

In some independently implementable technical solutions, after anti-fraud association analysis is performed on first scene-based office behavior data in a set number of sets of cloud office behavior information having anti-fraud association analysis requirements to obtain a first anti-fraud association analysis condition of the set number of sets of cloud office behavior information having anti-fraud association analysis requirements, based on the first anti-fraud association analysis condition, extracting a related collaboration item cluster from a clear record subset of the set number of sets of cloud office behavior information having anti-fraud association analysis requirements.

The related cooperation item set comprises a set number of target item sub-clusters, and each target item sub-cluster comprises a plurality of second active cooperation items extracted from a clear record subset of the cloud office behavior information with anti-fraud related analysis requirements.

For example, based on the first anti-fraud correlation analysis, the manner of extracting the correlation collaboration item cluster from the clear record subset of the set number of sets of cloud office behavior information having anti-fraud correlation analysis requirements may be to determine a second distribution description corresponding to the first distribution description in the clear record subset.

And extracting the target item sub-cluster which takes the second distribution description as a reference and is in a set size from the clear record subset to obtain the relevance cooperation item cluster. The number of target item sub-clusters contained in the associated behavior data set depends on the number of the cloud office behavior information with the anti-fraud associated analysis requirement.

Optionally, the set scale of the embodiment of the application needs to meet the requirement that the obtained associated collaboration item cluster only includes behavior data in one pair of associated behavior data sets, and does not include behavior data in other associated behavior data sets.

It can be understood that the target item sub-cluster obtained through the first anti-fraud association analysis case contains the distribution description of the association behavior data set in the cloud office behavior information with the anti-fraud association analysis requirement, so that the second anti-fraud association analysis case obtained by performing the anti-fraud association analysis on the target item sub-cluster of the scenized office behavior data also has the first distribution description information. A second distribution description is determined from the first distribution description, and target item sub-clusters of a set size based on the second distribution description are extracted to reduce the possibility of extracting erroneous target item sub-clusters.

In some independently implementable technical solutions, before performing anti-fraud association analysis on second scenario-based office behavior data corresponding to the associated collaborative item cluster to obtain a set number of sets of second anti-fraud association analysis conditions of cloud office behavior information having anti-fraud association analysis requirements, mapping the second active collaborative item in the target item sub-cluster to the second scenario-based office behavior data having migration office scenario data of the target item sub-cluster.

It can be understood that the manner of mapping the second active collaboration item in the target item sub-cluster to the second scenized office behavior data having the migration office scene data of the target item sub-cluster may be that the second active collaboration item is the first office business operation data, the second scenized office behavior data is set as the second office business operation data, and each target item sub-cluster is a specified constraint condition.

Specifically, the dynamic arrangement mode refers to a process of mapping the first active collaboration item to obtain first scenarized office behavior data of the migration office scene data carrying the cloud office behavior information, wherein the set threads used in the two processes may be the same or different. When the two setting threads are different, the difference lies in that the initial mapping adjustment quantity in the process is smaller than or equal to the initial mapping adjustment quantity used in the process of mapping the first active collaboration item to obtain the first scenarized office behavior data of the migration office scene data carrying the cloud office behavior information.

It can be understood that by extracting behavior key information visually described by high-quantization significance evaluation data, mapping the behavior key information into the scenized office behavior data with the target item sub-cluster migration office scene data, and performing anti-fraud association analysis on the scenized office behavior data through the scenized office behavior data, the office scene information of the target item sub-cluster can be considered in the anti-fraud association analysis process of the scenized office behavior data of the high-quantization significance evaluation data, so that the integrity and the accuracy of the anti-fraud association analysis condition of the cloud office behavior information are ensured.

And performing anti-fraud correlation analysis on the second scene office behavior data corresponding to the relevance collaboration item cluster to obtain a set number of second anti-fraud correlation analysis conditions of the cloud office behavior information with anti-fraud correlation analysis requirements.

The second scenario-based office behavior data is the second active collaboration item in the relevance collaboration item cluster, or is generated by mapping the second active collaboration item in the relevance collaboration item cluster.

In other words, the second scenized office activity data may be processed without content mapping by the mapping module, or may be processed with content mapping by the mapping module, and the second scenized office activity data is not further limited herein.

The way of performing anti-fraud association analysis on the second scenic office behavior data corresponding to the relevance collaborative item cluster to obtain the second anti-fraud association analysis condition of the set number of sets of cloud office behavior information having anti-fraud association analysis requirements may be to use any one target item sub-cluster of the relevance collaborative item cluster as a calibration sub-cluster, and use the second scenic office behavior data set and described in distribution in the calibration sub-cluster as template scene behavior data. The set distribution description herein may be a reference for calibrating the sub-clusters. Because the reference of the target item sub-cluster is one associated behavior data in the associated behavior data set, the associated behavior data is used as the template scene behavior data, so that the calculated visual association with each second scenized office behavior data in other target item sub-clusters is more accurate.

And searching for second scene office behavior data which has an anti-fraud association relation with the template scene behavior data in other target item sub-clusters of the relevance cooperation item cluster. For example, the manner of looking for the second scenized office behavior data having the anti-fraud association relationship with the template scene behavior data may be to obtain a visual association between the template scene behavior data and each second scenized office behavior data in other target item sub-clusters. For example, visual optimization processing is performed on the template scene behavior data (reference data) and second scenized office behavior data in other target matter sub-clusters, so as to obtain a correlation quantization matrix (such as strong and weak map expression of correlation). And the quantized data of different distribution descriptions in the relevance quantization matrix represents the anti-fraud relevance between the template scene behavior data and different second scene office behavior data. By obtaining the quantitative matrix of the relevance, the anti-fraud relevance between the template scene behavior data and each second scene office behavior data in other target item sub-clusters can be clearly reflected.

And searching for second scene office behavior data which has an anti-fraud association relation with the template scene behavior data from other target item sub-clusters based on the visual association condition. Exemplarily, the relevance quantization matrix is processed by setting a data conversion strategy, so that second scene office behavior data having an anti-fraud incidence relation with the template scene behavior data is obtained.

In this embodiment, the set data transformation policy may be a normalized data transformation policy. And obtaining the second anti-fraud association analysis condition based on the template scene behavior data and the second scene office behavior data with the anti-fraud association relation. Illustratively, a third distribution description of the template scene behavior data and second scene office behavior data searched for anti-fraud association relation with the template scene behavior data in at least two groups of cloud office behavior information with anti-fraud association analysis requirements is determined.

The second fraud prevention association analysis condition includes a third distribution description of the template scene behavior data and the second scenario office behavior data searched for the second scenario office behavior data having the fraud prevention association relationship with the template scene behavior data in the cloud office behavior information having the fraud prevention association analysis requirement, and fraud prevention association between the third distribution description and the second scenario office behavior data, of course, the third distribution description may not correspond to the information tag (the key information field having the key information identification capability) of the cloud office behavior information having the fraud prevention association analysis requirement, and may correspond to the content area of the non-information tag, so that fraud prevention association analysis of the accuracy of the non-information tag can be completed.

The specific expression form of the second anti-fraud association analysis case may be presented in the form of an item matching pair, or may be presented in the form of cloud office behavior information, and the following is that an embodiment of the big data office information analysis method for anti-fraud processing according to the present application shows related content of the second anti-fraud association analysis case. The in10 is the cloud office behavior information with the first anti-fraud correlation analysis requirement, and the in20 is the cloud office behavior information with the second anti-fraud correlation analysis requirement. The directed line segment between in10 and in20 can be used for representing the anti-fraud association analysis condition of two groups of cloud office behavior information. Regarding confidence indices, the presentation may be through the thickness of the directional line segments, such as using directional line segments of different thicknesses to represent confidence indices, or marking confidence indices directly near each directional line segment. Other manifestations of the second anti-fraud association analysis scenario are not further limited herein. It is understood that the fraud-resistant association analysis case in the embodiment of the present application may be expressed in the form of a knowledge graph.

It can be understood that the accuracy of the anti-fraud association analysis is further improved by performing the anti-fraud association analysis on the visual description of the low-quantitative saliency evaluation data first and then performing the anti-fraud association analysis on the visual description of the high-quantitative saliency evaluation data according to the anti-fraud association analysis on the visual description of the low-quantitative saliency evaluation data.

In order to more clearly describe the technical solutions proposed in the embodiments of the present application, the embodiments of the present application provide the following two examples for illustration.

In a first example, the big data office information analysis method for anti-fraud processing provided by the embodiment of the present application further includes the technical solution described in the following steps.

Step S21: and acquiring the first cloud office behavior information with the anti-fraud correlation analysis requirement and the second cloud office behavior information with the anti-fraud correlation analysis requirement.

The manner of obtaining the cloud office behavior information with the first fraud prevention correlation analysis requirement and the cloud office behavior information with the second fraud prevention correlation analysis requirement is referred to in step S11, and will not be further described here.

Step S22: the method comprises the steps of extracting a fuzzy record subset and a clear record subset of cloud office behavior information with anti-fraud association analysis requirements respectively, wherein the fuzzy record subset comprises a first active cooperation item, and the clear record subset comprises a second active cooperation item, and quantitative significance evaluation data of the fuzzy record subset is smaller than that of the clear record subset.

The method for extracting the fuzzy record subset and the clear record subset of the cloud office behavior information with the anti-fraud association analysis requirement may use a multi-dimensional intelligent thread, which may be specifically referred to in step S12, and will not be further described here.

Step S23: and importing the two groups of first active cooperation items into a set thread to obtain first scene office behavior data of the migration office scene data carrying the cloud office behavior information.

Of course, before executing step S23, the first activity cooperation item in the fuzzy record subset may be added to the distribution description mining process, and updated from the form of the two-dimensional list to the form of the linear queue, and the first activity cooperation item group in the form of the linear queue may be imported into the setting thread. Specifically, the process of importing the two groups of first active collaboration items into the setting thread to obtain the first scenarized office behavior data of the migration office scene data carrying the cloud office behavior information is referred to the step S13, and no further description is provided herein.

Step S24: and performing anti-fraud correlation analysis on the scenized office behavior data to obtain a first anti-fraud correlation analysis condition.

The manner of performing the anti-fraud correlation analysis on the scenized office behavior data with respect to the first scenized office behavior data is referred to the above step S14, and will not be further described here.

Step S25: and extracting the relevance collaboration item cluster from a set number of clear record subsets of the cloud office behavior information with anti-fraud association analysis requirements based on the first anti-fraud association analysis condition.

The process of extracting the relevance collaboration item cluster from the clear record subset of the set number of sets of cloud office behavior information with the anti-fraud association analysis requirement is referred to above, and no further description is provided here.

Step S26: and performing anti-fraud correlation analysis on the second scene office behavior data corresponding to the relevance collaboration item cluster to obtain a set number of second anti-fraud correlation analysis conditions of the cloud office behavior information with anti-fraud correlation analysis requirements.

Please refer to the above description for a manner of performing fraud prevention association analysis on the second scenario-based office behavior data corresponding to the relevance collaboration item cluster to obtain a set number of second fraud prevention association analysis conditions of the cloud office behavior information having fraud prevention association analysis requirements, which will not be further described herein.

In a second example, cloud office behavior information behaviour _ in _1 with a first anti-fraud association analysis requirement and cloud office behavior information behaviour _ in _2 with a second anti-fraud association analysis requirement are obtained. The cloud office behavior information behaviour _ in _1 with the first anti-fraud correlation analysis requirement and the cloud office behavior information behaviour _ in _2 with the second anti-fraud correlation analysis requirement can be the same or different in quantitative significance evaluation data. Inputting the cloud office behavior information behavior _ in _1 required by the first anti-fraud association analysis and the cloud office behavior information behavior _ in _2 required by the second anti-fraud association analysis into a multi-dimensional intelligent thread to extract diversified visual descriptions, for example, extracting fuzzy record subsets, namely, collection _11 and collection _21 of quantized significance evaluation data of the cloud office behavior information behavior _ in _1 required by the first anti-fraud association analysis and the cloud office behavior information behavior _ in _2 required by the second anti-fraud association analysis respectively, and extracting clear record subsets, namely, collection _12 and collection _22 of quantized significance evaluation data of the cloud office behavior information behavior _ in _1 required by the second anti-fraud association analysis and the office behavior information behavior _ in _2 required by the second anti-fraud association analysis respectively. It can be seen that the quantized significance evaluation data of the fuzzy record subset collection _11 is smaller than the quantized significance evaluation data of the clear record subset collection _12, and the quantized significance evaluation data of the fuzzy record subset collection _21 is smaller than the quantized significance evaluation data of the clear record subset collection _ 22.

And carrying out distribution description mining processing on the characteristics in the fuzzy record subsets, namely, collection _11 and collection _21, and simplifying the fuzzy record subsets, namely, collection _11 and collection _21 from a first statistical mode to a second statistical mode, namely, a one-dimensional linear queue. And leading the one-dimensional linear queues with distribution description mining processing into a set thread, in the set thread, respectively extracting the one-dimensional linear queues through an independent layering processing unit for characteristic dynamic arrangement, inputting the one-dimensional linear queues after dynamic arrangement into an interactive layering processing unit for characteristic dynamic arrangement between two groups of one-dimensional linear queues, taking the independent layering processing unit and the interactive layering processing unit as a primary mapping adjustment, wherein the number of the primary mapping adjustment is X, the output of the previous primary mapping adjustment is taken as the input of the next primary mapping adjustment, the output content of the last primary mapping adjustment is taken as the output content of the set thread, and the output content respectively comprises the one-dimensional linear queues. Illustratively, the independent hierarchical processing unit and the interactive hierarchical processing unit perform dynamic feature arrangement by extracting distribution description of features and active cooperation items on which local relevance features depend.

The data association evaluation index distribution between the one-dimensional linear queue and the fuzzy record subsets before simplification is obtained through the setting processing instruction, wherein the transverse boundary data of the data association evaluation index distribution is equal to (first ratio) 2 multiplied by the product of the transverse boundary data and the longitudinal boundary data of the cloud office behavior information behavior _ in _2 required by the second anti-fraud association analysis (namely, (first ratio) 2boundary1_2boundary2_ 2), and the longitudinal boundary data of the data association evaluation index distribution is equal to (first ratio) 2 multiplied by the product of the transverse boundary data and the longitudinal boundary data of the cloud office behavior information behavior _ in _1 required by the first anti-fraud association analysis (namely, (first ratio) 2boundary1_1boundary2_ 1).

And selecting anti-fraud association analysis groups (behaviour _ in _11 and scene _ in _ 21) with credibility indexes meeting conditions from the data association evaluation indexes, wherein the anti-fraud association analysis groups are not limited to one group and can be multiple groups. Finding out the features (behavour _ in _12, scene _ in _ 22) corresponding to the anti-fraud association analysis group (behavour _ in _11, scene _ in _ 21) from the clear record subsets collection _12 and collection _22, and extracting a target item sub-cluster group containing the features behavour _ in _12 or feature _ in _22, wherein the horizontal boundary data and the vertical boundary data of the target item sub-cluster in the target item sub-cluster group are both w. And inputting the target item sub-cluster group into another set thread to obtain the visual description after dynamic arrangement. Here, the setting thread may be the same as or different from the setting thread in the activity coordination matter mapping, and for example, the number of initial-level mapping adjustments in the setting thread may be smaller than the number of initial-level mapping adjustments in the activity coordination matter mapping.

On the basis of the above contents, the feature behavior _ in _12 described in the reference distribution of any one target item sub-cluster may be used as template scene behavior data, a visual optimization process is performed on all features in another target item sub-cluster to obtain a correlation quantization matrix, the correlation quantization matrix is input into a two-dimensional normalized data conversion strategy to calculate an expected matching distribution description in the target item sub-cluster, and the distribution description of behavior _ in _12 and the anti-fraud association relation with the distribution description is mapped to the first cloud office behavior information behavior _ in _1 and the second cloud office behavior information behavior _ in _2 which have the anti-fraud association analysis requirement, so as to obtain the anti-fraud association analysis condition of the office cloud behavior information which finally has the first anti-fraud association analysis requirement.

In some optional and independently implementable technical solutions, after obtaining the first fraud prevention correlation analysis condition of the set number of sets of cloud office behavior information having the fraud prevention correlation analysis requirement, the method may further include the following: analyzing the first anti-fraud correlation analysis condition to obtain a correlation analysis item corresponding to the first anti-fraud correlation analysis condition; judging whether the cloud office behavior information with the anti-fraud association analysis requirement exists in the set quantity group or not according to the association analysis items; and when the cloud office behavior information with the anti-fraud correlation analysis requirement in the set quantity group has a behavior safety risk, intercepting an operation behavior request corresponding to the cloud office behavior information with the anti-fraud correlation analysis requirement in the set quantity group.

In some selective and independently implementable technical solutions, analyzing the first anti-fraud association analysis condition to obtain an association analysis item corresponding to the first anti-fraud association analysis condition may include the following: obtaining a set of association analysis items for the first anti-fraud association analysis case, the set of association analysis items including at least two association analysis items; obtaining a security impact coefficient between each associated analysis item in the associated analysis item set and the first anti-fraud associated analysis condition; adjusting each correlation analysis item according to the safety influence coefficient corresponding to each correlation analysis item and the item content authority description of each correlation analysis item to obtain a corresponding correlation analysis item sorting result; generating a target security risk polarity record for the first anti-fraud association analysis case based on the association analysis project collation results, the security risk polarity record comprising at least two target risk polarity expressions.

In the embodiment of the application, the associated analysis items correspond to different analysis dimensions and analysis levels, including but not limited to office object addresses, office object identities, office business types, office business scenes, and the like. Further, the security impact coefficient may be understood as a degree of association or an impact degree between each associated analysis item and the first anti-fraud associated analysis condition, and based on this, the item content authority description may be understood as feature information of an authority security level corresponding to the associated analysis item, so that, by using the security impact coefficient corresponding to each associated analysis item and the item content authority description of each associated analysis item, sorting of each associated analysis item may be achieved, so as to obtain a target security risk polarity record highly matched with the first anti-fraud associated analysis condition, and a target risk polarity expression in the target security risk polarity record may represent a security level corresponding to the first anti-fraud associated analysis condition in a numerical form, so that, it may be determined by the security level whether there is a behavior security level in the set number of cloud office behavior information that has an anti-fraud associated analysis requirement And (4) total risk.

For example, the target risk polarity expression is subjected to weight distribution from high to low and is weighted, so that a comprehensive risk index corresponding to the cloud office behavior information can be obtained, and if the comprehensive safety index is greater than set safety knowledge, the cloud office behavior information is judged to have a behavior safety risk, so that the accuracy and the reliability of behavior safety risk judgment can be improved.

In some selective and independently implementable technical solutions, the adjusting, according to the security impact coefficient corresponding to each associated analysis item and the item content authority description of each associated analysis item, each associated analysis item to obtain a corresponding associated analysis item sorting result specifically includes: according to the safety influence coefficient corresponding to each correlation analysis item and the item content authority description of each correlation analysis item, disassembling each correlation analysis item to obtain at least two correlation analysis item subsets; and adjusting each association analysis item subset, and adjusting each association analysis item in each association analysis item subset respectively to obtain an association analysis item sorting result. By the design, the integrity of the sorting result of the associated analysis project can be ensured.

In some selective and independently implementable technical solutions, the parsing, according to the security impact coefficient corresponding to each associated analysis item and the item content authority description of each associated analysis item, each associated analysis item to obtain at least two associated analysis item subsets specifically includes: optimizing the item content permission description of each association analysis item according to the safety influence coefficient corresponding to each association analysis item to obtain the hot item content permission description of each association analysis item; and performing item analysis on each associated analysis item according to the hot item content authority description of each associated analysis item to obtain at least two associated analysis item subsets. By the design, the deletion of the association analysis item subset can be ensured as far as possible.

In some selective and independently implementable technical solutions, the adjusting each subset of the association analysis items, and respectively adjusting each association analysis item in each subset of the association analysis items to obtain the association analysis item sorting result specifically includes: adjusting each association analysis item subset according to the number of association analysis items contained in each association analysis item subset; and performing the following operations respectively for the various associated analysis item subsets: adjusting each associated analysis item in the associated analysis item subset according to the item content authority description of each associated analysis item in the associated analysis item subset and the correlation coefficient of the associated analysis item subset; and generating the associated analysis project arrangement result based on the adjustment result among the associated analysis project subsets and the adjustment result of each associated analysis project in the associated analysis project subsets. By the design, the integrity of the sorting result of the associated analysis project can be ensured.

Furthermore, the technical scheme provided by the embodiment of the application does not need to perform quality inspection of behavior information description, so that the influence of the precision of the quality inspection of the behavior information description on anti-fraud correlation analysis of the scenarized office behavior data is reduced, and the adaptability of the application to different office scenes is improved.

Based on the same inventive concept, there is also provided a big data office information analysis device 20 for anti-fraud processing, applied to a big data office information analysis system 10, the device comprising:

the information collection module 21 is configured to collect a set number of sets of cloud office behavior information with anti-fraud correlation analysis requirements;

the information analysis module 22 is configured to perform collaborative behavior item analysis on each group of cloud office behavior information one by one to obtain an office collaborative item record of each group of cloud office behavior information, where the office collaborative item record includes a plurality of first active collaborative items;

the data acquisition module 23 is configured to perform mapping processing on the first active collaboration item to obtain first scenarized office behavior data of the migratory office scene data that carries the cloud office behavior information;

the association analysis module 24 is configured to perform anti-fraud association analysis on the first scene-based office behavior data in the set number group of cloud office behavior information that requires the anti-fraud association analysis, so as to obtain a first anti-fraud association analysis condition of the set number group of cloud office behavior information that requires the anti-fraud association analysis.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus and method embodiments described above are illustrative only, as the flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, the big data office information analysis system 10, or a network device) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A big data office information analysis method for anti-fraud processing, applied to a big data office information analysis system, the method comprising:

2. The method of claim 1, wherein the office collaboration item record comprises a fuzzy record subset and a clear record subset, wherein the quantitative significance evaluation data of the fuzzy record subset is smaller than the clear record subset, the office collaboration item corresponding to the fuzzy record subset is the first active collaboration item, and the office collaboration item corresponding to the clear record subset is the second active collaboration item;

3. The method according to claim 2, wherein before performing anti-fraud association analysis on the second scenario-based office behavior data corresponding to the related collaborative event cluster to obtain a second anti-fraud association analysis condition of the set number of sets of cloud office behavior information for which anti-fraud association analysis needs to be performed, the method further comprises:

4. The method according to any one of claims 2 to 3, wherein the first scenario-based office behavior data with an anti-fraud association relationship in the set number of sets of cloud office behavior information with anti-fraud association analysis requirements is an association behavior data set, the distribution descriptions of the association behavior data set in the set number of sets of cloud office behavior information with anti-fraud association analysis requirements are first distribution descriptions, the first anti-fraud association analysis condition includes distribution description information expressing the first distribution descriptions, and the coverage range of the target item sub-cluster in the cloud office behavior information includes the first distribution description;

5. The method according to claim 4, wherein the extracting the relevance collaboration item cluster from the set number of clear record subsets of the cloud office behavior information for which anti-fraud correlation analysis requirements exist based on the first anti-fraud correlation analysis case comprises:

extracting the target item sub-cluster which takes the second distribution description as a reference and is of a set size from the clear record subset to obtain the relevance cooperation item cluster;

correspondingly, the set distribution is described as a reference of the calibration sub-cluster;

searching for the second scene office behavior data which has an anti-fraud association relation with the template scene behavior data from the other target item sub-clusters based on the visual association condition;

correspondingly, the obtaining of the visual association between the template scene behavior data and each second scenarized office behavior data in the other target item sub-clusters includes:

6. The method according to claim 1, wherein before the mapping process is performed on the first active collaboration item to obtain first scenarized office behavior data of migration office scene data carrying the cloud office behavior information, the method further comprises one or more of the following steps:

7. The method according to claim 1, wherein the performing anti-fraud association analysis on the first scenario-based office behavior data in the set number of sets of cloud office behavior information that requires anti-fraud association analysis to obtain a first anti-fraud association analysis condition of the set number of sets of cloud office behavior information that requires anti-fraud association analysis comprises:

obtaining the first anti-fraud association analysis condition based on the association behavior data set;

correspondingly, the obtaining of the data association evaluation index between the first scenized office behavior data with the difference in the set number of sets of cloud office behavior information with the anti-fraud association analysis requirement includes:

8. The method according to any one of claims 2 or 3, wherein the mapping the first active collaboration item to obtain first scenario-based office behavior data of migration office scenario data carrying the cloud office behavior information, or the mapping the second active collaboration item in the target item sub-cluster to obtain second scenario-based office behavior data of migration office scenario data covering the target item sub-cluster comprises:

obtaining second office business operation data based on dynamic arrangement of the first office business operation data in the same specified constraint condition and/or dynamic arrangement of the first office business operation data in the specified constraint condition with difference;

correspondingly, the obtaining the second office business operation data based on the dynamic arrangement of the first office business operation data in the same specified constraint condition and/or the dynamic arrangement of the first office business operation data in the specified constraint condition with a difference includes:

taking the fourth office business operation data as the second office business operation data on the premise that the content mapping processing of the current round is the last round;

correspondingly, the step of dynamically arranging the current office business operation data in the current specified constraint condition and the other first office business operation data is realized by an independent hierarchical processing unit in a set thread;

the step of dynamically arranging the third office business operation data of the current specified constraint condition and the third office business operation data of other specified constraint conditions is realized by an interactive hierarchical processing unit in the set thread;

correspondingly, the strategy executed by the independent hierarchical processing unit and/or the interactive hierarchical processing unit is a standardized hierarchical processing strategy.

9. A big data office information analysis system is characterized by comprising a processor, a communication bus and a memory; the processor and the memory communicate via the communication bus, the processor reading a computer program from the memory and operating to perform the method of any of claims 1-8.

10. A computer storage medium, characterized in that it stores a computer program which, when executed, implements the method of any one of claims 1-8.