CN116777629B - Online transaction management system - Google Patents
Online transaction management system Download PDFInfo
- Publication number
- CN116777629B CN116777629B CN202310825934.8A CN202310825934A CN116777629B CN 116777629 B CN116777629 B CN 116777629B CN 202310825934 A CN202310825934 A CN 202310825934A CN 116777629 B CN116777629 B CN 116777629B
- Authority
- CN
- China
- Prior art keywords
- transaction
- data
- error
- information
- unit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000011159 matrix material Substances 0.000 claims abstract description 77
- 238000000034 method Methods 0.000 claims abstract description 54
- 230000008569 process Effects 0.000 claims abstract description 46
- 238000012544 monitoring process Methods 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 16
- 238000013135 deep learning Methods 0.000 claims abstract description 15
- 238000004364 calculation method Methods 0.000 claims abstract description 12
- 238000013507 mapping Methods 0.000 claims abstract description 12
- 238000012795 verification Methods 0.000 claims abstract description 10
- 230000003416 augmentation Effects 0.000 claims description 26
- 238000000605 extraction Methods 0.000 claims description 12
- 230000002159 abnormal effect Effects 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 9
- 230000005856 abnormality Effects 0.000 claims description 8
- 238000006243 chemical reaction Methods 0.000 claims description 8
- 238000000638 solvent extraction Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000005516 engineering process Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 5
- 238000004806 packaging method and process Methods 0.000 claims description 5
- 230000003993 interaction Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 abstract description 7
- 239000000284 extract Substances 0.000 abstract description 6
- 238000012545 processing Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 12
- 239000000203 mixture Substances 0.000 description 5
- 230000009471 action Effects 0.000 description 3
- 230000006399 behavior Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000003321 amplification Effects 0.000 description 2
- 238000003199 nucleic acid amplification method Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005429 filling process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of monitoring of the Internet of things, and particularly discloses an online transaction management system, which comprises an error reporting module, a data processing module and a data processing module, wherein the error reporting module is used for receiving error transaction information and verifying the error transaction information; the data query module is used for querying transaction data in a preset storage database when verification passes; the deep learning module is used for determining transaction characteristics according to the transaction data and training the transaction characteristics to a shallow learning model of the error transaction information; and the model application module is used for monitoring the transaction flow in real time and outputting forecast error transaction information. The invention converts the error transaction information into a two-dimensional matrix according to time and an operating party, extracts the change characteristics in the two-dimensional matrix, generates a symbiotic matrix, extracts the characteristic values in the symbiotic matrix by means of a matrix calculation formula, obtains transaction characteristics, creates a mapping model by the transaction characteristics and the error transaction information, and can analyze and control the transaction process according to the mapping model.
Description
Technical Field
The invention relates to the technical field of monitoring of the Internet of things, in particular to an online transaction management system.
Background
With the advancement of society and the development of computer technology, online transactions are becoming popular as a new approach, and based on this, the variety of products is also increasing, such as virtual products, and in the face of virtual products, online transactions are becoming the mainstream.
However, when the online transaction brings convenience, the risk is also accompanied, and errors often occur in the online transaction process, for example, in the virtual product transaction process, the seller changes the product in the last second, and even if the seller makes a transaction, the buyer needs are difficult to meet; in the existing large number of error transaction flows, redundant operations are common to the operations, and how to monitor the operation process when online transaction is performed, so that the problem of reducing the error rate of online transaction is solved by the technical scheme of the invention.
Disclosure of Invention
(One) solving the technical problems
Aiming at the defects of the prior art, the invention provides an online transaction management system, which solves the problem of how to quickly locate the occurrence period of fluctuation in a plurality of data.
(II) technical scheme
In order to achieve the above purpose, the invention is realized by the following technical scheme:
there is provided an online transaction management system comprising:
the error reporting module is used for receiving the error transaction information uploaded by the user and verifying the error transaction information;
The data query module is used for querying transaction data in a preset storage database when verification passes;
The deep learning module is used for determining transaction characteristics according to the transaction data, creating a sample set and a test set according to the transaction characteristics, and training the transaction characteristics to a shallow learning model of the error transaction information;
The model application module is used for monitoring a transaction flow in real time based on a user request, determining fitting data according to the transaction flow, inputting the fitting data into a trained shallow learning model, and outputting forecast error transaction information;
and the instruction generation module is used for determining a process adjusting instruction according to the prediction error transaction information.
As a further aspect of the present invention, the error reporting module includes:
The request receiving unit is used for receiving the error report request containing the target object uploaded by the user;
the standard inquiring unit is used for inquiring standard price information of the target object based on the big data technology;
the payment inquiry unit is used for sending a payment information acquisition request to the user and receiving payment information fed back by the user;
And the comparison output unit is used for comparing the payment information with the standard price information and outputting a verification result according to the comparison result.
As a further aspect of the present invention, the data query module includes:
the connection unit is used for establishing a connection channel with the storage database when the verification is passed;
The data positioning unit is used for reading the identity information of both transaction sides, and positioning the transaction flow and fitting data thereof in the storage database according to the identity information; the method comprises the steps of opening a user-oriented authority control port in real time in the process of reading information;
the data packaging unit is used for packaging the positioned transaction flow and fitting data thereof according to the time sequence;
The data conversion unit is used for reading the occurrence time period of the error transaction information and converting the transaction flow and fitting data thereof into sample data according to the occurrence time period.
As a further aspect of the present invention, the data conversion unit includes:
The classifying subunit is used for reading the time point of the error transaction information, classifying the transaction flow and the fitting data thereof according to the time point and obtaining core data and auxiliary data;
the analysis subunit is used for inputting the auxiliary data into a trained shallow learning model to obtain forecast error transaction information;
the abnormality judging subunit is used for comparing the predicted error transaction information with the actual transaction information to determine the abnormality rate of the core data;
and the statistics subunit is used for counting the core data corresponding to all the error transaction information and the abnormal rate thereof to obtain sample data.
As a further aspect of the present invention, the deep learning module includes:
The data clustering unit is used for reading the sample data and clustering the core data according to the abnormal rate;
a data filling unit for filling the core data into a two-dimensional augmentation matrix based on time information; the number of rows and the number of columns of the two-dimensional augmentation matrix are both increasing functions of time;
the feature generation unit is used for converting the two-dimensional augmentation matrix into a symbiotic matrix in a preset direction, calculating a feature value of the symbiotic matrix and determining transaction features according to the feature value;
the sample determining unit is used for classifying transaction characteristics corresponding to core data in sample data according to a preset selection proportion, and creating a sample set and a test set;
The shallow training unit is used for training a shallow learning model based on the sample set and the test set; wherein the shallow learning model is a mapping from transaction characteristics to error transaction information; the anomaly rate is used to adjust the impact weight of the transaction characteristic in the training process.
As a further aspect of the present invention, the data stuffing unit includes:
The traversing subunit is used for traversing the core data based on the time information and acquiring the data source of the core data; the data source comprises two transaction parties; the data sources respectively correspond to rows and columns of the two-dimensional augmentation matrix;
the data transcoding sub-unit is used for inputting the core data into a preset data transcoding model to obtain a numerical value;
an inserting subunit for inserting the values into the two-dimensional augmentation matrix based on the time nodes; wherein the two-dimensional augmentation matrix contains row and column labels corresponding to the time nodes.
As a further aspect of the present invention, the feature generation unit includes:
the partitioning subunit is used for partitioning the two-dimensional augmentation matrix according to the time node to obtain a submatrix;
The scale determining subunit is used for obtaining the numerical value extremum of the submatrices and determining the matrix scale of the submatrices according to the numerical value extremum;
the extraction subunit is used for traversing the submatrices in a preset direction, acquiring two numerical values corresponding to adjacent elements, and respectively inserting the two numerical values into the submatrices as the row number and the column number;
And the calculating subunit is used for calculating the eigenvalue of the sub-symbiotic matrix based on a preset calculation formula and determining the transaction characteristic according to the eigenvalue.
As a further aspect of the present invention, the calculating the eigenvalue of the sub-symbiotic matrix based on the preset calculation formula, and determining the content of the transaction feature according to the eigenvalue includes:
calculating the eigenvalue of the sub-symbiotic matrix based on a preset calculation formula group;
Counting the characteristic values according to a preset counting rule, and determining transaction characteristics;
Wherein, the formula group includes:
A＝∑i∑jP(i,j)2；
C＝∑i∑j(i-j)2P(i,j)；
E＝-∑i∑jP(i,j)logP(i,j)；
wherein A is an energy value, C is a contrast ratio, and E is an entropy value; p (i, j) is the element value of the ith row and jth column in the subco-occurrence matrix.
As a further aspect of the present invention, the model application module includes:
The data monitoring unit is used for receiving monitoring requests sent by both transaction sides and monitoring the transaction flow in real time based on the monitoring requests; the transaction flow is the operation information of both transaction parties at each moment;
the data interception unit is used for recording a time node when a preset trigger instruction is received, and intercepting a transaction flow according to the time node;
The flow conversion unit is used for inputting the transaction flow into a preset data transcoding model to obtain a numerical value;
The feature extraction unit is used for generating a submatrix according to the numerical value and calculating the transaction feature of the submatrix;
The data fitting unit is used for counting transaction characteristics corresponding to each time node based on the time sequence to obtain fitting data;
The prediction execution unit is used for inputting the fitting data into a trained shallow learning model and outputting prediction error transaction information; wherein the predictive false transaction information contains a time stamp determined by a time node.
As a further aspect of the present invention, the instruction generating module includes:
the data interaction unit is used for sending the forecast error transaction information to the transaction parties and receiving the transaction intention degree fed back by the transaction parties;
And the process adjusting unit is used for determining a process adjusting instruction according to the transaction intent degree.
(III) beneficial effects
The invention provides an online transaction management system. Compared with the prior art, the method has the following beneficial effects:
The invention receives the error transaction information reported by the user in real time, converts the error transaction information into a two-dimensional matrix according to time and an operating party, extracts the change characteristics in the two-dimensional matrix, generates a symbiotic matrix, extracts the characteristic values in the symbiotic matrix by means of a matrix calculation formula, obtains the transaction characteristics, establishes a mapping model by the transaction characteristics and the error transaction information, and can analyze and control the transaction process according to the mapping model when receiving a monitoring request.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a block diagram of the components of an online transaction management system.
FIG. 2 is a block diagram showing the structure of an error reporting module in the online transaction management system.
FIG. 3 is a block diagram showing the structure of a data query module in the online transaction management system.
Fig. 4 is a block diagram of the composition and structure of a deep learning module in the online transaction management system.
Fig. 5 is a block diagram showing the construction of a model application module in an online transaction management system.
FIG. 6 is a block diagram of the constituent structures of the instruction generation module in the online transaction management system.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention are clearly and completely described, and it is obvious that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The embodiment of the application solves the problem of how to quickly locate the occurrence period of fluctuation in a plurality of data by providing an online transaction management system.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
FIG. 1 shows a block flow diagram of an online transaction management system, the present invention provides an online transaction management system, the system 10 comprising:
The error reporting module 11 is configured to receive error transaction information uploaded by a user, and verify the error transaction information;
The data query module 12 is configured to query transaction data in a preset storage database when verification is passed;
the deep learning module 13 is used for determining transaction characteristics according to the transaction data, creating a sample set and a test set according to the transaction characteristics, and training the transaction characteristics to a shallow learning model of the error transaction information;
The model application module 14 is configured to monitor a transaction flow in real time based on a user request, determine fitting data according to the transaction flow, input the fitting data into a trained shallow learning model, and output prediction error transaction information;
the instruction generating module 15 is configured to determine a process adjustment instruction according to the prediction error transaction information.
In an example of the technical solution of the present invention, the error reporting module 11, the data query module 12 and the deep learning module 13 are model training processes, and by acquiring transaction information (error transaction information) with abnormality uploaded by a user in real time, further extracting transaction characteristics, and processing the transaction characteristics and the error transaction information by means of the existing surface (shallow) recognition technology, a mapping from the transaction characteristics to the error transaction information can be established.
In the process, feature extraction rules and a generated shallow learning model are required to be stored, and in actual application, transaction features are identified by adopting the same feature extraction rules and the generated shallow learning model, so that prediction error information determined by a simulation process can be obtained.
In an example of the technical solution of the present invention, the model application module 14 and the instruction generation module 15 are specific application processes, and the feature extraction rules mentioned in the error reporting module 11, the data query module 12 and the deep learning module 13 are used to perform error prediction on the transaction flow monitored in real time, so that both sides of the transaction can adjust the transaction flow according to the error prediction result.
The transaction flow monitored in real time in the model application module 14 is raw data, which is stored in real time in a storage database mentioned in the data query module 12, and is used for adjusting the training process of the shallow recognition model, it is conceivable that the higher the true rate of the shallow recognition model is with the increase of the number of transaction flows.
In summary, the running processes of the model application module 14 and the instruction generation module 15 need to be implemented by means of the model training processes provided by the error reporting module 11, the data query module 12 and the deep learning module 13, and in addition, the error reporting module 11, the data query module 12 and the deep learning module 13 are affected by the model application module 14 and the instruction generation module 15 in real time, which are both recursive structures.
Fig. 2 is a block diagram of the composition and structure of an error reporting module in the online transaction management system, where the error reporting module 11 includes:
A request receiving unit 111, configured to receive an error report request containing a target object, which is uploaded by a user;
A standard inquiring unit 112 for inquiring standard price information of the target object based on the big data technology;
a payment inquiry unit 113, configured to send a payment information acquisition request to a user, and receive payment information fed back by the user;
and a comparison output unit 114 for comparing the payment information with the standard price information and outputting a verification result according to the comparison result.
The function of the error reporting module 11 is simple, and the error reporting information is acquired, and firstly, the error reporting request uploaded by the user is received.
Specifically, the error reporting request received by the technical scheme of the invention is generally only provided by one party, the other party is almost an offender, the other party is required to be monitored by the one party, and a verification process is needed to be additionally arranged, namely, the price information of the target object and the corresponding payment information are queried, and the price information and the corresponding payment information are compared to judge whether the abnormal condition exists or not.
Fig. 3 is a block diagram of the structure of a data query module in the online transaction management system, where the data query module 12 includes:
a connection unit 121 for establishing a connection channel with the storage database when the authentication passes;
The data positioning unit 122 is configured to read identity information of both sides of the transaction, and position a transaction flow and fitting data thereof in the storage database according to the identity information; the method comprises the steps of opening a user-oriented authority control port in real time in the process of reading information;
A data packaging unit 123, configured to package the located transaction flow and the fitting data thereof according to a time sequence;
the data conversion unit 124 is configured to read an occurrence period of the error transaction information, and convert the transaction flow and the fitting data thereof into sample data according to the occurrence period.
When both transaction parties use the platform, the model application module 14 monitors the transaction process based on the request, and data generated during monitoring is input into the storage database; the generated data includes the transaction flow and its fitting data, and details regarding the specific data generation process are found in the corresponding interpretations of the subsequent model application module 14.
All transaction behaviors generated between the two parties can be positioned in a storage database through the identity information of the two parties of the transaction, the error transaction information reported by one party (user) is queried, all the transaction behaviors can be classified according to the time characteristics of the error transaction information, the transaction behaviors corresponding to the error transaction information are data to be processed, and other data are auxiliary data.
Further, the reading the occurrence period of the error transaction information, and converting the transaction flow and the fitting data thereof into the content of the sample data according to the occurrence period includes:
The classifying subunit is used for reading the time point of the error transaction information, classifying the transaction flow and the fitting data thereof according to the time point and obtaining core data and auxiliary data;
the analysis subunit is used for inputting the auxiliary data into a trained shallow learning model to obtain forecast error transaction information;
the abnormality judging subunit is used for comparing the predicted error transaction information with the actual transaction information to determine the abnormality rate of the core data;
and the statistics subunit is used for counting the core data corresponding to all the error transaction information and the abnormal rate thereof to obtain sample data.
The above-mentioned content illustrates the application process of the data to be processed (core data) and auxiliary data, and it can be known from the above-mentioned content that the core data and auxiliary data are both transaction flow and fitting data thereof, and the difference is different in time; predicting auxiliary data by means of the generated shallow learning model, and then reading corresponding actual transaction information, wherein the actual transaction information is regarded as default data when no error reporting request is received; comparing the predicted result with the actual transaction information to obtain whether the historical transaction process of the two parties is stable or not, and further calculating the occurrence frequency (the number of errors/the total transaction number) of the errors, wherein the lower the frequency is, the higher the abnormality rate is; the method is that if one hundred transactions are error-free, an error is generated to be in an abnormal state, and the abnormal rate is high.
The error reporting module 11, the data query module 12 and the deep learning module 13 face all users, so that the number of the obtained error transaction information is not unique, corresponding generated core data and the abnormal rate thereof are also not unique, and all the core data and the abnormal rate thereof are counted, thus obtaining sample data.
Fig. 4 is a block diagram of the composition and structure of a deep learning module in the online transaction management system, where the deep learning module 13 includes:
A data clustering unit 131, configured to read sample data and cluster core data according to an anomaly rate;
A data filling unit 132 for filling the core data into a two-dimensional augmentation matrix based on time information; the number of rows and the number of columns of the two-dimensional augmentation matrix are both increasing functions of time;
A feature generating unit 133, configured to convert the two-dimensional augmentation matrix into a co-occurrence matrix in a preset direction, calculate a feature value of the co-occurrence matrix, and determine a transaction feature according to the feature value;
The sample determining unit 134 is configured to classify transaction features corresponding to core data in the sample data according to a preset selection ratio, and create a sample set and a test set;
A shallow training unit 135 for training a shallow learning model based on the sample set and the test set; wherein the shallow learning model is a mapping from transaction characteristics to error transaction information; the anomaly rate is used to adjust the impact weight of the transaction characteristic in the training process.
The above description specifically describes the function of the deep learning module 13, and first, the concept of depth and shallow layer needs to be described, so that, in popular terms, the depth is feature extraction+shallow layer identification, the shallow layer identification process is relatively simple, and can be completed by means of the existing classifier, and the key point is the feature extraction process.
Firstly, core data are clustered according to an anomaly rate, wherein the anomaly rate is used for representing analysis value of the core data and adjusting the influence duty ratio of corresponding transaction characteristics to a final model.
Then, after the core data are clustered according to the abnormal rate, the core data are sequentially converted into two-dimensional data, and the two-dimensional data are filled into an amplification matrix which can be continuously expanded, and the amplification process changes along with the change of time; according to the data span of the adjacent data, the two-dimensional augmentation matrix can be converted into a symbiotic matrix, and the symbiotic matrix is analyzed to obtain the characteristic value; the generated characteristic values can be non-unique, and all the characteristic values are counted to obtain the transaction characteristics.
Finally, a mapping of transaction characteristics to erroneous transaction information can be created by conventional shallow recognition techniques.
Further, the data filling unit includes:
The traversing subunit is used for traversing the core data based on the time information and acquiring the data source of the core data; the data source comprises two transaction parties; the data sources respectively correspond to rows and columns of the two-dimensional augmentation matrix;
the data transcoding sub-unit is used for inputting the core data into a preset data transcoding model to obtain a numerical value;
an inserting subunit for inserting the values into the two-dimensional augmentation matrix based on the time nodes; wherein the two-dimensional augmentation matrix contains row and column labels corresponding to the time nodes.
The above-mentioned contents define the data filling process, and can be obtained from the above-mentioned contents, the core data are transaction flow and fitting data thereof, the transaction flow is obtained in the transaction monitoring process (model application module 14), the transaction flow is used for characterizing the operations executed by both parties in the transaction process, the operations are very easy to obtain in the online transaction process, the numerical value labels corresponding to all the operations are preset and established, it is not difficult to convert the operation information monitored in real time into numerical values based on the established numerical value labels, and the obtained numerical values are inserted into the two-dimensional augmentation matrix.
The two-dimensional augmentation matrix is continuously expanded along with the change of time, and nodes need to be recorded in the expansion process.
It should be noted that the fitting data is a transaction characteristic generated based on the above procedure, and is different in that it occurs in the model application module 14, and if a corresponding memory is provided, the above content can be simplified into a data reading process, specifically, see the corresponding explanation part of the model application module 14.
In an example of the present invention, the feature generating unit includes:
the partitioning subunit is used for partitioning the two-dimensional augmentation matrix according to the time node to obtain a submatrix;
the two-dimensional augmentation matrix can be continuously segmented from the upper left to the lower right to obtain a plurality of submatrices, and the feature value of each time period can be obtained by analyzing each submatrix.
The scale determining subunit is used for obtaining the numerical value extremum of the submatrices and determining the matrix scale of the submatrices according to the numerical value extremum;
the extraction subunit is used for traversing the submatrices in a preset direction, acquiring two numerical values corresponding to adjacent elements, and respectively inserting the two numerical values into the submatrices as the row number and the column number;
The first operation in the analysis process is to convert the matrix into a co-occurrence matrix, for example, if there are 8 adjacent pairs of data with values (2, 3) in the matrix, then in the co-occurrence matrix, the values of the third column (the third row and the second column) in the second row are 8, so that the co-occurrence matrix can reflect the value change condition in the original matrix, which is the important point in the transaction monitoring process.
The calculating subunit is used for calculating the eigenvalue of the sub-symbiotic matrix based on a preset calculation formula and determining the transaction characteristic according to the eigenvalue;
and carrying out numerical operation on the symbiotic matrix to obtain a plurality of characteristic values, and counting all the characteristic values to obtain the characteristic values of each sub-matrix in the two-dimensional augmentation matrix.
Specifically, the calculating the eigenvalue of the sub-symbiotic matrix based on the preset calculation formula, and determining the content of the transaction characteristic according to the eigenvalue includes:
calculating the eigenvalue of the sub-symbiotic matrix based on a preset calculation formula group;
Counting the characteristic values according to a preset counting rule, and determining transaction characteristics;
Wherein, the formula group includes:
A＝∑i∑jP(i,j)2；
C＝∑i∑j(i-j)2P(i,j)；
E＝-∑i∑jP(i,j)logP(i,j)；
wherein A is an energy value, C is a contrast ratio, and E is an entropy value; p (i, j) is the element value of the ith row and jth column in the subco-occurrence matrix.
Further, the energy is the sum of squares of the elements of the gray level co-occurrence matrix, also called angular second order, and can represent the numerical distribution uniformity degree in the corresponding submatrix.
The contrast reflects the intensity of the change in the data in the sub-matrix.
The entropy value represents a measure of randomness of the submatrices. Entropy is maximum when all values in the sub-co-occurrence matrix are equal or the values in the sub-matrix exhibit maximum randomness.
Fig. 5 is a block diagram of the composition and structure of a model application module in an online transaction management system, and as a preferred embodiment of the present invention, the model application module 14 includes:
The data monitoring unit 141 is configured to receive a monitoring request sent by both parties of the transaction, and monitor a transaction flow in real time based on the monitoring request; the transaction flow is the operation information of both transaction parties at each moment;
the data intercepting unit 142 is configured to record a time node when a preset trigger instruction is received, and intercept a transaction flow according to the time node;
a flow conversion unit 143, configured to input the transaction flow into a preset data transcoding model to obtain a numerical value;
A feature extraction unit 144, configured to generate a submatrix according to the value, and calculate a transaction feature of the submatrix;
A data fitting unit 145, configured to calculate transaction characteristics corresponding to each time node based on a time sequence, and obtain fitting data;
the prediction execution unit 146 is configured to input the fitting data into a trained shallow learning model, and output prediction error transaction information; wherein the predictive false transaction information contains a time stamp determined by a time node.
The above-mentioned contents specifically describe the transaction monitoring process, firstly, the two transaction parties send monitoring requests together, the execution main body of the method can execute subsequent operations, and in general, in order to prevent errors, the two transaction parties send monitoring requests; along with the continuous operation of both transaction parties, some trigger instructions are continuously triggered, so that a plurality of time nodes are determined, for example, when the first party checks goods, a certain button is clicked, and the whole transaction flow can be segmented by the time nodes.
Then, the transaction flow is a collection of operation information, the process of converting the operation information into numerical values can be completed by means of a preset mapping table, and the transaction characteristics generated in the transaction flow are continuously extracted by means of the characteristic extraction process extracted from the content, so that fitting data are obtained; and inputting the fitting data into a trained shallow learning model, so that each time period can be predicted, prediction information is generated in real time, and both parties of the transaction are prompted.
Fig. 6 is a block diagram of the composition and structure of an instruction generating module in the online transaction management system, and the instruction generating module 15 includes:
a data interaction unit 151, configured to send the predicted error transaction information to both parties of the transaction, and receive the transaction intent degree fed back by both parties of the transaction;
and the process adjusting unit 152 is used for determining a process adjusting instruction according to the transaction intent degree.
In one example of the technical scheme of the invention, the instruction generation process is limited, the instruction generation process is jointly determined by two transaction parties, the transaction intention degree fed back by the two transaction parties is received, and the transaction flow is regulated by the transaction intention degree; in general, the intent of the transaction is not too low, and it is better to cancel the transaction directly once there is a party who does not want to continue the transaction.
In summary, compared with the prior art, the invention has the following beneficial effects:
The invention receives the error transaction information reported by the user in real time, converts the error transaction information into a two-dimensional matrix according to time and an operating party, extracts the change characteristics in the two-dimensional matrix, generates a symbiotic matrix, extracts the characteristic values in the symbiotic matrix by means of a matrix calculation formula, obtains the transaction characteristics, establishes a mapping model by the transaction characteristics and the error transaction information, and can analyze and control the transaction process according to the mapping model when receiving a monitoring request.
It should be noted that, from the above description of the embodiments, those skilled in the art will clearly understand that each embodiment may be implemented by means of software plus necessary general hardware platform. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the various embodiments or the system described in some parts of the embodiments. In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, system, 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, system, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, system, article, or apparatus that comprises the element.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (6)
1. An online transaction management system, the system comprising:
the error reporting module is used for receiving the error transaction information uploaded by the user and verifying the error transaction information;
The data query module is used for querying transaction data in a preset storage database when verification passes;
The deep learning module is used for determining transaction characteristics according to the transaction data, creating a sample set and a test set according to the transaction characteristics, and training the transaction characteristics to a shallow learning model of the error transaction information;
The model application module is used for monitoring a transaction flow in real time based on a user request, determining fitting data according to the transaction flow, inputting the fitting data into a trained shallow learning model, and outputting forecast error transaction information;
The instruction generation module is used for determining a process adjusting instruction according to the prediction error transaction information;
The error reporting module comprises:
The request receiving unit is used for receiving the error report request containing the target object uploaded by the user;
the standard inquiring unit is used for inquiring standard price information of the target object based on the big data technology;
the payment inquiry unit is used for sending a payment information acquisition request to the user and receiving payment information fed back by the user;
The comparison output unit is used for comparing the payment information with the standard price information and outputting a verification result according to a comparison result;
The data query module comprises:
the connection unit is used for establishing a connection channel with the storage database when the verification is passed;
The data positioning unit is used for reading the identity information of both transaction sides, and positioning the transaction flow and fitting data thereof in the storage database according to the identity information; the method comprises the steps of opening a user-oriented authority control port in real time in the process of reading information;
the data packaging unit is used for packaging the positioned transaction flow and fitting data thereof according to the time sequence;
The data conversion unit is used for reading the occurrence time period of the error transaction information and converting the transaction flow and fitting data thereof into sample data according to the occurrence time period;
the data conversion unit includes:
The classifying subunit is used for reading the time point of the error transaction information, classifying the transaction flow and the fitting data thereof according to the time point and obtaining core data and auxiliary data;
the analysis subunit is used for inputting the auxiliary data into a trained shallow learning model to obtain forecast error transaction information;
the abnormality judging subunit is used for comparing the predicted error transaction information with the actual transaction information to determine the abnormality rate of the core data;
The statistics subunit is used for counting core data corresponding to all the error transaction information and the abnormal rate of the core data to obtain sample data;
the deep learning module includes:
The data clustering unit is used for reading the sample data and clustering the core data according to the abnormal rate;
a data filling unit for filling the core data into a two-dimensional augmentation matrix based on time information; the number of rows and the number of columns of the two-dimensional augmentation matrix are both increasing functions of time;
the feature generation unit is used for converting the two-dimensional augmentation matrix into a symbiotic matrix in a preset direction, calculating a feature value of the symbiotic matrix and determining transaction features according to the feature value;
the sample determining unit is used for classifying transaction characteristics corresponding to core data in sample data according to a preset selection proportion, and creating a sample set and a test set;
The shallow training unit is used for training a shallow learning model based on the sample set and the test set; wherein the shallow learning model is a mapping from transaction characteristics to error transaction information; the anomaly rate is used to adjust the impact weight of the transaction characteristic in the training process.
2. The online transaction management system of claim 1, wherein the data populating unit includes:
The traversing subunit is used for traversing the core data based on the time information and acquiring the data source of the core data; the data source comprises two transaction parties; the data sources respectively correspond to rows and columns of the two-dimensional augmentation matrix;
the data transcoding sub-unit is used for inputting the core data into a preset data transcoding model to obtain a numerical value;
an inserting subunit for inserting the values into the two-dimensional augmentation matrix based on the time nodes; wherein the two-dimensional augmentation matrix contains row and column labels corresponding to the time nodes.
3. The online transaction management system according to claim 2, wherein the feature generation unit includes:
the partitioning subunit is used for partitioning the two-dimensional augmentation matrix according to the time node to obtain a submatrix;
The scale determining subunit is used for obtaining the numerical value extremum of the submatrices and determining the matrix scale of the submatrices according to the numerical value extremum;
the extraction subunit is used for traversing the submatrices in a preset direction, acquiring two numerical values corresponding to adjacent elements, and respectively inserting the two numerical values into the submatrices as the row number and the column number;
And the calculating subunit is used for calculating the eigenvalue of the sub-symbiotic matrix based on a preset calculation formula and determining the transaction characteristic according to the eigenvalue.
4. The online transaction management system of claim 3, wherein the calculating eigenvalues of the sub-symbiotic matrix based on a preset calculation formula, and determining contents of the transaction characteristics according to the eigenvalues comprises:
calculating the eigenvalue of the sub-symbiotic matrix based on a preset calculation formula group;
Counting the characteristic values according to a preset counting rule, and determining transaction characteristics;
Wherein, the formula group includes:
A＝∑i∑jP(i,j)2；
C＝∑i∑j(i-j)2P(i,j)；
E＝-∑i∑jP(i,j)logP(i,j)；
wherein A is an energy value, C is a contrast ratio, and E is an entropy value; p (i, j) is the element value of the ith row and jth column in the subco-occurrence matrix.
5. The online transaction management system of claim 1, wherein the model application module comprises:
The data monitoring unit is used for receiving monitoring requests sent by both transaction sides and monitoring the transaction flow in real time based on the monitoring requests; the transaction flow is the operation information of both transaction parties at each moment;
the data interception unit is used for recording a time node when a preset trigger instruction is received, and intercepting a transaction flow according to the time node;
The flow conversion unit is used for inputting the transaction flow into a preset data transcoding model to obtain a numerical value;
The feature extraction unit is used for generating a submatrix according to the numerical value and calculating the transaction feature of the submatrix;
The data fitting unit is used for counting transaction characteristics corresponding to each time node based on the time sequence to obtain fitting data;
The prediction execution unit is used for inputting the fitting data into a trained shallow learning model and outputting prediction error transaction information; wherein the predictive false transaction information contains a time stamp determined by a time node.
6. The online transaction management system of claim 5, wherein the instruction generation module comprises:
the data interaction unit is used for sending the forecast error transaction information to the transaction parties and receiving the transaction intention degree fed back by the transaction parties;
And the process adjusting unit is used for determining a process adjusting instruction according to the transaction intent degree.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310825934.8A CN116777629B (en) | 2023-07-06 | 2023-07-06 | Online transaction management system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310825934.8A CN116777629B (en) | 2023-07-06 | 2023-07-06 | Online transaction management system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116777629A CN116777629A (en) | 2023-09-19 |
| CN116777629B true CN116777629B (en) | 2024-05-03 |
Family
ID=87989451
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310825934.8A Active CN116777629B (en) | 2023-07-06 | 2023-07-06 | Online transaction management system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116777629B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101834260B1 (en) * | 2017-01-18 | 2018-03-06 | 한국인터넷진흥원 | Method and Apparatus for Detecting Fraudulent Transaction |
| CN110059802A (en) * | 2019-03-29 | 2019-07-26 | 阿里巴巴集团控股有限公司 | For training the method, apparatus of learning model and calculating equipment |
| CN115471309A (en) * | 2022-08-12 | 2022-12-13 | 北京神州新桥科技有限公司 | Transaction abnormity detection method and device, electronic equipment and readable storage medium |
| CN115689740A (en) * | 2022-09-07 | 2023-02-03 | 中国银行股份有限公司 | Transaction abnormity detection method and device based on deep learning |
| CN116308370A (en) * | 2021-12-16 | 2023-06-23 | 第四范式（北京）技术有限公司 | Training method of abnormal transaction recognition model, abnormal transaction recognition method and device |
-
2023
- 2023-07-06 CN CN202310825934.8A patent/CN116777629B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101834260B1 (en) * | 2017-01-18 | 2018-03-06 | 한국인터넷진흥원 | Method and Apparatus for Detecting Fraudulent Transaction |
| CN110059802A (en) * | 2019-03-29 | 2019-07-26 | 阿里巴巴集团控股有限公司 | For training the method, apparatus of learning model and calculating equipment |
| CN116308370A (en) * | 2021-12-16 | 2023-06-23 | 第四范式（北京）技术有限公司 | Training method of abnormal transaction recognition model, abnormal transaction recognition method and device |
| CN115471309A (en) * | 2022-08-12 | 2022-12-13 | 北京神州新桥科技有限公司 | Transaction abnormity detection method and device, electronic equipment and readable storage medium |
| CN115689740A (en) * | 2022-09-07 | 2023-02-03 | 中国银行股份有限公司 | Transaction abnormity detection method and device based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 面向银行业务的交易监控可视化系统设计与实现;任立男;段桂华;谭荻;王建新;;中南大学学报(自然科学版)(第10期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116777629A (en) | 2023-09-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11513869B2 (en) | Systems and methods for synthetic database query generation | |
| CN109325691B (en) | Abnormal behavior analysis method, electronic device and computer program product | |
| Newgard et al. | Advanced statistics: missing data in clinical research—part 2: multiple imputation | |
| WO2018183743A1 (en) | Composite machine-learning system for label prediction and training data collection | |
| US20190294990A1 (en) | Detecting false positives in statistical models | |
| CN111078544A (en) | Software defect prediction method, device, equipment and storage medium | |
| US20230091402A1 (en) | Systems and methods for expanding data classification using synthetic data generation in machine learning models | |
| CN109934268A (en) | Abnormal transaction detection method and system | |
| CN112767008A (en) | Enterprise revenue trend prediction method and device, computer equipment and storage medium | |
| CN115547466A (en) | Medical institution registration and review system and method based on big data | |
| CN114997916A (en) | Prediction method, system, electronic device and storage medium of potential user | |
| CN114819777A (en) | Enterprise sales business analysis and management system based on digital twin technology | |
| CN112990989B (en) | Value prediction model input data generation method, device, equipment and medium | |
| CN117035697B (en) | ITSM (integrated traffic simulation) platform optimization method and system based on historical dynamic analysis | |
| CN116777629B (en) | Online transaction management system | |
| CN116485185A (en) | Enterprise risk analysis system and method based on comparison data | |
| JP2021022051A (en) | Machine learning program, machine learning method, and machine learning apparatus | |
| KR20200071646A (en) | Detection apparatus for detecting anomaly log and operating method of same, and training apparatus and operating method of same | |
| CN114880590A (en) | Multi-language website currency automatic conversion system and method thereof | |
| CN112823502B (en) | Real-time feedback service for resource access rule configuration | |
| TWM601397U (en) | Customized marketing system with customer clustering service | |
| CN113239024B (en) | Bank abnormal data detection method based on outlier detection | |
| Wu et al. | A Collaborative Filtering Method for Operation Maintenance Behavior in Power Monitoring Systems | |
| US11714997B2 (en) | Analyzing sequences of interactions using a neural network with attention mechanism | |
| CN116955648B (en) | Knowledge graph analysis method based on non-privacy data association |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20240412 Address after: Room 1502-2, No. 1110 Jimei North Avenue, Software Park Phase III, Torch High tech Zone, Xiamen City, Fujian Province, 361000 Applicant after: Entrepreneurial Tree (Xiamen) Digital Technology Co.,Ltd. Country or region after: China Address before: 501, 5th Floor, Building 14, Yard 7, Dijin Road, Haidian District, Beijing, 100080 (F-195) Applicant before: Beijing Ruilin Light Technology Co.,Ltd. Country or region before: China |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |