CN117132371A - Method, apparatus, computer device and storage medium for predicting risk tolerance level - Google Patents
Method, apparatus, computer device and storage medium for predicting risk tolerance level Download PDFInfo
- Publication number
- CN117132371A CN117132371A CN202310682715.9A CN202310682715A CN117132371A CN 117132371 A CN117132371 A CN 117132371A CN 202310682715 A CN202310682715 A CN 202310682715A CN 117132371 A CN117132371 A CN 117132371A
- Authority
- CN
- China
- Prior art keywords
- data
- alliance chain
- behavior data
- risk
- information
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- 230000006399 behavior Effects 0.000 claims abstract description 185
- 238000013136 deep learning model Methods 0.000 claims abstract description 62
- 238000011156 evaluation Methods 0.000 claims abstract description 62
- 238000012216 screening Methods 0.000 claims abstract description 25
- 238000012549 training Methods 0.000 claims description 40
- 238000012502 risk assessment Methods 0.000 claims description 32
- 238000004590 computer program Methods 0.000 claims description 24
- 238000011176 pooling Methods 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 8
- 238000012795 verification Methods 0.000 claims description 8
- 239000012634 fragment Substances 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 4
- 230000003542 behavioural effect Effects 0.000 claims 1
- 230000008569 process Effects 0.000 description 11
- 238000012545 processing Methods 0.000 description 8
- 230000004044 response Effects 0.000 description 5
- 238000013500 data storage Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000008520 organization Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000013475 authorization Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 150000001875 compounds Chemical class 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000000977 initiatory effect Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 101100217298 Mus musculus Aspm gene Proteins 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 229910021389 graphene Inorganic materials 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000007246 mechanism Effects 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
- 230000009466 transformation Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2457—Query processing with adaptation to user needs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/27—Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The application relates to a method, a device, a computer device and a storage medium for predicting risk tolerance levels. The method comprises the following steps: responding to a risk evaluation instruction of an evaluation object, and acquiring object information of the evaluation object; generating a message digest of the object information through a message digest algorithm; accessing the alliance chain, and screening the blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to the object information; integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information; and inputting the object behavior data into the trained deep learning model for risk level prediction to obtain a risk bearing level corresponding to the evaluation object. The method can avoid the limitation caused by subjective arrangement of questions and answers, and further has higher accuracy of risk bearing grades obtained through model prediction.
Description
Technical Field
The present application relates to the field of artificial intelligence, and in particular, to a method, an apparatus, a computer device, a storage medium, and a computer program product for predicting risk tolerance levels.
Background
Risk Assessment (Risk Assessment) refers to the task of quantitatively assessing the likelihood of impact and loss of various aspects of life or property, before or after the occurrence of a Risk event, but not yet completed. The concept of risk assessment is also gradually introduced into other industries or fields; taking the financial industry as an example, with the development of smart phones and internet technology, personal financial products have widely come into view of the masses, and individuals should complete the necessary steps of risk assessment before investment financial institutions such as banks, insurance and securities perform investment financial institutions.
However, in the conventional risk assessment method or method, characteristic attribute information of a target object is collected by adopting a risk measurement and assessment method, and the information collection method generally uses question options in the risk measurement and assessment method, selects close answers from limited question options according to the target object, and collects and sorts the answers as attribute information of the target object. The accuracy of the final risk assessment results is low due to the limitation of the questionnaire in forming question options.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a method, an apparatus, a computer device, a computer-readable storage medium, and a computer program product for predicting a risk tolerance level with higher accuracy.
In a first aspect, the present application provides a method for predicting a risk tolerance level. The method comprises the following steps:
responding to a risk evaluation instruction of an evaluation object, and acquiring object information of the evaluation object;
generating a message digest of the object information through a message digest algorithm;
accessing a alliance chain, and screening blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to object information, wherein historical behavior data corresponding to the object information is stored in the alliance chain blocks;
integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information;
and inputting the object behavior data into a trained deep learning model for risk level prediction to obtain a risk bearing level corresponding to the evaluation object, wherein the trained deep learning model is trained based on training sample data, and the training sample data comprises historical behavior data of different objects and corresponding risk levels.
In one embodiment, before filtering the blocks of the coalition chain according to the information abstract and the ciphertext abstract in the coalition chain blocks to obtain a plurality of coalition chain blocks corresponding to the object information, the method further includes:
Acquiring a node digital certificate carried by the risk assessment instruction;
checking the access permission of the node digital certificate according to a certificate trust chain stored in a alliance chain node;
screening the blocks of the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to the object information, wherein the steps comprise:
and if the node digital certificate passes the access authority verification, screening the blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain block to obtain a plurality of alliance chain blocks corresponding to the object information.
In one embodiment, the integrating the historical behavior data in the plurality of alliance chain blocks to obtain the object behavior data corresponding to the object information includes:
acquiring ciphertext data in a plurality of alliance chain blocks corresponding to the object information;
according to plaintext data obtained by decrypting the ciphertext data;
and generating object behavior data corresponding to the object information based on the plain text data obtained by decryption.
In one embodiment, the plaintext data obtained by decrypting the ciphertext data includes:
Acquiring a private key corresponding to the risk assessment instruction;
decrypting the ciphertext data through a private key corresponding to the risk assessment instruction to obtain storage address information of historical behavior data corresponding to the object information;
addressing nodes accessed to the alliance chain according to the storage address information to obtain a plurality of storage nodes;
retrieving a plurality of historical data segments from a local store of the storage node according to the object information;
and combining the plurality of historical data fragments to form plaintext data.
In one embodiment, the inputting the object behavior data into the trained deep learning model for risk level prediction, before obtaining the risk bearing level corresponding to the evaluation object, includes:
acquiring a deep learning model to be trained and a training data set of historical behavior data, wherein the training data set comprises a risk level label and a behavior data sample;
inputting the behavior data sample into the deep learning model to be trained to predict risk level, and obtaining a prediction level label;
and if the similarity value between the prediction grade label and the risk grade label is not smaller than a preset similarity threshold value, obtaining a trained deep learning model.
In one embodiment, inputting the behavior data sample to the deep learning model to be trained for prediction, and obtaining a prediction grade label includes:
carrying out convolution pooling treatment on the behavior data samples through a convolution layer and a pooling layer which are preset in a deep learning model to be trained to obtain behavior characteristic data;
and calculating the behavior characteristic data through a convolution function preset in the deep learning model to be trained to obtain a prediction grade label.
In a second aspect, the application further provides a device for predicting the risk tolerance level. The device comprises:
the information acquisition module is used for responding to a risk assessment instruction of an assessment object and acquiring object information of the assessment object;
the abstract generating module is used for generating an abstract of the object information through an abstract algorithm;
the block screening module accesses the alliance chain, screens the block in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain block to obtain a plurality of alliance chain blocks corresponding to the object information, and the alliance chain blocks store historical behavior data corresponding to the object information;
the data acquisition module is used for integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information;
The level prediction module is used for inputting the object behavior data into a trained deep learning model to perform risk level prediction to obtain a risk bearing level corresponding to the evaluation object, the trained deep learning model is trained based on training sample data, and the training sample data comprises historical behavior data of different objects and corresponding risk levels.
In a third aspect, the present application also provides a computer device. The computer device comprises a memory storing a computer program and a processor which when executing the computer program performs the steps of:
responding to a risk evaluation instruction of an evaluation object, and acquiring object information of the evaluation object;
generating a message digest of the object information through a message digest algorithm;
accessing a alliance chain, and screening blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to object information, wherein historical behavior data corresponding to the object information is stored in the alliance chain blocks;
integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information;
And inputting the object behavior data into a trained deep learning model for risk level prediction to obtain a risk bearing level corresponding to the evaluation object, wherein the trained deep learning model is trained based on training sample data, and the training sample data comprises historical behavior data of different objects and corresponding risk levels.
In a fourth aspect, the present application also provides a computer-readable storage medium. The computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of:
responding to a risk evaluation instruction of an evaluation object, and acquiring object information of the evaluation object;
generating a message digest of the object information through a message digest algorithm;
accessing a alliance chain, and screening blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to object information, wherein historical behavior data corresponding to the object information is stored in the alliance chain blocks;
integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information;
And inputting the object behavior data into a trained deep learning model for risk level prediction to obtain a risk bearing level corresponding to the evaluation object, wherein the trained deep learning model is trained based on training sample data, and the training sample data comprises historical behavior data of different objects and corresponding risk levels.
In a fifth aspect, the present application also provides a computer program product. The computer program product comprises a computer program which, when executed by a processor, implements the steps of:
responding to a risk evaluation instruction of an evaluation object, and acquiring object information of the evaluation object;
generating a message digest of the object information through a message digest algorithm;
accessing a alliance chain, and screening blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to object information, wherein historical behavior data corresponding to the object information is stored in the alliance chain blocks;
integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information;
And inputting the object behavior data into a trained deep learning model for risk level prediction to obtain a risk bearing level corresponding to the evaluation object, wherein the trained deep learning model is trained based on training sample data, and the training sample data comprises historical behavior data of different objects and corresponding risk levels.
The application provides a method, a device, computer equipment and a storage medium for predicting risk tolerance level; the method stores the historical behavior data in advance in a alliance chain mode, can package the historical behavior data from different sources into blocks in the alliance chain, and enables the historical behavior data to be stored more safely and comprehensively based on the characteristic that the data in the alliance chain blocks cannot be tampered; when the prediction is performed, the risk assessment instruction is responded, the block corresponding to the object information is obtained based on the comparison screening of the information abstract carrying the object information in the instruction and the ciphertext abstract in the alliance chain, and the historical behavior data of various sources in the alliance chain can be more comprehensively and accurately invoked in a manner of comparison screening by the ciphertext abstract, so that the data content input into the deep learning model is more detailed and comprehensive, the limitation caused by subjective arrangement of questions and answers is avoided, and the accuracy of the risk bearing grade obtained through model prediction is higher.
Drawings
FIG. 1 is an application environment diagram of a method of predicting risk tolerance levels in one embodiment;
FIG. 2 is a flow chart of a method of predicting risk tolerance level in one embodiment;
FIG. 3 is a flow chart illustrating the steps of the admission permission verification sub-step in one embodiment;
FIG. 4 is a flow chart of another embodiment of a method for predicting risk tolerance;
FIG. 5 is a block diagram showing a structure of a risk tolerance level predicting apparatus according to an embodiment;
fig. 6 is an internal structural diagram of a computer device in one embodiment.
Detailed Description
Before the description of the specific embodiments is developed, the following definitions and descriptions of the related terms related to the present application are needed:
a federated chain is a cluster of multiple private chains, a blockchain that is jointly participated in management by multiple organizations, each organization or organization managing one or more nodes whose data only allows different organizations within the system to read, write and send. Each node of the federation chain typically has an entity organization corresponding thereto that can join and leave the network after authorization. Organizations constitute benefit-related federations that collectively maintain healthy operation of blockchains.
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
The method for predicting the risk tolerance level provided by the embodiment of the application can be applied to an application environment shown in fig. 1. Wherein the terminal 102 communicates with the server 104 via a network. The data storage system may store data that the server 104 needs to process. The data storage system may be integrated on the server 104 or may be located on a cloud or other network server. In this implementation environment, the data storage manner in the server 104 may store the historical behavior data related to the evaluation object in a coalition chain manner. The server 104 serves as a node in the alliance chain, after the node passes authentication and authorization of other nodes in the alliance chain, the behavior data of the evaluation object received or temporarily stored in the server 104 is packaged based on a consensus mechanism in the alliance chain to form a new block in the alliance chain, and meanwhile, the information of the new block is sent to other nodes in the alliance chain in a broadcast mode. In this implementation environment, the server 104 starts prediction of the risk tolerance level of the evaluation object in response to the risk evaluation instruction transmitted from the terminal 102. The server 104 first determines an evaluation object to be subjected to risk tolerance level evaluation based on the received risk evaluation instruction, and synchronously acquires object information of the evaluation object. Server 104 generates a message digest of the object information by a message digest algorithm. After obtaining the information digest, server 104 may filter the historical behavior data stored in the block in the federated chain according to the information digest by directly comparing the obtained information digest with the ciphertext digest in the federated chain block. If the abstract content is the same, the historical behavior data of the evaluation object in the alliance chain block is determined. And integrating the historical behavior data in all the blocks according to the screening result to form object behavior data of the evaluation object. The server 104 further loads a trained deep learning model, which is trained based on training sample data, wherein the training sample data adopted in the training process comprises historical behavior data of different objects and corresponding risk levels. And (3) inputting the object behavior data into a trained deep learning model to predict the risk level, and finally outputting the risk bearing level of the evaluation object by the deep learning model. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, internet of things devices, and portable wearable devices, where the internet of things devices may be smart speakers, smart televisions, smart air conditioners, smart vehicle devices, and the like. The portable wearable device may be a smart watch, smart bracelet, headset, or the like. The server 104 may be implemented as a stand-alone server or as a server cluster of multiple servers.
In one embodiment, as shown in fig. 2, a method for predicting risk tolerance level is provided, and the method is applied to the server 104 in fig. 1 for illustration, and includes the following steps:
step 202, in response to a risk assessment instruction of an assessment object, object information of the assessment object is acquired.
In an embodiment, the risk assessment instruction is instruction information for performing sustainable risk level prediction on the current assessment object in response to an operation of the target object in the terminal interactive interface. The evaluation object is a target object which needs to be determined to be able to bear the risk level in a specific scene, and the specific scene includes but is not limited to a scene of pushing financial products, intelligent auxiliary decision and the like. The object information refers to information content capable of uniquely describing the evaluation object, including, but not limited to, information such as a target name and an ID.
In an embodiment, the server obtains an interaction instruction of the operation object through an interaction operation interface of the terminal, triggers a corresponding risk assessment instruction, and sends the risk assessment instruction to the server. The server performs necessary analysis processing on the risk assessment instruction to obtain object information of the assessment object carried in the risk assessment instruction, for example, an ID of the assessment object. Based on the ID of the evaluation object, the server can extract the historical behavior data corresponding to the evaluation object from the alliance chain block.
In step 204, a message digest of the object information is generated by a message digest algorithm.
The information summarization algorithm is also called a hash algorithm and a hash algorithm, and converts data with any length into a data string with fixed length through a function; for example, a string of 16 digits; the main characteristic of the message digest algorithm is that the encryption process does not need a key, and the encrypted data cannot be decrypted, and the same ciphertext can be obtained only by inputting the same plaintext data through the same message digest algorithm. Correspondingly, the information digest in the embodiment is a fixed-length one-way Ha Xisan column value generated by the digest algorithm for the object information.
Specifically, the information digest algorithm that may be employed in an embodiment includes, but is not limited to, MD5 encryption algorithm, SHA1 encryption algorithm, and SHA256 encryption algorithm. Taking the process of generating the information abstract by using the MD5 encryption algorithm as an example, in the embodiment, after the server analyzes the risk assessment instruction to obtain the object ID of the assessment object, when the MD5 is used for encryption, the information content of the object ID is first filled, so that the bit length of the result obtained by summing up 512 is over 448. Thus, the bit Length (Bits Length) of the information will be extended to n×512+448, where N is a non-negative integer and N may be zero. After the padding process, the bit length of the information=n×512+448+64= (n+1) ×512, i.e. the length is exactly an integer multiple of 512. In the information after the padding expansion, the initial 128-bit values are initial link variables, and the parameters are used for the first round of operation and are expressed by a big-end byte order, and are respectively: a=0x01234567, b=0x89abcdef, c=0xfedcba98, d=0x 76543210. Packet data is then processed, the first packet requiring copying of the four link variables above into the other four variables: a to a, B to B, C to C, D to D. The variables from the second packet are the result of the operation of the previous packet, i.e., a=a, b=b, c=c, d=d. The main cycle has four wheels (only three wheels in MD 4), and each cycle is very similar. The first round was operated 16 times. Each operation performs a nonlinear function operation on three of a, b, c, and d, and then adds the result to a fourth variable, a subgroup of text, and a constant. The result is then shifted to the left by an indefinite number and one of a, b, c or d is added. Finally, one of a, b, c or d is replaced with the result. The final output result is a cascade of a, b, c and d, and the cascade result is the information abstract of the object information.
And 206, accessing the alliance chain, and screening the blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to the object information, wherein the alliance chain blocks store historical behavior data corresponding to the object information.
In an embodiment, the structure of a federated chain block is similar to that of a common blockchain, and the block also comprises a block head and a block body, wherein historical behavior data of a plurality of objects are stored in the block body. The block header includes the version number of the block, the hash value of the previous block, the timestamp, the difficulty target, and the hash value of the block header itself. The ciphertext abstract may be abstract information obtained by performing encryption operation on historical behavior data stored in a coalition chain block or object information to which the historical behavior data belongs through an information abstract algorithm. In addition, the historical behavior data corresponding to the object information in the embodiment is a data record corresponding to the business behavior generated by the object at different times and in different scenes.
Illustratively, in the embodiment, the server is used as a node of the federation chain, and after passing the authentication of other nodes in the federation chain, the server can access the federation chain and acquire the historical behavior data corresponding to the object information from the blocks in the federation chain. In order to obtain the historical behavior data corresponding to the evaluation object, the server needs to filter the historical behavior data stored in each block in the alliance chain after successfully accessing to the alliance chain. Because the alliance chain node in the embodiment distinguishes the behavior data belonging to different objects in the block when the historical behavior data is encrypted and packed to construct a new block; for example, the method comprises the steps of performing simple classification processing according to historical behavior data generated by different objects to form a behavior data set of the objects, generating a data set label based on the object information of the objects, and giving corresponding behavior data sets. More specifically, the data set label marked in the embodiment may be the whole content or part of the content of the ciphertext abstract generated by using the information abstract algorithm to subject information. Furthermore, after the embodiment and obtaining the information abstract corresponding to the object information of the behavior data to be called, the matching search is performed with the ciphertext abstract in the alliance chain block, so as to obtain the alliance chain block storing the historical behavior data corresponding to the object information. It should be noted that, in the embodiment, the nodes accessing to the federation chain may all implement the acquisition of the behavior data, and package the acquired behavior data to form a block of the federation chain, so that the behavior data corresponding to a single object may be stored in the blocks of multiple federation chains. Therefore, when matching search is performed between the information abstract of the object information and the ciphertext abstract in the alliance chain block, a plurality of alliance chain blocks with an indefinite number may be obtained. After a plurality of alliance chain blocks are obtained after matching search, historical behavior data in the blocks are extracted, and a plurality of pieces of historical behavior data of an evaluation object are obtained.
Step 208, integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information.
The object behavior data is the sum of the behavior data stored in the alliance chain block by the current object. Historical behavior data for the same object may be stored in multiple federated chain blocks, which may be historical behavior data for multiple objects due to the historical behavior data stored in the federated chain blocks in embodiments; meanwhile, historical behavior data generated by a single object in different periods and different scenes can be stored in different blocks in a blockchain; for example, the background servers of the A system platform and the B system platform adopt the same alliance chain to store the consumption behavior data of the consumption objects, so that the consumption behaviors of the same consumption objects in the A system platform are packaged to form blocks in the alliance chain; similarly, the consumption behavior of the consumption object in the B-system platform is also packed into blocks in the federation chain. Thus, after passing the matching filtering, the historical behavior data in the plurality of federated chain blocks needs to be integrated.
For example, after the server performs the matching search through step 206 in the embodiment, a record of a plurality of pieces of historical behavior data of the evaluation object may be obtained. Further, the server integrates a plurality of pieces of historical behavior data to form object behavior data of the evaluation object. It will be appreciated that in embodiments the object behavior data, which may be in the form of data of a dataset; in addition, in the process of integrating a plurality of pieces of history behavior data to form object behavior data, necessary data cleansing work may be performed on the history data behavior, for example, repeated history behavior data, behavior data with missing contents, and other noise data may be removed.
Step 210, inputting object behavior data into a trained deep learning model for risk level prediction to obtain a risk bearing level corresponding to an evaluation object; the trained deep learning model is trained based on training sample data, wherein the training sample data comprises historical behavior data of different objects and corresponding risk levels.
The deep learning model refers to a machine learning model based on an artificial neural network, and learns and processes data through a multi-layer neural network. The core idea of the deep learning model is to learn the characteristic representation of data through multi-layer nonlinear transformation, so as to realize the identification and classification of complex modes. In an embodiment, the trained deep learning model is obtained by training sample data formed by behavior data; the behavior data constituting the training sample data may be a plurality of pieces of historical behavior data including a plurality of different objects.
When predicting an evaluation object with an unknown risk tolerance level, the object behavior data obtained in the previous step is input into a trained deep learning model for prediction, the risk tolerance level corresponding to the evaluation object is obtained, and the risk tolerance level is sent to a mobile terminal for visual display.
According to the prediction method of the risk tolerance level, the historical behavior data are stored in advance in a alliance chain mode, the historical behavior data from different sources can be packaged to form blocks in the alliance chain, and the historical behavior data are stored more safely and comprehensively based on the characteristic that the data in the alliance chain blocks cannot be tampered; when the prediction is performed, the risk assessment instruction is responded, the block corresponding to the object information is obtained based on the comparison screening of the information abstract carrying the object information in the instruction and the ciphertext abstract in the alliance chain, and the historical behavior data of various sources in the alliance chain can be more comprehensively and accurately invoked in a manner of comparison screening by the ciphertext abstract, so that the data content input into the deep learning model is more detailed and comprehensive, the limitation caused by subjective arrangement of questions and answers is avoided, and the accuracy of the risk bearing grade obtained through model prediction is higher.
In one embodiment, as shown in fig. 3, before screening the blocks of the federated chain according to the information abstract and the ciphertext abstract in the blocks of the federated chain to obtain a plurality of blocks of the federated chain corresponding to the object information, the method further includes the following steps:
step 302, a node digital certificate carried by a risk assessment instruction is obtained.
In contrast to blockchains, because the blockchain is not completely decentralised, the blockchain only allows authorized entities to participate in the blockchain network, thus increasing the number of authorized processes in the blockchain; and the authorized operation is completed in the form of issuing digital certificates. The digital certificate which is issued by a certificate issuing organization in the alliance chain, namely an authentication center and stored in the local node of the alliance chain is recorded as a node digital certificate. From the data structure, the node digital certificate is a file digitally signed by a certificate authority that contains public key owner information and a public key.
Illustratively, in the federation chain of the embodiment, the server is required to acquire the corresponding node digital certificate to join the federation chain to become a federation chain node. Firstly, generating a public-private key pair in a server through a preset asymmetric encryption algorithm, then, initiating an authentication application to an authentication center in a alliance chain by the server, and transmitting the generated public key and node information of the server to the authentication center; the authentication center verifies the authenticity of the information provided by the application node through various means such as online, offline and the like. After passing the authenticity verification, the authentication center issues an authentication file, namely a node electronic certificate, to the applied server. When the server needs to predict the risk bearing grade of the evaluation object, the server needs to acquire the historical behavior data of the evaluation object from the alliance chain, then a request instruction of risk evaluation needs to be initiated to the alliance chain, and a node digital certificate of the current server is attached to the request instruction so that other nodes in the alliance chain can carry out admission verification on the current server.
And step 304, checking the access permission of the node digital certificate according to the certificate trust chain stored in the alliance chain node.
Since trust relationships can exist between certificates, one certificate can prove that another certificate is also truly trusted, i.e. the trust relationships between the certificates can be nested. For example, the C certificate trusts A and B, then A trusts A1 and A2, and B trusts B1 and B2. The certificates form a tree relation, and the tree relation is a certificate trust chain.
For example, when the federation chain node applies for authentication to the authentication center, the private key of the authentication center is required to carry out asymmetric encryption on the signature digest of the whole certificate, namely the node digital certificate can be decrypted by the public key of the authentication center to obtain the signature digest of the node certificate. When the same digest algorithm needs to be used again in the embodiment, that is, the algorithm used for saving the node digital certificate is stored in the node digital certificate, the whole certificate is signed, and if the obtained signature is consistent with the signature on the certificate, the certificate is trusted. Based on the above principle embodiment, when the federation chain node performs admission permission verification on the server initiating the risk assessment instruction, the federation chain node reads the authentication center of the node digital certificate of the server, and performs inquiry of the certificate trust chain in the authentication center, if the certificate trust chain is not the root certificate, the recursion is continued until the root certificate is obtained. Then, decrypting and verifying the legitimacy of the upper layer of certificate by using the public key of the root certificate, and then taking the public key of the upper layer of certificate to verify the legitimacy of the upper layer of certificate; and (5) recursively backtracking. And finally, verifying that the node digital certificate of the server is trusted, and finishing the checking of the admission permission.
And 306, if the node digital certificate passes the access authority verification, screening the blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to the object information.
In an embodiment, after the server node passes the checking of the admission authority in the federation chain, the historical behavior data stored in the federation chain block is screened based on the information abstract and the ciphertext abstract generated in the previous steps, and the obtained federation chain area storing the historical behavior data corresponding to the evaluation object is obtained. The specific screening process, as described in the previous embodiments, is not described here in detail.
In one embodiment, the process of integrating the historical behavior data in the plurality of alliance chain blocks to obtain the object behavior data corresponding to the object information in the method may include the following steps:
step one, ciphertext data in a plurality of alliance chain blocks corresponding to object information is obtained.
And step two, according to plaintext data obtained by decrypting the ciphertext data.
And thirdly, generating object behavior data corresponding to the object information based on the plaintext data obtained through decryption.
Specifically, in order to avoid leakage of object information caused by direct storage in a plaintext manner, an asymmetric encryption algorithm is introduced in the embodiment, and historical behavior data in the alliance chain block is stored in an encrypted manner.
In the embodiment, a public and private key pair capable of encrypting and decrypting data in the alliance chain node is formed through an asymmetric encryption algorithm, when the history behavior data is stored in a uplink mode, the alliance chain node needs to encrypt the history behavior data through a public key, and ciphertext data formed after encryption is packaged to form a block in the alliance chain and is accessed to the alliance chain. Correspondingly, in the extraction and integration stage of the object behavior data, after the alliance chain block storing the historical behavior data of the evaluation object is determined through matching screening, the encrypted historical behavior data, namely ciphertext data, in the alliance chain block is extracted. Then, based on a private key generated by asymmetric encryption, decrypting the ciphertext data to obtain decrypted historical behavior data, namely plaintext data; and integrating the plaintext data obtained after decrypting the plurality of ciphertext data to obtain object behavior data corresponding to the object information of the evaluation object. Based on an asymmetric algorithm, historical data sharing is performed through the alliance chain nodes, meanwhile, data can be prevented from being tampered maliciously, and the safety of data storage is improved.
In one embodiment, the method according to the process of decrypting the ciphertext data to obtain plaintext data may include the following steps:
step one, a private key corresponding to the risk assessment instruction is obtained.
And secondly, decrypting the ciphertext data through a private key corresponding to the risk assessment instruction to obtain storage address information of the historical behavior data corresponding to the object information.
And thirdly, addressing the nodes accessed to the alliance chain according to the storage address information to obtain a plurality of storage nodes.
And step four, a plurality of historical data fragments are called from the local storage of the storage node according to the object information.
And fifthly, combining the plurality of historical data fragments to form plaintext data.
In an embodiment, taking financial software as an example, an object may generate business behaviors in platforms of a plurality of different financial institutions, and historical behavior data corresponding to the business behaviors are all packed to form blocks in a coalition chain for storage. However, if the generated historical behavior data is processed through encryption and then is packaged in blocks, the data processing response is certainly slow, and serious computing power resource occupation is caused. In order to improve the efficiency of data processing, in the embodiment, specific historical behavior data content can be selectively stored to nodes of a alliance chain to be locally stored, and then the specific historical behavior data content is associated and bound with a storage address of the historical behavior data based on object information of a business behavior object; for example, in an embodiment, the character string of the object information may be directly spliced with the character string of the storage address to form the compound character string. And when the data is stored in a uplink mode, the compound character strings formed by combination are directly packaged to form a new alliance chain block to be added into the alliance chain. When the historical behavior data is extracted based on the object information, after the object information is screened to obtain the alliance chain blocks related to the object information, firstly decrypting the data content in the blocks according to the private key generated by the symmetric encryption, extracting the storage address information related to the object information in the decrypted data, addressing based on the storage address information, determining the alliance chain node in which the evaluation object is stored, then directly calling the historical behavior data from the storage space of the alliance chain node, and finally integrating the historical behavior data fragments extracted from the alliance chain nodes to finally obtain the data content containing all the historical behavior data of the evaluation object.
In one embodiment, the method inputs the object behavior data into the trained deep learning model to predict the risk level, and before obtaining the risk tolerance level corresponding to the evaluation object, the method further includes the following steps:
step one, obtaining a deep learning model to be trained and a training data set of historical behavior data, wherein the training data set comprises risk level labels and behavior data samples.
And secondly, inputting the behavior data sample into a deep learning model to be trained to predict the risk level, and obtaining a prediction level label.
And thirdly, if the similarity value between the prediction grade label and the risk grade label is not smaller than a preset similarity threshold value, obtaining a trained deep learning model.
Taking a supervised deep learning model as an example, the embodiment performs indifferently obtaining historical behavior data from a alliance chain block in advance, and assigns risk level labels to a plurality of historical behavior data by means of manual labeling or automatic labeling. And inputting the historical behavior data in the training sample data into a pre-built deep learning model, and predicting the predicted risk level of the historical behavior data in the training sample data through the deep learning model. Similarity comparison is carried out between the predicted risk level and a risk level mark which is pre-assigned with historical behavior data; and when the similarity value of the two is smaller than a preset similarity threshold value, optimizing and adjusting parameters of the deep learning model until the similarity value of the two is larger than or equal to the preset similarity threshold value, and outputting to obtain the trained deep learning model.
In one embodiment, the method for inputting the behavior data sample into the deep learning model to be trained to predict, obtaining the prediction grade label comprises the following steps:
step one, carrying out convolution pooling treatment on the behavior data samples through a convolution layer and a pooling layer preset in a deep learning model to be trained to obtain behavior characteristic data.
And secondly, calculating the behavior characteristic data through a convolution function preset in the deep learning model to be trained to obtain a prediction grade label.
Taking a supervised deep learning model as an example, carrying out convolution pooling operation on the behavior data samples according to preset convolution pooling times to obtain a feature set; then, the embodiment calculates the feature set by using a preset activation function to obtain a predicted risk level, and calculates by using a pre-constructed loss function according to the predicted risk level and a pre-labeled risk level label to obtain a loss value. Further comparing the loss value with a preset loss threshold value, and returning to the behavior data sample when the loss value is greater than or equal to the preset threshold value; and stopping training when the loss value is smaller than a preset threshold value, and obtaining the deep learning model after training is completed.
With reference to fig. 4 of the specification, taking a specific application scenario of risk assessment of a financial product as an example, the method for predicting risk tolerance level provided by the present application is described in more complete and detailed as follows:
before the risk tolerance level prediction is performed, the embodiment needs to construct a alliance chain between financial institutions, and can record the financial business behaviors of the target object and support reading. Among other things, financial business activities include, but are not limited to: credit card amount, income condition, consumption condition of large amount of commodity, past investment and financial condition, etc.
When the risk tolerance level is predicted, the desensitized historical financial business behaviors are required to be collected, and the data are subjected to characteristic extraction to serve as input data in a training set; collecting the risk bearing grades corresponding to the target objects at present as output data in a training set; an end-to-end deep learning model is trained using the data to predict the risk tolerance level of the new target object. Wherein, the risk classification includes: r1 discreet (or conservative), R2 robust, R3 balanced, R4 aggressive, R5 aggressive.
When a risk assessment condition is triggered, firstly, reading historical financial business behavior data of the target object stored on a alliance chain; then sending the data to a deep learning module for prediction; finally, the predicted risk level result is returned to the target object; and recommending different financial products according to the risk level result, wherein the recommendation rule is that a target object rated as Rn risk level cannot purchase financial products with the risk level higher than Cn risk level. For example, a target subject rated as an R2 risk level can only purchase financial products of C1, C2 risk levels, and cannot purchase other levels.
It should be understood that, although the steps in the flowcharts related to the above embodiments are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.
Based on the same inventive concept, the embodiment of the present application further provides a risk tolerance level prediction apparatus for implementing the above-mentioned risk tolerance level prediction method. The implementation of the solution provided by the apparatus is similar to the implementation described in the above method, so the specific limitation in the embodiments of the apparatus for predicting risk tolerance level or levels provided below may refer to the limitation of the method for predicting risk tolerance level hereinabove, and will not be repeated herein.
In one embodiment, as shown in fig. 5, there is provided a risk tolerance level prediction apparatus 500, including: an information acquisition module 501, a summary generation module 502, a block screening module 503, a data acquisition module 504, and a level prediction module 505, wherein:
the information acquisition module 501 acquires object information of an evaluation object in response to a risk evaluation instruction of the evaluation object.
The summary generation module 502 generates a message summary of the object information by a message summary algorithm.
The block filtering module 503 accesses the federation chain, and filters the block in the federation chain according to the information abstract and the ciphertext abstract in the block in the federation chain, so as to obtain a plurality of block in the federation chain corresponding to the object information, where the block in the federation chain stores the historical behavior data corresponding to the object information.
The data acquisition module 504 integrates the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information.
The level prediction module 505 inputs the object behavior data into a trained deep learning model to perform risk level prediction, so as to obtain a risk bearing level corresponding to the evaluated object, and the trained deep learning model is trained based on training sample data, wherein the training sample data comprises historical behavior data of different objects and corresponding risk levels.
In one embodiment, the block screening module 503 may further be capable of obtaining a node digital certificate carried by the risk assessment instruction; checking the access permission of the node digital certificate according to the certificate trust chain stored in the alliance chain node; screening the blocks of the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to the object information, wherein the steps comprise: if the node digital certificate passes the access authority verification, the blocks in the alliance chain are screened according to the information abstract and the ciphertext abstract in the alliance chain blocks, so that a plurality of alliance chain blocks corresponding to the object information are obtained.
In one embodiment, the data acquisition module 504 is further capable of acquiring ciphertext data in a plurality of federated chain blocks corresponding to the object information; according to the plaintext data obtained by decrypting the ciphertext data; and generating object behavior data corresponding to the object information based on the plain text data obtained by decryption.
In one embodiment, the data obtaining module 504 is further capable of obtaining a private key corresponding to the risk assessment instruction; decrypting the ciphertext data through a private key corresponding to the risk assessment instruction to obtain storage address information of historical behavior data corresponding to the object information; addressing nodes accessed to the alliance chain according to the storage address information to obtain a plurality of storage nodes; retrieving a plurality of historical data segments from a local store of the storage node according to the object information; a plurality of historical data segments are combined to form plaintext data.
In one embodiment, the risk tolerance level prediction apparatus 500 further includes a model training module, where the model training module is capable of obtaining a deep learning model to be trained and a training data set of historical behavior data, where the training data set includes a risk level tag and a behavior data sample; inputting the behavior data sample into a deep learning model to be trained to predict the risk level, and obtaining a prediction level label; and if the similarity value between the prediction grade label and the risk grade label is not smaller than a preset similarity threshold value, obtaining a trained deep learning model.
In one embodiment, the model training module can also perform convolution pooling processing on the behavior data sample through a convolution layer and a pooling layer preset in the deep learning model to be trained to obtain behavior characteristic data; and calculating the behavior characteristic data through a convolution function preset in the deep learning model to be trained to obtain a prediction grade label.
The respective modules in the above-described risk tolerance level prediction apparatus may be implemented in whole or in part by software, hardware, or a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 6. The computer device includes a processor, a memory, an Input/Output interface (I/O) and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The computer device can be used as a alliance chain link point to access to an alliance chain, and the obtained historical behavior data can be packaged to form blocks in the alliance chain for storage. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for communicating with an external terminal through a network connection. The computer program, when executed by a processor, implements a method of predicting a risk tolerance level.
It will be appreciated by those skilled in the art that the structure shown in FIG. 6 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the computer device to which the present inventive arrangements may be applied, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.
In an embodiment, there is also provided a computer device comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the method embodiments described above when the computer program is executed.
In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored which, when executed by a processor, carries out the steps of the method embodiments described above.
In an embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of the method embodiments described above.
It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and the data (including but not limited to data for analysis, stored data, presented data, etc.) related to the present application are information and data authorized by the user or sufficiently authorized by each party, and the collection, use and processing of the related data need to comply with the related laws and regulations and standards of the related country and region.
Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magnetic random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (Phase Change Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in the form of a variety of forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), and the like. The databases referred to in the embodiments provided herein may include at least one of a relational database and a non-relational database. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processor referred to in the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, or the like, but is not limited thereto.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The foregoing examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of the application should be assessed as that of the appended claims.
Claims (10)
1. A method of predicting a risk tolerance level, the method comprising:
responding to a risk evaluation instruction of an evaluation object, and acquiring object information of the evaluation object;
generating a message digest of the object information through a message digest algorithm;
accessing a alliance chain, and screening blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to object information, wherein historical behavior data corresponding to the object information is stored in the alliance chain blocks;
Integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information;
and inputting the object behavior data into a trained deep learning model for risk level prediction to obtain a risk bearing level corresponding to the evaluation object, wherein the trained deep learning model is trained based on training sample data, and the training sample data comprises historical behavior data of different objects and corresponding risk levels.
2. The method of claim 1, wherein before filtering the blocks of the federated chain to obtain a plurality of federated chain blocks corresponding to the object information according to the information digest and the ciphertext digest in the federated chain blocks, further comprising:
acquiring a node digital certificate carried by the risk assessment instruction;
checking the access permission of the node digital certificate according to a certificate trust chain stored in a alliance chain node;
screening the blocks of the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain blocks to obtain a plurality of alliance chain blocks corresponding to the object information, wherein the steps comprise:
and if the node digital certificate passes the access authority verification, screening the blocks in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain block to obtain a plurality of alliance chain blocks corresponding to the object information.
3. The method of claim 1, wherein integrating the historical behavior data in the plurality of federated chain blocks to obtain object behavior data corresponding to object information comprises:
acquiring ciphertext data in a plurality of alliance chain blocks corresponding to the object information;
according to plaintext data obtained by decrypting the ciphertext data;
and generating object behavior data corresponding to the object information based on the plain text data obtained by decryption.
4. The method of claim 3, wherein the plaintext data resulting from decrypting the ciphertext data comprises:
acquiring a private key corresponding to the risk assessment instruction;
decrypting the ciphertext data through a private key corresponding to the risk assessment instruction to obtain storage address information of historical behavior data corresponding to the object information;
addressing nodes accessed to the alliance chain according to the storage address information to obtain a plurality of storage nodes;
retrieving a plurality of historical data segments from a local store of the storage node according to the object information;
and combining the plurality of historical data fragments to form plaintext data.
5. The method according to any one of claims 1 to 4, wherein before inputting the object behavior data into a trained deep learning model to perform risk level prediction, obtaining a risk tolerance level corresponding to the evaluation object, the method comprises:
Acquiring a deep learning model to be trained and a training data set of historical behavior data, wherein the training data set comprises a risk level label and a behavior data sample;
inputting the behavior data sample into the deep learning model to be trained to predict risk level, and obtaining a prediction level label;
and if the similarity value between the prediction grade label and the risk grade label is not smaller than a preset similarity threshold value, obtaining a trained deep learning model.
6. The method of claim 5, wherein inputting the behavioral data samples into the deep learning model to be trained for prediction, obtaining a prediction rank label comprises:
carrying out convolution pooling treatment on the behavior data samples through a convolution layer and a pooling layer which are preset in a deep learning model to be trained to obtain behavior characteristic data;
and calculating the behavior characteristic data through a convolution function preset in the deep learning model to be trained to obtain a prediction grade label.
7. A risk tolerance level prediction apparatus, the apparatus comprising:
the information acquisition module is used for responding to a risk assessment instruction of an assessment object and acquiring object information of the assessment object;
The abstract generating module is used for generating an abstract of the object information through an abstract algorithm;
the block screening module accesses the alliance chain, screens the block in the alliance chain according to the information abstract and the ciphertext abstract in the alliance chain block to obtain a plurality of alliance chain blocks corresponding to the object information, and the alliance chain blocks store historical behavior data corresponding to the object information;
the data acquisition module is used for integrating the historical behavior data in the plurality of alliance chain blocks to obtain object behavior data corresponding to the object information;
the level prediction module is used for inputting the object behavior data into a trained deep learning model to perform risk level prediction to obtain a risk bearing level corresponding to the evaluation object, the trained deep learning model is trained based on training sample data, and the training sample data comprises historical behavior data of different objects and corresponding risk levels.
8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310682715.9A CN117132371A (en) | 2023-06-09 | 2023-06-09 | Method, apparatus, computer device and storage medium for predicting risk tolerance level |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310682715.9A CN117132371A (en) | 2023-06-09 | 2023-06-09 | Method, apparatus, computer device and storage medium for predicting risk tolerance level |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117132371A true CN117132371A (en) | 2023-11-28 |
Family
ID=88849848
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310682715.9A Pending CN117132371A (en) | 2023-06-09 | 2023-06-09 | Method, apparatus, computer device and storage medium for predicting risk tolerance level |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117132371A (en) |
-
2023
- 2023-06-09 CN CN202310682715.9A patent/CN117132371A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220230071A1 (en) | Method and device for constructing decision tree | |
| Liu et al. | A blockchain-based framework of cross-border e-commerce supply chain | |
| Tanwar et al. | Machine learning adoption in blockchain-based smart applications: The challenges, and a way forward | |
| Kumar et al. | Artificial intelligence and blockchain integration in business: trends from a bibliometric-content analysis | |
| Ekramifard et al. | A systematic literature review of integration of blockchain and artificial intelligence | |
| WO2018041066A1 (en) | Data processing method, apparatus and system based on block chain technology | |
| CN112132198B (en) | Data processing method, device and system and server | |
| Miglani et al. | Blockchain management and machine learning adaptation for IoT environment in 5G and beyond networks: A systematic review | |
| CN111666460A (en) | User portrait generation method and device based on privacy protection and storage medium | |
| CN112465627B (en) | Financial loan auditing method and system based on block chain and machine learning | |
| US9235814B2 (en) | Machine learning memory management and distributed rule evaluation | |
| CN110046156A (en) | Content Management System and method, apparatus, electronic equipment based on block chain | |
| Kumar et al. | Development of a cloud-assisted classification technique for the preservation of secure data storage in smart cities | |
| Klaine et al. | A privacy-preserving blockchain platform for a data marketplace | |
| Kummar et al. | Blockchain based big data solutions for internet of things (IoT) and smart cities | |
| Silva et al. | Towards federated learning: An overview of methods and applications | |
| Sheeba et al. | Digital Hash Data Encryption for IoT Financial Transactions using Blockchain Security in the Cloud | |
| Ukanah et al. | Blockchain application in healthcare | |
| CN117132371A (en) | Method, apparatus, computer device and storage medium for predicting risk tolerance level | |
| Vlahavas et al. | Unsupervised clustering of bitcoin transactions | |
| Nguyen et al. | The benefits and challenges of applying Blockchain technology into Big Data: A literature review | |
| US20230418921A1 (en) | Intelligent authentication of users in metaverse leveraging non-fungible tokens and behavior analysis | |
| Swanthana et al. | An intelligent homomorphic blockchain approach for securing stock market data | |
| CN115314513B (en) | Trust twinning method based on block chain and related equipment | |
| Forestal et al. | Text mining-based patent analysis of Blockchain technology applications |
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 |