CN111639706A - Personal risk portrait generation method based on image set and related equipment - Google Patents
Personal risk portrait generation method based on image set and related equipment Download PDFInfo
- Publication number
- CN111639706A CN111639706A CN202010474298.5A CN202010474298A CN111639706A CN 111639706 A CN111639706 A CN 111639706A CN 202010474298 A CN202010474298 A CN 202010474298A CN 111639706 A CN111639706 A CN 111639706A
- Authority
- CN
- China
- Prior art keywords
- risk
- image
- classification model
- labels
- test
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/55—Clustering; Classification
-
- 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
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Business, Economics & Management (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Databases & Information Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of big data, and provides a personal risk portrait generation method and device based on an image set, an electronic device and a storage medium, wherein the method comprises the following steps: acquiring an image data set of an object to be evaluated; inputting each image in the image data set into a rough classification model for identification to obtain a plurality of primary labels; inputting images corresponding to a plurality of target primary labels in the plurality of primary labels into a plurality of fine classification models for identification to obtain a plurality of secondary labels; calculating a risk factor according to the plurality of secondary labels; calculating a risk score from the plurality of secondary labels; and generating a risk portrait report of the object to be evaluated based on the risk factors and the risk scores. Furthermore, the invention may also relate to the field of blockchain technology, the coarse and fine classification models may be stored in blockchains. The risk of the object to be evaluated can be objectively reflected according to the image data set of the object to be evaluated, and the generated risk portrait is higher in accuracy and more comprehensive.
Description
Technical Field
The invention relates to the technical field of big data, in particular to a personal risk portrait generation method and device based on an image set, electronic equipment and a storage medium.
Background
In the field of credit wind control, a large amount of valuable data is needed to evaluate the generation of personal risk profiles for credit applicants. Currently, people credit investigation reports, third party blacklists, multi-head loans, part of internet consumption data and the like are fully mined and utilized. However, the quality and saturation of these data still make it difficult to accurately and comprehensively describe the personal risk representation.
With the popularization of smart phones and 4G technologies, image information is more and more common, for example, a large number of images may be stored in mobile phones, friend circles and cloud disks of people, and the images have a large amount of valuable information.
Therefore, it is important to mine newly available supplemental data based on images to generate a personal risk profile.
Disclosure of Invention
In view of the above, there is a need for a method, an apparatus, an electronic device and a storage medium for generating a personal risk profile based on an image set, which can objectively reflect the risk of an object to be evaluated according to an image data set of the object to be evaluated, and generate a risk profile with higher accuracy and more comprehensive performance.
The invention provides a personal risk portrait generation method based on image sets, which comprises the following steps:
acquiring an image data set of an object to be evaluated;
inputting each image in the image data set into a rough classification model for identification to obtain a plurality of primary labels;
inputting images corresponding to a plurality of target primary labels in the plurality of primary labels into a plurality of fine classification models for identification to obtain a plurality of secondary labels;
calculating a risk factor according to the plurality of secondary labels;
calculating a risk score from the plurality of secondary labels;
and generating a risk portrait report of the object to be evaluated based on the risk factor and the risk score.
According to an alternative embodiment of the present invention, the training process of the coarse classification model includes:
determining a plurality of target classes;
crawling a plurality of sample images for each target category;
constructing a plurality of training sample pairs, wherein each training sample pair comprises a sample image and a corresponding target class identifier;
constructing a plurality of training sample sets and a test sample set according to the plurality of training sample pairs;
loading each training sample set to an inclusion V3 network for migration learning to obtain a plurality of candidate rough classification models, and loading the test sample set to each candidate rough classification model for testing to obtain the test passing rate of each candidate rough classification model;
determining a candidate coarse classification model corresponding to the highest test passing rate in the test passing rates as an optimal coarse classification model;
wherein the coarse classification model is stored on a blockchain.
According to an optional embodiment of the present invention, the loading the test sample set into the candidate rough classification model for testing to obtain a test passing rate includes:
loading each sample image in the test sample set to the candidate rough classification model for testing to obtain a test class identifier;
judging whether the test type identification of the sample image in the test sample set is the same as the corresponding target type identification;
when the test category identification of the sample image in the test sample set is the same as the corresponding target category identification, determining that the sample image passes the test;
calculating the proportion of the number of the sample images passing the test in the test sample set;
and determining the proportion as the test passing rate of the candidate rough classification model.
According to an optional embodiment of the present invention, the inputting the images corresponding to the target primary labels in the plurality of primary labels into a plurality of fine classification models to obtain a plurality of secondary labels by recognition includes:
inputting an image corresponding to the first target first-level label into a first fine classification model for identification to obtain a first second-level label;
inputting the image corresponding to the second target primary label into a second fine classification model for identification to obtain a second secondary label;
inputting the image corresponding to the third target primary label into a third fine classification model for identification to obtain a third secondary label;
wherein the plurality of fine classification models are stored on a blockchain.
According to an alternative embodiment of the present invention, said calculating a risk factor from said plurality of secondary labels comprises:
calculating a first number of the first secondary labels, calculating a second number of the second secondary labels, and calculating a third number of the third secondary labels;
determining target secondary labels according to the first quantity, the second quantity and the third quantity;
determining the occupation of the object to be evaluated according to the target secondary label;
adding the occupation as a risk factor to a set of risk factors.
According to an alternative embodiment of the invention, said calculating a risk score from said plurality of secondary labels comprises:
calculating a first ratio of the first secondary label, a second ratio of the second secondary label and a third ratio of the third secondary label;
determining a first score corresponding to the first ratio, determining a second score corresponding to the second ratio, and determining a third score corresponding to the third ratio;
and calculating the sum of the first score, the second score and the third score to obtain a risk score.
According to an optional embodiment of the present invention, after acquiring the image data set of the object to be evaluated, the method for generating a personal risk representation based on the image set further comprises:
the image data set is normalized.
A second aspect of the present invention provides an image-set-based personal risk representation generation apparatus, the apparatus comprising:
the image acquisition module is used for acquiring an image data set of an object to be evaluated;
the first input module is used for inputting each image in the image data set into the rough classification model for identification to obtain a plurality of primary labels;
the second input module is used for inputting images corresponding to a plurality of target primary labels in the plurality of primary labels into a plurality of fine classification models for recognition to obtain a plurality of secondary labels;
the first calculation module is used for calculating risk factors according to the secondary labels;
a second calculation module for calculating a risk score according to the plurality of secondary labels;
and the report generation module is used for generating a risk portrait report of the object to be evaluated based on the risk factors and the risk scores.
A third aspect of the invention provides an electronic device comprising a processor for implementing the image set based personal risk representation generation method when executing a computer program stored in a memory.
A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the image-set-based personal risk representation generation method.
In summary, according to the personal risk profile generation method, the personal risk profile generation device, the electronic device and the storage medium based on the image set, the image data set is applied to the wind control field initiatively, valuable information capable of reflecting risks is extracted from the image data set of the object to be evaluated in a targeted manner, the risk profile is generated through expert rules and a scoring model, the risks of the object to be evaluated can be reflected objectively, and the generated risk profile is high in accuracy and comprehensive.
Drawings
FIG. 1 is a flowchart of a method for generating a personal risk representation based on an image set according to an embodiment of the present invention.
Fig. 2 is a block diagram of a personal risk representation generation apparatus based on image sets according to a second embodiment of the present invention.
Fig. 3 is a schematic structural diagram of an electronic device according to a third embodiment of the present invention.
The following detailed description will further illustrate the invention in conjunction with the above-described figures.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, a detailed description of the present invention will be given below with reference to the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and features of the embodiments may be combined with each other without conflict.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
FIG. 1 is a flowchart of a method for generating a personal risk representation based on an image set according to an embodiment of the present invention. The personal risk representation generation method based on the image set specifically comprises the following steps, and the sequence of the steps in the flow chart can be changed and some steps can be omitted according to different requirements.
S11, an image data set of the object to be evaluated is acquired.
A large number of images may be stored in the mobile phone, the friend circle, and the cloud disk of the object to be evaluated. After the object to be evaluated is determined, images in a mobile phone, a friend circle, a microblog, a cloud disk and the like of the object to be evaluated can be obtained to form an image data set.
In an optional embodiment, after acquiring the image data set of the object to be evaluated, the method for generating a personal risk representation based on the image set further includes:
the image data set is normalized.
The normalization process may include a combination of one or more of the following: size normalization processing, image format normalization processing and gray level normalization processing.
In this alternative embodiment, the size normalization process and the image format normalization process are performed on each image in the image dataset to facilitate the batch import of images into the rough classification model. The gray level normalization processing is performed on each image in the image data set so as to convert all the images into gray level images, and the identification efficiency of the coarse classification model and the fine classification model on the images is improved conveniently.
S12, inputting each image in the image data set to a rough classification model for identification to obtain a plurality of primary labels.
The image data set of the object to be evaluated may contain images of multiple categories, such as a landscape category, a people category, a food category, a certificate category, a screenshot category, a traffic category, a pet category, and the like. Because of the multiple image categories, a classification model is required to be used for preliminary classification to determine which large category each image belongs to, and then the classification model is used for further subdivision to determine which small category the image specifically belongs to. This ensures a higher accuracy of classification.
The rough classification model may be trained offline in advance, used online to identify the category of each image, and output a primary label. The primary label is used for representing the category identification of the large category to which the image belongs.
In an alternative embodiment, the training process of the coarse classification model includes:
(1) determining a plurality of target classes;
(2) crawling a plurality of sample images for each target category;
(3) constructing a plurality of training sample pairs, wherein each training sample pair comprises a sample image and a corresponding target class identifier;
(4) constructing a plurality of training sample sets and a test sample set according to the plurality of training sample pairs;
(5) loading each training sample set to an inclusion V3 network for migration learning to obtain a plurality of candidate rough classification models, and loading the test sample set to each candidate rough classification model for testing to obtain the test passing rate of each candidate rough classification model;
(6) and determining the candidate rough classification model corresponding to the highest test passing rate in the test passing rates as the optimal rough classification model.
The coarse classification model may employ Tensorflow as a training framework.
Illustratively, assume that the determined plurality of categories includes: the method comprises the steps of setting a target category identification corresponding to a landscape class to be 1, setting a target category identification corresponding to a figure class to be 2, setting a target category identification corresponding to a gourmet class to be 3, setting a target category identification corresponding to a certificate class to be 4, setting a target category identification corresponding to a screenshot class to be 5, setting a target category identification corresponding to a traffic class to be 6 and setting a target category identification corresponding to a pet class to be 7, and acquiring a large number of sample images from the world wide web or a special image data set aiming at each category.
A sample image A in the landscape class and an object class identifier "1" are constructed into a sample pair (sample image A, 1). Firstly, acquiring a first number of sample pairs from the constructed sample pairs as a first sample set and acquiring a second number of sample pairs as a second sample set; obtaining a third number of sample pairs from the first sample set as a training sample set randomly and repeatedly, so as to obtain a plurality of training sample sets; and finally, randomly acquiring a third number of sample pairs from the second sample set to serve as a test sample set.
In the optional embodiment, the training sample set is obtained randomly and in a replaced manner, so that the generalization capability of the rough classification model obtained based on the training of the training sample set is strong. And simultaneously training a plurality of coarse classification models, and determining the coarse classification model with the highest test passing rate as the optimal coarse classification model, so that the classification accuracy can be ensured.
The training process for transfer learning is prior art and the present invention is not described in detail herein.
In an optional embodiment, the loading the test sample set into the candidate rough classification model for testing to obtain a test passing rate includes:
loading each sample image in the test sample set to the candidate rough classification model for testing to obtain a test class identifier;
judging whether the test type identification of the sample image in the test sample set is the same as the corresponding target type identification;
when the test category identification of the sample image in the test sample set is the same as the corresponding target category identification, determining that the sample image passes the test;
calculating the proportion of the number of the sample images passing the test in the test sample set;
and determining the proportion as the test passing rate of the candidate rough classification model.
For example, assuming that a sample image F in the test sample set has a corresponding target class identifier of "1", and a test class identifier obtained through the candidate rough classification model 1 is "1", it indicates that the candidate rough classification model 1 predicts correctly and the test passes. Assuming that the corresponding target class identifier of the sample image M in the test sample set is "2", and the test class identifier obtained through the candidate rough classification model 1 is "1", it indicates that the candidate rough classification model 1 has a wrong prediction, and the test does not pass. And calculating the number of the test samples passing the test of the candidate rough classification model 1 in the test sample set, so as to calculate the test passing rate of the candidate rough classification model 1.
And S13, inputting the images corresponding to the target primary labels in the primary labels into a plurality of fine classification models for recognition to obtain a plurality of secondary labels.
And if the personal risk portrait of the object to be evaluated is generated, determining the type of the image in the personal image data set of the object to be evaluated, and estimating the risk of the object to be evaluated according to the type of the image.
The target primary label is a category identification which is defined in advance and used for representing risk correlation of an object to be evaluated, such as a character class, a screenshot class and a certificate class. The images in the character class, the screenshot class and the certificate class can depict information such as occupation, family structure, assets and the like of the object to be evaluated to a certain extent.
And determining a primary label of each image in the image data set of the object to be evaluated through the rough classification model, and when the primary label of the image is a target primary label, marking the target primary label on the image and inputting the target primary label into a plurality of fine classification models to further obtain a plurality of secondary labels.
In an optional embodiment, the inputting the images corresponding to the target primary labels in the primary labels into a plurality of fine classification models for recognition to obtain a plurality of secondary labels includes:
inputting an image corresponding to the first target first-level label into a first fine classification model for identification to obtain a first second-level label;
inputting the image corresponding to the second target primary label into a second fine classification model for identification to obtain a second secondary label;
and inputting the image corresponding to the third target primary label into a third fine classification model for identification to obtain a third secondary label.
In this optional embodiment, assuming that the first target primary label is a person class, the second target primary label is a screenshot class, and the third target primary label is a certificate class, when it is determined by the rough classification model that the category identifier of a certain image is the first target primary label, the image is recognized in the fine classification model of the person class to obtain the first and second primary labels of the image, for example, a baby, a child, an adult, an old person, and a group photo.
And the second subdivision classification model is used for further predicting the screenshot images and outputting subdivision screenshot categories such as games, chats, shopping and net credits.
And the third subdivision classification model is used for further predicting the certificate class images and outputting subdivision certificate classes such as identity cards, academic degree cards, driving licenses, marriage certificates, house property cards, passports and the like.
The training process of the first, second and third fine classification models is the same as the training process of the coarse classification model, and the present invention is not repeated herein.
And S14, calculating a risk factor according to the plurality of secondary labels.
And an expert rule base is created in advance according to expert experience, and the corresponding relation between the label and the occupation is recorded in the expert rule base.
In an optional embodiment, said calculating a risk factor from said plurality of secondary labels comprises:
calculating a first number of the first secondary labels, calculating a second number of the second secondary labels, and calculating a third number of the third secondary labels;
determining target secondary labels according to the first quantity, the second quantity and the third quantity;
determining the occupation of the object to be evaluated according to the target secondary label;
adding the occupation as a risk factor to a set of risk factors.
Illustratively, if idcard _ varlist [0] is initialized to the number of identification cards, zj _ bank _ num [0] is the number of bank cards, and if idcard _ varlist > -10 or zj _ bank _ num > -10 or idcard _ varlist + zj _ bank _ num >15, then [ suspected credit broker ] is added to the risk factor set.
And S15, calculating a risk score according to the plurality of secondary labels.
And giving different scores to each secondary label according to different label values, and finally accumulating the scores of the plurality of secondary labels to obtain the risk score.
In an alternative embodiment, said calculating a risk score from said plurality of secondary labels comprises:
calculating a first ratio of the first secondary label, a second ratio of the second secondary label and a third ratio of the third secondary label;
determining a first score corresponding to the first ratio, determining a second score corresponding to the second ratio, and determining a third score corresponding to the third ratio;
and calculating the sum of the first score, the second score and the third score to obtain a risk score.
For example, taking the [ identification card ] secondary label as an example, when the proportion is in the range of [ 0-1 ], the corresponding score is + 5; when the proportion is in the range of (2-5), the corresponding fraction is-5; when the proportion is in the range of 6-10, the corresponding fraction is-10; when the proportion is in the range of [ 10+ ], the corresponding fraction is-20.
The calculated risk score is an integer between [0, 100], and the higher the score is, the lower the individual risk level is represented; the lower the score, the higher the individual risk level.
And S16, generating a risk portrait report of the object to be evaluated based on the risk factor and the risk score.
In this embodiment, a plurality of risk score ranges and a risk level corresponding to each risk score range are set. As shown in the following table:
| serial number | Risk score Range | Risk rating |
| 1 | 0-20 | Height of |
| 2 | 21-60 | In |
| 3 | 61-100 | Is low in |
After the risk score of the object to be evaluated is determined, the risk score range where the risk score is located can be determined, and therefore the risk level is determined.
And generating a risk portrait of the object to be evaluated based on the risk factors and the risk levels by adopting a preset template, wherein information such as the risk levels, the risk factors and the like is displayed in a generated risk portrait report.
In the credit wind control field, the prior art mainly depends on structured data such as a third party blacklist, multi-head credit data, user identity information and the like to carry out risk assessment, and the traditional structured data is very limited and is difficult to accurately and comprehensively describe personal risk portraits. If a supplier a purchases a lot of loan data and then a supplier B purchases a lot of loan data, the data for a and B will be similar, although they will be different, and generally will only bring a little marginal improvement.
The images are universally existed in the environments of mobile phones, computers, friend circles and the like, almost all people have the images, the saturation is high, the images are easy to obtain, and the image data are not overlapped with the structural data, so that the risk degree of the risk portrait can be improved. The image data set is initiatively applied to the wind control field, valuable information capable of reflecting risks is extracted from the image data set of the object to be evaluated in a targeted mode, the risk portrait is generated through expert rules and a scoring model, the risks of the object to be evaluated can be objectively reflected, and the generated risk portrait is higher in accuracy and more comprehensive; the image data sets of different evaluation objects have high differentiation degree, so the marginal improvement effect is obvious.
It is emphasized that, to further ensure the privacy and security of the coarse classification model and/or the fine classification models, the coarse classification model and/or the fine classification models may also be stored in nodes of a block chain.
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
Fig. 2 is a block diagram of a personal risk representation generation apparatus based on image sets according to a second embodiment of the present invention.
In some embodiments, the image set-based personal risk representation generating device 20 may include a plurality of functional modules comprising program code segments. Program code for various program segments of the image collection based personal risk representation generation apparatus 20 may be stored in a memory of the electronic device and executed by the at least one processor to perform (see detailed description of FIG. 1) the functions of image collection based personal risk representation generation.
In this embodiment, the image set-based personal risk representation generating device 20 may be divided into a plurality of functional modules according to the functions to be performed. The functional module may include: an image acquisition module 201, an image processing module 202, a first input module 203, a model training module 204, a second input module 205, a first calculation module 206, a second calculation module 207, and a report generation module 208. The module referred to herein is a series of computer program segments capable of being executed by at least one processor and capable of performing a fixed function and is stored in memory. In the present embodiment, the functions of the modules will be described in detail in the following embodiments.
The image obtaining module 201 is configured to obtain an image data set of an object to be evaluated.
A large number of images may be stored in the mobile phone, the friend circle, and the cloud disk of the object to be evaluated. After the object to be evaluated is determined, images in a mobile phone, a friend circle, a microblog, a cloud disk and the like of the object to be evaluated can be obtained to form an image data set.
The image processing module 202 is configured to perform normalization processing on the image data set.
The normalization process may include a combination of one or more of the following: size normalization processing, image format normalization processing and gray level normalization processing.
In this alternative embodiment, the size normalization process and the image format normalization process are performed on each image in the image dataset to facilitate the batch import of images into the rough classification model. The gray level normalization processing is performed on each image in the image data set so as to convert all the images into gray level images, and the identification efficiency of the coarse classification model and the fine classification model on the images is improved conveniently.
The first input module 203 is configured to input each image in the image data set into a rough classification model for identification to obtain a plurality of primary labels.
The image data set of the object to be evaluated may contain images of multiple categories, such as a landscape category, a people category, a food category, a certificate category, a screenshot category, a traffic category, a pet category, and the like. Because of the multiple image categories, a classification model is required to be used for preliminary classification to determine which large category each image belongs to, and then the classification model is used for further subdivision to determine which small category the image specifically belongs to. This ensures a higher accuracy of classification.
The rough classification model may be trained offline in advance, used online to identify the category of each image, and output a primary label. The primary label is used for representing the category identification of the large category to which the image belongs.
The model training module 204 is used for training a coarse classification model and a fine classification module. The training process of the coarse classification model is the same as that of the fine classification model, and the invention is only explained by taking the coarse classification model as an example.
In an alternative embodiment, the training process of the coarse classification model includes:
(1) determining a plurality of target classes;
(2) crawling a plurality of sample images for each target category;
(3) constructing a plurality of training sample pairs, wherein each training sample pair comprises a sample image and a corresponding target class identifier;
(4) constructing a plurality of training sample sets and a test sample set according to the plurality of training sample pairs;
(5) loading each training sample set to an inclusion V3 network for migration learning to obtain a plurality of candidate rough classification models, and loading the test sample set to each candidate rough classification model for testing to obtain the test passing rate of each candidate rough classification model;
(6) and determining the candidate rough classification model corresponding to the highest test passing rate in the test passing rates as the optimal rough classification model.
The coarse classification model may employ Tensorflow as a training framework.
Illustratively, assume that the determined plurality of categories includes: the method comprises the steps of setting a target category identification corresponding to a landscape class to be 1, setting a target category identification corresponding to a figure class to be 2, setting a target category identification corresponding to a gourmet class to be 3, setting a target category identification corresponding to a certificate class to be 4, setting a target category identification corresponding to a screenshot class to be 5, setting a target category identification corresponding to a traffic class to be 6 and setting a target category identification corresponding to a pet class to be 7, and acquiring a large number of sample images from the world wide web or a special image data set aiming at each category.
A sample image A in the landscape class and an object class identifier "1" are constructed into a sample pair (sample image A, 1). Firstly, acquiring a first number of sample pairs from the constructed sample pairs as a first sample set and acquiring a second number of sample pairs as a second sample set; obtaining a third number of sample pairs from the first sample set as a training sample set randomly and repeatedly, so as to obtain a plurality of training sample sets; and finally, randomly acquiring a third number of sample pairs from the second sample set to serve as a test sample set.
In the optional embodiment, the training sample set is obtained randomly and in a replaced manner, so that the generalization capability of the rough classification model obtained based on the training of the training sample set is strong. And simultaneously training a plurality of coarse classification models, and determining the coarse classification model with the highest test passing rate as the optimal coarse classification model, so that the classification accuracy can be ensured.
The training process for transfer learning is prior art and the present invention is not described in detail herein.
In an optional embodiment, the loading the test sample set into the candidate rough classification model for testing to obtain a test passing rate includes:
loading each sample image in the test sample set to the candidate rough classification model for testing to obtain a test class identifier;
judging whether the test type identification of the sample image in the test sample set is the same as the corresponding target type identification;
when the test category identification of the sample image in the test sample set is the same as the corresponding target category identification, determining that the sample image passes the test;
calculating the proportion of the number of the sample images passing the test in the test sample set;
and determining the proportion as the test passing rate of the candidate rough classification model.
For example, assuming that a sample image F in the test sample set has a corresponding target class identifier of "1", and a test class identifier obtained through the candidate rough classification model 1 is "1", it indicates that the candidate rough classification model 1 predicts correctly and the test passes. Assuming that the corresponding target class identifier of the sample image M in the test sample set is "2", and the test class identifier obtained through the candidate rough classification model 1 is "1", it indicates that the candidate rough classification model 1 has a wrong prediction, and the test does not pass. And calculating the number of the test samples passing the test of the candidate rough classification model 1 in the test sample set, so as to calculate the test passing rate of the candidate rough classification model 1.
The second input module 205 is configured to input images corresponding to a plurality of target primary labels in the plurality of primary labels into a plurality of fine classification models to identify the images to obtain a plurality of secondary labels.
And if the personal risk portrait of the object to be evaluated is generated, determining the type of the image in the personal image data set of the object to be evaluated, and estimating the risk of the object to be evaluated according to the type of the image.
The target primary label is a category identification which is defined in advance and used for representing risk correlation of an object to be evaluated, such as a character class, a screenshot class and a certificate class. The images in the character class, the screenshot class and the certificate class can depict information such as occupation, family structure, assets and the like of the object to be evaluated to a certain extent.
And determining a primary label of each image in the image data set of the object to be evaluated through the rough classification model, and when the primary label of the image is a target primary label, marking the target primary label on the image and inputting the target primary label into a plurality of fine classification models to further obtain a plurality of secondary labels.
In an optional embodiment, the inputting, by the second input module 205, the images corresponding to the target primary labels in the plurality of primary labels into the plurality of fine classification models to obtain a plurality of secondary labels includes:
inputting an image corresponding to the first target first-level label into a first fine classification model for identification to obtain a first second-level label;
inputting the image corresponding to the second target primary label into a second fine classification model for identification to obtain a second secondary label;
and inputting the image corresponding to the third target primary label into a third fine classification model for identification to obtain a third secondary label.
In this optional embodiment, assuming that the first target primary label is a person class, the second target primary label is a screenshot class, and the third target primary label is a certificate class, when it is determined by the rough classification model that the category identifier of a certain image is the first target primary label, the image is recognized in the fine classification model of the person class to obtain the first and second primary labels of the image, for example, a baby, a child, an adult, an old person, and a group photo.
And the second subdivision classification model is used for further predicting the screenshot images and outputting subdivision screenshot categories such as games, chats, shopping and net credits.
And the third subdivision classification model is used for further predicting the certificate class images and outputting subdivision certificate classes such as identity cards, academic degree cards, driving licenses, marriage certificates, house property cards, passports and the like.
The training process of the first, second and third fine classification models is the same as the training process of the coarse classification model, and the present invention is not repeated herein.
The first calculating module 206 is configured to calculate a risk factor according to the plurality of secondary labels.
And an expert rule base is created in advance according to expert experience, and the corresponding relation between the label and the occupation is recorded in the expert rule base.
In an alternative embodiment, the first calculation module 206 calculating the risk factor according to the plurality of secondary labels comprises:
calculating a first number of the first secondary labels, calculating a second number of the second secondary labels, and calculating a third number of the third secondary labels;
determining target secondary labels according to the first quantity, the second quantity and the third quantity;
determining the occupation of the object to be evaluated according to the target secondary label;
adding the occupation as a risk factor to a set of risk factors.
Illustratively, if idcard _ varlist [0] is initialized to the number of identification cards, zj _ bank _ num [0] is the number of bank cards, and if idcard _ varlist > -10 or zj _ bank _ num > -10 or idcard _ varlist + zj _ bank _ num >15, then [ suspected credit broker ] is added to the risk factor set.
The second calculating module 207 is configured to calculate a risk score according to the plurality of secondary labels.
And giving different scores to each secondary label according to different label values, and finally accumulating the scores of the plurality of secondary labels to obtain the risk score.
In an alternative embodiment, the second calculation module 207 calculating the risk score according to the plurality of secondary labels comprises:
calculating a first ratio of the first secondary label, a second ratio of the second secondary label and a third ratio of the third secondary label;
determining a first score corresponding to the first ratio, determining a second score corresponding to the second ratio, and determining a third score corresponding to the third ratio;
and calculating the sum of the first score, the second score and the third score to obtain a risk score.
For example, taking the [ identification card ] secondary label as an example, when the proportion is in the range of [ 0-1 ], the corresponding score is + 5; when the proportion is in the range of (2-5), the corresponding fraction is-5; when the proportion is in the range of 6-10, the corresponding fraction is-10; when the proportion is in the range of [ 10+ ], the corresponding fraction is-20.
The calculated risk score is an integer between [0, 100], and the higher the score is, the lower the individual risk level is represented; the lower the score, the higher the individual risk level.
The report generation module 208 is configured to generate a risk profile report of the subject to be evaluated based on the risk factor and the risk score.
In this embodiment, a plurality of risk score ranges and a risk level corresponding to each risk score range are set. As shown in the following table:
| serial number | Risk score Range | Risk rating |
| 1 | 0-20 | Height of |
| 2 | 21-60 | In |
| 3 | 61-100 | Is low in |
After the risk score of the object to be evaluated is determined, the risk score range where the risk score is located can be determined, and therefore the risk level is determined.
And generating a risk portrait of the object to be evaluated based on the risk factors and the risk levels by adopting a preset template, wherein information such as the risk levels, the risk factors and the like is displayed in a generated risk portrait report.
In the credit wind control field, the prior art mainly depends on structured data such as a third party blacklist, multi-head credit data, user identity information and the like to carry out risk assessment, and the traditional structured data is very limited and is difficult to accurately and comprehensively describe personal risk portraits. If a supplier a purchases a lot of loan data and then a supplier B purchases a lot of loan data, the data for a and B will be similar, although they will be different, and generally will only bring a little marginal improvement.
The images are universally existed in the environments of mobile phones, computers, friend circles and the like, almost all people have the images, the saturation is high, the images are easy to obtain, and the image data are not overlapped with the structural data, so that the risk degree of the risk portrait can be improved. The image data set is initiatively applied to the wind control field, valuable information capable of reflecting risks is extracted from the image data set of the object to be evaluated in a targeted mode, the risk portrait is generated through expert rules and a scoring model, the risks of the object to be evaluated can be objectively reflected, and the generated risk portrait is higher in accuracy and more comprehensive; the image data sets of different evaluation objects have high differentiation degree, so the marginal improvement effect is obvious.
It is emphasized that, to further ensure the privacy and security of the coarse classification model and/or the fine classification models, the coarse classification model and/or the fine classification models may also be stored in nodes of a block chain.
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
Fig. 3 is a schematic structural diagram of an electronic device according to a third embodiment of the present invention. In the preferred embodiment of the present invention, the electronic device 3 comprises a memory 31, at least one processor 32, at least one communication bus 33 and a transceiver 34.
It will be appreciated by those skilled in the art that the configuration of the electronic device shown in fig. 3 does not constitute a limitation of the embodiment of the present invention, and may be a bus-type configuration or a star-type configuration, and the electronic device 3 may include more or less other hardware or software than those shown, or a different arrangement of components.
In some embodiments, the electronic device 3 includes an electronic device capable of automatically performing numerical calculation and/or information processing according to preset or stored instructions, and the hardware includes, but is not limited to, a microprocessor, an application specific integrated circuit, a programmable gate array, a digital processor, an embedded device, and the like. The electronic device 3 may also include a client device, which includes, but is not limited to, any electronic product that can interact with a client through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device, for example, a personal computer, a tablet computer, a smart phone, a digital camera, and the like.
It should be noted that the electronic device 3 is only an example, and other existing or future electronic products, such as those that can be adapted to the present invention, should also be included in the scope of the present invention, and are included herein by reference.
In some embodiments, the memory 31 is used for storing program codes and various data, such as devices installed in the electronic device 3, and realizes high-speed and automatic access to programs or data during the operation of the electronic device 3. The Memory 31 includes a Read-Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), a One-time Programmable Read-Only Memory (OTPROM), an electronically Erasable rewritable Read-Only Memory (Electrically-Erasable Programmable Read-Only Memory (EEPROM)), an optical Read-Only Memory (CD-ROM) or other optical disk Memory, a magnetic disk Memory, a tape Memory, or any other medium readable by a computer that can be used to carry or store data.
In some embodiments, the at least one processor 32 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or may be composed of a plurality of integrated circuits packaged with the same or different functions, including one or more Central Processing Units (CPUs), microprocessors, digital Processing chips, graphics processors, and combinations of various control chips. The at least one processor 32 is a Control Unit (Control Unit) of the electronic device 3, connects various components of the electronic device 3 by using various interfaces and lines, and executes various functions and processes data of the electronic device 3 by running or executing programs or modules stored in the memory 31 and calling data stored in the memory 31.
In some embodiments, the at least one communication bus 33 is arranged to enable connection communication between the memory 31 and the at least one processor 32 or the like.
Although not shown, the electronic device 3 may further include a power supply (such as a battery) for supplying power to each component, and preferably, the power supply may be logically connected to the at least one processor 32 through a power management device, so as to implement functions of managing charging, discharging, and power consumption through the power management device. The power supply may also include any component of one or more dc or ac power sources, recharging devices, power failure detection circuitry, power converters or inverters, power status indicators, and the like. The electronic device 3 may further include various sensors, a bluetooth module, a Wi-Fi module, and the like, which are not described herein again.
It is to be understood that the described embodiments are for purposes of illustration only and that the scope of the appended claims is not limited to such structures.
The integrated unit implemented in the form of a software functional module may be stored in a computer-readable storage medium. The software functional module is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, an electronic device, or a network device) or a processor (processor) to execute the parts of the image-set-based personal risk representation generation method according to the embodiments of the present invention.
In a further embodiment, in conjunction with fig. 2, the at least one processor 32 may execute operating devices of the electronic device 3 as well as installed various types of applications, program codes, and the like, such as the various modules described above.
The memory 31 has program code stored therein, and the at least one processor 32 can call the program code stored in the memory 31 to perform related functions. For example, the respective modules illustrated in fig. 2 are program codes stored in the memory 31 and executed by the at least one processor 32, thereby implementing the functions of the respective modules.
In one embodiment of the present invention, the memory 31 stores a plurality of instructions that are executed by the at least one processor 32 to implement all or a portion of the steps of the image set based personal risk representation generation method of the present invention.
Specifically, the at least one processor 32 may refer to the description of the relevant steps in the embodiment corresponding to fig. 1, and details are not repeated here.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and other divisions may be realized in practice.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
In addition, functional modules in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional module.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned. Furthermore, it is obvious that the word "comprising" does not exclude other elements or that the singular does not exclude the plural. A plurality of units or means recited in the apparatus claims may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.
Finally, it should be noted that the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (10)
1. A personal risk representation generation method based on image sets is characterized by comprising the following steps:
acquiring an image data set of an object to be evaluated;
inputting each image in the image data set into a rough classification model for identification to obtain a plurality of primary labels;
inputting images corresponding to a plurality of target primary labels in the plurality of primary labels into a plurality of fine classification models for identification to obtain a plurality of secondary labels;
calculating a risk factor according to the plurality of secondary labels;
calculating a risk score from the plurality of secondary labels;
and generating a risk portrait report of the object to be evaluated based on the risk factor and the risk score.
2. The method of generating a personal risk representation based on an image set of claim 1, wherein the training process of the coarse classification model comprises:
determining a plurality of target classes;
crawling a plurality of sample images for each target category;
constructing a plurality of training sample pairs, wherein each training sample pair comprises a sample image and a corresponding target class identifier;
constructing a plurality of training sample sets and a test sample set according to the plurality of training sample pairs;
loading each training sample set to an inclusion V3 network for migration learning to obtain a plurality of candidate rough classification models, and loading the test sample set to each candidate rough classification model for testing to obtain the test passing rate of each candidate rough classification model;
determining a candidate coarse classification model corresponding to the highest test passing rate in the test passing rates as an optimal coarse classification model;
wherein the coarse classification model is stored on a blockchain.
3. The method of claim 2, wherein said loading said set of test samples into said candidate rough classification model for testing to obtain a test pass rate comprises:
loading each sample image in the test sample set to the candidate rough classification model for testing to obtain a test class identifier;
judging whether the test type identification of the sample image in the test sample set is the same as the corresponding target type identification;
when the test category identification of the sample image in the test sample set is the same as the corresponding target category identification, determining that the sample image passes the test;
calculating the proportion of the number of the sample images passing the test in the test sample set;
and determining the proportion as the test passing rate of the candidate rough classification model.
4. The method for generating a personal risk profile based on an image set according to any one of claims 1 to 3, wherein the inputting the images corresponding to the target primary labels of the primary labels into a plurality of fine classification models for identification to obtain a plurality of secondary labels comprises:
inputting an image corresponding to the first target first-level label into a first fine classification model for identification to obtain a first second-level label;
inputting the image corresponding to the second target primary label into a second fine classification model for identification to obtain a second secondary label;
inputting the image corresponding to the third target primary label into a third fine classification model for identification to obtain a third secondary label;
wherein the plurality of fine classification models are stored on a blockchain.
5. The method of generating a personal risk representation based on a collection of images as claimed in claim 4 wherein said calculating a risk factor from said plurality of secondary labels comprises:
calculating a first number of the first secondary labels, calculating a second number of the second secondary labels, and calculating a third number of the third secondary labels;
determining target secondary labels according to the first quantity, the second quantity and the third quantity;
determining the occupation of the object to be evaluated according to the target secondary label;
adding the occupation as a risk factor to a set of risk factors.
6. The method of generating a personal risk representation based on an image collection of claim 4, wherein said calculating a risk score based on said plurality of secondary labels comprises:
calculating a first ratio of the first secondary label, a second ratio of the second secondary label and a third ratio of the third secondary label;
determining a first score corresponding to the first ratio, determining a second score corresponding to the second ratio, and determining a third score corresponding to the third ratio;
and calculating the sum of the first score, the second score and the third score to obtain a risk score.
7. The method of claim 6, wherein after acquiring the image data set of the subject to be evaluated, the method further comprises:
the image data set is normalized.
8. An image-set-based personal risk representation generation apparatus, comprising:
the image acquisition module is used for acquiring an image data set of an object to be evaluated;
the first input module is used for inputting each image in the image data set into the rough classification model for identification to obtain a plurality of primary labels;
the second input module is used for inputting images corresponding to a plurality of target primary labels in the plurality of primary labels into a plurality of fine classification models for recognition to obtain a plurality of secondary labels;
the first calculation module is used for calculating risk factors according to the secondary labels;
a second calculation module for calculating a risk score according to the plurality of secondary labels;
and the report generation module is used for generating a risk portrait report of the object to be evaluated based on the risk factors and the risk scores.
9. An electronic device, comprising a processor configured to implement the image set based personal risk representation generation method of any one of claims 1 to 7 when executing a computer program stored in a memory.
10. A computer-readable storage medium, on which a computer program is stored, the computer program, when being executed by a processor, implementing a method for generating a personal risk representation based on a set of images as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010474298.5A CN111639706A (en) | 2020-05-29 | 2020-05-29 | Personal risk portrait generation method based on image set and related equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010474298.5A CN111639706A (en) | 2020-05-29 | 2020-05-29 | Personal risk portrait generation method based on image set and related equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111639706A true CN111639706A (en) | 2020-09-08 |
Family
ID=72329450
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010474298.5A Pending CN111639706A (en) | 2020-05-29 | 2020-05-29 | Personal risk portrait generation method based on image set and related equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111639706A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112365189A (en) * | 2020-11-30 | 2021-02-12 | 支付宝(杭州)信息技术有限公司 | Case distribution method and device |
| CN112529074A (en) * | 2020-12-09 | 2021-03-19 | 平安科技（深圳）有限公司 | Service information processing method and related equipment |
| CN112990792A (en) * | 2021-05-11 | 2021-06-18 | 北京智源人工智能研究院 | Method and device for automatically detecting infringement risk and electronic equipment |
| CN113849760A (en) * | 2021-12-02 | 2021-12-28 | 云账户技术(天津)有限公司 | Sensitive information risk assessment method, system and storage medium |
-
2020
- 2020-05-29 CN CN202010474298.5A patent/CN111639706A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112365189A (en) * | 2020-11-30 | 2021-02-12 | 支付宝(杭州)信息技术有限公司 | Case distribution method and device |
| CN112529074A (en) * | 2020-12-09 | 2021-03-19 | 平安科技（深圳）有限公司 | Service information processing method and related equipment |
| CN112529074B (en) * | 2020-12-09 | 2024-04-26 | 平安科技（深圳）有限公司 | Service information processing method and related equipment |
| CN112990792A (en) * | 2021-05-11 | 2021-06-18 | 北京智源人工智能研究院 | Method and device for automatically detecting infringement risk and electronic equipment |
| CN113849760A (en) * | 2021-12-02 | 2021-12-28 | 云账户技术(天津)有限公司 | Sensitive information risk assessment method, system and storage medium |
| CN113849760B (en) * | 2021-12-02 | 2022-07-22 | 云账户技术(天津)有限公司 | Sensitive information risk assessment method, system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111639706A (en) | Personal risk portrait generation method based on image set and related equipment | |
| CN108351877A (en) | The adaptive classification of data item | |
| CN112507936A (en) | Image information auditing method and device, electronic equipment and readable storage medium | |
| CN111950625A (en) | Risk identification method and device based on artificial intelligence, computer equipment and medium | |
| CN113946690A (en) | Potential customer mining method and device, electronic equipment and storage medium | |
| CN112328646B (en) | Multitask course recommendation method and device, computer equipment and storage medium | |
| CN113516417A (en) | Service evaluation method and device based on intelligent modeling, electronic equipment and medium | |
| CN114399212A (en) | Ecological environment quality evaluation method and device, electronic equipment and storage medium | |
| CN114862140A (en) | Behavior analysis-based potential evaluation method, device, equipment and storage medium | |
| CN114880449A (en) | Reply generation method and device of intelligent question answering, electronic equipment and storage medium | |
| CN112818028B (en) | Data index screening method and device, computer equipment and storage medium | |
| CN113591881A (en) | Intention recognition method and device based on model fusion, electronic equipment and medium | |
| CN113837631A (en) | Employee evaluation method and device, electronic device and readable storage medium | |
| CN117235633A (en) | Mechanism classification method, mechanism classification device, computer equipment and storage medium | |
| CN115577983B (en) | Enterprise task matching method based on block chain, server and storage medium | |
| CN116843483A (en) | Vehicle insurance claim settlement method, device, computer equipment and storage medium | |
| CN112465632A (en) | New financial AI intelligent wind control decision method and system | |
| CN113435746B (en) | User workload scoring method and device, electronic equipment and storage medium | |
| CN113781247B (en) | Protocol data recommendation method, device, computer equipment and storage medium | |
| CN112084408B (en) | List data screening method, device, computer equipment and storage medium | |
| CN113657546A (en) | Information classification method and device, electronic equipment and readable storage medium | |
| CN113987351A (en) | Artificial intelligence based intelligent recommendation method and device, electronic equipment and medium | |
| CN113610504A (en) | Data processing method and device, computer equipment and storage medium | |
| CN112365051A (en) | Agent retention prediction method and device, computer equipment and storage medium | |
| CN110362981A (en) | The method and system of abnormal behaviour are judged based on credible equipment fingerprint |
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 |