CN116308748A - Knowledge graph-based user fraud judgment system - Google Patents
Knowledge graph-based user fraud judgment system Download PDFInfo
- Publication number
- CN116308748A CN116308748A CN202310264236.5A CN202310264236A CN116308748A CN 116308748 A CN116308748 A CN 116308748A CN 202310264236 A CN202310264236 A CN 202310264236A CN 116308748 A CN116308748 A CN 116308748A
- Authority
- CN
- China
- Prior art keywords
- user
- fraud
- knowledge
- module
- identity
- 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.)
- Granted
Links
- 238000004364 calculation method Methods 0.000 claims abstract description 18
- 230000000694 effects Effects 0.000 claims abstract description 17
- 238000012795 verification Methods 0.000 claims description 10
- 230000001815 facial effect Effects 0.000 claims description 9
- 238000012546 transfer Methods 0.000 claims description 7
- 230000003542 behavioural effect Effects 0.000 claims description 3
- 230000009467 reduction Effects 0.000 description 44
- 238000012545 processing Methods 0.000 description 14
- 238000000034 method Methods 0.000 description 12
- 230000008569 process Effects 0.000 description 9
- 238000011161 development Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010225 co-occurrence analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- 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
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention belongs to the field of fraud judgment, and discloses a user fraud judgment system based on a knowledge graph, which comprises a login module, a storage module, a generation module and a calculation module; the login module is used for identifying the identity of the staff and judging whether the staff has working permission or not; the storage module is used for storing personal data of a user and blacklist data of a financial structure; the generating module is used for inputting the identity information of the user needing to judge the fraudulent activity by the staff with the working authority and generating a knowledge graph of the user according to the identity information; the calculation module is used for calculating and obtaining a fraud risk score of the user based on the knowledge graph, and if the fraud risk score is higher than a set score threshold value, fraud is indicated to exist; and if the fraud risk score is smaller than or equal to a set score threshold, indicating that no fraud exists. The invention improves the accuracy of judging the fraudulent activity.
Description
Technical Field
The invention relates to the field of fraud judgment, in particular to a user fraud judgment system based on a knowledge graph.
Background
The knowledge graph combines theory and method of subjects such as application mathematics, graphics, information visualization technology, information science and the like with metering quotation analysis and co-occurrence analysis and the like, and visualizes the core structure, development course, front edge field and whole knowledge structure of the subjects in a visual graph mode, so as to achieve the aim of multi-subject fusion. The complex knowledge field is displayed through methods such as data mining, information processing, knowledge measurement, graphic drawing and the like, the dynamic development rule of the knowledge field is revealed, and a practical and valuable reference is provided for discipline research.
With the rapid development of the internet financial market, the internet finance is continuously upgraded by the fraud means in the process of business development, and the fraud mode is more scene, specialized and intelligent. The difficulty in identifying fraud risk by credit risk practitioners is increasing, and the demands of credit businesses on fraud risk prevention are also increasing.
In the process of fraud recognition, conventional fraud recognition generally aims at personal information of a user to judge whether fraud is recognized, but the judgment mode is inaccurate. In the process of fraud identification, group behavior is also an important reference information, so that a fraud judgment system capable of combining group behavior and personal information is needed.
Disclosure of Invention
The invention aims to disclose a user fraudulent behavior judging system based on a knowledge graph, which solves the problem of how to improve the accuracy of fraudulent identification.
In order to achieve the above purpose, the invention adopts the following technical scheme:
the invention provides a knowledge graph-based user fraudulent behavior judgment system, which comprises a login module, a storage module, a generation module and a calculation module, wherein the login module is used for logging in a user;
the login module is used for identifying the identity of the staff and judging whether the staff has working permission or not;
the storage module is used for storing personal data of a user and blacklist data of a financial structure;
the generating module is used for inputting the identity information of the user needing to judge the fraudulent activity by the staff with the working authority and generating a knowledge graph of the user according to the identity information;
the calculation module is used for calculating and obtaining a fraud risk score of the user based on the knowledge graph, and if the fraud risk score is higher than a set score threshold value, fraud is indicated to exist;
and if the fraud risk score is smaller than or equal to a set score threshold, indicating that no fraud exists.
Preferably, the login module comprises a face image acquisition unit, a verification information storage unit and an identification unit;
the face image acquisition unit is used for acquiring face images of staff;
the verification information storage unit is used for storing face images of all personnel with working rights;
the identification unit is used for judging whether the facial images with the similarity between the facial images obtained by the facial image acquisition unit in the verification information storage unit is larger than a set threshold value, if yes, the staff has the working authority, and if not, the staff does not have the working authority.
Preferably, the personal data of the user includes identity data and behavior data;
the identity data comprises the name, reserved telephone, address, bank card number and identity card number of the user;
the behavioral data includes transfer records, device login records, and business transaction records.
Preferably, the blacklist data includes the names, cell phone numbers, addresses and login records of the blacklisted persons.
Preferably, the identity information of the user includes any one of an identification card number, a bank card number and a reserved phone.
Preferably, generating a knowledge graph of the user according to the identity information includes:
acquiring identity data corresponding to the identity information;
and generating a knowledge graph of the user based on the identity data and the blacklist data.
Preferably, the fraud risk score of the user is calculated based on the knowledge graph, including:
acquiring a set M of risk identification rules which are met by a user based on a knowledge graph;
calculating a fraud risk score based on the elements in set M:
where rsksco represents fraud risk score, weight m For the weight of risk identification rule m, sco m A score for rule m is identified for risk.
In the invention, the identification of the fraudulent activity of the user is judged by the fraudulent risk score obtained by calculation of the knowledge graph, wherein the knowledge graph not only contains personal data of the user, but also contains blacklist data, namely, the group activity information related to the user is considered in the calculation process, thereby improving the accuracy of judging the fraudulent activity.
Drawings
Fig. 1 is a schematic diagram of a system for determining fraudulent use of a user based on a knowledge graph according to the present invention.
Fig. 2 is a schematic diagram of a face image after noise reduction obtained according to the present invention.
Detailed Description
The present invention is described in detail below with reference to the accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate the principles of the invention and, by way of example, other aspects, features and advantages of the invention will be apparent from the detailed description.
The invention provides a knowledge-graph-based user fraud judging system, which is shown in an embodiment of FIG. 1, and comprises a login module, a storage module, a generation module and a calculation module;
the login module is used for identifying the identity of the staff and judging whether the staff has working permission or not;
the storage module is used for storing personal data of a user and blacklist data of a financial structure;
the generating module is used for inputting the identity information of the user needing to judge the fraudulent activity by the staff with the working authority and generating a knowledge graph of the user according to the identity information;
the calculation module is used for calculating and obtaining a fraud risk score of the user based on the knowledge graph, and if the fraud risk score is higher than a set score threshold value, fraud is indicated to exist;
and if the fraud risk score is smaller than or equal to a set score threshold, indicating that no fraud exists.
In the invention, the identification of the fraudulent activity of the user is judged by the fraudulent risk score obtained by calculation of the knowledge graph, wherein the knowledge graph not only contains personal data of the user, but also contains blacklist data, namely, the group activity information related to the user is considered in the calculation process, thereby improving the accuracy of judging the fraudulent activity.
Preferably, the login module comprises a face image acquisition unit, a verification information storage unit and an identification unit;
the face image acquisition unit is used for acquiring face images of staff;
the verification information storage unit is used for storing face images of all personnel with working rights;
the identification unit is used for judging whether the facial images with the similarity between the facial images obtained by the facial image acquisition unit in the verification information storage unit is larger than a set threshold value, if yes, the staff has the working authority, and if not, the staff does not have the working authority.
Because the data involved in the fraudulent identification process are private data of the user and confidential data of the financial institution, the authority identification needs to be carried out on the staff to ensure the data security.
Preferably, determining whether there is a face image in the authentication information storage unit in which the similarity between the face images obtained by the face image acquisition unit is greater than a set threshold value includes:
carrying out noise reduction treatment on the face image obtained by the face image acquisition unit to obtain a face image after noise reduction;
respectively calculating the similarity between the face image after noise reduction and each face image stored in the verification information storage unit;
judging whether the obtained multiple similarities are greater than a set threshold value or not, if yes, indicating that the information storage unit is provided with the face images with the similarities greater than the set threshold value, and if not, indicating that the information storage unit is not provided with the face images with the similarities greater than the set threshold value.
Preferably, as shown in fig. 2, the noise reduction processing is performed on the face image obtained by the face image collecting unit, so as to obtain a noise-reduced face image, which includes:
carrying out recognition of representative pixel points on the face image to obtain a representative pixel point set;
partitioning the face image, and dividing the face image into K areas with identical areas;
acquiring the noise reduction processing sequence of all areas;
sequentially carrying out noise reduction treatment on each region according to the noise reduction treatment sequence;
and forming a face image after noise reduction by using all the areas subjected to noise reduction.
In the conventional noise reduction processing algorithm, the noise reduction processing is generally performed directly on the face image, but this noise reduction processing method cannot further improve the noise reduction effect by using the pixel information of the representative pixels in the image and the distribution information among the representative pixels. In the invention, after the face image is divided into a plurality of areas, the areas are sorted by the distribution characteristics of the representative pixel points, and the noise reduction processing is sequentially performed according to the sorting result, so that the noise reduction processing effect can be improved.
Preferably, the face region is identified by representing pixels to obtain a set of representing pixels, including
Calculating the representative score of each pixel point;
and taking the pixel points with the representative scores larger than the set score threshold as representative pixel points, and storing the representative pixel points into a representative pixel point set.
Preferably, the representative score is obtained by the following function:
in the above function, repsco k Alpha, beta and l are respectively gray scale weight, first quantity weight and second quantity weight, and are the representative fraction of the pixel point k k Gravuel is a set of pixel points in the neighborhood in the b×b range of pixel point k k Gravuel is the gray value of pixel k i Is the gray value of pixel i, hvey i Is a representative degree parameter of the pixel point i,
(x i ,y i ) Is the coordinates of pixel i, (x) k ,y k ) Is the coordinate of the pixel point k, phi 1 For the set pixel value coefficient, Φ 2 For the set distance coefficient, topvuel is the maximum gray value in the face image, numofsue is the sum k The number of elements in sam k ，numofgrad k Is a sue k The number of the pixel points with the gradient direction difference between the middle pixel point and the pixel point k smaller than the set angle threshold value is the sum k Sets of pixels having gradient direction differences between the middle and pixel points k smaller than a set angle threshold, gravuel j Gravuerlsd is a set comparison constant for the gradation value of the pixel j.
The representative score is calculated from the gray value difference, gradient direction and sam between the neighborhood k The three directions of the gray value difference of the pixel points in the (a) can be calculated, and the representing score can represent the states of the pixel points from the three aspects of the pixel value, the gradient direction and the gray difference, so that the pixel points with large difference with surrounding pixel points and small gray difference between the pixel points similar to the gradient direction can be selected, the pixel points can represent the contour direction in the face area, and the subsequent noise reduction processing sequence of the area is acquired based on the pixel points. In calculating the gray value difference, i.e., the portion corresponding to alpha, the present invention does not directly apply k Gray value average value and gravuel of pixel points in (a) k The comparison is performed by calculating the product of the representative degree parameter and the gray value, and the calculated representative degree parameter is used for the comparisonIn the course, in addition to the gray value difference, the difference of coordinates is also considered, so that the representative degree parameter can comprehensively represent the item from the two aspects k Pairs of pixels in gravuel k The influence degree of noise on gray value difference calculation can be reduced, and the accuracy of gray value difference is improved.
Preferably, the acquiring the noise reduction processing order of all the areas includes:
respectively calculating the sequence score of each region;
the regions are ordered according to the order of the order score from high to low to obtain a noise reduction processing order,
wherein for the nth region block n The calculation function of the sequence score is as follows:
in the above function, the function is that,
is block n Is a block n The number of representative pixel points contained in +.>
For the total number of pixel points contained in the pixel points, eta is the sum ratio, eta is E (0, 1), xydst is a set coordinate coefficient larger than 0, min { mstx, msty } represents taking the smaller value of mstx and msty, and dist (u, u-1) represents taking block n The representative pixel points contained in the pixel array are ordered according to the descending order of the abscissa, and then the u-th representative pixel point and the u-1-th representative pixel point are arrangedThe distance between points, mstx is the abscissa variance, msty is the ordinate variance, x m And y m Respectively represent ublock n Abscissa and ordinate of representative pixel point m, ublock n Is block n The collection of representative pixel points contained in the image, n E [1, K ]]。
In the calculation process, on one hand, the number of the representative pixel points is considered, the more the number of the representative pixel points is, the more the pixel information of the representative pixel points contained in the region and the distribution information among the representative pixel points are represented, and on the other hand, the distribution characteristics of the representative pixel points are considered, the better the continuity of the representative pixel points is, the more linear the distribution is, the smaller the product of min { mstx, msty } and dist (u, u-1) is, and the larger the sequence score is.
Therefore, the larger the order score is, the more the noise reduction calculation is needed to be performed on the area, and the noise reduction result of the area with the front noise reduction order can be subjected to noise reduction processing by the area with the rear noise reduction order due to the change of the pixel points after noise reduction, so that the pixel information and the distribution information can be transmitted to the subsequent noise reduction process, and the noise reduction processing effect is improved.
Preferably, the noise reduction processing is sequentially performed on each region according to the noise reduction processing sequence, including:
obtaining a median numid:
for the regions sequenced in the front numid bit, adopting a wavelet noise reduction algorithm to respectively carry out noise reduction treatment on each region according to the noise reduction treatment sequence;
and for other areas, adopting NML algorithm to respectively perform noise reduction treatment on each area according to the noise reduction treatment sequence.
In the invention, in the process of carrying out noise reduction according to the noise reduction sequence, different noise reduction methods are adopted for the areas with different sequences to carry out noise reduction, and the more the sequence score is, the more the contour information is contained in the area, so in order to ensure the accuracy of similarity calculation, the invention adopts a wavelet noise reduction algorithm with longer running time, but better contour information is reserved to carry out noise reduction, and the importance of the information contained in the area with the sequence behind the median is reduced, therefore, the invention utilizes the algorithm with shorter running time to carry out noise reduction, so that the invention can balance the noise reduction effect and the time consumed by noise reduction. The noise reduction time is shortened while the effect of the noise reduction processing is ensured.
Preferably, the personal data of the user includes identity data and behavior data;
the identity data comprises the name, reserved telephone, address, bank card number and identity card number of the user;
the behavioral data includes transfer records, device login records, and business transaction records.
Specifically, the device login record may include a mac address, login IP, login location, and the like of the login device. The business transaction record includes the name of the network site responsible for the transaction, the transaction time, the transaction project, the signed file, etc.
Preferably, the blacklist data includes the names, cell phone numbers, addresses and login records of the blacklisted persons.
The blacklist data is data of persons who have been blacklisted by the financial institution. The risk of fraud is analyzed by analyzing the relationship between the user and these blacklisted persons, the closer the relationship the higher the risk of fraud.
Preferably, the identity information of the user includes any one of an identification card number, a bank card number and a reserved phone.
In particular, in addition to the above-listed types of identity information, other types of information capable of determining the identity of a user are possible.
Preferably, generating a knowledge graph of the user according to the identity information includes:
acquiring identity data corresponding to the identity information;
and generating a knowledge graph of the user based on the identity data and the blacklist data.
Specifically, the association between the blacklist person and the user may be performed through the identity data and the blacklist data, for example, if the user transfers to the blacklist person and the data such as the transfer amount and the transfer frequency conform to the set rules, the association is established through the transfer record. For another example, the reserved phone of the customer belongs to the mobile phone number of the blacklist person, and the reserved phone of the customer is associated with the mobile phone number of the blacklist person. In the knowledge graph, the users and the blacklisted people associated with the users are points in the graph, and the relationship between the users and the blacklisted people forms edges in the graph.
When the knowledge graph is generated, a distance rule can be set, and if the distance between a certain blacklist person and a user is larger than a set distance threshold value, the blacklist person is not included in the knowledge graph.
For example, when a point a representing a user is directly connected to other points, the distance between the two is 1, and if the distance between the point B and the point a is 1, and the point C is connected to the point B and not to the point a, the distance between the point C and the point a is 2.
If point C is not directly connected to point a, but there are multiple paths that are indirectly connected to point a, the minimum distance among the multiple paths is selected as the distance between point a and point C.
Preferably, the fraud risk score of the user is calculated based on the knowledge graph, including:
acquiring a set M of risk identification rules which are met by a user based on a knowledge graph;
calculating a fraud risk score based on the elements in set M:
where rsksco represents fraud risk score, weight m For the weight of risk identification rule m, sco m A score for rule m is identified for risk.
Specifically, the risk identification rule is a preset rule, for example, a reserved phone of a user corresponds to a plurality of users within a set time period, for example, within one year; the mac address of the login device of the user is the same as the mac address of the device used when the blacklisted person logs in.
By weighting and fusing the scores of the risk identification rules, a score that comprehensively represents the risk of fraud for the user can be obtained.
The technical scheme, the working process and the implementation effect of the invention are described in detail, and it is to be noted that the description is only a typical example of the invention, and besides, the invention can also have other various specific embodiments, and all the technical schemes formed by adopting equivalent substitution or equivalent transformation fall within the scope of the invention claimed.
Claims (7)
1. The system for judging the fraudulent use behavior of the user based on the knowledge graph is characterized by comprising a login module, a storage module, a generation module and a calculation module;
the login module is used for identifying the identity of the staff and judging whether the staff has working permission or not;
the storage module is used for storing personal data of a user and blacklist data of a financial structure;
the generating module is used for inputting the identity information of the user needing to judge the fraudulent activity by the staff with the working authority and generating a knowledge graph of the user according to the identity information;
the calculation module is used for calculating and obtaining a fraud risk score of the user based on the knowledge graph, and if the fraud risk score is higher than a set score threshold value, fraud is indicated to exist;
and if the fraud risk score is smaller than or equal to a set score threshold, indicating that no fraud exists.
2. The knowledge-graph-based user fraud judgment system according to claim 1, wherein the login module comprises a face image acquisition unit, a verification information storage unit and an identification unit;
the face image acquisition unit is used for acquiring face images of staff;
the verification information storage unit is used for storing face images of all personnel with working rights;
the identification unit is used for judging whether the facial images with the similarity between the facial images obtained by the facial image acquisition unit in the verification information storage unit is larger than a set threshold value, if yes, the staff has the working authority, and if not, the staff does not have the working authority.
3. The knowledge-graph-based user fraud determination system of claim 1, wherein the personal data of the user includes identity data and behavior data;
the identity data comprises the name, reserved telephone, address, bank card number and identity card number of the user;
the behavioral data includes transfer records, device login records, and business transaction records.
4. The knowledge-graph-based user fraud determination system of claim 1, wherein the blacklist data includes names, cell phone numbers, addresses, and login records of blacklisted persons.
5. The knowledge-graph-based user fraud determination system of claim 1, wherein the identity information of the user includes any one of an identification card number, a bank card number, and a reserved phone.
6. The knowledge-graph-based user fraud determination system of claim 1, wherein generating the knowledge graph of the user based on the identity information comprises:
acquiring identity data corresponding to the identity information;
and generating a knowledge graph of the user based on the identity data and the blacklist data.
7. The knowledge-based user fraud determination system of claim 1, wherein the knowledge-based calculation of the fraud risk score for the user includes:
acquiring a set M of risk identification rules which are met by a user based on a knowledge graph;
calculating a fraud risk score based on the elements in set M:
where rsksco represents fraud risk score, weight m For the weight of risk identification rule m, sco m A score for rule m is identified for risk.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310264236.5A CN116308748B (en) | 2023-03-19 | 2023-03-19 | Knowledge graph-based user fraud judgment system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310264236.5A CN116308748B (en) | 2023-03-19 | 2023-03-19 | Knowledge graph-based user fraud judgment system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116308748A true CN116308748A (en) | 2023-06-23 |
| CN116308748B CN116308748B (en) | 2023-10-20 |
Family
ID=86828401
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310264236.5A Active CN116308748B (en) | 2023-03-19 | 2023-03-19 | Knowledge graph-based user fraud judgment system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116308748B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117057641A (en) * | 2023-07-27 | 2023-11-14 | 武汉威克睿特科技有限公司 | Digital scoring system for evaluating employee performance |
Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663696A (en) * | 2012-03-31 | 2012-09-12 | 广东威创视讯科技股份有限公司 | Denoising method of enlarged image and system thereof |
| CN104168402A (en) * | 2013-05-17 | 2014-11-26 | 浙江大华技术股份有限公司 | Method and device for video frame image defogging |
| CN109191281A (en) * | 2018-08-21 | 2019-01-11 | 重庆富民银行股份有限公司 | A kind of group's fraud identifying system of knowledge based map |
| CN109446977A (en) * | 2018-10-25 | 2019-03-08 | 平安科技（深圳）有限公司 | Image processing method, device, storage medium and terminal based on recognition of face |
| CN110111110A (en) * | 2019-04-01 | 2019-08-09 | 北京三快在线科技有限公司 | The method and apparatus of knowledge based map detection fraud, storage medium |
| CN110246110A (en) * | 2018-03-01 | 2019-09-17 | 腾讯科技（深圳）有限公司 | Image evaluation method, device and storage medium |
| CN110297912A (en) * | 2019-05-20 | 2019-10-01 | 平安科技（深圳）有限公司 | Cheat recognition methods, device, equipment and computer readable storage medium |
| CN111242481A (en) * | 2020-01-13 | 2020-06-05 | 北京银地信息科技有限公司 | E-government platform based on big data |
| CN111259167A (en) * | 2020-02-11 | 2020-06-09 | 广州众赢科技有限公司 | User request risk identification method and device |
| CN111652596A (en) * | 2020-06-15 | 2020-09-11 | 深圳前海微众银行股份有限公司 | Credit service anti-fraud method, device, terminal equipment and storage medium |
| CN112053221A (en) * | 2020-08-14 | 2020-12-08 | 百维金科(上海)信息科技有限公司 | Knowledge graph-based internet financial group fraud detection method |
| CN112053222A (en) * | 2020-08-14 | 2020-12-08 | 百维金科(上海)信息科技有限公司 | Knowledge graph-based internet financial group fraud detection method |
| CN113298837A (en) * | 2021-07-27 | 2021-08-24 | 南昌工程学院 | Image edge extraction method and device, storage medium and equipment |
| CN113807723A (en) * | 2021-09-24 | 2021-12-17 | 重庆富民银行股份有限公司 | Risk identification method for knowledge graph |
| CN113850665A (en) * | 2021-09-14 | 2021-12-28 | 江苏中交车旺科技有限公司 | Logistics financial knowledge graph-based fraud prevention and control method and system |
| CN114648456A (en) * | 2022-03-17 | 2022-06-21 | 青岛海信医疗设备股份有限公司 | Ultrasonic image noise reduction method and electronic equipment |
| CN114998282A (en) * | 2022-06-16 | 2022-09-02 | 平安科技（深圳）有限公司 | Image detection method, image detection device, electronic equipment and storage medium |
-
2023
- 2023-03-19 CN CN202310264236.5A patent/CN116308748B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663696A (en) * | 2012-03-31 | 2012-09-12 | 广东威创视讯科技股份有限公司 | Denoising method of enlarged image and system thereof |
| CN104168402A (en) * | 2013-05-17 | 2014-11-26 | 浙江大华技术股份有限公司 | Method and device for video frame image defogging |
| CN110246110A (en) * | 2018-03-01 | 2019-09-17 | 腾讯科技（深圳）有限公司 | Image evaluation method, device and storage medium |
| CN109191281A (en) * | 2018-08-21 | 2019-01-11 | 重庆富民银行股份有限公司 | A kind of group's fraud identifying system of knowledge based map |
| CN109446977A (en) * | 2018-10-25 | 2019-03-08 | 平安科技（深圳）有限公司 | Image processing method, device, storage medium and terminal based on recognition of face |
| CN110111110A (en) * | 2019-04-01 | 2019-08-09 | 北京三快在线科技有限公司 | The method and apparatus of knowledge based map detection fraud, storage medium |
| CN110297912A (en) * | 2019-05-20 | 2019-10-01 | 平安科技（深圳）有限公司 | Cheat recognition methods, device, equipment and computer readable storage medium |
| CN111242481A (en) * | 2020-01-13 | 2020-06-05 | 北京银地信息科技有限公司 | E-government platform based on big data |
| CN111259167A (en) * | 2020-02-11 | 2020-06-09 | 广州众赢科技有限公司 | User request risk identification method and device |
| CN111652596A (en) * | 2020-06-15 | 2020-09-11 | 深圳前海微众银行股份有限公司 | Credit service anti-fraud method, device, terminal equipment and storage medium |
| CN112053221A (en) * | 2020-08-14 | 2020-12-08 | 百维金科(上海)信息科技有限公司 | Knowledge graph-based internet financial group fraud detection method |
| CN112053222A (en) * | 2020-08-14 | 2020-12-08 | 百维金科(上海)信息科技有限公司 | Knowledge graph-based internet financial group fraud detection method |
| CN113298837A (en) * | 2021-07-27 | 2021-08-24 | 南昌工程学院 | Image edge extraction method and device, storage medium and equipment |
| CN113850665A (en) * | 2021-09-14 | 2021-12-28 | 江苏中交车旺科技有限公司 | Logistics financial knowledge graph-based fraud prevention and control method and system |
| CN113807723A (en) * | 2021-09-24 | 2021-12-17 | 重庆富民银行股份有限公司 | Risk identification method for knowledge graph |
| CN114648456A (en) * | 2022-03-17 | 2022-06-21 | 青岛海信医疗设备股份有限公司 | Ultrasonic image noise reduction method and electronic equipment |
| CN114998282A (en) * | 2022-06-16 | 2022-09-02 | 平安科技（深圳）有限公司 | Image detection method, image detection device, electronic equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 孙佳男;孙俊;: "融合深度特征表达与似物性采样的目标跟踪算法", 小型微型计算机系统, no. 01, pages 86 - 90 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117057641A (en) * | 2023-07-27 | 2023-11-14 | 武汉威克睿特科技有限公司 | Digital scoring system for evaluating employee performance |
| CN117057641B (en) * | 2023-07-27 | 2024-03-08 | 武汉威克睿特科技有限公司 | Digital scoring system for evaluating employee performance |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116308748B (en) | 2023-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9639769B2 (en) | Liveness detection | |
| Yue et al. | A review of data mining-based financial fraud detection research | |
| Banker et al. | Strategic contributions of information technology: an empirical study of ATM networks | |
| CN108665159A (en) | A kind of methods of risk assessment, device, terminal device and storage medium | |
| CN110413707A (en) | The excavation of clique's relationship is cheated in internet and checks method and its system | |
| CN108090830B (en) | Credit risk rating method and device based on facial portrait | |
| CN113240509A (en) | Loan risk assessment method based on multi-source data federal learning | |
| CN116308748B (en) | Knowledge graph-based user fraud judgment system | |
| Chamboko et al. | Modelling and forecasting recurrent recovery events on consumer loans | |
| US20210224588A1 (en) | Recruitment process graph based unsupervised anomaly detection | |
| KR102005733B1 (en) | Block chain-based person-to-person financial service offering system using credit rating assessment result drawn on online big data analysis | |
| CN116738444B (en) | Xia Puli value-based multi-party contribution degree evaluation method for data security sharing platform | |
| CN111768230A (en) | Label recommendation method and device for client portrait system and computer equipment | |
| CN114297447B (en) | Electronic certificate marking method and system based on epidemic prevention big data and readable storage medium | |
| Magno et al. | Stereotypes in search engine results: understanding the role of local and global factors | |
| Diaz et al. | Investigating the common authorship of signatures by off-line automatic signature verification without the use of reference signatures | |
| Kiruthika et al. | Image quality assessment based fake face detection | |
| CN116881472B (en) | Funds penetration and character relation analysis method based on graph database technology | |
| JP2004054769A (en) | Loan assets management system, loan assets management method, and recording medium and program therefor | |
| Krysovatyy et al. | Classification Method of Fictitious Enterprises Based on Gaussian Naive Bayes | |
| Apparao et al. | Financial statement fraud detection by data mining | |
| Ordouei et al. | Integrating data mining and knowledge management to improve customer relationship management in banking industry (Case study of Caspian Credit Institution) | |
| CN111260372A (en) | Resource transfer user group determination method, device, computer equipment and storage medium | |
| CN111986034B (en) | Medical insurance group fraud monitoring method, system and storage medium | |
| Yuan¹ et al. | DenseNet Convolutional Neural Network for Breast Cancer Diagnosis |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |