CN108681936B - Fraud group identification method based on modularity and balanced label propagation - Google Patents
Fraud group identification method based on modularity and balanced label propagation Download PDFInfo
- Publication number
- CN108681936B CN108681936B CN201810382121.5A CN201810382121A CN108681936B CN 108681936 B CN108681936 B CN 108681936B CN 201810382121 A CN201810382121 A CN 201810382121A CN 108681936 B CN108681936 B CN 108681936B
- Authority
- CN
- China
- Prior art keywords
- community
- node
- fraud
- modularity
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0609—Buyer or seller confidence or verification
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a cheating group identification method based on modularity and balanced label propagation, which comprises the following steps: calculating pairwise similarity of all users by using the ID characteristics and combining known fraud identifications of the users, establishing a similarity matrix, and establishing a correlation diagram through the similarity matrix; running a Louvain algorithm on the established graph to obtain community and level information of each node; and taking the community, the level information and the fraud identifier to which each node belongs as initial community information of each node, operating a balanced label propagation process to obtain the final community to which each node belongs, dividing the network according to whether the node belongs to the common community, and dividing the fraud group according to the fraud identifier obtained by propagation. The invention applies the fraud group recognition method based on modularity and balanced label propagation to the fields of application anti-fraud and transaction anti-fraud for the first time, utilizes information such as transaction correlation and the like to construct a correlation map, integrates community modularity information, and utilizes a balanced label propagation algorithm to detect fraud communities so as to prevent potential fraud transactions.
Description
Technical Field
The invention belongs to the field of transaction anti-fraud and application anti-fraud, and particularly relates to a fraud group identification method based on modularity and balance label propagation
Background
With the explosive growth of online services such as electronic commerce, third-party payment and the like, online fraud cases are rampant increasingly, trends of variable methods and diversified fields are presented, and how to effectively and timely identify frequently occurring online fraud behaviors becomes a problem which needs to be solved urgently. The traditional online fraud detection method usually implements fraud detection on relevant characteristics of constructed service flow aiming at each online transaction or merchant entity modeling, has excellent fraud effect on obvious characteristics of the transaction, ignores the group-partner relevance behind the fraud transaction, and has poor group-partner fraud identification capability on forging normal user information.
Community discovery is a technology for finding out potential contact rules of nodes in a complex network structure by identifying communities or sub-networks with specific rules in the complex network structure and further dividing the complex network. In the field of transaction anti-fraud and application anti-fraud, a client can construct a complex network for highlighting abnormal behaviors through transaction related information and application related information, and the network is analyzed and mined by utilizing a community discovery technology, so that fraudulent groups in the network can be effectively identified, and fraudulent behaviors are prevented.
Disclosure of Invention
The invention aims to provide a fraud group identification method based on modularity and balanced label propagation aiming at the defects of the prior art.
The purpose of the invention is realized by the following technical scheme: a fraudulent group identification method based on modularity and balanced tag propagation, the method comprising the steps of:
step 1, extracting ID characteristics in a scene of transaction anti-fraud or application anti-fraud;
step 2, calculating pairwise similarity of all users (including fraud blacklists and normal users) by using ID features extracted from transaction data or application data and combining known fraud identifications of the users, establishing a similarity matrix, and establishing a correlation diagram through the similarity matrix;
step 3, operating a Louvain algorithm on the established association graph to obtain community and level information of each node;
and 4, taking the community, the level information and the fraud identifier of each node as the initial community information of each node, operating a balanced label propagation process to obtain the final community to which each node belongs, dividing the network according to whether the node belongs to the common community, and dividing the fraud group according to the fraud identifier obtained by propagation.
Further, in step 2, the ID features include a card number, an account number, ip, and a device fingerprint.
Further, in the step 2, let n characteristics of the user i be Xi,1,Xi,2,Xi,3…Xi,nSimilarity w between user i and user ji,jThe definition can refer to the actual service situation, recommend to use common attributes, cosine distances, and the like, and can be optionally defined as follows:
common attributes:
wi,j＝∑ku(Xi,k,Xj,k)(k＝1…·n)/k
cosine distance:
wi,j＝Cos(Xi,Xj)
for m users, the following similarity matrix is formed:
further, in step 2, setting 0 in the similarity matrix lower than the threshold p, establishing edge contact for user nodes not being 0, and constructing a graph structure, where the similarity between nodes is the weight of an edge.
Further, the step 3 comprises:
(1) initially, each node in the association graph is assumed to belong to an independent community;
(2) sequentially traversing all neighbor nodes of each node i in the association graph, and calculating the modularity variable quantity delta Q before and after distributing the node i to the community to which the neighbor node belongs; updating the maximum value max delta Q of the modularity change, wherein max _ j is a neighbor node corresponding to the maximum value max delta Q, when max delta Q is larger than 0, the node i is distributed to a community where max _ j is located, otherwise, the node i is kept unchanged;
(3) re-executing the step (2) until the node home community is not changed any more;
(4) merging nodes belonging to the same community in the association diagram into a super node to reconstruct a network, wherein the weight of the super node is converted by the weight of edges among nodes in the community, and the weight of the edges among the super nodes is converted by the weight of the edges among the community intervals, so that the compression of the association diagram is realized;
(5) and (4) re-executing the step (1) until the set iteration number or the modularity of the association diagram is not changed any more, and finally obtaining the community of each node in each level.
Further, the step 4 comprises:
(1) setting the initial attribution of each node in the association diagram established in the step 2 as a community to which each level obtained by a Louvain algorithm belongs to obtain < community id, probability value > information of each node, wherein the initial probability is equal to 1/(the number of communities to which the node belongs), and the community id is composed of the level, the community to which the level belongs and a fraud identifier;
(2) traversing all the neighbor nodes of each node, adding the probabilities corresponding to the same community id, and marking as b and bmaxAfter the probabilities are added<Community id, probability value>The maximum value of the median probability; according to the formula
Filtering the community id of each node in the association graph, wherein q is an adjustable parameter, and the value range of q is [0,1]]To (c) to (d);
(3) normalizing < community id, probability value > information of each node;
(4) repeating the step (2) until reaching the specified iteration number;
(5) and dividing the network according to whether the network belongs to the common community, and dividing the cheating group according to the cheating identification obtained by propagation.
The invention has the beneficial effects that: the invention applies the cheating group recognition method based on modularity and balance label propagation to the fields of application anti-cheating and transaction anti-cheating for the first time, combines the advantages of Louvain and balance label propagation algorithm, not only utilizes the associated information of the seed cheating node, but also considers the requirement of the community modularity to be optimal, and finally recognizes the cheating group with suspicious transaction behaviors and application actions, thereby having better community structure and excellent accuracy. The method has great research significance and use value in the fields of transaction anti-fraud and application anti-fraud.
Drawings
FIG. 1 is a schematic diagram of a correlation diagram established using a similarity matrix;
FIG. 2 is a schematic diagram of community and level information of each node according to the association graph;
FIG. 3 is the result of tag propagation in conjunction with a hierarchy and a rogue tag.
Detailed Description
The following detailed description of the present invention will be made with reference to the accompanying drawings and examples, which are provided for illustration of the present invention and are not intended to limit the scope of the present invention.
The invention provides a fraud group identification method based on modularity and balanced label propagation, which comprises the following steps:
step 1, extracting characteristics such as card numbers, account numbers, ip and device fingerprints; as shown in Table 1
TABLE 1 transaction characteristics Table
And 2, calculating pairwise similarity of all users (including a fraud blacklist and normal users) by using the features extracted from the transaction data, establishing a similarity matrix, and establishing a correlation diagram through the matrix, as shown in fig. 1, wherein circles in the diagram represent user nodes, numbers represent user IDs, and numbers on edges in the diagram represent edge weights calculated by the edge similarity matrix.
And 3, operating a Louvain algorithm on the established association graph to obtain community and level information of each node, wherein as shown in FIG. 2, a plurality of nodes at the upper left corner are divided into a community, and graph compression is performed for 3 times.
And 4, taking the community, the hierarchy information and the fraud identifier of each node as initial community information of each node, operating a balanced label propagation process to obtain a final community to which each node belongs, dividing a network according to whether the node belongs to a common community, and dividing fraud groups according to the fraud identifier obtained by propagation, wherein as shown in fig. 3, three communities are identified in total, wherein the black community at the upper left corner is a fraud group, the white community at the upper right corner is a normal user community, and the gray community at the lower part is a suspicion group.
Wherein the step 2 is implemented according to the following steps:
let the characteristic of user i be Xi,1,Xi,2,Xi,3…·Xi,nThe similarity definition between the user i and the user j can refer to the actual service condition, recommend to use common attributes, cosine distances and the like, and can be optionally defined as follows:
common attributes:
wi,j＝∑ku(Xi,k,Xj,k)(k＝1…n)/k
cosine distance:
wi,j＝Cos(Xi,Xj)
for m users, the following similarity matrix is formed:
setting 0 in the similarity matrix lower than the threshold p, establishing edge relation for user nodes which are not 0, and constructing a graph structure, wherein the similarity between nodes is the weight of the edge, as shown in fig. 1, wherein a circle in the graph represents a user node, a number represents a user ID, and a number on the edge in the graph represents the edge weight calculated by the edge similarity matrix.
Wherein the step 3 is implemented according to the following steps:
(1) initially, each node in the association graph is assumed to belong to an independent community;
(2) sequentially traversing all neighbor nodes of each node i in the association graph, and calculating the modularity variable quantity delta Q before and after distributing the node i to the community to which the neighbor node belongs; updating the maximum value max delta Q of the modularity change, wherein max _ j is a neighbor node corresponding to the maximum value max delta Q, when max delta Q is larger than 0, the node i is distributed to a community where max _ j is located, otherwise, the node i is kept unchanged;
(3) re-executing the step (2) until the node home community is not changed any more;
(4) merging nodes belonging to the same community in the association diagram into a super node to reconstruct a network, wherein the weight of the super node is converted by the weight of edges among nodes in the community, and the weight of the edges among the super nodes is converted by the weight of the edges among the community intervals, so that the compression of the association diagram is realized;
(5) and (3) re-executing the step (1) until the set iteration number or the modularity of the association graph is not changed any more, and finally obtaining the community of each node in each level, wherein as shown in fig. 2, a plurality of nodes at the upper left corner are divided into communities, and graph compression is performed for 3 times.
The modularity measures the good or bad of a community network partition by calculating the difference between the edge number of the internal node of the community and the edge number under the random condition, and the value range of the community network partition is [ -1/2,1], which is defined as follows:
wherein A isi,jIs the weight of the edge between node i and node j; k is a radical ofi＝∑jAi,jRepresents the sum of the weights of all edges connected to node i; c. CiRepresenting the community to which the node i belongs;
representing the sum of the weights of all edges.
In the formula
The probability that node j is connected to any one node is
Now node i has kiSo that the edges of nodes i and j are at random
Wherein Δ Q is defined as follows:
in the formula, sigma in represents the sum of the side weights of nodes with key points in the community, sigma tot represents the sum of the side weights in the incident community, and ki,inRepresenting the weighted sum of the incident communities of node i.
Wherein the step 4 is implemented according to the following steps:
(1) setting the initial attribution of each node in the association diagram established in the step 2 as a community to which each level obtained by a Louvain algorithm belongs to obtain < community id, probability value > information of each node, wherein the initial probability is equal to 1/(the number of communities to which the node belongs), and the community id is composed of the level, the community to which the level belongs and a fraud identifier;
(2) traversing all the neighbor nodes of each node, adding the probabilities corresponding to the same community id, and marking as b and bmaxAfter the probabilities are added<Community id, probability value>The maximum value of the median probability; according to the formula
Filtering the community id of each node in the association graph, wherein q is an adjustable parameter, and the value range of q is [0,1]]To (c) to (d);
(3) normalizing < community id, probability value > information of each node;
(4) repeating the step (2) until reaching the specified iteration number;
(5) and dividing a cheating group according to the cheating identification obtained by propagation and dividing the network according to whether the community belongs to the common community, wherein three communities are identified in total as shown in figure 3, wherein a black community at the upper left corner is a cheating group, a white community at the upper right corner is a normal user community, and a gray community at the lower part is an in-doubt group.
The invention provides a cheating group partner identification method based on modularity and balanced label propagation, which combines the advantages of Louvain and balanced label propagation algorithm, not only utilizes the associated information of seed cheating nodes, but also considers the requirement of optimal community modularity, finally identifies the cheating group partner with suspicious transaction behaviors and application actions, and has better community structure and excellent accuracy. The method has great research significance and use value in the fields of transaction anti-fraud and application anti-fraud.
The above embodiments are only for illustrating the invention and are not to be construed as limiting the invention, and those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention, therefore, all equivalent technical solutions also fall into the scope of the invention, and the scope of the invention is defined by the claims.
Claims (4)
1. A fraudulent group identification method based on modularity and balanced tag propagation, the method comprising the steps of:
step 1, extracting ID characteristics in a scene of transaction anti-fraud or application anti-fraud;
step 2, calculating pairwise similarity of all users including a fraud blacklist and normal users by using ID features extracted from transaction data or application data and combining known fraud identifications of the users, establishing a similarity matrix, and establishing a correlation diagram through the similarity matrix;
step 3, operating a Louvain algorithm on the established association graph to obtain community and level information of each node; the method comprises the following specific steps:
(1) initially, each node in the association graph is assumed to belong to an independent community;
(2) sequentially traversing all neighbor nodes of each node i in the association graph, and calculating the modularity variable quantity delta Q before and after distributing the node i to the community to which the neighbor node belongs; updating the maximum value max delta Q of the modularity change, wherein max _ j is a neighbor node corresponding to the maximum value max delta Q, when max delta Q is larger than 0, the node i is distributed to a community where max _ j is located, otherwise, the node i is kept unchanged;
(3) re-executing the step (2) until the node home community is not changed any more;
(4) merging nodes belonging to the same community in the association diagram into a super node to reconstruct a network, wherein the weight of the super node is converted by the weight of edges among nodes in the community, and the weight of the edges among the super nodes is converted by the weight of the edges among the community intervals, so that the compression of the association diagram is realized;
(5) re-executing the step (1) until the set iteration number or the modularity of the association diagram is not changed, and finally obtaining the community of each node at each level;
step 4, taking the community, the level information and the fraud identification of each node as the initial community information of each node, operating a balanced label propagation process to obtain the final community of each node, dividing the network according to whether the node belongs to the common community, and dividing the fraud group according to the fraud identification obtained by propagation; the method comprises the following specific steps:
(1) setting the initial attribution of each node in the association diagram established in the step 2 as a community to which each level obtained by a Louvain algorithm belongs to obtain < community id, probability value > information of each node, wherein the initial probability is equal to 1/(the number of communities to which the node belongs), and the community id is composed of the level, the community to which the level belongs and a fraud identifier;
(2) traversing all the neighbor nodes of each node, adding the probabilities corresponding to the same community id, and marking as b and bmaxAfter the probabilities are added<Community id, probability value>The maximum value of the median probability; according to the formula
Filtering the community id of each node in the association graph, wherein q is an adjustable parameter, and the value range of q is [0,1]]To (c) to (d);
(3) normalizing < community id, probability value > information of each node;
(4) repeating the step (2) until reaching the specified iteration number;
(5) and dividing the network according to whether the network belongs to the common community, and dividing the cheating group according to the cheating identification obtained by propagation.
2. A method of fraud group identification based on modularity and balanced tag propagation as claimed in claim 1, wherein in step 2 said ID features include card number, account number, ip and device fingerprint.
3. The method as claimed in claim 1, wherein in step 2, let n characteristics of user i be Xi,1,Xi,2,Xi,3…Xi,nThe similarity definition between user i and user j may refer to the actual service condition and adopt a common attribute or cosine distance.
4. The method for identifying fraudulent groups based on modularity and balanced label propagation as claimed in claim 1, wherein in step 2, setting 0 in the similarity matrix lower than the threshold p, where p is an adjustable parameter, the value range of p is between [0 and 1], and the user nodes not 0 establish edge relations, and construct a graph structure, and the similarity between nodes is the weight of the edge.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810382121.5A CN108681936B (en) | 2018-04-26 | 2018-04-26 | Fraud group identification method based on modularity and balanced label propagation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810382121.5A CN108681936B (en) | 2018-04-26 | 2018-04-26 | Fraud group identification method based on modularity and balanced label propagation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108681936A CN108681936A (en) | 2018-10-19 |
| CN108681936B true CN108681936B (en) | 2021-11-02 |
Family
ID=63802082
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810382121.5A Active CN108681936B (en) | 2018-04-26 | 2018-04-26 | Fraud group identification method based on modularity and balanced label propagation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108681936B (en) |
Families Citing this family (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109949046B (en) * | 2018-11-02 | 2023-06-09 | 创新先进技术有限公司 | Identification method and device for risk group partner |
| CN109816535A (en) * | 2018-12-13 | 2019-05-28 | 中国平安财产保险股份有限公司 | Cheat recognition methods, device, computer equipment and storage medium |
| CN109784636A (en) * | 2018-12-13 | 2019-05-21 | 中国平安财产保险股份有限公司 | Fraudulent user recognition methods, device, computer equipment and storage medium |
| CN109685132B (en) * | 2018-12-20 | 2020-12-11 | 四川新网银行股份有限公司 | Label propagation method based on bipartite graph |
| CN109754258B (en) * | 2018-12-24 | 2023-05-12 | 同济大学 | Online transaction fraud detection method based on individual behavior modeling |
| CN110020866B (en) * | 2019-01-22 | 2023-06-13 | 创新先进技术有限公司 | Training method and device for recognition model and electronic equipment |
| CN111666346A (en) * | 2019-03-06 | 2020-09-15 | 京东数字科技控股有限公司 | Information merging method, transaction query method, device, computer and storage medium |
| CN110070364A (en) * | 2019-03-27 | 2019-07-30 | 北京三快在线科技有限公司 | Method and apparatus, storage medium based on the fraud of graph model detection clique |
| CN110135890A (en) * | 2019-04-15 | 2019-08-16 | 深圳壹账通智能科技有限公司 | The product data method for pushing and relevant device of knowledge based relation excavation |
| CN111951021A (en) * | 2019-05-15 | 2020-11-17 | 财付通支付科技有限公司 | Method and device for discovering suspicious communities, storage medium and computer equipment |
| CN110232524A (en) * | 2019-06-14 | 2019-09-13 | 哈尔滨哈银消费金融有限责任公司 | Social networks cheats the construction method of model, antifraud method and apparatus |
| CN110245875A (en) * | 2019-06-21 | 2019-09-17 | 深圳前海微众银行股份有限公司 | Risk of fraud appraisal procedure, device, equipment and storage medium |
| CN110490730B (en) * | 2019-08-21 | 2022-07-26 | 北京顶象技术有限公司 | Abnormal fund aggregation behavior detection method, device, equipment and storage medium |
| CN110717828B (en) * | 2019-09-09 | 2022-09-30 | 中国科学院计算技术研究所 | Abnormal account detection method and system based on frequent transaction mode |
| CN110929509B (en) * | 2019-10-16 | 2023-09-15 | 上海大学 | Domain event trigger word clustering method based on louvain community discovery algorithm |
| CN110766091B (en) * | 2019-10-31 | 2024-02-27 | 上海观安信息技术股份有限公司 | Method and system for identifying trepanning loan group partner |
| CN111241297B (en) * | 2020-01-06 | 2024-05-14 | 广州拉卡拉信息技术有限公司 | Atlas data processing method and apparatus based on label propagation algorithm |
| CN113129010A (en) * | 2020-01-10 | 2021-07-16 | 联洋国融(北京)科技有限公司 | Fraud group mining system and method based on complex network model |
| CN111325350B (en) * | 2020-02-19 | 2023-09-29 | 第四范式（北京）技术有限公司 | Suspicious tissue discovery system and method |
| CN111325578B (en) * | 2020-02-20 | 2023-10-31 | 深圳市腾讯计算机系统有限公司 | Sample determination method and device of prediction model, medium and equipment |
| CN111428217B (en) * | 2020-04-12 | 2023-07-28 | 中信银行股份有限公司 | Fraudulent party identification method, apparatus, electronic device and computer readable storage medium |
| CN111612038B (en) * | 2020-04-24 | 2024-04-26 | 平安直通咨询有限公司上海分公司 | Abnormal user detection method and device, storage medium and electronic equipment |
| CN111641517A (en) * | 2020-04-29 | 2020-09-08 | 深圳壹账通智能科技有限公司 | Community division method and device for homogeneous network, computer equipment and storage medium |
| CN111581530A (en) * | 2020-05-07 | 2020-08-25 | 北京互金新融科技有限公司 | Community division method and device |
| CN111523831B (en) * | 2020-07-03 | 2020-11-03 | 支付宝(杭州)信息技术有限公司 | Risk group identification method and device, storage medium and computer equipment |
| CN111598713B (en) * | 2020-07-24 | 2021-12-14 | 北京淇瑀信息科技有限公司 | Cluster recognition method and device based on similarity weight updating and electronic equipment |
| CN112148767A (en) * | 2020-09-11 | 2020-12-29 | 支付宝(杭州)信息技术有限公司 | Group mining method, abnormal group identification method and device and electronic equipment |
| CN112184299A (en) * | 2020-09-23 | 2021-01-05 | 中国建设银行股份有限公司 | Arbitrage user identification method, apparatus, electronic device and medium |
| CN112910888A (en) * | 2021-01-29 | 2021-06-04 | 杭州迪普科技股份有限公司 | Illegal domain name registration group mining method and device |
| CN112785320B (en) * | 2021-02-01 | 2023-09-19 | 北京互金新融科技有限公司 | Credit risk determination method and device, storage medium and electronic equipment |
| CN113033966A (en) * | 2021-03-03 | 2021-06-25 | 携程旅游信息技术（上海）有限公司 | Risk target identification method and device, electronic equipment and storage medium |
| CN113326178A (en) * | 2021-06-22 | 2021-08-31 | 北京奇艺世纪科技有限公司 | Abnormal account number propagation method and device, electronic equipment and storage medium |
| CN113344697A (en) * | 2021-06-29 | 2021-09-03 | 平安普惠企业管理有限公司 | Group partner detection method, device, equipment and medium |
| CN114820219B (en) * | 2022-05-23 | 2022-09-20 | 杭银消费金融股份有限公司 | Complex network-based fraud community identification method and system |
| CN117575782B (en) * | 2024-01-15 | 2024-05-07 | 杭银消费金融股份有限公司 | Leiden community discovery algorithm-based group fraud identification method |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100185935A1 (en) * | 2009-01-21 | 2010-07-22 | Nec Laboratories America, Inc. | Systems and methods for community detection |
| CN103729475B (en) * | 2014-01-24 | 2016-10-26 | 福州大学 | Multi-tag in a kind of social networks propagates overlapping community discovery method |
| CN107392559A (en) * | 2016-05-16 | 2017-11-24 | 阿里巴巴集团控股有限公司 | Institutional framework determination methods and device |
| CN107292424B (en) * | 2017-06-01 | 2020-01-21 | 四川新网银行股份有限公司 | Anti-fraud and credit risk prediction method based on complex social network |
| CN107590504A (en) * | 2017-07-31 | 2018-01-16 | 阿里巴巴集团控股有限公司 | Abnormal main body recognition methods and device, server |
| CN107730262B (en) * | 2017-10-23 | 2021-09-24 | 创新先进技术有限公司 | Fraud identification method and device |
| CN107943879A (en) * | 2017-11-14 | 2018-04-20 | 上海维信荟智金融科技有限公司 | Fraud group detection method and system based on social networks |
-
2018
- 2018-04-26 CN CN201810382121.5A patent/CN108681936B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN108681936A (en) | 2018-10-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108681936B (en) | Fraud group identification method based on modularity and balanced label propagation | |
| Liang et al. | Data fusion approach for collaborative anomaly intrusion detection in blockchain-based systems | |
| WO2022121145A1 (en) | Ethereum phishing scam detection method and apparatus based on graph classification | |
| CN113240509B (en) | Loan risk assessment method based on multi-source data federal learning | |
| CN108492173A (en) | A kind of anti-Fraud Prediction method of credit card based on dual-mode network figure mining algorithm | |
| CN110413707A (en) | The excavation of clique's relationship is cheated in internet and checks method and its system | |
| CN109754258B (en) | Online transaction fraud detection method based on individual behavior modeling | |
| Xie et al. | A feature extraction method for credit card fraud detection | |
| CN111738817B (en) | Method and system for identifying risk community | |
| CN109284988A (en) | Data analysis system and method | |
| CN108764943B (en) | Suspicious user monitoring and analyzing method based on fund transaction network | |
| CN109886334A (en) | A kind of shared nearest neighbor density peak clustering method of secret protection | |
| CN113159866A (en) | Method for building network user trust evaluation model in big data environment | |
| CN110458094B (en) | Equipment classification method based on fingerprint similarity | |
| CN110445772B (en) | Internet host scanning method and system based on host relationship | |
| CN115859187A (en) | Object identification method and device, electronic equipment and storage medium | |
| Wang et al. | Local outlier detection based on information entropy weighting | |
| Valehi et al. | A graph matching algorithm for user authentication in data networks using image-based physical unclonable functions | |
| CN116843342A (en) | Financial fraud detection method, device, equipment and medium based on graph neural network | |
| CN108984773B (en) | Method and system for verifying blacklist multidimensional information under data missing condition, readable storage medium and device | |
| CN110889467A (en) | Company name matching method and device, terminal equipment and storage medium | |
| CN115965466A (en) | Sub-graph comparison-based Ethernet room account identity inference method and system | |
| Chu et al. | Exploiting spatial-temporal behavior patterns for fraud detection in telecom networks | |
| CN112069392B (en) | Method and device for preventing and controlling network-related crime, computer equipment and storage medium | |
| Petrov | Identification of unusual wallets on ethereum platform |
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 | ||
| CP01 | Change in the name or title of a patent holder |
Address after: Room ABCD, 17th floor, building D, Paradise Software Park, No.3 xidoumen Road, Xihu District, Hangzhou City, Zhejiang Province, 310012 Patentee after: Zhejiang Bangsheng Technology Co.,Ltd. Address before: Room ABCD, 17th floor, building D, Paradise Software Park, No.3 xidoumen Road, Xihu District, Hangzhou City, Zhejiang Province, 310012 Patentee before: ZHEJIANG BANGSUN TECHNOLOGY Co.,Ltd. |
|
| CP01 | Change in the name or title of a patent holder |