CN111489190A - Anti-cheating method and system based on user relationship - Google Patents
Anti-cheating method and system based on user relationship Download PDFInfo
- Publication number
- CN111489190A CN111489190A CN202010182551.XA CN202010182551A CN111489190A CN 111489190 A CN111489190 A CN 111489190A CN 202010182551 A CN202010182551 A CN 202010182551A CN 111489190 A CN111489190 A CN 111489190A
- Authority
- CN
- China
- Prior art keywords
- user relationship
- user
- graph
- chain
- normal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000010586 diagram Methods 0.000 claims abstract description 121
- 230000004044 response Effects 0.000 claims abstract description 5
- 238000004458 analytical method Methods 0.000 claims description 26
- 238000004364 calculation method Methods 0.000 claims description 16
- 238000013480 data collection Methods 0.000 claims description 7
- 230000008676 import Effects 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 8
- 230000000694 effects Effects 0.000 description 13
- 230000006399 behavior Effects 0.000 description 5
- 238000009434 installation Methods 0.000 description 5
- 230000014509 gene expression Effects 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 102100038367 Gremlin-1 Human genes 0.000 description 2
- 101001032872 Homo sapiens Gremlin-1 Proteins 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000004888 barrier function Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005242 forging Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0225—Avoiding frauds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Finance (AREA)
- Development Economics (AREA)
- General Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- Probability & Statistics with Applications (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Game Theory and Decision Science (AREA)
- Marketing (AREA)
- Economics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to an anti-cheating method and system based on user relationship, wherein the method comprises the following steps: collecting user relationship data reflecting the relationship among users in a preset time period; obtaining a user relationship graph based at least in part on the user relationship data; and determining that the user in the user relationship diagram is a cheating user in response to the difference between the user relationship diagram and the user relationship diagram in the normal form between one or more dimensional characteristics exceeding an allowable range. The method is based on the graph database, generates the user relationship graph according to the user relationship data, can accurately and effectively identify the cheating user by analyzing the user relationship graph, and has the advantages of simple and efficient process, simple system structure and good expandability.
Description
Technical Field
The invention relates to the technical field of internet application, in particular to an anti-cheating method and an anti-cheating system based on user relationship.
Background
With the rapid development of the mobile internet, various application programs (apps, abbreviated as applications) have appeared. Apps play an important role in various fields from eating and wearing to learning in daily life. Correspondingly, application developers actively develop new apps, improving their performance, to obtain more users. Typically, application developers also acquire users through some means of promotion. These promotional means include, for example, providing additional points or cash, etc. However, some users use the promotion means of the application developers to cheat the points or the red envelope by forging the users and the like. These users are called cheating users.
In some cases, the cheating user directly compromises the economic interests of the application developer. For example, many application developers have attracted users in the form of cash red packs, check-in red packs, bonus points, etc. in order to encourage users to use the services provided by the App. The cheating behavior of these cheating users to cheat the red packet or the score brings huge loss to the application developer. For another example, application developers use distribution channels to provide for installation downloads of apps and pay for the number of applications installed by the channels. In this scenario, many cheating installation behaviors occur, for example, cheating installation is performed by means of a farm with a flashing machine, and the like, which also brings huge losses to application developers.
In some cases, the behavior of the cheating user disturbs the analysis and judgment of the application developer to improve App performance using user data. For an online App, the application developer will continuously monitor its performance, the experience it brings to the user, in order to constantly improve App performance. The user data is the basis of monitoring, analysis and improvement, and the data of cheating users disturb the analysis process and create barriers for the improvement of App.
In some cases, the cheating user also distracts the data statistics and analysis of the third party platform. For example, in some App providing platforms, in order to attract users and rank apps provided by the apps, like an installation ranking list and a hit ranking list of similar applications, a user who downloads and installs apps in order to obtain red packages and scores provided by apps is not a real user of the apps, but the downloading and installation behaviors of the user are undoubtedly basic data of the ranking, so that the ranking of a third-party platform is not accurate, and the third-party platform needs to discriminate cheating users when performing data statistics and analysis in order to obtain accurate and as real ranking as possible, so that the difficulty of the data statistics and analysis is increased.
Thus, identifying the cheating user plays an important role in the benign functioning of the App. In the existing identification of some cheating users, operation data of the users are generally collected, cheating analysis dimensions and methods are set, and the operation data of the users are utilized to analyze in the preset analysis dimensions according to a preset analysis method, so that whether the users are the cheating users or not is determined. However, the user operation data used in the analysis by the method is various, such as the user ID, specific operation content, type, method, operation time, and the like, and thus, when setting the cheating analysis dimension and method, the user operation data needs to be respectively set according to the specific App type, the provided service data, and the like, which cannot give consideration to different App types, and the analysis process is too complicated and inefficient.
Disclosure of Invention
Aiming at the technical problems in the prior art, the invention provides an anti-cheating method and an anti-cheating system based on user relations, which are used for effectively identifying cheating users and reducing the complexity of identification.
In order to solve the above technical problem, according to an aspect of the present invention, the present invention provides a user relationship-based anti-cheating method, including the steps of: collecting user relationship data reflecting the relationship among users in a preset time period; obtaining a user relationship graph based at least in part on the user relationship data; and determining that the user in the user relationship diagram is a cheating user in response to the difference between the user relationship diagram and the user relationship diagram in the normal form between one or more dimensional characteristics exceeding an allowable range.
The relationships among the users comprise introduction user relationships, teacher-apprentice user relationships, sharing user relationships or rewarding user relationships. Wherein the user relationship graph comprises: nodes representing users and edges between nodes representing user relationships. Further, the edges between the nodes further include directions representing the level relationship between the users.
Preferably, in the aforementioned method, the one or more dimensional features comprise one or more of: graph, total number of nodes, total number of edges, chain level length, number of edges for a single node, and number of edges/number of nodes.
Preferably, the method further comprises: carrying out similarity calculation on the plurality of user relationship graphs; classifying the plurality of user relationship graphs based at least in part on the calculated similarity to obtain a plurality of user relationship graph sets; counting the number of the user relationship graphs in each user relationship graph set; and taking the user in the set with the largest number of the user relationship graphs as a normal user, and taking the users in the rest sets as suspected users.
Preferably, the method further includes extracting a user relationship diagram in a normal form from the user relationship diagram set of the normal user and determining the allowable range of the cheating user. And acquiring the allowable range according to the suspected user relationship diagram and/or the normal user relationship diagram.
Preferably, the step of obtaining the allowable range according to the suspected user relationship diagram and/or the normal user relationship diagram specifically includes: acquiring one or more typical dimension characteristics of a suspected user relationship diagram; obtaining typical dimension characteristics corresponding to a normal user relation graph; and acquiring the difference of the suspected user relationship diagram and one or more typical dimension characteristics of the user relationship diagram, wherein the difference is an allowable range.
In one embodiment, the user relationship graph in the normal form is one or more preset user relationship graphs determined according to historical data; the allowable range is a threshold range of one or more dimension characteristics determined when the user relationship graph of the normal form is determined according to historical data.
In another embodiment, the normal form user relationship graph is one or more user relationship graphs determined according to a preset user relationship calculation model; the allowable range is a threshold range of one or more dimension characteristics determined when the user relationship graph of the normal form is determined according to the user relationship calculation model.
Wherein the threshold range of dimensional features includes one or more of: a node total number difference threshold range, an edge total number difference threshold range, a chain level length difference threshold range, an edge number difference threshold range of a single node, and an edge number/node number difference threshold range.
In one embodiment, the user relationship graph has a pattern of a star graph, a chain graph or a star-chain graph.
Preferably, the normal form user relationship graph comprises one or more of the following patterns: the star-shaped graph, the star-shaped graph with the total number of nodes smaller than the threshold value of the total number of nodes, the chain graph with the chain level smaller than the threshold value of the chain level, and the star-chain graph with the chain level of the longest chain smaller than the threshold value of the chain level.
Preferably, the change in the chain level threshold is proportional to time.
Preferably, the dimension features for calculating the difference between the user relationship diagram and the user relationship diagram of the normal form comprise a user relationship diagram pattern and/or a chain level length, and the allowable range for determining the cheating user is a chain level length difference threshold range. Preferably, the variation of the chain step length difference range is proportional to time.
According to another aspect of the present invention, there is also provided a user relationship-based anti-cheating system, comprising a data collection module, a graph generation module, and an analysis module, wherein the data collection module is configured to collect user relationship data reflecting user relationships within a preset time period; the graph generation module is configured to obtain a user relationship graph based at least in part on user relationship data; the analysis module is configured to determine that a user in the user relationship graph is a cheating user in response to a difference between the user relationship graph and a normal-morphology user relationship graph between one or more dimensional features exceeding an allowable range.
Preferably, the graph generation module is further configured to generate the user relationship graph with the user as the node and the user relationship as the edge between the nodes.
Preferably, the one or more dimensional features comprise one or more of: graph, total number of nodes, total number of edges, chain level length, number of edges for a single node, and number of edges/number of nodes.
Preferably, the analysis module further comprises: a comparison data acquisition unit configured to acquire a user relationship diagram in a normal form and an allowable range for confirming a cheating user; the comparison unit is configured to compare the current user relationship graph with the user relationship graph in the normal form so as to obtain the difference of the current user relationship graph and the user relationship graph in the corresponding dimension characteristics; and the determining unit is configured to determine that the user in the user relationship diagram is a cheating user when the difference between the corresponding dimension characteristics of the current user relationship diagram and the user relationship diagram in the normal form exceeds the allowable range.
Preferably, the contrast data acquisition unit is further configured to include: the similarity calculation subunit is configured to perform similarity calculation on the plurality of user relationship graphs in the current preset time period; a classification subunit configured to classify the plurality of user relationship graphs based at least in part on the calculated similarities, and determine a set with a largest number of user relationship graphs as a normal user relationship graph set, and determine the remaining sets as suspected user relationship graph sets; the graph determining subunit is configured to extract the user relationship graph type from the normal user relationship graph set as a user relationship graph in a normal form; and the allowed range determining subunit is configured to acquire the allowed range for determining the cheating user according to the user relationship graph in the normal user relationship graph set and/or the user relationship graph in the suspected user relationship graph set.
Preferably, the allowable range determining subunit is further configured to obtain one or more typical dimension features of the normal user relationship diagram and the suspected user relationship diagram, respectively, and use a difference between the suspected normal user relationship diagram and the suspected user relationship diagram as an allowable range.
Preferably, the data collection module is further configured to collect user relationship data reflecting user relationships within a preset time period in a relationship database.
Preferably, the graph generation module is configured to include: a graph definition unit configured to define nodes of a graph and edges between the nodes; a graph generation unit configured to import user relationship data into a graph database, generate a user relationship graph according to defined nodes and edges between the nodes; and a query unit configured to query the user relationship graph in the graph database within a predetermined time period to determine a graph, a total number of nodes, a total number of edges, a chain level length, a number of edges of a single node, and a number of edges/nodes of each user relationship graph.
In a preferred embodiment, the analysis module is further configured to use the star map, the star map with the total number of nodes smaller than the threshold value of the total number of nodes, the chain map with the chain level length smaller than the chain level threshold value, and the star chain map with the chain level length of the longest chain smaller than the chain level threshold value as the user relationship map of the normal form.
In a preferred embodiment, the analysis module is configured to determine that the user in the user relationship graph is a cheating user when the difference between the chain level lengths of the user relationship graph and the user relationship graph in the normal form exceeds the chain level length difference threshold range by taking the graph and/or the chain level length of the user relationship graph as the dimension characteristic when comparing the user relationship graph with the user relationship graph in the normal form.
The method is based on the graph database, generates the user relation graph according to the relation between the users, can accurately and effectively identify the cheating users by analyzing the user relation graph, and has the advantages of simple and efficient process, simple system structure and good expandability.
Drawings
Preferred embodiments of the present invention will now be described in further detail with reference to the accompanying drawings, in which:
FIG. 1 is a flowchart of a user relationship-based anti-cheating method according to an embodiment of the present invention;
FIGS. 2A-2C are diagrams of a star-like user relationship based on a teacher-apprentice relationship, according to one embodiment of the invention;
FIGS. 3A-3C are diagrams of a catenated user relationship based on an introductory relationship, according to one embodiment of the present invention;
4A-4C are diagrams of a star-chain user relationship based on a reward relationship, according to one embodiment of the invention;
FIG. 5 is a flow diagram of a method for determining a normal profile of a user relationship graph from current data, according to one embodiment of the invention;
FIG. 6 is a flow diagram of a user invitation relationship based anti-cheating method according to one embodiment of the invention;
FIG. 7 is a presentation of a user relationship graph over a predetermined period of time derived from user relationship data, according to one embodiment of the invention;
FIG. 8 is a functional block diagram of a user relationship based anti-cheating system, according to one embodiment of the present invention;
FIG. 9 is a functional block diagram of a graph generation module according to one embodiment of the present invention;
FIG. 10 is a functional block diagram of an analysis module according to one embodiment of the present invention; and
FIG. 11 is a functional block diagram of a contrast data acquisition unit according to one embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the following detailed description, reference is made to the accompanying drawings that form a part hereof and in which is shown by way of illustration specific embodiments of the application. In the drawings, like numerals describe substantially similar components throughout the different views. Various specific embodiments of the present application are described in sufficient detail below to enable those skilled in the art to practice the teachings of the present application. It is to be understood that other embodiments may be utilized and structural, logical or electrical changes may be made to the embodiments of the present application.
For an App providing a certain service, the amount of users will increase with time and promotion activities after the App is online. And after the user downloads and installs the App, the personal information of the user is stored in a user database by the App server. The personal information of the user comprises user registration information, such as a user name, a user ID, a terminal ID, a mailbox address and the like. The user personal information also includes activity participation information, such as an activity to participate when becoming a user, an activity to participate after becoming a user. The activities refer to certain activities initiated by the App, such as introducing a user to earn points, sharing App advertisements to earn points, and accumulating points to obtain rewards. The user personal information also includes the relationship between the user and other users, for example, the introduction and the introduced relationship, the teacher-apprentice relationship with other users, the sharing relationship with other users, and the like. The present invention determines the cheating users by analyzing these user relationships.
Fig. 1 is a flowchart of a user relationship-based anti-cheating method according to an embodiment of the present invention. The method comprises the following steps:
in step S1, user relationship data reflecting the relationship between users within a preset time period is collected. For example, user relationship data registered within a day, such as an introductory relationship between users, e.g., user B is a user introduced by user a, is collected. And for example, a teacher-apprentice user relationship, such as the fact that the user B is the apprentice of the user A. Two users with a teacher-apprentice user relationship have a certain reward distribution relationship in some activities with a reward mechanism, and thus the reward distribution relationship is also extended. For example, the method includes sharing a user relationship, for example, the user a shares a link to the user B, where the link is a content link such as an advertisement or a commodity. Also for example, the relationship between users is a reward user relationship, such as in some activities where user B receives a reward that is partially assigned to user A. The user relationship described in the present invention includes not only the foregoing various specific examples, but also other user relationships that are not illustrated.
At step S2, a user relationship graph is obtained based at least in part on the user relationship data. As shown in fig. 2-4. In fig. 2A-2C, according to the apprentice mechanism, the user a has respectively acquired 5 apprentices, so as to obtain users B1-B5, the relationship is as shown in fig. 2A, according to the relationship between the user a and the users B1-B5 in fig. 2A, the user is taken as a node, the relationship is taken as an edge, so as to obtain fig. 2B, and fig. 2C is obtained by simplification. The node attribute includes user information, and the edge attribute includes a direction of a level relationship between users, for example, each edge in fig. 2C includes a master relationship from a user a to a user B, that is, the user a is a master of the user B. In fig. 3A-3C, according to the introduction relationship, the user a introduces the user B, the user B introduces the user C, and so on, and with the user as a node and the introduction relationship as an edge, fig. 3B is generated, and fig. 3C can be obtained by simplification. As shown in fig. 4A-4C, according to the bonus mechanism, the user a and the users B1-B5 are in bonus level 1 relationship, the user B3 and the user B31 are in bonus level 2 relationship, the user B31 and the user B32 are in bonus level 3 relationship, according to the user relationship of fig. 4A, the users are used as nodes, and the relationship between the users is used as a relationship, so that fig. 4B is generated, and fig. 4C can be obtained by simplification. A plurality of user relationship maps within a predetermined period of time are obtained by step S2. Whether the users in the user relationship graph are cheating users is identified one by one.
In step S3, a user relationship diagram is obtained.
And step S4, comparing the difference between the user relationship graph and the user relationship graph in the normal form in one or more dimension characteristics.
And step S5, judging whether the difference is in an allowable range, if so, determining that the user in the user relationship diagram is a normal user in step S6. If the difference is outside the allowable range, it is determined in step S7 that the user in the user relationship diagram is a cheating user.
In step S8, it is determined whether there is any user relationship graph that has not been compared, and if so, the process returns to step S3. If all the user relationship graphs are judged to be finished, the flow is ended.
The feature described in step S4 may be a dimension, and for example, the feature may be any one of a graph, a total number of nodes, a total number of edges, a chain level length, a number of edges of a single node, or a number of edges/nodes of a user relationship graph. The features described may also be of multiple dimensions, as may combinations of features previously described.
In one embodiment, the normal form user relationship graph is one or more user relationship graphs determined according to historical data. When the user relation graph of the normal form is determined, the threshold range of each dimension characteristic is also determined. For example, by analyzing the user relationship graph generated from the historical user relationship data, the user relationship graph generated from the user relationship of the normal user is determined as the user relationship graph of the normal form for comparison. For example, the user relationship graph of the normal form may be a star graph/star chain graph with the total number of nodes or edges being 3-10, or a chain graph with the chain level length being less than 10. This is because the normal number of users having correlation with each other is not too large at the time of the predetermined time period, and if the number of users in a certain user relationship diagram increases abruptly within the predetermined time period, the possibility that the users in the user relationship diagram are cheating users is high.
In another embodiment, a user relationship calculation model is adopted to determine a user relationship graph in a normal form and a threshold range of each dimension characteristic of the user relationship graph. And training the model by utilizing a certain algorithm, such as various classification algorithms, according to the known normal user relationship data and the cheating user relationship data. And obtaining a user relation graph in a normal form and a corresponding threshold range of the dimension characteristics through the model.
In another embodiment, as illustrated in fig. 5, the user relationship graph of the normal form is determined according to the current data, and the allowable range for determining the cheating user is obtained. The method specifically comprises the following steps:
in step S41, similarity calculation is performed on the plurality of generated user relationship diagrams. When the similarity is calculated, the similarity is calculated by taking any one of the user relationship graph patterns, the total number of nodes, the total number of edges, the chain level length, the number of edges of a single node or the number of edges/nodes as a feature, for example, when the total number of nodes is taken as the feature, different total numbers of nodes are respectively set as different clustering centers, then the total number of nodes of each user relationship graph is counted, and the distance between each user relationship graph and the different clustering centers is calculated.
And step S42, classifying the multiple user relationship graphs based on the similarity to obtain multiple user relationship graph sets. Continuing with the example in step S41, the distance to the cluster center is classified according to the user relationship graph, and the cluster center having the shortest distance to the user relationship graph is classified. Thereby dividing all the user relationship graphs into a plurality of sets.
Step S43, count the number of user relationship graphs in each set.
Step S44, the user in the user relationship graph set with the largest number of user relationship graphs is taken as the normal user, and the users in the remaining user relationship graph sets are suspected users.
In step S45, the user relationship diagram with the largest number of user relationship diagrams is extracted as the user relationship diagram in the normal form. The extracted user relation graph comprises relation data such as a graph, the number of nodes, the number of edges, chain-level length and the like.
In step S46, an allowable range for distinguishing the user relationship diagram of the normal form from the cheating user relationship diagram is determined. The allowed range may be a characteristic threshold range, such as a total number of nodes value threshold range, a chain level length threshold range, a single node edge number threshold range edge number/node number threshold range, and so on. The threshold range may be determined according to the user relationship graph in the normal form, for example, when the threshold range of the total node value is determined, the difference value is used as the threshold range of the total node value according to the minimum total node number and the maximum total node number in the user relationship graph in the normal form.
The allowable range may also be a node total number difference threshold range, an edge total number difference threshold range, a chain level length difference threshold range, an edge number difference threshold range of a single node, and an edge number/node number difference threshold range, which are determined by the normal-form user relationship diagram and the suspected user relationship diagram. For example, one or more corresponding typical dimension features of the normal user relationship diagram and the suspected user relationship diagram are obtained respectively. The typical dimension feature is a feature that can represent a feature of the user relationship graph most, for example, for the user relationship graph shown in fig. 2C, the user relationship graph is star-shaped, and the total number of nodes or edges is the typical dimension feature thereof, and for the user relationship graph shown in fig. 3C, the user relationship graph is chain-shaped, and the total number of levels is the typical dimension feature thereof; for the user relationship diagram shown in fig. 4C, which includes a star shape and a chain shape, it needs to determine whether there are more users forming the star shape or the chain shape, and the shape with the most users is taken as its typical shape, as shown in fig. 4C, if the star user relationship diagram is taken as its typical shape, the total number of nodes is the typical dimension characteristic of the user relationship diagram. Since there are multiple user relationship graphs in the user relationship graph set in the normal form, as a specific embodiment, an average value of typical dimension features of all user relationship graphs is taken as a typical dimension feature value of the user relationship graph in the normal form. And similarly, obtaining the typical dimension characteristics of the suspected user relationship graph. And comparing the difference between the two, such as the difference between the total number of nodes, wherein the difference is the threshold range for judging the cheating user, namely the allowable range.
FIG. 6 is a flowchart of a method for anti-cheating based on user-invitation relationships, according to one embodiment of the invention. In this embodiment, in order to attract users, the App launches an activity in which a user can obtain a reward by inviting, in which the user can send an invitation to his relatives, friends, etc. by sending a link, a page, a mail, etc., and when the invited person becomes the App user, both can receive corresponding rewards. The reward is related to the service provided by the App, for example, when the App is applied to reading class novels, the reward can be paid reading data for providing extra time for free, and can also be reading points, reading grades and the like. In order to identify whether the currently added user is a cheating user, the present embodiment provides the following method:
step S100, collecting user relation data reflecting the relation between users in a preset time period. In this embodiment, the App stores the user relationship data in a relationship database, as shown in the following table:
the user data in the relational database includes the ID of the user who sent the invitation and the ID of the user who was invited. In addition, other information is also included, such as registration time of the user who sends the invitation, registration time of the user who is invited and the like. The preset time period may be the entire activity period, for example, when the activity is 3 days, the preset time period is the 3 days. The preset time period may also be divided into different time periods, for example, when the activity is 7 days, the identification is performed three times, such as setting the first 2 days as a time period, the middle 2 days as a time period, and the last 1 day as a time period.
Step S101, defining nodes and edges, and importing user relationship data into a graph database. In which a user ID is defined as a node and an invitation relation is defined as an edge, and attributes of the node and the edge may be defined, for example, user information such as a user ID, an activity ID, etc. are defined as attributes of the node and an invitation relation is defined as attributes of the edge.
And S102, obtaining a user relation graph by inquiring various relations of the obtained nodes and edges. The user relationship data imported in the preset time period can be queried through a Gremlin graph query language to obtain the total number of nodes, the total number of edges, the chain level length, the number of edges of a single node, the number of edges/nodes and the like, so that the user relationship graph is obtained. As shown in the display view of fig. 7. The user relationship diagram 71 in fig. 7 is a typical star diagram, the user relationship diagram 72 is a typical chain diagram, and the user relationship diagram 73 is a typical star chain diagram.
And step S103, analyzing the plurality of user relationship graphs and determining cheating users. In one embodiment, according to historical data, a star chart is used as a user relation chart in a normal form, a chain chart with a chain level length smaller than a + bx level is used as a user relation chart in a normal form, and for the star chain chart including both the star chart and the chain chart, the star chain chart with a chain level length of the longest chain smaller than a + bx is used as a user relation chart in a normal form. Wherein the chain level refers to an edge connecting two nodes, as shown in fig. 3C. x is a time parameter, such as 1 day, 2 days, etc., which corresponds to a preset period of time for collecting data. a and b are empirical values, which are different according to different types and operation situations of apps, for example, according to historical data, in a 1-day time period, a user in the user relationship diagram with a chain level threshold of 5 is a normal user, in a 2-day time period, a user in the user relationship diagram with a chain level threshold of 7 is a normal user, and verification is performed through a large amount of historical data, so that the following results can be obtained: the relation when a is 5, x is N-1, N is a time unit, and b is 2 can determine the user relation graph of the normal user. Correspondingly, the allowable range is also set to a + bx, i.e., the allowable range becomes larger as the period of time increases. The relational expression which is proportional to time and is adopted when the chain-level length threshold value and the allowable range are determined is obtained only according to one App, different Apps have different rules and correspond to different relational expressions, and a person skilled in the art can obtain the relational expression which accords with the rules through historical data analysis.
In step S102, the user relationship maps for 1 day obtained in step S101 are compared with the user relationship map of the normal form described above. For example, if the preset time period is set to 1 day, the chain level threshold in this embodiment is 5, and correspondingly, the allowable range is 5. If the user relationship diagram is a star diagram like the user relationship diagram 71, it indicates that there is no difference between the user relationship diagram and the normal user relationship diagram, and in step S1031a, it is determined that the user in the user relationship diagram is a normal user. If the user relationship diagram is a chain diagram, in step S1031b, comparing the chain level with the chain level threshold value 5, in step S1032b, determining whether the chain level of the user relationship diagram is greater than or equal to the chain level threshold value 5, and if the chain level is less than 5 as in the user relationship diagram 72, in step S1033b, determining that the user in the user relationship diagram is a normal user; if the chain level is greater than 5 levels, it is determined as the suspected user relationship diagram in step S1034b, and it is determined in step S1035b whether the difference between the chain level of the suspected user relationship diagram and the chain level threshold 5 exceeds the allowable range 5, that is, whether the chain level of the suspected user relationship diagram exceeds 10, and if so, it is determined in step S1036b that the user in the suspected user relationship diagram is a cheating user. If the difference between the chain level of the suspected user relationship diagram and the chain level threshold 5 is smaller than the allowable range 5, in step S107b, it is determined that the user in the user relationship diagram is a normal user, or the user in the user relationship diagram is marked as a suspected user, and the suspected user is left to a subsequent processing flow, where the subsequent processing flow includes further examining the behavior data of the user in the user relationship diagram, such as browsing data, network data, etc., within a set examination period (e.g., 3 days, 1 week) to determine whether the user is a cheating user. Since it is not a content of the present invention, it will not be described herein again.
If the user relationship diagram is a star-chain diagram, that is, includes both a star-chain diagram and a chain diagram, as in the user relationship diagrams 73, 74 and 75 in fig. 7, the step S101c determines the chain level of the longest chain, and then the process goes to step S101b, and the subsequent process flow is the same as the chain diagram. The longest chain of the user relationship graph 73 in fig. 7 is 3 levels, which is less than the chain level threshold, and thus the user relationship graph 73 is determined to be a normal user relationship graph. Whereas the user relationship graph 74 of fig. 7, with the longest chain having a chain level length of 22 levels, differs from the chain level threshold 5 by 17, well beyond the allowed range 5. Thus, the user in the user relationship diagram 74 can be determined to be a cheating user, and similarly, the user in the user relationship diagram 75 can be determined to be a cheating user.
In order to improve the discretion and discretion of the judgment, when determining the cheating user, a plurality of thresholds may be set for determining the normal user, the suspected user, and the cheating user. In the foregoing embodiment, the allowable range may be subdivided into two segments, for example, the resetting allowable range is set to 5 and 10, if the difference between the chain level length of the longest chain in the user relationship diagram and the chain level threshold 5 is between 5 and 10, the user may be considered as a suspected user, and the user is left in the examination period and determined by combining with other data. And if the difference value of the chain level length of the longest chain of the user relationship graph and the chain level threshold value 5 exceeds 10, determining the user as a cheating user.
Fig. 8 is a schematic diagram of a user relationship-based anti-cheating system according to an embodiment of the present invention, the system including a data collection module 1, a graph generation module 2, and an analysis module 3. The data collection module 1 is configured to collect user relationship data reflecting a relationship between users in a preset time period. For example, the user data in the preset time period is found from the relational database, and the relationships among the users, such as introduction relationships, teacher-apprentice relationships, sharing relationships, and the like, are determined.
The graph generation module 2 is configured to obtain a user relationship graph based at least in part on the user relationship data. For example, defining nodes and edges in the graph, defining node attributes and edge attributes, importing the user relationship data into a graph database, and obtaining each user relationship graph through query. Specifically, the graph generating module 2 is shown in fig. 9 and includes a graph defining unit 20, a data importing unit 21, and a querying unit 22, where the graph defining unit 20 is configured to define nodes of a graph and edges between the nodes, for example, a user ID is used as a node, a relationship between users is used as an edge, and specific attributes of the user ID can also be defined in the edges of the nodes. The data import unit 21 imports user relationship data and definition data into a graph database. The query unit 22 queries the user relationship data in the graph database within a predetermined time period using graph query languages such as Gremlin, so as to obtain the user relationship graph according to the relationships such as the total number of nodes, the total number of edges, the length of chain levels, the number of edges of a single node, and the number of edges/nodes obtained by query. In one embodiment, the resulting user relationship graph is shown in FIG. 7.
The analysis module 3 is configured to analyze the user relationship graph, and determine that a user in the user relationship graph is a cheating user when a difference between one or more dimensional features of the user relationship graph and the user relationship graph in the normal form exceeds an allowable range. Wherein the dimensional features include one or more of: graph, total number of nodes, total number of edges, chain level length, number of edges for a single node, and number of edges/number of nodes.
Specifically, as shown in fig. 10, a schematic diagram of an analysis module according to an embodiment of the present invention is shown. The analysis module 3 includes a comparison data acquisition unit 31, a comparison unit 32, and a determination unit 33. The comparison data acquiring unit 31 is used for acquiring a normal user relationship diagram and an allowable range for confirming a cheating user. For example, when the user relationship is a teacher-apprentice relationship, a star map or a star map in which the total number of nodes is less than a certain threshold may be used as the normal user relationship map, and the corresponding allowable range may be a node total number difference threshold. Or, a chain graph with a chain level length smaller than the chain level threshold value is used as the user relationship graph in the normal form, and the corresponding allowable range may be a preset chain level length difference threshold value.
In another embodiment, the normal user relationship graph and the allowable range are determined by a user relationship calculation model obtained through pre-training. For example, historical user relationship data of the App is transmitted to the model, and a user relationship graph of a normal form and an allowable range thereof are obtained.
In another embodiment, a normal user relationship graph is determined according to all current user relationship graphs. Specifically, as shown in fig. 11, a schematic block diagram of a comparison data obtaining unit according to an embodiment of the present invention is shown. The comparison data acquisition unit 31 includes a similarity degree operator unit 310, a classification subunit 311, a map determination subunit 312, and an allowable range determination subunit 313. The similarity calculation subunit 310 is configured to perform similarity calculation on the multiple user relationship graphs in the current predetermined time period. For example, by using any one of the graph, the total number of nodes, the total number of edges, the total number of levels, the number of edges of a single node, or the number of edges/nodes as a feature, a corresponding cluster center is set, and the distance from each user relationship graph to the cluster center is calculated. The classification subunit 311 classifies the plurality of user relationship graphs based on the calculated similarity, and determines the set with the largest number of user relationship graphs as a normal user relationship graph set, and the remaining sets are suspected user relationship graph sets; the graph determining subunit 312 is configured to extract the user relationship graph type from the normal user relationship graph set as the user relationship graph in the normal form. The allowed range determining subunit 313 is configured to obtain an allowed range for determining the cheating user according to the user relationship diagram in the normal user relationship diagram set and/or the user relationship diagram in the suspected user relationship diagram set. For example, one or more typical dimension features of the normal user relationship diagram and the suspected user relationship diagram are respectively obtained, and the difference between the suspected user relationship diagram and the one or more typical dimension features of the user relationship diagram is used as the allowable range.
In one embodiment, a star chart is used as a user relation chart of a normal form, a chain chart with a chain level length smaller than a + bx is used as a user relation chart of a normal form, and for the star chain chart including both the star chart and the chain chart, a star chain chart with a chain level length of the longest chain smaller than a + bx is used as a user relation chart of a normal form, wherein x is a time parameter, such as 1 day, 2 days and the like, which corresponds to a preset time period for collecting data. a and b are empirical values. In one embodiment, a is 5, x is N-1; n is the number of days; b is 2.
The comparing unit 32 is configured to compare the current user relationship diagram with the normal user relationship diagram to obtain a difference between the two in the corresponding dimension characteristics. The dimension characteristic difference comprises any one or any several of the difference of the graph, the difference of the total number of nodes, the difference of the total number of edges, the difference of chain level length, the difference of the number of single-node edges or the number of edges/node number. When comparing multiple dimensional features, the order of comparison may be set. For example, the difference of the patterns is compared, if the star map with the total number of nodes less than 20 is the user relationship map with the normal form, the comparison of the total number of nodes/the total number of edges is performed when the current user relationship map is different from the user relationship map with the normal form.
The determining unit 33 is configured to determine according to a difference between the corresponding dimension characteristics of the current user relationship diagram and the normal user relationship diagram and the corresponding allowable range. For example, the detailed judgment is performed according to a preset or calculated user relationship diagram in a normal form and the allowable difference range of the corresponding dimension characteristics. In one embodiment, the star map is a normal user relationship map, the chain map with the chain level less than 5 is a normal user relationship map, the star map with the chain level length less than 5 of the longest chain is a normal user relationship map, and the cheating user relationship map is obtained when the difference between the chain level length of the longest chain of the corresponding user relationship map and the chain level threshold is greater than or equal to 5. When the comparing unit 32 compares the user relationship graph 74 in fig. 7 with the relationship graph of the normal form, the obtained difference is the graph difference, that is, the user relationship graph 74 includes a chain graph in addition to a star graph, and then compares the difference between the chain level length and the chain level threshold 5, and the difference between the longest chain level length and the chain level threshold 5 is 17, which exceeds the allowable range 5. Thus, the determination unit 33 determines that the user in the user relationship diagram 74 is a cheating user, and similarly, the user in the user relationship diagram 75 is a cheating user, and the user in the user relationship diagram 73 is a normal user.
The invention utilizes the structuralization and high-efficiency data modeling and query capability of the graph database to the data, can accurately and effectively identify cheating users by analyzing the difference between the user relationship graph generated according to the user relationship data and the user relationship graph with the normal form, and has simple and high-efficiency process, simple system structure and good expansibility.
The above embodiments are provided only for illustrating the present invention and not for limiting the present invention, and those skilled in the art can make various changes and modifications without departing from the scope of the present invention, and therefore, all equivalent technical solutions should fall within the scope of the present invention.
Claims (29)
1. An anti-cheating method based on user relationships, comprising:
collecting user relationship data reflecting the relationship among users in a preset time period;
obtaining a user relationship graph based at least in part on the user relationship data; and
and determining that the user in the user relationship diagram is a cheating user in response to the difference between the user relationship diagram and the user relationship diagram in the normal form between one or more dimensional characteristics exceeding an allowable range.
2. The method of claim 1, wherein the user relationship comprises an introduction user relationship, a teacher-apprentice user relationship, a sharing user relationship, or a rewarding user relationship.
3. The method of claim 1, wherein the user relationship graph comprises:
a node representing a user; and
edges between nodes representing user relationships.
4. The method of claim 3, wherein the edges between the nodes further comprise directions representing a level relationship between users.
5. The method of claim 3, wherein the one or more dimensional features comprise one or more of: graph, total number of nodes, total number of edges, chain level length, number of edges for a single node, and number of edges/number of nodes.
6. The method of claim 1, further comprising:
carrying out similarity calculation on the plurality of user relationship graphs;
classifying the plurality of user relationship graphs based at least in part on the calculated similarity to obtain a plurality of user relationship graph sets;
counting the number of the user relationship graphs in each user relationship graph set; and
and taking the user in the set with the largest number of the user relationship graphs as a normal user, and taking the users in the rest sets as suspected users.
7. The method of claim 6, further comprising extracting a normal form of user relationship graph and the allowable range for determining cheating users from the set of user relationship graphs of normal users.
8. The method of claim 7, further comprising: and acquiring the allowable range according to the suspected user relationship diagram and/or the normal user relationship diagram.
9. The method of claim 8, further comprising:
acquiring one or more typical dimension characteristics of a suspected user relationship diagram;
obtaining typical dimension characteristics corresponding to a user relation graph in a normal form; and
and acquiring the difference of one or more typical dimension characteristics of the suspected user relationship diagram and the normal-form user relationship diagram, wherein the difference is an allowable range.
10. The method according to claim 1, wherein the normal form user relationship graph is one or more preset user relationship graphs determined according to historical data; the allowable range is a threshold range of one or more dimension characteristics determined when the user relationship graph of the normal form is determined according to historical data.
11. The method according to claim 1, wherein the normal form user relationship graph is one or more user relationship graphs determined according to a preset user relationship calculation model; the allowable range is a threshold range of one or more dimension characteristics determined when the user relationship graph of the normal form is determined according to the user relationship calculation model.
12. The method of claim 10 or 11, wherein the threshold range of dimensional features comprises one or more of:
a node total number difference threshold range, an edge total number difference threshold range, a chain level length difference threshold range, an edge number difference threshold range of a single node, and an edge number/node number difference threshold range.
13. The method of claim 3, wherein the pattern of the user relationship graph is a star graph, a chain graph, or a star-chain graph.
14. The method of claim 13, wherein the normal-form user relationship graph comprises one or more of the following: the star-shaped graph comprises a star-shaped graph, a star-shaped graph with the total number of nodes smaller than a threshold value of the total number of nodes, a chain graph with the chain level length smaller than a chain level threshold value and a star-chain graph with the chain level length of the longest chain smaller than the chain level threshold value.
15. The method of claim 14, wherein the change in the chain level threshold is proportional to time.
16. The method of claim 15, wherein the dimensional features used to calculate the difference between the user relationship graph and the normal-morphology user relationship graph comprise a user relationship graph pattern and/or a chain-level length, the allowable range for determining cheating users being a chain-level length difference threshold range.
17. The method of claim 16, wherein the range of chain level length difference threshold values varies in direct proportion to time.
18. A user relationship based anti-cheating system, comprising:
a data collection module configured to collect user relationship data reflecting a user relationship within a preset time period;
a graph generation module configured to obtain a user relationship graph based at least in part on the user relationship data; and
an analysis module configured to determine that a user in the user relationship graph is a cheating user in response to a difference between one or more dimensional features of the user relationship graph and a normal-morphology user relationship graph exceeding an allowable range.
19. The system of claim 18, the graph generation module further configured to generate a user relationship graph with users as nodes and user relationships as edges between nodes.
20. The system of claim 19, the one or more dimensional features comprising one or more of: graph, total number of nodes, total number of edges, chain level length, number of edges for a single node, and number of edges/number of nodes.
21. The system of claim 18, the analysis module further comprising:
a comparison data acquisition unit configured to acquire a user relationship diagram in a normal form and an allowable range for confirming a cheating user;
the comparison unit is configured to compare the current user relationship graph with the user relationship graph in the normal form so as to obtain the difference of the current user relationship graph and the user relationship graph in the corresponding dimension characteristics; and
the determining unit is configured to determine that the user in the user relationship diagram is a cheating user when the difference between the corresponding dimension characteristics of the current user relationship diagram and the user relationship diagram in the normal form exceeds an allowable range.
22. The system of claim 21, wherein the normal form user relationship graph is one or more preset user relationship graphs determined according to historical data; the allowable range is a threshold range of one or more dimension characteristics determined when the user relationship graph of the normal form is determined according to historical data.
23. The system of claim 21, wherein the normal form user relationship graph is a user relationship graph determined according to a preset user relationship estimation model; the allowable range is a threshold range of one or more dimension characteristics determined when the user relationship graph of the normal form is determined according to the user relationship calculation model.
24. The system of claim 21, the contrast data acquisition unit further configured to comprise:
the similarity calculation subunit is configured to perform similarity calculation on the plurality of user relationship graphs in the current preset time period;
a classification subunit configured to classify the plurality of user relationship graphs based at least in part on the calculated similarities, and determine a set with a largest number of user relationship graphs as a normal user relationship graph set, and determine the remaining sets as suspected user relationship graph sets;
the graph determining subunit is configured to extract the user relationship graph type from the normal user relationship graph set as a user relationship graph in a normal form; and
and the allowed range determining subunit is configured to acquire an allowed range for determining the cheating user according to the user relationship graph in the normal user relationship graph set and/or the user relationship graph in the suspected user relationship graph set.
25. The system of claim 24, wherein the allowed range determining subunit is further configured to obtain one or more typical dimension features of the normal user relationship graph and the suspected user relationship graph, respectively, and use a difference between the suspected normal user relationship graph and the suspected user relationship graph as an allowed range.
26. The system of claim 18, the data collection module further configured to collect user relationship data reflecting user relationships over a preset time period in a relational database.
27. The system of claim 26, the graph generation module configured to comprise:
a graph definition unit configured to define nodes of a graph and edges between the nodes;
a data import unit configured to import user relationship data and definition data into a graph database; and
the query unit is configured to query the user relationship in the graph database within a preset time period, and obtain a user relationship graph according to one or more of the following relationships obtained by query: total number of nodes, total number of edges, chain level length, number of edges of a single node, and number of edges/number of nodes.
28. The system of claim 26, the analysis module further configured to treat the star map, the star map with the total number of nodes less than the threshold value of the total number of nodes, the chain map with the chain level length less than the threshold value of the chain level, and the star map with the chain level length of the longest chain less than the threshold value of the chain level as the user relationship map of the normal morphology.
29. The system of claim 28, the analysis module configured to determine a user in a user relationship graph as a cheating user when a difference in chain level lengths of the user relationship graph and a normal form user relationship graph exceeds a chain level length difference threshold range, with a pattern and/or chain level length of the user relationship graph as dimensional features when comparing the user relationship graph to the normal form user relationship graph.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010182551.XA CN111489190A (en) | 2020-03-16 | 2020-03-16 | Anti-cheating method and system based on user relationship |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010182551.XA CN111489190A (en) | 2020-03-16 | 2020-03-16 | Anti-cheating method and system based on user relationship |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111489190A true CN111489190A (en) | 2020-08-04 |
Family
ID=71811695
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010182551.XA Pending CN111489190A (en) | 2020-03-16 | 2020-03-16 | Anti-cheating method and system based on user relationship |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111489190A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113205129A (en) * | 2021-04-28 | 2021-08-03 | 五八有限公司 | Cheating group identification method and device, electronic equipment and storage medium |
| CN114553738A (en) * | 2022-02-25 | 2022-05-27 | 支付宝实验室(新加坡)有限公司 | Abnormality detection method and apparatus |
| CN114596111A (en) * | 2022-03-03 | 2022-06-07 | 浙江吉利控股集团有限公司 | Risk identification model generation method, device, equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106485559A (en) * | 2015-08-19 | 2017-03-08 | 阿里巴巴集团控股有限公司 | Cheating recognition methods and device for on-line shop |
| CN108615119A (en) * | 2018-05-09 | 2018-10-02 | 平安普惠企业管理有限公司 | A kind of recognition methods of abnormal user and equipment |
| CN110032583A (en) * | 2019-03-12 | 2019-07-19 | 平安科技（深圳）有限公司 | A kind of recognition methods of fraud clique, device, readable storage medium storing program for executing and terminal device |
| CN110046929A (en) * | 2019-03-12 | 2019-07-23 | 平安科技（深圳）有限公司 | A kind of recognition methods of fraud clique, device, readable storage medium storing program for executing and terminal device |
| CN110177094A (en) * | 2019-05-22 | 2019-08-27 | 武汉斗鱼网络科技有限公司 | A kind of user community recognition methods, device, electronic equipment and storage medium |
| CN110807129A (en) * | 2019-09-29 | 2020-02-18 | 北京淇瑀信息科技有限公司 | Method and device for generating multilayer user relationship graph set and electronic equipment |
-
2020
- 2020-03-16 CN CN202010182551.XA patent/CN111489190A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106485559A (en) * | 2015-08-19 | 2017-03-08 | 阿里巴巴集团控股有限公司 | Cheating recognition methods and device for on-line shop |
| CN108615119A (en) * | 2018-05-09 | 2018-10-02 | 平安普惠企业管理有限公司 | A kind of recognition methods of abnormal user and equipment |
| CN110032583A (en) * | 2019-03-12 | 2019-07-19 | 平安科技（深圳）有限公司 | A kind of recognition methods of fraud clique, device, readable storage medium storing program for executing and terminal device |
| CN110046929A (en) * | 2019-03-12 | 2019-07-23 | 平安科技（深圳）有限公司 | A kind of recognition methods of fraud clique, device, readable storage medium storing program for executing and terminal device |
| CN110177094A (en) * | 2019-05-22 | 2019-08-27 | 武汉斗鱼网络科技有限公司 | A kind of user community recognition methods, device, electronic equipment and storage medium |
| CN110807129A (en) * | 2019-09-29 | 2020-02-18 | 北京淇瑀信息科技有限公司 | Method and device for generating multilayer user relationship graph set and electronic equipment |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113205129A (en) * | 2021-04-28 | 2021-08-03 | 五八有限公司 | Cheating group identification method and device, electronic equipment and storage medium |
| CN113205129B (en) * | 2021-04-28 | 2023-04-07 | 五八有限公司 | Cheating group identification method and device, electronic equipment and storage medium |
| CN114553738A (en) * | 2022-02-25 | 2022-05-27 | 支付宝实验室(新加坡)有限公司 | Abnormality detection method and apparatus |
| CN114596111A (en) * | 2022-03-03 | 2022-06-07 | 浙江吉利控股集团有限公司 | Risk identification model generation method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9245252B2 (en) | Method and system for determining on-line influence in social media | |
| CN111489190A (en) | Anti-cheating method and system based on user relationship | |
| US20190355058A1 (en) | Method and apparatus for processing credit score real-time adjustment, and processing server | |
| CN108363821A (en) | A kind of information-pushing method, device, terminal device and storage medium | |
| CN103345512A (en) | Online advertising click-through rate forecasting method and device based on user attribute | |
| CN108304853B (en) | Game correlation obtaining method and device, storage medium and electronic device | |
| CN105608179A (en) | Method and device for determining relevance of user identification | |
| CN109978033A (en) | The method and apparatus of the building of biconditional operation people's identification model and biconditional operation people identification | |
| WO2005104436A1 (en) | Method and system for generating a population representative of a set of users of a communication network | |
| CN108416684B (en) | Method and device for evaluating credibility of account main body and server | |
| CN111522724B (en) | Method and device for determining abnormal account number, server and storage medium | |
| CN109858965A (en) | A kind of user identification method and system | |
| CN111882420A (en) | Generation method of response rate, marketing method, model training method and device | |
| CN109949089A (en) | A kind of method, apparatus and terminal of determining displaying rate | |
| CN111986027A (en) | Abnormal transaction processing method and device based on artificial intelligence | |
| CN114254867A (en) | Telecommunication fraud victim risk assessment system and method | |
| CN110213209B (en) | Cheating detection method and device for push information click and storage medium | |
| CN111914645A (en) | Method and device for identifying false information, electronic equipment and storage medium | |
| CN111340540A (en) | Monitoring method, recommendation method and device of advertisement recommendation model | |
| CN111159541A (en) | Method and device for determining account behavior preference | |
| CN114245185A (en) | Video recommendation method, model training method, device, electronic equipment and medium | |
| CN113468394A (en) | Data processing method and device, electronic equipment and storage medium | |
| CN112819499A (en) | Information transmission method, information transmission device, server and storage medium | |
| JP2004094383A (en) | Recommendation device and advertisement distribution method | |
| CN113743619A (en) | Cheating user identification method and device based on associated network behaviors |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200804 |
|
| WD01 | Invention patent application deemed withdrawn after publication |