CN113742538B - Service analysis method and device based on graph level, electronic equipment and storage medium - Google Patents
Service analysis method and device based on graph level, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN113742538B CN113742538B CN202110809121.0A CN202110809121A CN113742538B CN 113742538 B CN113742538 B CN 113742538B CN 202110809121 A CN202110809121 A CN 202110809121A CN 113742538 B CN113742538 B CN 113742538B
- Authority
- CN
- China
- Prior art keywords
- resource
- node
- target
- loop
- network
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/901—Indexing; Data structures therefor; Storage structures
- G06F16/9024—Graphs; Linked lists
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention provides a business analysis method based on a graph level, which comprises the following steps: acquiring a target resource association network of a target service, wherein the target resource association network comprises resource nodes, resource weight edges and resource directions; according to the resource nodes and the resource weight edges, the target resource association network is subjected to ring removal to obtain a target resource loop-free network; performing random walk on the target resource loop-free network based on the resource direction, and determining the hierarchical information of each resource node in the target resource loop-free network; and carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule. The target resource association network can be de-looped, the target resource loop-free network can be randomly walked through the resource direction, and then the flow path in the target resource loop-free network is traversed, so that the graph level information containing the whole flow direction is obtained for carrying out service analysis, and the analysis efficiency and the analysis accuracy can be improved.
Description
Technical Field
The present invention relates to the field of artificial intelligence, and in particular, to a business analysis method, apparatus, electronic device, and storage medium based on graph hierarchy.
Background
The graph hierarchy analysis refers to an analysis method for performing hierarchy definition on nodes or edges in a graph, and the graph refers to ordered triples consisting of the nodes, the edges and connection relations. Through hierarchical division, nodes in the graph can be classified, so that related personnel can have more visual and clear knowledge on the graph, and the hierarchical analysis method has wide application in different fields, including fund hierarchical division in finance, social network hierarchical division in the Internet field and the like. Currently existing algorithms for graph level partitioning include a cluster algorithm community detection algorithm and a deep learning method, however, these methods have some drawbacks or limitations. The biggest defect of the clustering algorithm is that the clustering algorithm can only conduct hierarchical division on the respective information of each node, and the whole information of the whole network cannot be learned in a whole way; in addition, the community detection algorithm pays attention to the overall communication relation too much, and local differences among nodes are ignored; deep learning algorithms often require a large amount of data, and due to the lack of a sufficient amount of data, the deep learning algorithms cannot be landed in most real scenes. In addition, the most important point of the existing method is that when the whole directional edge flow direction among the whole graph nodes needs to be explored in a complex network, the problem cannot be solved by the method. Learning of graph flow information is important in many scenarios, for example, taking a fund flow analysis scenario as an example, most accounts are only a certain fund transfer node, mining itself is not significant, and only mining the flow direction of the whole fund can actually find the final inflow node of the fund. It can be seen that the existing graph-level analysis cannot analyze the overall graph flow information.
Disclosure of Invention
In a first aspect, an embodiment of the present invention provides a graph-level-based service analysis method capable of improving analysis efficiency and analysis accuracy, where the method includes:
acquiring a target resource association network of a target service, wherein the target resource association network comprises a resource node, a resource weight edge and a resource direction;
According to the resource nodes and the resource weight edges, the target resource association network is subjected to ring removal to obtain a target resource loop-free network;
performing random walk on the target resource loop-free network based on the resource direction, and determining the hierarchical information of each resource node in the target resource loop-free network;
And carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule.
Optionally, the performing, according to the resource node and the resource weight edge, loop removal on the target resource association network to obtain a target resource loop-free network includes:
Calculating the number S1 of strong communication components of a single resource node in the target resource association network;
traversing a target resource association network to remove one resource weight edge, and calculating the number S2 of strong communication components of a single resource node;
And de-looping the target resource association network based on the strong communication component number S1 of the single resource node and the strong communication component number S2 of the single resource node to obtain a target resource loop-free network.
Optionally, the de-looping the target resource association network based on the number of strong connected components S1 of the single resource node and the number of strong connected components S2 of the single resource node to obtain a target resource loop-free network, including:
Calculating the difference value between the number S2 of the strong communication components of the single resource node and the number S1 of the strong communication components of the single resource node;
Deleting the resource weight edge with the minimum resource weight in the resource weight edges corresponding to the maximum difference value from the target resource association network;
and when all the strong connected components in the target resource association network are formed by single resource nodes, obtaining the target resource loop-free network.
Optionally, the resource direction includes a resource outflow direction, the determining, based on the resource direction, hierarchical information of each resource node in the target resource loop-free network by performing random walk on the target resource loop-free network includes:
constructing a starting node set according to the target resource loop-free network, wherein the starting node set comprises all resource nodes in the target resource loop-free network;
constructing an exit direction associated node set of each resource node according to the target resource loop-free network and the initial node set, wherein the exit direction associated node set comprises all resource nodes connected with the corresponding resource node in the target resource loop-free network in a resource exit direction;
And carrying out random walk based on the initial node set and the exit direction association node set, and calculating to obtain the hierarchical information of each resource node in the target resource loop-free network.
Optionally, the performing random walk based on the initial node set and the exit direction association node set, and calculating to obtain the hierarchical information of each resource node in the target resource loop-free network, includes:
Randomly taking out a resource node from the initial node set as an initial resource node, determining the hierarchy of the initial resource node as i=0, and recording the hierarchy result of the initial resource node in a first dictionary;
Randomly taking out a resource node from the output direction associated node set corresponding to the initial resource node as an output direction associated resource node, and judging whether the output direction associated resource node is recorded in a first dictionary or not;
if the exit-direction associated resource node is not recorded in the first dictionary, determining that the level of the exit-direction associated resource node is I+1, and recording the level result of the exit-direction associated resource node in the first dictionary;
Traversing the exit direction associated node set to obtain the hierarchy information of the initial resource node;
Traversing the initial node set to obtain the level information of all initial resource nodes;
And calculating the hierarchy information of each resource node in the target resource loop-free network based on the hierarchy information of all the initial resource nodes.
Optionally, after traversing the set of starting nodes to obtain the hierarchical information of all the starting resource nodes, the method further includes:
Recording the hierarchy results in the first dictionary in a second dictionary, and clearing the hierarchy results in the first dictionary;
traversing the initial node set to obtain the level information of all initial resource nodes, wherein the level information comprises:
And randomly traversing the initial resource nodes for n times to obtain n groups of hierarchical information of each initial resource node.
Optionally, the calculating, based on the hierarchical information of all the starting resource nodes, the hierarchical information of each resource node in the target resource loop-free network includes:
Calculating the average level and the level standard deviation of each initial resource node according to the n groups of level information of each initial resource node;
And determining the hierarchy information of each resource node in the target resource loop-free network based on the average hierarchy and the hierarchy standard deviation of each initial resource node.
Optionally, the target business rule includes business levels, each business level corresponds to a graph level, and the graph level analysis is performed on the target business by using the level information and a preset target business rule, including:
dividing the hierarchy information of each resource node in the target resource loop-free network according to service hierarchy to obtain a service hierarchy chart;
And carrying out graph hierarchy analysis on the target service through the service hierarchy graph.
In a second aspect, an embodiment of the present invention provides a graph-level-based traffic analysis apparatus capable of improving analysis efficiency and analysis accuracy, the apparatus including:
The system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring a target resource association network of a target service, and the target resource association network comprises a resource node, a resource weight edge and a resource direction;
the ring removing module is used for removing rings from the target resource association network according to the resource nodes and the resource weight edges to obtain a target resource ring-free network;
The wandering module is used for carrying out random wandering on the target resource loop-free network based on the resource direction and determining the level information of each resource node in the target resource loop-free network;
and the analysis module is used for carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule.
In a third aspect, an embodiment of the present invention provides an electronic device, including: the system comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the steps in the business analysis method based on the graph level provided by the embodiment of the invention when executing the computer program.
In a fourth aspect, an embodiment of the present invention provides a computer readable storage medium, where a computer program is stored, where the computer program when executed by a processor implements the steps in the graph-level based service analysis method provided by the embodiment of the present invention.
In the embodiment of the invention, a target resource association network of a target service is acquired, wherein the target resource association network comprises resource nodes, resource weight edges and resource directions; according to the resource nodes and the resource weight edges, the target resource association network is subjected to ring removal to obtain a target resource loop-free network; performing random walk on the target resource loop-free network based on the resource direction, and determining the hierarchical information of each resource node in the target resource loop-free network; and carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule. According to the embodiment of the invention, the ring-removing can be carried out on the target resource association network, the complicated annular loop in the target resource association network can be removed, the target resource acyclic network can be obtained, node level analysis based on the resource direction in the target resource acyclic network can be clearer, the target resource acyclic network can be randomly walked through the resource direction, and then the flow path in the target resource acyclic network can be traversed, so that the graph level information containing the whole flow direction is obtained for carrying out service analysis, and the analysis efficiency and the analysis accuracy can be improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a business analysis method based on a graph level according to an embodiment of the present invention;
FIG. 2 is a flow chart of a method of de-looping according to an embodiment of the present invention;
FIG. 3 is a flow chart of another method of decyclization provided by embodiments of the present invention;
fig. 4 is a flowchart of a hierarchical information determining method according to an embodiment of the present invention;
FIG. 5 is a flowchart of another hierarchical information determination method according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of a service analysis device based on a graph level according to an embodiment of the present invention;
FIG. 7 is a schematic structural view of a de-ring module according to an embodiment of the present invention;
FIG. 8 is a schematic structural diagram of a decyclization sub-module according to one embodiment of the present invention;
FIG. 9 is a schematic structural view of a wander module according to an embodiment of the present invention;
FIG. 10 is a schematic diagram of a third computing sub-module according to an embodiment of the present invention;
FIG. 11 is a schematic diagram of a third computing sub-module according to an embodiment of the present invention;
FIG. 12 is a schematic diagram of a computing unit according to an embodiment of the present invention;
FIG. 13 is a schematic structural diagram of an analysis module according to an embodiment of the present invention;
fig. 14 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, fig. 1 is a flowchart of a service analysis method based on a graph level according to an embodiment of the present invention, as shown in fig. 1, the service analysis method based on the graph level includes the following steps:
101. And acquiring a target resource association network of the target service.
In the embodiment of the present invention, the target resource-related network includes a resource node, a resource weight edge, and a resource direction. The target resource association network can be obtained from the service database directly, or can be built by the user.
The process of constructing the target resource association network is a process of constructing target resource data into a weighted directed ring graph (WEIGHTED DIRECTED CYCLIC GRAPH, abbreviated as WDCG). In the weighted directed graph, the graph includes nodes, edges and directions, where two nodes are connected by an edge having a direction, e.g., node a and node B are connected by an edge C, and the direction is a to B. Because of the complexity of the nodes and edges, most of the nodes are in the same complicated loop, and the hierarchy of each node cannot be determined in one loop.
The target business can be hierarchical analysis of a fund flow network, hierarchical analysis of a social network and the like, and specifically can be constructed by bank customer portrait analysis, and can also be member analysis of fund flow analysis and fund operation organization in the public security field.
In the embodiment of the invention, with the hierarchical analysis of the fund flow network as a target service, the constructed target resource association network can also be called a fund association network, in the fund association network, each resource node is a related person or company in a platform, each resource weight edge represents the size of a fund net transaction between two resource nodes, and the resource direction represents the fund net flow direction (or called a fund net flow direction). In the embodiment of the invention, other funds transaction details, such as time, times and the like, can be ignored when constructing the funds correlation network. Conventional funds association networks often build more transaction details, which can result in the funds flow network focusing more on node details and making it difficult to learn the overall funds flow. Therefore, in the embodiment of the invention, the net value transaction size and the net value flow direction are obtained through calculation through housing other funds transaction details, so that the whole funds flow direction can be focused and analyzed more purposefully.
Of course, the hierarchical analysis of the funds flow network as a targeted service is merely exemplary and should not be construed as limiting of embodiments of the invention. The built target resource association network can also be called a social network, and the hierarchical analysis of the fund flow network can also be migrated into the hierarchical analysis of the social network, for example, in the social network, each resource node is a related person or company in a platform, each resource weight edge represents the social relationship value between two resource nodes, and the resource direction represents social initiative (who actively contacts). The social relation value can be obtained by calculating data such as the number of times of connection, the duration of connection, the number of times of the same line, the duration of the same line and the like, and the social initiative can be determined by the number of times of actively initiating the connection and the number of times of actively participating in the same line.
It should be noted that, the target resource association network is a graph structure, and the graph structure includes ordered triples formed by nodes, edges and connection relations.
102. And according to the resource nodes and the resource weight edges, the target resource association network is de-looped to obtain the target resource loop-free network.
In the embodiment of the present invention, the target resource-related network includes more resource nodes and resource weight edges, so that a plurality of loop-shaped loops are formed, so that most of the resource nodes are located in the same loop-shaped loop, and in the loop-shaped loop, all the resource nodes are connected with each other, and the hierarchical information of the resource nodes cannot be determined. It should be noted that, just because the hierarchy information of each resource node cannot be determined in the loop, the existing graph hierarchy analysis method often abandons the flow direction of the mined resources, changes from the resources of a single resource node to the mode to analyze the hierarchy thereof, but the method cannot really find the resource node with a problem, and mostly only does nonsensical search in the intermediate resource node.
In the embodiment of the invention, platform personnel often use tens of accounts to continuously transfer money to each other in order to avoid the discovery of the supervision or law enforcement departments, funds obtained by the platform illegally are washed for several times or even tens of times in the crowd, and then the funds are transferred out from different account nodes, so that a large number and large-scale loop loops appear, and analysis and evidence obtaining of the supervision or law enforcement departments are hindered and disturbed.
Specifically, unimportant resource weight edges in the target resource association network can be removed, for example, resource weight edges with low resource weight can be removed in the loop, so that the loop is disconnected, and the weighted directed loop graph is converted into a weighted directed acyclic graph (WEIGHTED DIRECTED ACYCLIC GRAPH, abbreviated as WDAG), thereby facilitating the division of graph levels.
It should be noted that, the problem of converting the weighted directed loop graph into the weighted directed loop-free graph is a minimum feedback arc, which is an NP-hard (Non-DETERMINISTIC POLYNOMIAL-hard, chinese is an unresolved in polynomial time), so that the approximate solution can be obtained only by some methods, for example, the approximate solution can be obtained by a greedy algorithm, which is an algorithm that finds a local optimum choice in each step.
Optionally, referring to fig. 2, fig. 2 is a flowchart of a method for performing ring removal according to an embodiment of the present invention, as shown in fig. 2, on the basis of the embodiment of fig. 1, the method for performing ring removal on a target resource association network may include the following steps:
201. the number of strongly connected components S1 of a single resource node in the target resource-associated network is calculated.
In the embodiment of the present invention, the strong communication component refers to a plurality of resource nodes that are reachable in pairs, for example, a resource node a to a resource node B have a resource weight edge C, and a resource node B to a resource node a have a resource weight edge D, which may be referred to as a strong communication component. Further, if every two resource nodes of the directed graph are strongly connected, the directed graph is called a strongly connected graph, and a maximum strongly connected subgraph of the directed non-strongly connected graph is called a strongly connected component. The strongly connected components include the number of loop-part integral and loop-free part nodes, so the fewer the number of loops in a graph, the greater the number of strongly connected components of his individual resource nodes.
202. And traversing the target resource association network to remove one resource weight edge, and calculating the number S2 of the strong communication components of the single resource node.
In the embodiment of the invention, each resource weight edge can be traversed and removed, and the number S2 of strong connected components corresponding to a single resource node in the target resource association network when each resource weight edge is removed is calculated, for example, when one resource weight edge C is removed, the number S2C of strong connected components of a single resource node in the target resource association network without the resource weight edge C is calculated, and when one resource weight edge D is removed, the number S2D of strong connected components of a single resource node in the target resource association network without the resource weight edge D is calculated.
In one possible embodiment, each resource weight edge in the removal loop may be traversed to calculate the number of strongly connected components S2 of the corresponding individual resource node in the target resource-associated network when each resource weight edge is removed. Therefore, the target resource association network can be more specifically de-looped, and the calculation speed of de-looping is improved.
203. And de-looping the target resource association network based on the strong connected component number S1 of the single resource node and the strong connected component number S2 of the single resource node to obtain the target resource loop-free network.
In the embodiment of the present invention, after subtracting one resource weight edge, the number of strong connected components S2 of a single resource node in the target resource-associated network may be increased or unchanged (may be decreased in the case of three nodes) compared to the number of strong connected components S1 of a single resource node in the target resource-associated network without subtracting the resource weight edge, when the subtracted resource weight edge is one resource weight edge in the loop, the number of strong connected components S2 of a single resource node in the target resource-associated network is increased compared to the number of strong connected components S1 of a single resource node in the target resource-associated network without subtracting the resource weight edge, and when the subtracted resource weight edge is not one resource weight edge in the loop, the number of strong connected components S2 of a single resource node in the target resource-associated network is unchanged compared to the number of strong connected components S1 of a single resource node in the target resource-associated network without subtracting the resource weight edge. And when the number of the strong connected components S2 of the single resource node is the largest compared with the number of the strong connected components S1 of the single resource node in the target resource association network without subtracting the resource weight edge, determining to remove the corresponding resource weight edge to obtain an intermediate target resource association network, and iterating the process until the number of the strong connected components S2 of the single resource node in the target resource association network is not increased any more, thereby obtaining the target resource loop-free network.
Specifically, referring to fig. 3, fig. 3 is a flowchart of another method for performing ring removal according to an embodiment of the present invention, as shown in fig. 3, on the basis of the embodiment of fig. 2, the step of performing ring removal on the target resource association network may include the following steps:
301. The difference between the number of strongly connected components S2 of a single resource node and the number of strongly connected components S1 of a single resource node is calculated.
In the embodiment of the present invention, the difference between the number S2 of the strong connected components of the single resource node and the number S1 of the strong connected components of the single resource node mainly refers to the increment of the number of the strong connected components of the single resource node after the target resource association network removes an edge.
302. And deleting the resource weight edge with the minimum resource weight in the resource weight edges corresponding to the maximum difference value from the target resource association network.
In the embodiment of the present invention, the maximum difference is the maximum difference between the number of strong connected components S2 of a single resource node and the number of strong connected components S1 of the single resource node, and at this time, the maximum difference indicates that the destruction of the loop-type loop in the target resource association network is the maximum by removing the corresponding resource weight edge.
Finding a resource weight edge or a set of resource weight edges can maximize S2-S1, and if only one resource weight edge exists, the resource weight edge is removed from the target resource association network. If a group of resource weight edges exist, the group of resource weight edges comprise a plurality of resource weight edges, and one resource weight edge with the smallest resource weight in the plurality of resource weight edges is found out and removed from the target resource association network.
303. And when all the strong connected components in the target resource association network are formed by single resource nodes, obtaining the target resource loop-free network.
In the embodiment of the invention, after one resource weight edge is removed from the target resource association network, the target resource association network with one less resource weight edge is obtained, the strong communication component number S2 of the single resource node at the moment can be used as the strong communication component number S1 of the single resource node, one resource weight edge is removed by traversing on the basis of the target resource association network with one less resource weight edge, and the strong communication component number of the single resource node is calculated to be used as the strong communication component number S2 of the new single resource node. And iterating the steps 301 and 302 until all the strong connected components in the target resource association network are formed by single resource nodes, then the target resource association network can be considered to have no loop, and the iteration can be stopped, so that the target resource loop-free network is obtained.
By the embodiments of fig. 2 and 3 described above, most unimportant resource weight edges can be removed and a loop-like loop exists to convert a weighted directed loop graph to a weighted directed loop-free graph.
103. And carrying out random walk on the target resource loop-free network based on the resource direction, and determining the hierarchical information of each resource node in the target resource loop-free network.
In the embodiment of the invention, the resource direction refers to the resource flowing direction, and in the hierarchical analysis of the fund flow network, the resource direction refers to the fund flowing direction, and for one resource node, the resource flowing direction comprises a resource outflow direction and a resource inflow direction.
In the embodiment of the invention, the starting point can be randomly determined in the target resource loop-free network, the random walk is carried out according to the resource outflow direction of the starting point, the random walk path is obtained, and the hierarchical information of each resource node in the random walk path is determined. Traversing all starting points of the target resource loop-free network to obtain all random walk paths of the target resource loop-free network, thereby determining the hierarchical information of each resource node in the target resource loop-free network.
It should be noted that, the target resource loop-free network is a weighted directed loop-free graph, the weighted directed loop-free graph has no loop structure, no hierarchy definition barrier has been existed, and the hierarchy of each resource node can be defined. However, when traversing the weighted directed acyclic graph, there is a problem of hierarchy differences for the same resource node, i.e., the hierarchy defined by the same resource node in different paths is different. For example, if a is transferred to B, B is transferred to C, a is transferred to C, there are path AB, path BC, path AC, and path ABC, where C is at the third level in terms of path ABC, and where C is at the second level in terms of path AC, where the level of C is difficult to define directly and exactly.
Optionally, referring to fig. 4, fig. 4 is a flowchart of a method for determining hierarchical information, where the resource direction includes a resource outflow direction, as shown in fig. 4, and the method for determining hierarchical information includes the following steps based on the embodiment of fig. 1:
401. And constructing a starting node set according to the target resource loop-free network.
In the embodiment of the present invention, the set of starting nodes includes all resource nodes in the target resource loop-free network. And adding all the resource nodes in the target resource loop-free network into a node set, so as to construct and obtain a starting node set.
402. And constructing an exit direction associated node set of each resource node according to the target resource loop-free network and the initial node set.
In the embodiment of the present invention, the outgoing direction associated node set includes all resource nodes connected to a corresponding resource node in the target resource loop-free network in a resource outgoing direction. It should be noted that, each resource node in the initial node set corresponds to one output direction associated node set.
403. And carrying out random walk based on the initial node set and the exit direction association node set, and calculating to obtain the hierarchical information of each resource node in the target resource loop-free network.
In the embodiment of the invention, a resource node in a starting node set is used as a starting resource node, an exit direction associated node set is used as an exit direction associated resource node, the starting resource node is randomly taken from the starting resource node set, the exit direction associated resource node is randomly taken from the corresponding exit direction associated node set, a plurality of travel paths of the starting resource node are obtained through random travel, and the level information of each resource node is determined according to the connection condition of each resource node in the travel paths.
Optionally, referring to fig. 5, fig. 5 is a flowchart of another method for determining hierarchical information according to an embodiment of the present invention, as shown in fig. 5, where the method for determining hierarchical information further includes the following steps based on the embodiment of fig. 4:
501. And randomly taking out one resource node from the initial node set as an initial resource node, determining the hierarchy of the initial resource node as I=0, and recording the hierarchy result of the initial resource node in the first dictionary.
In the embodiment of the invention, the randomly fetched resource node is locked or deleted in the initial node set, so that the resource node is prevented from being repeatedly fetched. The level result of the initial resource node is i=0, the first dictionary may be understood as an indexed set, and the initial resource node and the corresponding level result may be recorded.
502. And randomly taking one resource node from the output direction associated node set corresponding to the initial resource node as an output direction associated resource node, and judging whether the direction associated resource node is recorded in the first dictionary.
In the embodiment of the invention, the randomly fetched resource node is locked or deleted in the exit direction associated node set, so that the resource node is prevented from being repeatedly fetched. Whether the direction-associated resource node is recorded in the first dictionary is determined in order to determine whether the direction-associated resource node has determined the same hierarchical result.
503. If the output direction associated resource node is not recorded in the first dictionary, determining that the hierarchy of the output direction associated resource node is I+1, and recording the hierarchy result of the output direction associated resource node in the first dictionary.
In the embodiment of the present invention, if the exit direction associated resource node is not recorded in the first dictionary, it is indicated that there is no exit direction associated resource node or no hierarchy result corresponding to the exit direction associated resource node in the first dictionary, where the hierarchy of the exit direction associated resource node may be i+1, and the hierarchy result of the exit direction associated resource node may be recorded in the first dictionary.
If the output direction associated resource node is already recorded in the first dictionary, the output direction associated resource node does not need to be recorded, and the next resource node is randomly taken out from the output direction associated node set corresponding to the initial resource node to serve as the output direction associated resource node.
504. Traversing the direction-exiting associated node set to obtain the hierarchy information of the initial resource node.
In the initial example of the present invention, the set of direction-associated nodes may be traversed by iterating steps 502 and 503 described above.
505. Traversing the initial node set to obtain the level information of all initial resource nodes.
In an embodiment of the present invention, the set of starting nodes may be traversed by iterating steps 501 through 504 described above.
506. And calculating the hierarchy information of each resource node in the target resource loop-free network based on the hierarchy information of all the initial resource nodes.
In the embodiment of the present invention, through the above step 505, the hierarchy information of all the initial resource nodes may be obtained, where each initial resource node corresponds to one resource node in the target resource loop-free network, so that the hierarchy information of each resource node in the target resource loop-free network may be obtained by calculation according to the hierarchy information of all the initial resource nodes.
Optionally, after step 505, the above-mentioned hierarchical information determining method further includes: and recording the hierarchy result in the first dictionary in a second dictionary, and clearing the hierarchy result in the first dictionary.
The step 505 includes: and randomly traversing the initial resource nodes for n times to obtain n groups of hierarchical information of each initial resource node.
Further, the average hierarchy and the standard deviation of the hierarchy of each initial resource node may be calculated according to the n groups of hierarchy information of each initial resource node; and determining the hierarchy information of each resource node in the target resource loop-free network based on the average hierarchy and the hierarchy standard deviation of each initial resource node. Specifically, the average level of each initial resource node is shown in the following formula:
the level standard deviation of each initial resource node is shown in the following formula:
in the above two formulas, n is the number of random walks, and l i is the level to which the ith walk node belongs.
In the embodiment of the invention, the information of the belonging hierarchy of each resource node can be obtained, and when the resource flows to the final node in the target resource association network, the number of hierarchies and the range stability of the belonging hierarchy are required to pass on average.
104. And carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule.
In the embodiment of the present invention, the target service rule may include service levels, each service level corresponds to a graph level, and the level information of each resource node in the target resource loop-free network is divided according to the service levels to obtain a service level graph; and carrying out graph hierarchy analysis on the target service through the service hierarchy graph.
Specifically, after the hierarchical information of the resource nodes is obtained, a rule model can be built by combining target business rules according to the local information of the nodes such as the access degree information, the weight information, the net value information and the like of the resource nodes, so that the resource nodes can be clearly and accurately classified. Thus, the whole resource flow direction information in the graph is learned, and the local characteristic condition of the resource node is learned.
In the embodiment of the invention, a target resource association network of a target service is acquired, wherein the target resource association network comprises resource nodes, resource weight edges and resource directions; according to the resource nodes and the resource weight edges, the target resource association network is subjected to ring removal to obtain a target resource loop-free network; performing random walk on the target resource loop-free network based on the resource direction, and determining the hierarchical information of each resource node in the target resource loop-free network; and carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule. The method can remove the complicated annular loop in the target resource association network by means of de-looping the target resource association network to obtain a target resource loop-free network, node level analysis based on the resource direction in the target resource loop-free network can be clearer, the target resource loop-free network is subjected to random walk through the resource direction, and then the flow direction path in the target resource loop-free network is traversed to obtain graph level information containing the whole flow direction for service analysis, so that analysis efficiency and analysis accuracy can be improved.
It should be noted that, the business analysis method based on the graph level provided by the embodiment of the invention can be applied to devices such as a smart phone, a computer, a server and the like which can perform business analysis of the graph level.
Optionally, referring to fig. 6, fig. 6 is a schematic structural diagram of a service analysis device based on a graph level according to an embodiment of the present invention, as shown in fig. 6, where the device includes:
An obtaining module 601, configured to obtain a target resource-related network of a target service, where the target resource-related network includes a resource node, a resource weight edge, and a resource direction;
the de-looping module 602 is configured to de-loop the target resource-associated network according to the resource node and the resource weight edge to obtain a target resource loop-free network;
A walk module 603, configured to perform random walk on the target resource loop-free network based on the resource direction, and determine hierarchical information of each resource node in the target resource loop-free network;
and the analysis module 604 is configured to perform graph level analysis on the target service according to the level information and a preset target service rule.
Alternatively, as shown in fig. 7, the ring removal module 602 includes:
a first calculation submodule 6021, configured to calculate the number S1 of strong connected components of a single resource node in the target resource-associated network;
A second calculation submodule 6022, configured to traverse the target resource-associated network to remove one of the resource weight edges, and calculate the number S2 of strong-connectivity components of a single resource node;
And the ring removing sub-module 6023 is configured to perform ring removing on the target resource association network based on the number of strong connected components S1 of the single resource node and the number of strong connected components S2 of the single resource node, so as to obtain a target resource ring-free network.
Optionally, as shown in fig. 8, the ring removing sub-module 6023 includes:
a calculating unit 60231, configured to calculate a difference value between the number of strong-connectivity components S2 of the single resource node and the number of strong-connectivity components S1 of the single resource node;
a deleting unit 60232, configured to delete, from the target resource-associated network, a resource weight edge with a minimum resource weight among resource weight edges corresponding to the maximum difference value;
And the determining unit 60233 is configured to obtain a target resource loop-free network when all the strong connected components in the target resource association network are formed by a single resource node.
Optionally, as shown in fig. 9, the resource direction includes a resource outflow direction, and the walk module 603 includes:
A first construction submodule 6031, configured to construct a starting node set according to the target resource loop-free network, where the starting node set includes all resource nodes in the target resource loop-free network;
A second building submodule 6032, configured to build an outgoing direction association node set of each resource node according to the target resource loop-free network and the starting node set, where the outgoing direction association node set includes all resource nodes connected with the corresponding resource node in the target resource loop-free network in a resource outgoing direction;
and a third computing sub-module 6033, configured to perform random walk based on the initial node set and the exit direction association node set, and calculate to obtain hierarchical information of each resource node in the target resource loop-free network.
Optionally, as shown in fig. 10, the third computing sub-module 6033 includes:
a first processing unit 60331, configured to randomly take out a resource node from the initial node set as an initial resource node, determine that a level of the initial resource node is i=0, and record a level result of the initial resource node in a first dictionary;
A second processing unit 60332, configured to randomly take one resource node from the output direction association node set corresponding to the initial resource node as an output direction association resource node, and determine whether the output direction association resource node is recorded in a first dictionary;
A third processing unit 60333, configured to determine that the hierarchy of the exit-direction associated resource node is i+1 if the exit-direction associated resource node is not recorded in the first dictionary, and record the hierarchy result of the exit-direction associated resource node in the first dictionary;
The first traversing unit 60334 is configured to traverse the exit direction association node set to obtain the hierarchy information of the initial resource node;
A second traversing unit 60335, configured to traverse the initial node set to obtain hierarchy information of all initial resource nodes;
And a calculating unit 60336, configured to calculate, based on the hierarchical information of all the starting resource nodes, the hierarchical information of each resource node in the target resource loop-free network.
Optionally, as shown in fig. 11, the third computing sub-module 6033 further includes:
a fourth processing unit 60337, configured to record the hierarchy result in the first dictionary in a second dictionary, and empty the hierarchy result in the first dictionary;
The computing unit 60336 is further configured to randomly traverse the starting resource node to n times, to obtain n sets of hierarchical information of each starting resource node.
Optionally, as shown in fig. 12, the calculating unit 60336 includes:
A calculating subunit 603361, configured to calculate, according to the n sets of hierarchy information of each starting resource node, an average hierarchy and a hierarchy standard deviation of each starting resource node;
a determining subunit 603362, configured to determine, based on the average hierarchy and the standard deviation of the hierarchy of each starting resource node, hierarchy information of each resource node in the target resource loop-free network.
Optionally, as shown in fig. 13, the target business rule includes business levels, each business level corresponds to one graph level, and the analyzing module 604 includes:
a dividing submodule 6041, configured to divide the hierarchy information of each resource node in the target resource loop-free network according to a service hierarchy, to obtain a service hierarchy graph;
an analysis submodule 6042, configured to perform graph level analysis on the target service through the service level graph.
It should be noted that, the service analysis device based on the graph level provided by the embodiment of the invention can be applied to devices such as a smart phone, a computer, a server and the like which can perform service analysis of the graph level.
The business analysis device based on the graph level provided by the embodiment of the invention can realize each process realized by the business analysis method based on the graph level in the method embodiment, and can achieve the same beneficial effects. In order to avoid repetition, a description thereof is omitted.
Referring to fig. 14, fig. 14 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, as shown in fig. 14, including: memory 1402, processor 1401 and a computer program stored on the memory 1402 and executable on the processor 1401 for a graph-level based business analysis method, wherein:
The processor 1401 is configured to call a computer program stored in the memory 1402, and execute the following steps:
acquiring a target resource association network of a target service, wherein the target resource association network comprises a resource node, a resource weight edge and a resource direction;
According to the resource nodes and the resource weight edges, the target resource association network is subjected to ring removal to obtain a target resource loop-free network;
performing random walk on the target resource loop-free network based on the resource direction, and determining the hierarchical information of each resource node in the target resource loop-free network;
And carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule.
Optionally, the performing, by the processor 1401, the de-looping the target resource-associated network according to the resource node and the resource weight edge to obtain a target resource loop-free network, including:
Calculating the number S1 of strong communication components of a single resource node in the target resource association network;
traversing a target resource association network to remove one resource weight edge, and calculating the number S2 of strong communication components of a single resource node;
And de-looping the target resource association network based on the strong communication component number S1 of the single resource node and the strong communication component number S2 of the single resource node to obtain a target resource loop-free network.
Optionally, the de-looping the target resource association network based on the number of strong connected components S1 of the single resource node and the number of strong connected components S2 of the single resource node by the processor 1401 to obtain a target resource loop-free network, including:
Calculating the difference value between the number S2 of the strong communication components of the single resource node and the number S1 of the strong communication components of the single resource node;
Deleting the resource weight edge with the minimum resource weight in the resource weight edges corresponding to the maximum difference value from the target resource association network;
and when all the strong connected components in the target resource association network are formed by single resource nodes, obtaining the target resource loop-free network.
Optionally, the resource direction includes a resource outflow direction, the determining, by the processor 1401, the hierarchy information of each resource node in the target resource loop-free network by performing random walk on the target resource loop-free network based on the resource direction includes:
constructing a starting node set according to the target resource loop-free network, wherein the starting node set comprises all resource nodes in the target resource loop-free network;
constructing an exit direction associated node set of each resource node according to the target resource loop-free network and the initial node set, wherein the exit direction associated node set comprises all resource nodes connected with the corresponding resource node in the target resource loop-free network in a resource exit direction;
And carrying out random walk based on the initial node set and the exit direction association node set, and calculating to obtain the hierarchical information of each resource node in the target resource loop-free network.
Optionally, the performing, by the processor 1401, random walk based on the starting node set and the exit direction association node set, and calculating to obtain hierarchical information of each resource node in the target resource loop-free network includes:
Randomly taking out a resource node from the initial node set as an initial resource node, determining the hierarchy of the initial resource node as i=0, and recording the hierarchy result of the initial resource node in a first dictionary;
Randomly taking out a resource node from the output direction associated node set corresponding to the initial resource node as an output direction associated resource node, and judging whether the output direction associated resource node is recorded in a first dictionary or not;
if the exit-direction associated resource node is not recorded in the first dictionary, determining that the level of the exit-direction associated resource node is I+1, and recording the level result of the exit-direction associated resource node in the first dictionary;
Traversing the exit direction associated node set to obtain the hierarchy information of the initial resource node;
Traversing the initial node set to obtain the level information of all initial resource nodes;
And calculating the hierarchy information of each resource node in the target resource loop-free network based on the hierarchy information of all the initial resource nodes.
Optionally, after the traversing the set of starting nodes and obtaining the hierarchical information of all starting resource nodes, the method executed by the processor 1401 further includes:
Recording the hierarchy results in the first dictionary in a second dictionary, and clearing the hierarchy results in the first dictionary;
The traversing the set of starting nodes performed by the processor 1401 obtains the hierarchical information of all the starting resource nodes, including:
And randomly traversing the initial resource nodes for n times to obtain n groups of hierarchical information of each initial resource node.
Optionally, the calculating, by the processor 1401, the hierarchical information of each resource node in the target resource loop-free network based on the hierarchical information of all the starting resource nodes includes:
Calculating the average level and the level standard deviation of each initial resource node according to the n groups of level information of each initial resource node;
And determining the hierarchy information of each resource node in the target resource loop-free network based on the average hierarchy and the hierarchy standard deviation of each initial resource node.
Optionally, the target business rule includes business levels, each business level corresponds to a graph level, and the performing, by the processor 1401, graph level analysis on the target business by using the level information and a preset target business rule includes:
dividing the hierarchy information of each resource node in the target resource loop-free network according to service hierarchy to obtain a service hierarchy chart;
And carrying out graph hierarchy analysis on the target service through the service hierarchy graph.
It should be noted that, the electronic device provided by the embodiment of the present invention may be applied to devices such as a smart phone, a computer, and a server, which may perform service analysis at a graph level.
The electronic equipment provided by the embodiment of the invention can realize each process realized by the business analysis method based on the graph level in the embodiment of the method, and can achieve the same beneficial effects. In order to avoid repetition, a description thereof is omitted.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements each process of the graph-level-based service analysis method or the graph-level-based service analysis method at the application end provided by the embodiment of the invention, and can achieve the same technical effect, so that repetition is avoided and redundant description is omitted.
Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM) or the like.
The foregoing disclosure is illustrative of the present invention and is not to be construed as limiting the scope of the invention, which is defined by the appended claims.
Claims (9)
1. A business analysis method based on graph level, comprising the steps of:
acquiring a target resource association network of a target service, wherein the target resource association network comprises a resource node, a resource weight edge and a resource direction;
Calculating the number S1 of strong communication components of a single resource node in the target resource association network; traversing a target resource association network to remove one resource weight edge, and calculating the number S2 of strong communication components of a single resource node; calculating the difference value between the number S2 of the strong communication components of the single resource node and the number S1 of the strong communication components of the single resource node; deleting the resource weight edge with the minimum resource weight in the resource weight edges corresponding to the maximum difference value from the target resource association network; when all the strong communication components in the target resource association network are formed by single resource nodes, obtaining a target resource loop-free network;
performing random walk on the target resource loop-free network based on the resource direction, and determining the hierarchical information of each resource node in the target resource loop-free network;
And carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule.
2. The method of claim 1, wherein the resource direction comprises a resource outflow direction, wherein the randomly walking the target resource loop-free network based on the resource direction, determining hierarchical information for each resource node in the target resource loop-free network comprises:
constructing a starting node set according to the target resource loop-free network, wherein the starting node set comprises all resource nodes in the target resource loop-free network;
constructing an exit direction associated node set of each resource node according to the target resource loop-free network and the initial node set, wherein the exit direction associated node set comprises all resource nodes connected with the corresponding resource node in the target resource loop-free network in a resource exit direction;
And carrying out random walk based on the initial node set and the exit direction association node set, and calculating to obtain the hierarchical information of each resource node in the target resource loop-free network.
3. The method of claim 2, wherein the calculating, based on the starting node set and the outgoing direction association node set, the hierarchy information of each resource node in the target resource loop-free network includes:
Randomly taking out a resource node from the initial node set as an initial resource node, determining the hierarchy of the initial resource node as i=0, and recording the hierarchy result of the initial resource node in a first dictionary;
Randomly taking out a resource node from the output direction associated node set corresponding to the initial resource node as an output direction associated resource node, and judging whether the output direction associated resource node is recorded in a first dictionary or not;
if the exit-direction associated resource node is not recorded in the first dictionary, determining that the level of the exit-direction associated resource node is I+1, and recording the level result of the exit-direction associated resource node in the first dictionary;
Traversing the exit direction associated node set to obtain the hierarchy information of the initial resource node;
Traversing the initial node set to obtain the level information of all initial resource nodes;
And calculating the hierarchy information of each resource node in the target resource loop-free network based on the hierarchy information of all the initial resource nodes.
4. The method of claim 3, wherein after said traversing the set of starting nodes to obtain hierarchical information for all starting resource nodes, the method further comprises:
Recording the hierarchy results in the first dictionary in a second dictionary, and clearing the hierarchy results in the first dictionary;
traversing the initial node set to obtain the level information of all initial resource nodes, wherein the level information comprises:
And randomly traversing the initial resource nodes for n times to obtain n groups of hierarchical information of each initial resource node.
5. The method of claim 4, wherein calculating the hierarchy information for each resource node in the target resource loop-free network based on the hierarchy information for all starting resource nodes comprises:
Calculating the average level and the level standard deviation of each initial resource node according to the n groups of level information of each initial resource node;
And determining the hierarchy information of each resource node in the target resource loop-free network based on the average hierarchy and the hierarchy standard deviation of each initial resource node.
6. The method according to any one of claims 1 to 5, wherein the target business rules include business levels, each business level corresponding to a graph level, and the performing graph level analysis on the target business by using the level information and a preset target business rule includes:
dividing the hierarchy information of each resource node in the target resource loop-free network according to service hierarchy to obtain a service hierarchy chart;
And carrying out graph hierarchy analysis on the target service through the service hierarchy graph.
7. A graph-level based traffic analysis device, the device comprising:
The system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring a target resource association network of a target service, and the target resource association network comprises a resource node, a resource weight edge and a resource direction;
The de-looping module is used for calculating the number S1 of strong communication components of a single resource node in the target resource association network; traversing a target resource association network to remove one resource weight edge, and calculating the number S2 of strong communication components of a single resource node; calculating the difference value between the number S2 of the strong communication components of the single resource node and the number S1 of the strong communication components of the single resource node; deleting the resource weight edge with the minimum resource weight in the resource weight edges corresponding to the maximum difference value from the target resource association network; when all the strong communication components in the target resource association network are formed by single resource nodes, obtaining a target resource loop-free network;
The wandering module is used for carrying out random wandering on the target resource loop-free network based on the resource direction and determining the level information of each resource node in the target resource loop-free network;
and the analysis module is used for carrying out graph hierarchical analysis on the target service through the hierarchical information and a preset target service rule.
8. An electronic device, comprising: memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps in the graph-hierarchy based business analysis method of any one of claims 1 to 6 when the computer program is executed.
9. A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, implements the steps in the graph-hierarchy based traffic analysis method as claimed in any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110809121.0A CN113742538B (en) | 2021-07-16 | 2021-07-16 | Service analysis method and device based on graph level, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110809121.0A CN113742538B (en) | 2021-07-16 | 2021-07-16 | Service analysis method and device based on graph level, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113742538A CN113742538A (en) | 2021-12-03 |
| CN113742538B true CN113742538B (en) | 2024-09-27 |
Family
ID=78728731
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110809121.0A Active CN113742538B (en) | 2021-07-16 | 2021-07-16 | Service analysis method and device based on graph level, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113742538B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118132245A (en) * | 2022-12-01 | 2024-06-04 | 马上消费金融股份有限公司 | Resource allocation method and device and electronic equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109039786A (en) * | 2018-09-28 | 2018-12-18 | 山东中创软件商用中间件股份有限公司 | A kind of topology drawing generating method, device, equipment and readable storage medium storing program for executing |
| CN110704694A (en) * | 2019-09-29 | 2020-01-17 | 哈尔滨工业大学（威海） | Organization hierarchy dividing method based on network representation learning and application thereof |
| CN111314138A (en) * | 2020-02-19 | 2020-06-19 | 腾讯科技（深圳）有限公司 | Detection method of directed network, computer readable storage medium and related equipment |
Family Cites Families (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2612967B2 (en) * | 1991-01-18 | 1997-05-21 | 松下電器産業株式会社 | Network diagram automatic generation method and system |
| US8005770B2 (en) * | 2008-06-09 | 2011-08-23 | Microsoft Corporation | Parallel generation of a bayesian network |
| GB201810392D0 (en) * | 2015-11-25 | 2018-08-08 | Teamifier Inc | Methods for the augmentation, exploration, and maintenance of project hierarchies |
| CN105528296B (en) * | 2016-01-13 | 2018-10-23 | 东北大学 | A kind of class cluster test method of object-oriented software |
| CN106067139A (en) * | 2016-05-31 | 2016-11-02 | 北京红马传媒文化发展有限公司 | A kind of mthods, systems and devices of construction feature database |
| CN106130899B (en) * | 2016-07-14 | 2019-09-13 | 华为技术有限公司 | A method and device for generating service paths |
| CN106776858A (en) * | 2016-11-28 | 2017-05-31 | 华中科技大学 | A kind of figure based on SCC DAG calculates iterative processing method |
| US10713088B2 (en) * | 2017-03-23 | 2020-07-14 | Amazon Technologies, Inc. | Event-driven scheduling using directed acyclic graphs |
| CN107453928B (en) * | 2017-09-22 | 2019-06-14 | 冀北电力有限公司电力经济技术研究院 | A method and device for evaluating the importance of nodes in a power communication network |
| US10496468B2 (en) * | 2018-01-21 | 2019-12-03 | EMC IP Holding Company LLC | Root cause analysis for protection storage devices using causal graphs |
| US11188850B2 (en) * | 2018-03-30 | 2021-11-30 | Derek Alexander Pisner | Automated feature engineering of hierarchical ensemble connectomes |
| CN110765215A (en) * | 2019-09-30 | 2020-02-07 | 深圳云天励飞技术有限公司 | Query method and device for personnel common relationship, electronic equipment and storage medium |
| CN110798467B (en) * | 2019-10-30 | 2021-10-08 | 腾讯科技（深圳）有限公司 | Target object identification method and device, computer equipment and storage medium |
| CN111147377B (en) * | 2019-12-05 | 2020-09-11 | 连连银通电子支付有限公司 | Method, device, equipment and medium for determining service channel |
| CN112989137B (en) * | 2021-04-21 | 2021-08-06 | 军事科学院系统工程研究院网络信息研究所 | Entity relationship analysis method for complex network information system construction |
-
2021
- 2021-07-16 CN CN202110809121.0A patent/CN113742538B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109039786A (en) * | 2018-09-28 | 2018-12-18 | 山东中创软件商用中间件股份有限公司 | A kind of topology drawing generating method, device, equipment and readable storage medium storing program for executing |
| CN110704694A (en) * | 2019-09-29 | 2020-01-17 | 哈尔滨工业大学（威海） | Organization hierarchy dividing method based on network representation learning and application thereof |
| CN111314138A (en) * | 2020-02-19 | 2020-06-19 | 腾讯科技（深圳）有限公司 | Detection method of directed network, computer readable storage medium and related equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113742538A (en) | 2021-12-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109615116B (en) | A kind of telecommunication fraud incident detection method and detection system | |
| Ramaki et al. | A systematic mapping study on intrusion alert analysis in intrusion detection systems | |
| Rinzivillo et al. | The purpose of motion: Learning activities from individual mobility networks | |
| Zhang et al. | An influence-based approach for root cause alarm discovery in telecom networks | |
| Selim et al. | Towards shortest path identification on large networks | |
| Blasch et al. | Pattern activity clustering and evaluation (PACE) | |
| KR102086936B1 (en) | User data sharing method and device | |
| Sun et al. | An entropy‐based self‐adaptive node importance evaluation method for complex networks | |
| CN113742538B (en) | Service analysis method and device based on graph level, electronic equipment and storage medium | |
| CN117668826A (en) | Event association analysis method and system based on graph model | |
| Zawalski et al. | Fast and precise: Adjusting planning horizon with adaptive subgoal search | |
| Paramasivam et al. | Cor-ENTC: correlation with ensembled approach for network traffic classification using SDN technology for future networks | |
| CN115577119A (en) | Knowledge graph inference model training method, device and storage medium | |
| Li et al. | Exploration of a network security situational awareness model based on multisource data fusion | |
| Brandusoiu et al. | Churn prediction in the telecommunications sector using Bayesian networks | |
| Acerbi et al. | Continuous time bayesian networks for gene network reconstruction: a comparative study on time course data | |
| Gokulakrishan et al. | An advancing method for web service reliability and scalability using ResNet convolution neural network optimized with Zebra Optimization Algorithm | |
| Amin et al. | City hotspot identification using smart cyber-physical social system | |
| Uddin et al. | Assembly queries: Planning and discovering assemblies of moving objects using partial information | |
| Chen et al. | MBAN-MLC: A multi-label classification method and its application in automating fault diagnosis | |
| CN113158229A (en) | Data auditing method based on knowledge graph | |
| Jena et al. | Predictive Framework for Advanced Customer Churn Prediction using Machine Learning | |
| Amor et al. | Be a collaborator and a competitor in crowdsourcing system | |
| Fafoutellis et al. | A Causal Deep Learning Framework for Traffic Forecasting | |
| Whitsitt et al. | On the Extraction and Analysis of a Social Network with Partial Organizational Observation |
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 |