A routing network expression method, device, electronic device and storage medium

By filtering and iteratively modifying the routing node set in complex routing networks, the problem of low expression efficiency of routing networks in the prior art is solved, and efficient routing network expression is achieved.

CN115914084BActive Publication Date: 2024-11-29CHONGQING SATELLITE NETWORK SYSTEM CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202211615767.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-12-15
Publication Date
2024-11-29
Estimated Expiration
2042-12-15

AI Technical Summary

Technical Problem

In complex routing networks, it is difficult for the prior art to efficiently express the routing network, resulting in inefficiency. Especially when the number of candidate routing nodes is large, it takes a lot of time to compare node similarity, and it is impossible to complete the routing network expression in a timely manner.

Method used

By determining the basic routing node set based on the node identification interval of the target routing node, candidate routing nodes that meet the similarity and quantity conditions of attribute characteristics are selected, and the initial associated routing node set is modified through multiple iterations until the target associated routing node set is obtained. This improves the efficiency of routing network expression.

Benefits of technology

Through the methods of global structure analysis and local relationship discovery, the time consumption caused by the large number of candidate routing nodes is avoided, the efficiency of routing network expression is improved, and the routing network expression is ensured in a timely manner.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115914084B_ABST
    Figure CN115914084B_ABST
Patent Text Reader

Abstract

The present application discloses a routing network expression method, device, electronic device and storage medium, and relates to the field of communication network technology. In the present application, based on the node identification of the target routing node in the routing network, the basic routing node set is determined, and the initial associated routing node set is combined with the attribute feature similarity condition and the routing node quantity condition, so that the initial associated routing node set is iteratively modified multiple times, and finally the target associated routing node set is obtained. In this way, the technical drawbacks of the routing network expression in the related technology when the number of candidate routing nodes contained in the routing node set is large are avoided, and the efficiency of the routing network expression is improved through the routing network expression method of global structure analysis and local relationship discovery.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of communication network technology, and in particular to a routing network expression method, device, electronic device and storage medium. Background Art

[0002] With the rapid development of information technology, routing networks are widely used in information transmission between computer networks to ensure the effective sharing of network resources and data, and to achieve coordination between software and hardware. Routing networks include but are not limited to: routers, switches, hubs, hosts and other devices.

[0003] Furthermore, with the development of communication network technology, the scale of the network structure of routing networks has increased exponentially. Therefore, in order to clearly display the network structure of complex routing networks, how to accurately express the network structure of complex routing networks has become an important issue that needs to be solved urgently.

[0004] At present, when a complex routing network is expressed, at least one candidate routing node (i.e., a candidate routing node subset) that satisfies a routing node similarity condition with a target routing node is usually screened out from a routing node set corresponding to the complex routing network, so that each candidate routing node included in the obtained candidate routing node subset is used as an associated routing node of the target routing node.

[0005] However, when the routing network expression method is used, when the number of candidate routing nodes contained in the routing node set is large, it will take a lot of time to compare the node similarity between the target routing node and each candidate routing node, and thus it is impossible to ensure that the routing network expression is completed in a timely manner.

[0006] Therefore, using the above method, the efficiency of routing network expression is low. Summary of the invention

[0007] The embodiments of the present application provide a routing network expression method, device, electronic device and storage medium to improve the efficiency of routing network expression.

[0008] In a first aspect, an embodiment of the present application provides a routing network expression method, the method comprising:

[0009] Determine a basic routing node set corresponding to the target routing node based on the node identification interval to which the node identification of the target routing node in the routing network belongs; wherein at least one basic routing node included in the basic routing node set is a routing node of the same node level;

[0010] From the basic routing node set, at least one candidate routing node that meets the attribute feature similarity condition with the target routing node and the routing node quantity condition is selected, and the at least one candidate routing node is saved in the initial associated routing node set;

[0011] Iterate and modify the initial associated routing node set multiple times until a target associated routing node set that meets the associated routing node condition is obtained;

[0012] At least one candidate routing node included in the target associated routing node set is used as an associated routing node of the target routing node.

[0013] In a second aspect, an embodiment of the present application further provides a routing network expression device, the device comprising:

[0014] A determination module, configured to determine a basic routing node set corresponding to a target routing node based on a node identification interval to which a node identification of the target routing node belongs in a routing network; wherein at least one basic routing node included in the basic routing node set is a routing node of the same node level;

[0015] A screening module is used to screen out at least one candidate routing node from the basic routing node set that satisfies the attribute feature similarity condition with the target routing node and meets the routing node quantity condition, and save the at least one candidate routing node to the initial associated routing node set;

[0016] A modification module, used for iteratively modifying the initial associated routing node set for multiple times until a target associated routing node set that meets the associated routing node condition is obtained;

[0017] The processing module is used to use at least one candidate routing node included in the target associated routing node set as an associated routing node of the target routing node.

[0018] In an optional embodiment, before determining the basic routing node set corresponding to the target routing node based on the node identification interval to which the node identification of the target routing node belongs in the routing network, the determination module is further used to:

[0019] Based on the neighborhood feature intervals to which the node neighborhood features of the multiple basic routing nodes belong, the multiple basic routing nodes are layered to obtain multiple node levels;

[0020] The multiple basic routing nodes are saved in the basic routing node sets respectively set corresponding to the multiple node levels.

[0021] In an optional embodiment, when screening out at least one candidate routing node that satisfies the attribute feature similarity condition with the target routing node and meets the routing node quantity condition from the basic routing node set, the screening module is specifically used to:

[0022] Determine respectively the vector similarity between the attribute feature vector of at least one basic routing node and the attribute feature vector of the target routing node;

[0023] Based on the obtained at least one vector similarity and the routing node quantity threshold, at least one candidate routing node that meets the attribute feature similarity condition and the routing node quantity condition is screened out from the basic routing node set.

[0024] In an optional embodiment, when respectively determining the vector similarity between the attribute feature vector of at least one basic routing node and the attribute feature vector of the target routing node, the screening module is specifically used to:

[0025] Based on the number of basic routing nodes in the routing network, determine the corresponding binary tree depth;

[0026] For at least one basic routing node, perform the following operations respectively:

[0027] From the parent node sequence set associated with the basic routing node, filter out the parent node sequence that meets the node similarity condition with the target routing node and whose sequence length is the depth of the binary tree;

[0028] Based on the vector similarity between the attribute feature vector of the parent node sequence containing the sequence routing node and the attribute feature vector of the target routing node, the vector similarity between the attribute feature vector of the basic routing node and the attribute feature vector of the target routing node is obtained.

[0029] In an optional embodiment, when at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition is screened out from the basic routing node set based on the obtained at least one vector similarity and routing node quantity threshold, the screening module is specifically used to:

[0030] Based on at least one vector similarity, determining the joint similarity of at least one candidate routing node combination in the basic routing node set; wherein the number of candidate routing nodes included in each candidate routing node combination is a routing node quantity threshold;

[0031] Based on the obtained at least one joint similarity, determining a joint similarity ranking order of at least one candidate routing node combination;

[0032] The candidate routing nodes included in the candidate routing node combination that meets the joint similarity arrangement order condition are used as at least one candidate routing node that meets the attribute feature similarity condition and reaches the routing node quantity condition.

[0033] In an optional embodiment, when at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition is screened out from the basic routing node set based on the obtained at least one vector similarity and routing node quantity threshold, the screening module is specifically used to:

[0034] Based on at least one vector similarity, determining a vector similarity arrangement order of at least one candidate routing node;

[0035] From the basic routing node set, select candidate routing nodes that meet the vector similarity arrangement order condition and reach the routing node quantity threshold;

[0036] The obtained at least one candidate routing node is used as at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition.

[0037] In an optional embodiment, when the initial associated routing node set is modified iteratively for multiple times until a target associated routing node set that meets the associated routing node condition is obtained, the modification module is specifically used to:

[0038] During an iterative modification, the following operations are performed:

[0039] Based on the routing node association information of at least one candidate routing node currently included in the initial associated routing node set, obtaining a subset of associated routing nodes adjacent to the node level corresponding to the routing node association information;

[0040] Based on the attribute feature similarity between at least one candidate routing node included in the associated routing node subset and the target routing node, the initial associated routing node set is modified to obtain a modified initial associated routing node set.

[0041] In a third aspect, an embodiment of the present application further proposes an electronic device, comprising a processor and a memory, wherein the memory stores a program code, and when the program code is executed by the processor, the processor executes the steps of the routing network expression method described in the first aspect above.

[0042] In a fourth aspect, an embodiment of the present application further proposes a computer-readable storage medium, which includes a program code. When the program code is run on an electronic device, the program code is used to enable the electronic device to execute the steps of the routing network expression method described in the first aspect above.

[0043] In a fifth aspect, an embodiment of the present application further provides a computer program product, which, when called by a computer, enables the computer to execute the steps of the routing network expression method as described in the first aspect.

[0044] The beneficial effects of this application are as follows:

[0045] In the routing network expression method provided in the embodiment of the present application, an initial associated routing node set is obtained according to the basic routing node set corresponding to the target routing node, the attribute feature similarity condition, and the routing node quantity condition, and then the initial associated routing node set is iteratively modified multiple times to finally obtain the target associated routing node set. This avoids the technical drawbacks in the related art that when the number of candidate routing nodes included in the routing node set is large, a large amount of time is required to perform node similarity comparison between the target routing node and all candidate routing nodes, and thus the routing network expression cannot be completed in a timely manner. The routing network expression method through global structure analysis and local relationship discovery improves the efficiency of routing network expression.

[0046] In addition, other features and advantages of the present application will be described in the subsequent description, and partly become apparent from the description, or be understood by practicing the present application. The purpose and other advantages of the present application can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described here are used to provide a further understanding of the present application, constitute a part of the present application, and do not constitute an improper limitation on the present application. In the drawings:

[0048] Figure 1 A schematic diagram of the structure of a routing network provided in an embodiment of the present application;

[0049] Figure 2 An optional schematic diagram of a system architecture applicable to embodiments of the present application;

[0050] Figure 3 A schematic diagram of an implementation flow of a hierarchical classification method of basic routing nodes provided in an embodiment of the present application;

[0051] Figure 4 A schematic diagram of a specific application scenario of a K-core decomposition method provided in an embodiment of the present application;

[0052] Figure 5A A schematic diagram of a specific application scenario of a local collapse method provided in an embodiment of the present application;

[0053] Figure 5B A schematic diagram of a specific application scenario of a local collapse method provided in an embodiment of the present application;

[0054] Figure 6 A schematic diagram of an implementation flow of a routing network expression method provided in an embodiment of the present application;

[0055] Figure 7 A schematic diagram of an implementation flow of a method for screening candidate routing nodes provided in an embodiment of the present application;

[0056] Figure 8 A schematic diagram of an implementation flow of a method for screening candidate routing nodes provided in an embodiment of the present application;

[0057] Fig. 9 A method based on the embodiment of the present application is provided Figure 6 Schematic diagram of the

[0058] Fig.10 The schematic diagram of the structure of a device provided by an embodiment of the present application is exemplarily shown;

[0059] Fig.11 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is exemplarily shown. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the technical solution of the present application, rather than all of the embodiments. Based on the embodiments recorded in the application documents, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the technical solution of the present application.

[0061] It should be noted that in the description of this application, "multiple" is understood as "at least two". "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent three situations: A exists alone, A and B exist at the same time, and B exists alone. A and B are connected, which can represent two situations: A and B are directly connected and A and B are connected through C. In addition, in the description of this application, words such as "first" and "second" are only used for the purpose of distinguishing descriptions, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying order.

[0062] In addition, the collection, dissemination, and use of data in the technical solution of this application comply with the requirements of relevant national laws and regulations.

[0063] Some technical terms in the embodiments of the present application are explained below to facilitate understanding by those skilled in the art.

[0064] (1) K-core decomposition: It is a good method to identify specific subsets of a graph by recursively removing all nodes with degree less than K until the degree of the remaining nodes is greater than or equal to K. A node is said to have a K-core (or equivalently, a K-shell) if it belongs to the K-core but not to the (K+1)-core.

[0065] (2) Local collapse: Its core strategy is to merge the edge nodes and edges of the original network graph and then divide it into a series of hierarchical network graphs.

[0066] (3) Unsupervised learning model Skip-Gram: Also known as the skip-word model, Skip-Gram is a cutting-edge feature learning method in natural language processing. It can generate meaningful word representations based on the context of a sentence, that is, infer words within a certain window of the context through the central word.

[0067] (4) Hierarchical softmax: A Huffman tree is constructed for each word in the dictionary according to its frequency, ensuring that words with higher frequency are in a relatively shallow layer, and words with lower frequency are in the leaf nodes at a deeper level of the Huffman tree. Each word is in a leaf node on this Huffman tree.

[0068] In addition, the original |V| classification problem can be transformed into a log|V| binary classification problem. For example, when the probability of the target word is originally calculated, because the ordinary Softmax is used, the probability of each word in the dictionary must be required. In order to reduce the amount of calculation in this step, in the hierarchical Softmax, the probability of the current word in its context is also calculated. It only needs to be transformed into a path prediction problem in the Huffman tree. Because the current word corresponds to a path in the Huffman tree, this path starts from the root node in the binary tree, passes through a series of intermediate parent nodes, and finally reaches the leaf node of the current word. Then, at each parent node, it corresponds to a binary classification problem (essentially a logistic regression (LR) classifier), and the construction process of the Huffman tree ensures that the depth of the tree is log|V|, so only two classifications are needed to obtain the probability of the target word, which is greatly reduced compared to the original |V| times of calculation.

[0069] (5) Sigmiod function: It can also be written as σ(). Its range is (0, 1). It is often used for binary classification problems. The function is smooth and easy to differentiate. Therefore, it is usually used as the activation function of neural networks to map variables between 0 and 1.

[0070] Further, based on the above-mentioned nouns and related terminology explanations, the design concept of the embodiments of the present application is briefly introduced below:

[0071] At present, there are two basic methods of network expression: discrete representation and distributed representation; among them, the traditional topology-based network expression usually directly uses the adjacency matrix (discrete representation), but the adjacency matrix may contain noise or redundant information. The embedding-based network expression method (distributed representation) aims to learn the dense and continuous representation of network nodes in low-dimensional space, which not only reduces noise or redundant information, but also retains the inherent network information.

[0072] However, the current embedding-based network representation methods have not yet been used to analyze routing networks, that is, to express routing networks. Moreover, most network representation learning methods are basically based on network modeling of the original network, without considering the extraction and refinement of network information. Routing networks usually contain multiple devices such as routers, switches, hubs, and hosts (see Figure 1 As shown, each device can be regarded as a routing node), and the functions and importance of different devices are different, so the routing network itself carries "granularity" information.

[0073] For example, based on the characteristic that the routing network itself contains "granularity" information, a single routing node (device) can be regarded as a basic particle, the routing node and its neighbors can form a relatively coarse neighborhood granularity, and more nodes with similar attributes can form a coarser community granularity.

[0074] It can be seen that the network information of the original routing network is large and messy, that is, the positions and neighborhood relationships of different devices in the routing network are basically different. Therefore, it will be difficult to accurately extract and refine these network information by directly performing network representation learning on the original routing network.

[0075] Therefore, if we adopt the idea of ​​"divide and conquer" when modeling the routing network, that is, divide the original network of the routing network into different granular layers, each granular layer corresponds to nodes or edges with certain similar characteristics. In this way, we can analyze and parallelize the network at each granular layer, which can not only reduce the impact of other redundant information, but also reduce the complexity of the model through parallel computing.

[0076] In view of this, in an embodiment of the present application, in order to solve the above-mentioned problem, a (hierarchical) routing network expression method is proposed, which specifically includes: determining a basic routing node set corresponding to the target routing node based on the node identifier interval to which the node identifier of the target routing node in the routing network belongs; wherein at least one basic routing node contained in the basic routing node set is a routing node of the same node level; then, from the basic routing node set, screening out candidate routing nodes that meet the attribute feature similarity condition with the target routing node and meet the routing node quantity condition, and saving at least one candidate routing node to the initial associated routing node set; further, performing multiple iterations on the initial associated routing node set until a target associated routing node set that meets the associated routing node condition is obtained; finally, using at least one candidate routing node contained in the target associated routing node set as the associated routing node of the target routing node, which also avoids to a certain extent the technical disadvantage that when the number of candidate routing nodes contained in the routing node set is large, a large amount of time is required to perform node similarity comparison between the target routing node and each candidate routing node, and thus the routing network expression of the routing network cannot be completed in a timely manner, thereby improving the efficiency of the routing network expression.

[0077] See also Figure 2 As shown, it is a schematic diagram of a system architecture applicable to an embodiment of the present application, and the system architecture includes: a target terminal 201 and a server 202. The target terminal 201 and the server 202 can exchange information through a communication network, wherein the communication mode adopted by the communication network may include: a wireless communication mode and a wired communication mode.

[0078] Exemplarily, the target terminal 201 can access the network through cellular mobile communication technology and communicate with the server 202, wherein the cellular mobile communication technology, for example, includes the fifth generation mobile communication (5th Generation Mobile Networks, 5G) technology.

[0079] Optionally, the target terminal 201 may access the network and communicate with the server 202 via a short-range wireless communication method, wherein the short-range wireless communication method, for example, includes Wireless Fidelity (Wi-Fi) technology.

[0080] The embodiment of the present application does not impose any restriction on the number of communication devices involved in the above system architecture. For example, there may be multiple target terminals, or no target terminal, or other network devices may be included, such as Figure 2 As shown, only the target terminal 201 and the server 202 are described as examples, and the above-mentioned devices and their respective functions are briefly introduced below.

[0081] The target terminal 201 is a device that can provide voice and / or data connectivity to the user, and can be a device that supports wired and / or wireless connection.

[0082] Exemplarily, the target terminal 201 includes, but is not limited to: mobile phones, tablet computers, laptop computers, PDAs, mobile Internet devices (MID), wearable devices, virtual reality (VR) devices, augmented reality (AR) devices, wireless terminal devices in industrial control, wireless terminal devices in unmanned driving, wireless terminal devices in smart grids, wireless terminal devices in transportation safety, wireless terminal devices in smart cities, or wireless terminal devices in smart homes, etc.

[0083] In addition, the target terminal 201 may be installed with a relevant client, which may be software, such as an application (Application, APP), a browser, a short video software, etc., or a web page, a small program, etc. In the embodiment of the present application, the target terminal 201 may be used to send the network structure information of the routing network to the server 202 before the routing network is expressed, and to send a routing network expression request carrying the node identifier of the target routing node to the server 202 when the routing network is expressed.

[0084] Server 202 can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), as well as big data and artificial intelligence platforms.

[0085] It is worth mentioning that in the embodiment of the present application, the server 202 is used to determine the basic routing node set corresponding to the target routing node based on the node identification interval to which the node identification of the target routing node in the routing network belongs, and then screen out at least one candidate routing node from the basic routing node set that meets the attribute feature similarity condition with the target routing node and meets the routing node quantity condition, and save the at least one candidate routing node to the initial associated routing node set, thereby performing multiple iterative modifications on the initial associated routing node set until a target associated routing node set that meets the associated routing node condition is obtained, and then at least one candidate routing node contained in the target associated routing node set is used as the associated routing node of the target routing node.

[0086] The following describes the routing network expression method provided by the exemplary embodiment of the present application in combination with the above-mentioned system architecture and with reference to the accompanying drawings. It should be noted that the above-mentioned system architecture is only shown to facilitate understanding of the spirit and principles of the present application, and the implementation of the present application is not limited in this regard.

[0087] It should be noted that before the server determines the basic routing node set corresponding to the target routing node based on the node identifier of the target routing node in the routing network and the node identifier interval to which it belongs, it is necessary to classify the multiple basic routing nodes included in the routing network into node hierarchies and save each basic routing node into the corresponding basic routing node set. Figure 3 As shown, it is an implementation flow chart of a hierarchical classification method of a basic routing node provided in an embodiment of the present application. The execution subject takes a server as an example. The specific implementation process of the method is as follows:

[0088] S301: Based on the neighborhood feature intervals to which the node neighborhood features of the multiple basic routing nodes belong, multiple basic routing nodes are layered to obtain multiple node levels.

[0089] The node neighborhood characteristics of each basic routing node represent: the position of the corresponding basic routing node in the routing network, and the connection relationship between the corresponding basic routing node and other basic routing nodes.

[0090] Exemplarily, assuming that 5 basic routing nodes and 3 characteristic neighborhood intervals are taken as an example, the larger the characteristic neighborhood interval of the basic routing node is, the richer the node neighborhood characteristics of the basic routing node are, that is, the more central the routing network is, the higher the "importance" is, and the node neighborhood characteristics of the above 5 basic routing nodes and their corresponding neighborhood characteristic intervals are shown in Table 1:

[0091] Table 1

[0092] Basic routing node Base.No1 Base.No2 Base.No3 Base.No4 Base.No5 Node neighborhood characteristics Nei.Char1 Nei.Char2 Nei.Char3 Nei.Char4 Nei.Char5 Neighborhood feature interval 1 2 1 3 2

[0093] Therefore, based on the node neighborhood features of the basic routing nodes recorded in the above table and their respective corresponding neighborhood feature intervals, the above 5 basic routing nodes are classified into node levels to obtain the node levels of the above 5 basic routing nodes. For example, if the server determines that the node neighborhood feature Nei.Char1 of the basic routing node Base.No1 belongs to the neighborhood feature interval 1, then the node level of the basic routing node Base.No1 can be determined as A according to the correspondence between the preset neighborhood feature intervals and the node levels (for example, the correspondence is: 1-A, 2-B, 3-C).

[0094] In an optional embodiment, the node neighborhood features and their corresponding neighborhood feature intervals are determined according to the K-core decomposition method. For example, an undirected unweighted routing network is defined, G = (V, E) where V represents the set of all routing nodes in the routing network, E represents the set of edges, and different types of closed lines are used to highlight different cores. Figure 4 As shown, it is obvious Figure 4 Each routing node in belongs to core 1. The thick solid line surrounds all routing nodes in the entire routing network graph. Then, nodes with degree 1 are recursively deleted. The remaining network represents core 2 and is surrounded by thin solid circles. Further deletion of nodes can obtain the central core 3 and is surrounded by dotted circles. It should be noted that if a routing node belongs to core 3, it also belongs to core 2 and core 1.

[0095] Obviously, when the K-core decomposition method is used to determine the above-mentioned node neighborhood characteristics and their corresponding neighborhood characteristic intervals, the node neighborhood characteristics are the degree, and the neighborhood characteristic interval corresponding to the node neighborhood characteristics is the number of cores. To a certain extent, it is suitable for routing networks divided by device type.

[0096] In an optional embodiment, the node neighborhood features and their corresponding neighborhood feature intervals are determined by local collapse. For example, an undirected unweighted routing network is defined, G = (V, E), where V represents the set of all routing nodes in the routing network, and E represents the set of edges. Through a certain contraction strategy, multiple graphs G smaller than G are found. s =(V s , E s ), it should be noted that G s Number of routing nodes in |V s | is much smaller than the number of routing nodes in G |V|, denoted as: |V s |<<|V|,G s The number of edges in the routing network |E s | is much smaller than the number of edges in the routing network in G |E|, denoted as: |E s |<<|E|, that is, while reducing the scale of the routing network, the basic structure of the graph can be preserved to a great extent.

[0097] See also Figure 5A As shown in Figure 1, it is a merging process of hosts and routers in a routing network. In a routing network, edge routing nodes are mostly hosts, and more central routing nodes are mostly routers. Since a router may connect to multiple hosts, the host can be "collapsed" into the router and merged into a larger routing node, thereby reducing the size of the routing network; see Figure 5BAs shown in FIG. 1 , it is the process of merging routers in a routing network. The center of a routing network is mostly composed of routers. Through center merging, routing nodes that have a common center point as a neighbor can be merged two by two to further reduce the network scale of the routing network.

[0098] It should be noted that after repeated iterations of the above method, the server can obtain multiple routing networks of different sizes, among which the merged devices can be described by "coarse-grained nodes" and the original devices can be described by "fine-grained nodes". The above-mentioned merger of each level represents the corresponding neighborhood feature interval.

[0099] S302: multiple basic routing nodes are saved in basic routing node sets respectively set corresponding to multiple node levels.

[0100] Exemplarily, when executing step S302, still taking the above-mentioned 5 basic routing nodes and 3 node levels as an example, after obtaining the node levels of the above-mentioned 5 basic routing nodes in the routing network, the server can save the above-mentioned 5 basic routing nodes to the basic routing node sets set for each corresponding node level, wherein each node level and its corresponding basic routing node set are shown in Table 2:

[0101] Table 2

[0102]

[0103] Obviously, based on the above table, after the server obtains the node level A of the basic routing node Base.No1 in the routing network, it can determine the basic routing node set Base.Rou.Set1 set corresponding to the node level A based on the correspondence between the above node level and the basic routing node set, thereby saving the basic routing node to the basic routing node set Base.Rou.Set1.

[0104] In an optional embodiment, after the server saves the multiple basic routing nodes included in the routing network to the basic routing node sets set at the corresponding node levels, it can also establish the node identification of the basic routing node, the corresponding relationship between the identification interval and the basic routing node set, so that the corresponding basic routing node set can be directly determined by the node identification of the target routing node in the future, thereby improving the expression efficiency of the routing network to a certain extent.

[0105] Based on the above method steps, after the server saves multiple basic routing nodes in the routing network to the corresponding basic routing node set, it can express the routing network. Figure 6As shown, it is an implementation flow chart of a routing network expression method provided in an embodiment of the present application. The execution subject is still taken as an example of a server. The specific implementation process of the method is as follows:

[0106] S601: Determine a basic routing node set corresponding to a target routing node based on a node identifier interval to which a node identifier of a target routing node belongs in a routing network.

[0107] Among them, at least one basic routing node included in the basic routing node set is a routing node of the same node level.

[0108] Exemplarily, when executing step S601, the server may receive a routing network expression request from the target terminal, obtain the node identifier of the target routing node from the routing network expression request, and determine the identifier interval to which the node identifier belongs, and then based on the correspondence between the identifier interval and the basic routing node set, filter out the basic routing node set corresponding to the target routing node from each basic routing node set. It should be noted that the basic routing nodes corresponding to each node identifier in the same identifier interval have the same node level, that is, they belong to the same basic routing node set.

[0109] S602: Filter out at least one candidate routing node from the basic routing node set that satisfies the attribute feature similarity condition with the target routing node and meets the routing node quantity condition, and save the at least one candidate routing node to the initial associated routing node set.

[0110] Exemplarily, the attribute characteristics may represent basic attribute characteristics such as the device type of the routing node or the community to which the routing node belongs.

[0111] In an optional implementation, when executing step S602, after determining the set of basic routing nodes corresponding to the target routing node, the server can respectively determine the attribute feature vector of at least one basic routing node and the vector similarity between the attribute feature vector of the target routing node. It should be noted that since the core idea of ​​Skip-Gram is to maximize the possibility of context in the sequence, the attribute feature vectors of the above-mentioned basic routing nodes, the attribute feature vector of the target routing node, and each vector similarity can be obtained by Skip-Gram, wherein the context possibility calculation formula of Skip-Gram is defined as follows:

[0112]

[0113] Where w represents the size of the window, i.e., the length of the previous context or the length of the following context, Φ(i) represents the vector representation of node i, i.e., the (low-dimensional) attribute feature vector, which can be written as: Φ(i)∈R d , R drepresents the vector dimension of Φ(i), P r ({iw,...,i+w}\i|Φ(i)) represents the conditional probability that the sequence {iw,...,i+w} excluding node i is the context sequence of node i, P r (j|Φ(i)) represents the conditional probability that any node in the sequence {iw,...,i+w} that does not include node i is the context node of node i.

[0114] It can be seen that the above-set routing node number threshold can be expressed as 2w, which can be determined according to the actual situation. In addition, the vector similarity between the attribute feature vector of the above-mentioned basic routing node and the attribute feature vector of the target routing node can be expressed as conditional probability P r (j|Φ(i)) represents.

[0115] In an optional implementation, the server can determine the corresponding binary tree depth according to the number of basic routing nodes in the routing network when determining the vector similarity between the attribute feature vector of at least one basic routing node and the attribute feature vector of the target routing node, and thus perform the following operations for the at least one basic routing node: from the set of parent node sequences associated with the basic routing nodes, screen out the parent node sequence that meets the node similarity condition with the target routing node and whose sequence length is the binary tree depth, and then based on the vector similarity between the attribute feature vector of the sequence routing node included in the parent node sequence and the attribute feature vector of the target routing node, obtain the vector similarity between the attribute feature vector of the basic routing node and the attribute feature vector of the target routing node.

[0116] Based on the above binary tree method steps for calculating vector similarity, the prediction problem can be transformed into a problem of calculating the probability of maximizing a specific path in the hierarchy, that is, assigning routing nodes to the leaves of the binary tree, and then the binary tree classifier can be used to calculate the above P r (j|Φ(i)), illustratively, for each node j∈V, the hierarchical Softmax is assigned a set of tree nodes (T1, T2, ..., T h ), optionally, the specific path composed of tree nodes can be obtained based on parallel random walks, in which case the conditional probability P r (j|Φ(i)) can be defined as:

[0117]

[0118] Among them, P r (j|Φ(i)) represents the conditional probability that any node in the sequence {iw,...,i+w} that does not include node i is the context node of node i, T uis the parent node of node j, P r (T u |Φ(i)) represents a tree node (T1, T2, ..., T h ) any node T u is the conditional probability of the context node of the node, Represents a logical operation of rounding up.

[0119] Given that there are two cases on the path from the root node to the leaf node, left branch or right branch, which just corresponds to the logistic regression problem, the conditional probability P can be r (T u |Φ(i)) is defined as follows:

[0120]

[0121] Among them, σ() is the Sigmoid activation function, Ψ(T u ) represents node T u The vector representation of node i, Φ(i) represents the vector representation of node i, S u =0 means select the left branch, S u =1 means to select the right branch. It should be noted that since S u It can only take two values, 0 and 1, so the conditional probability P r (T u |Φ(i)) can be redefined as:

[0122] It should be noted that the depth of the above binary tree can be The vector similarity between the attribute feature vector of the above sequence routing node and the attribute feature vector of the target routing node can be expressed as conditional probability P r (T u |Φ(i)) indicates that the above parent node sequence set contains one or more tree nodes (T1, T2, ..., T h ), that is, the parent node sequence, so the node similarity condition can be maximized conditional probability P r (j|Φ(i)), which can be expressed as:

[0123]

[0124] That is, the product of the vector similarity between the sequence routing node and the target routing node in a specific path is the largest.

[0125] Next, after determining the vector similarity between the attribute feature vector of at least one basic routing node and the attribute feature vector of the target routing node, the server can screen out at least one candidate routing node that meets the attribute feature similarity condition and the routing node quantity condition from the basic routing node set based on the obtained at least one vector similarity and the routing node quantity threshold; wherein the routing node quantity condition is the routing node quantity threshold, and the routing node quantity threshold is set according to actual conditions.

[0126] In an alternative implementation, see Figure 7 As shown, the server can adopt the following method to screen out at least one candidate routing node that meets the attribute feature similarity condition and the routing node quantity condition from the basic routing node set based on the obtained at least one vector similarity and the routing node quantity threshold. The specific implementation process of this method is as follows:

[0127] S701: Determine a joint similarity of at least one candidate routing node combination in a basic routing node set based on at least one vector similarity.

[0128] The number of candidate routing nodes included in each candidate routing node combination is a routing node quantity threshold, and in at least one candidate routing node combination, there is at least one different candidate routing node between any two of them.

[0129] For example, if the set routing node quantity threshold is 2w, the joint similarity of the candidate routing node combinations can be expressed as follows:

[0130]

[0131] It should be noted that node i here represents the target routing node, P r ({iw,...,i+w}\i|Φ(i)) represents the joint similarity of the candidate routing node combination, Φ(i) represents the attribute feature vector of the target routing node i, P r (j|Φ(i) represents the vector similarity corresponding to the basic routing node j.

[0132] S702: Determine a joint similarity arrangement order of at least one candidate routing node combination based on the obtained at least one joint similarity.

[0133] For example, if five joint similarities are taken as an example, after determining the five joint similarities, the server can determine the joint similarity arrangement order of the five candidate routing node combinations, as shown in Table 3:

[0134] Table 3

[0135]

[0136]

[0137] Obviously, from the above table, it can be seen that the joint similarity arrangement order of the above five candidate routing node combinations (Can.Rou.Com.1, Can.Rou.Com.2, Can.Rou.Com.3, Can.Rou.Com.4 and Can.Rou.Com.5) is: 51432.

[0138] S703: taking the candidate routing nodes included in the candidate routing node combination that meets the joint similarity arrangement order condition as at least one candidate routing node that meets the attribute feature similarity condition and reaches the routing node quantity condition.

[0139] The candidate routing node combination that meets the joint similarity ranking condition is represented as: the candidate routing node combination with the highest joint similarity ranking or the largest joint similarity.

[0140] Exemplarily, still taking Table 3 as an example, it can be seen that the candidate routing node combination Can.Rou.Com.2 is a candidate routing node combination that meets the joint similarity arrangement order condition. Therefore, the candidate routing nodes with the routing node quantity threshold contained in the candidate routing node combination Can.Rou.Com.2 are taken as candidate routing nodes that meet the attribute feature similarity condition and reach the routing node quantity condition.

[0141] In an alternative implementation, see Figure 8 As shown, the server can also adopt the following method, based on the obtained at least one vector similarity and the routing node quantity threshold, to screen out at least one candidate routing node that meets the attribute feature similarity condition and the routing node quantity condition from the basic routing node set. The specific implementation process of this method is as follows:

[0142] S801: Determine a vector similarity arrangement order of at least one candidate routing node based on at least one vector similarity.

[0143] Exemplarily, taking 6 vector similarities as an example, after determining the above 6 vector similarities, the server can determine the vector similarity arrangement order of the 6 candidate routing nodes included in the basic routing node set, as shown in Table 4:

[0144] Table 4

[0145] Candidate routing nodes Vector Similarity Vector similarity sorting order Can.Rou.No.1 0.86 3 Can.Rou.No.2 0.22 6 Can.Rou.No.3 0.98 1 Can.Rou.No.4 0.55 5 Can.Rou.No.5 0.72 4 Can.Rou.No.6 0.94 2

[0146] Obviously, from the above table, it can be seen that the vector similarity arrangement order of the above 6 candidate routing nodes (Can.Rou.No.1, Can.Rou.No.2, Can.Rou.No.3, Can.Rou.No.4, Can.Rou.No.5 and Can.Rou.No.6) is: 361542.

[0147] S802: Filter out candidate routing nodes that meet the vector similarity arrangement order condition and reach the routing node quantity threshold from the basic routing node set.

[0148] Exemplarily, assuming that the above-set routing node quantity threshold is 4, the above-mentioned vector similarity arrangement order condition is: the top 3 vector similarities or the 3 with the highest vector similarities. If Table 4 is still taken as an example, the candidate routing nodes that meet the vector similarity arrangement order condition and reach the set routing node quantity threshold can be screened out from the 6 candidate routing nodes included in the basic routing node set: Can.Rou.No.1, Can.Rou.No.3 and Can.Rou.No.6.

[0149] S803: taking the obtained at least one candidate routing node as at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition.

[0150] Furthermore, after obtaining at least one candidate routing node, the server may save the obtained at least one candidate routing node into the initial associated routing node set.

[0151] Obviously, based on the above method steps S701 to S703 and S801 to S803, the attribute feature similarity condition can be maximized conditional probability P r ({iw,...,i+w}\i|Φ(i)), for the node set V of the routing network, it can be expressed as:

[0152]

[0153] Accordingly, the loss function can be defined as minimizing the inverse of its logarithmic function, and the smaller the result obtained by the loss function, the more accurate the characterization of the acquired initial associated routing node set, wherein the loss function can be defined as follows:

[0154]

[0155] It is worth noting that the server updates the node representation vector Φ by minimizing the loss function L, and finally obtains the optimized node vector representation. Moreover, since the hierarchical Softmax runs on the binary tree, the computational complexity is reduced from O(|V|) to O(log|V|).

[0156] S603: performing multiple iterations of modification on the initial associated routing node set until a target associated routing node set that meets the associated routing node condition is obtained.

[0157] In an optional implementation, when executing step S603, after obtaining the initial associated routing node set, the server may perform multiple iterative modifications on the initial associated routing node set, wherein in one round of iteration, the following operations are performed: based on the routing node association information of at least one candidate routing node currently contained in the initial associated routing node set, a subset of associated routing nodes adjacent to the node level corresponding to the initial associated routing node information is obtained, thereby modifying the initial associated routing node set based on the similarity in attribute features between at least one candidate routing node contained in the associated routing node subset and the target routing node, to obtain a modified initial associated routing node set.

[0158] It should be noted that, in each round of iteration, the basic routing nodes can be continuously screened based on the above method steps, and finally at least one candidate routing node that meets the attribute feature similarity conditions and the routing node quantity conditions is obtained, that is, a target associated routing node set that meets the associated routing node conditions is obtained; in addition, in the embodiment of the present application, the above method can be used for any basic routing node in the routing network, thereby completing the network expression of the entire routing network.

[0159] S604: Use at least one candidate routing node included in the target associated routing node set as an associated routing node of the target routing node.

[0160] Specifically, when executing step S604, after obtaining the target associated routing node set, the server can use at least one candidate routing node included in the target associated routing node set as an associated routing node of the target routing node, thereby completing the routing network expression of the target routing node.

[0161] In an optional implementation, based on the routing network expression method of S601 to S604 above, refer to Fig. 9As shown, it is a schematic diagram of a specific application scenario of a routing network expression method provided by an embodiment of the present application. The server determines a basic routing node set (e.g., a second basic routing node set) corresponding to the target routing node Dest.Rou.Node based on the node identification interval (e.g., a second node identification interval) to which the node identification Node.ID of the target routing node Dest.Rou.Node in the routing network Rou.Network belongs; then, from the basic routing node set (the second basic routing node set), select the basic routing node set (e.g., the second basic routing node set) that meets the attribute feature similarity condition with the target routing node Dest.Rou.Node and meets the routing node quantity condition (e.g., 4) (e.g., candidate routing node 1, candidate routing node 2, candidate routing node 3, and candidate routing node 4), and save each obtained candidate routing node into an initial associated routing node set Init.Ass.Set; further, the initial associated routing node set Init.Ass.Set is iteratively modified multiple times until a target associated routing node set Tar.Ass.Set that meets the associated routing node condition is obtained; finally, at least one candidate routing node included in the target associated routing node set Tar.Ass.Set is used as an associated routing node of the target routing node Dest.Rou.Node.

[0162] To summarize, in the routing network expression method provided in the embodiment of the present application, an initial associated routing node set is obtained based on the basic routing node set corresponding to the target routing node, the attribute feature similarity condition, and the routing node quantity condition, and then the initial associated routing node set is iteratively modified multiple times to finally obtain the target associated routing node set.

[0163] This approach avoids the technical drawbacks in related technologies that, when the routing node set contains a large number of candidate routing nodes, it takes a lot of time to compare the node similarities between the target routing node and all candidate routing nodes, and thus it is impossible to ensure that the routing network expression is completed in a timely manner. The routing network expression method based on global structure analysis and local relationship discovery improves the efficiency of routing network expression.

[0164] Further, based on the same technical concept, the present application embodiment provides a routing network expression device, which is used to implement the above method flow of the present application embodiment. Fig.10 As shown, the routing network expression device includes: a determination module 1001, a screening module 1002, a modification module 1003 and a processing module 1004, wherein:

[0165] The determination module 1001 is used to determine a basic routing node set corresponding to the target routing node based on the node identification interval to which the node identification of the target routing node belongs in the routing network; wherein at least one basic routing node included in the basic routing node set is a routing node of the same node level;

[0166] A screening module 1002 is used to screen out at least one candidate routing node from the basic routing node set that satisfies the attribute feature similarity condition with the target routing node and meets the routing node quantity condition, and save the at least one candidate routing node to the initial associated routing node set;

[0167] A modification module 1003 is used to perform multiple iterations of modification on the initial associated routing node set until a target associated routing node set that meets the associated routing node condition is obtained;

[0168] The processing module 1004 is configured to use at least one candidate routing node included in the target associated routing node set as an associated routing node of the target routing node.

[0169] In an optional embodiment, before determining the basic routing node set corresponding to the target routing node based on the node identification interval to which the node identification of the target routing node belongs in the routing network, the determination module 1001 is further used to:

[0170] Based on the neighborhood feature intervals to which the node neighborhood features of the multiple basic routing nodes belong, the multiple basic routing nodes are layered to obtain multiple node levels;

[0171] The multiple basic routing nodes are saved in the basic routing node sets respectively set for each node level.

[0172] In an optional embodiment, when screening out at least one candidate routing node that satisfies the attribute feature similarity condition with the target routing node and meets the routing node quantity condition from the basic routing node set, the screening module 1002 is specifically used to:

[0173] Determine respectively the vector similarity between the attribute feature vector of at least one basic routing node and the attribute feature vector of the target routing node;

[0174] Based on the obtained at least one vector similarity and the routing node quantity threshold, at least one candidate routing node that meets the attribute feature similarity condition and the routing node quantity condition is screened out from the basic routing node set.

[0175] In an optional embodiment, when respectively determining the vector similarity between the attribute feature vector of at least one basic routing node and the attribute feature vector of the target routing node, the screening module 1002 is specifically used to:

[0176] Based on the number of basic routing nodes in the routing network, determine the corresponding binary tree depth;

[0177] For at least one basic routing node, perform the following operations respectively:

[0178] From the parent node sequence set associated with the basic routing node, filter out the parent node sequence that meets the node similarity condition with the target routing node and whose sequence length is the depth of the binary tree;

[0179] Based on the vector similarity between the attribute feature vector of the parent node sequence containing the sequence routing node and the attribute feature vector of the target routing node, the vector similarity between the attribute feature vector of the basic routing node and the attribute feature vector of the target routing node is obtained.

[0180] In an optional embodiment, when selecting candidate routing nodes that meet the attribute feature similarity condition and at least one routing node quantity condition from the basic routing node set based on the obtained at least one vector similarity and routing node quantity threshold, the screening module 1002 is specifically used to:

[0181] Based on at least one vector similarity, determining the joint similarity of at least one candidate routing node combination in the basic routing node set; wherein the number of candidate routing nodes included in each candidate routing node combination is a routing node quantity threshold;

[0182] Based on the obtained at least one joint similarity, determining a joint similarity ranking order of at least one candidate routing node combination;

[0183] The candidate routing nodes included in the candidate routing node combination that meets the joint similarity arrangement order condition are used as at least one candidate routing node that meets the attribute feature similarity condition and reaches the routing node quantity condition.

[0184] In an optional embodiment, when at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition is screened out from the basic routing node set based on the obtained at least one vector similarity and routing node quantity threshold, the screening module 1002 is specifically used to:

[0185] Based on at least one vector similarity, determining a vector similarity arrangement order of at least one candidate routing node;

[0186] From the basic routing node set, select candidate routing nodes that meet the vector similarity arrangement order condition and reach the routing node quantity threshold;

[0187] The obtained at least one candidate routing node is used as at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition.

[0188] In an optional embodiment, when the initial associated routing node set is modified iteratively for multiple times until a target associated routing node set that meets the associated routing node condition is obtained, the modification module 1003 is specifically used to:

[0189] During an iterative modification, the following operations are performed:

[0190] Based on the routing node association information of at least one candidate routing node currently included in the initial associated routing node set, obtaining a subset of associated routing nodes adjacent to the node level corresponding to the routing node association information;

[0191] Based on the attribute feature similarity between at least one candidate routing node included in the associated routing node subset and the target routing node, the initial associated routing node set is modified to obtain a modified initial associated routing node set.

[0192] Based on the same technical concept, the embodiment of the present application also provides an electronic device, which can implement the routing network expression method process provided in the above embodiment of the present application. In one embodiment, the electronic device can be a server, or a terminal device or other electronic device. Fig.11 As shown, the electronic device may include:

[0193] At least one processor 1101, and a memory 1102 connected to the at least one processor 1101. The specific connection medium between the processor 1101 and the memory 1102 is not limited in the embodiment of the present application. Fig.11 In the example, the processor 1101 and the memory 1102 are connected via the bus 1100. The bus 1100 is Fig.11 The bus 1100 is represented by a bold line, and the connection between other components is only for schematic illustration and is not intended to be limiting. The bus 1100 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig.11 Only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. Alternatively, the processor 1101 can also be called a controller, and there is no limitation on the name.

[0194] In the embodiment of the present application, the memory 1102 stores instructions that can be executed by at least one processor 1101. The at least one processor 1101 can execute a routing network expression method discussed above by executing the instructions stored in the memory 1102. The processor 1101 can implement Fig.10 The functions of each module in the device shown.

[0195] Among them, the processor 1101 is the control center of the device, and can use various interfaces and lines to connect the various parts of the entire control device. By running or executing instructions stored in the memory 1102 and calling the data stored in the memory 1102, the various functions of the device and processing data, the device can be monitored as a whole.

[0196] In an optional design, the processor 1101 may include one or more processing units, and the processor 1101 may integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the modem processor may not be integrated into the processor 1101. In some embodiments, the processor 1101 and the memory 1102 may be implemented on the same chip, and in some embodiments, they may also be implemented separately on separate chips.

[0197] Processor 1101 may be a general-purpose processor, such as a CPU, a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component, and may implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of a routing network expression method disclosed in the embodiments of the present application may be directly embodied as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0198] Memory 1102, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer executable programs and modules. Memory 1102 may include at least one type of storage medium, such as flash memory, hard disk, multimedia card, card-type memory, random access memory (Random Access Memory, RAM), static random access memory (Static Random Access Memory, SRAM), programmable read-only memory (Programmable Read Only Memory, PROM), read-only memory (Read Only Memory, ROM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), magnetic memory, disk, optical disk, etc. Memory 1102 is any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory 1102 in the embodiment of the present application can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.

[0199] By designing and programming the processor 1101, the code corresponding to the routing network expression method described in the above embodiment can be fixed into the chip, so that the chip can execute Figure 6 The steps of a routing network expression method in the embodiment shown are as follows: How to design and program the processor 1101 is a technology well known to those skilled in the art and will not be described in detail here.

[0200] Based on the same inventive concept, an embodiment of the present application further provides a storage medium, which stores computer instructions. When the computer instructions are executed on a computer, the computer executes a routing network expression method discussed above.

[0201] In some optional embodiments, the present application also provides various aspects of a routing network expression method, which can also be implemented in the form of a program product, which includes program code. When the program product is run on an apparatus, the program code is used to enable the control device to execute the steps of a routing network expression method according to various exemplary embodiments of the present application described above in this specification.

[0202] It should be noted that, although several units or subunits of the device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided into multiple units to be embodied.

[0203] In addition, although the operations of the method of the present application are described in a specific order in the drawings, this does not require or imply that the operations must be performed in this specific order, or that all the operations shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0204] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0205] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a server, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0206] Program code for performing the operations of the present application may be written using any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a stand-alone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0207] Where a remote computing device is involved, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0208] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0209] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0210] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents.

Claims

1. A method for expressing a routing network, characterized in that: include: Based on the node identification interval to which the node identification of the target routing node in the routing network belongs, determine the basic routing node set corresponding to the target routing node; wherein at least one basic routing node included in the basic routing node set is a routing node of the same node level; From the basic routing node set, at least one candidate routing node that meets the attribute feature similarity condition with the target routing node and the routing node quantity condition is selected, and the at least one candidate routing node is saved in the initial associated routing node set; The initial associated routing node set is modified iteratively for multiple times until a target associated routing node set that meets the associated routing node condition is obtained; wherein, in one round of iterative modification, the following operations are performed: based on routing node association information of at least one candidate routing node currently included in the initial associated routing node set, a subset of associated routing nodes adjacent to the node level corresponding to the routing node association information is obtained, and based on the attribute feature similarity between at least one candidate routing node included in the associated routing node subset and the target routing node, the initial associated routing node set is modified to obtain a modified initial associated routing node set; At least one candidate routing node included in the target associated routing node set is used as an associated routing node of the target routing node.

2. The method according to claim 1, characterized in that Before determining the basic routing node set corresponding to the target routing node based on the node identification interval to which the node identification of the target routing node in the routing network belongs, the method further includes: Based on the neighborhood feature intervals to which the node neighborhood features of the multiple basic routing nodes belong, the multiple basic routing nodes are hierarchically classified to obtain multiple node levels; The multiple basic routing nodes are saved in basic routing node sets respectively set corresponding to the multiple node levels.

3. The method according to claim 1 or 2, characterized in that The step of selecting at least one candidate routing node from the basic routing node set that satisfies the attribute feature similarity condition with the target routing node and meets the routing node quantity condition includes: respectively determining the vector similarity between the attribute feature vector of the at least one basic routing node and the attribute feature vector of the target routing node; Based on the obtained at least one vector similarity and routing node quantity threshold, at least one candidate routing node that meets the attribute feature similarity condition and the routing node quantity condition is screened out from the basic routing node set.

4. The method according to claim 3, characterized in that The respectively determining the vector similarity between the attribute feature vector of the at least one basic routing node and the attribute feature vector of the target routing node comprises: Determining a corresponding binary tree depth based on the number of basic routing nodes of the routing network; For the at least one basic routing node, the following operations are performed respectively: Filter out, from the parent node sequence set associated with the basic routing node, a parent node sequence that satisfies the node similarity condition with the target routing node and whose sequence length is the depth of the binary tree; Based on the vector similarity between the attribute feature vector of the sequence routing node included in the parent node sequence and the attribute feature vector of the target routing node, the vector similarity between the attribute feature vector of the basic routing node and the attribute feature vector of the target routing node is obtained.

5. The method according to claim 3, characterized in that The step of selecting, based on the obtained at least one vector similarity and the routing node quantity threshold, at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition from the basic routing node set comprises: Based on the at least one vector similarity, determining the joint similarity of at least one candidate routing node combination in the basic routing node set; wherein the number of candidate routing nodes included in each candidate routing node combination is the routing node quantity threshold; Based on the obtained at least one joint similarity, determining a joint similarity arrangement order of the at least one candidate routing node combination; The candidate routing nodes included in the candidate routing node combination that meets the joint similarity arrangement order condition are used as at least one candidate routing node that meets the attribute feature similarity condition and reaches the routing node quantity condition.

6. The method according to claim 3, characterized in that The step of selecting, based on the obtained at least one vector similarity and the routing node quantity threshold, at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition from the basic routing node set comprises: Based on the at least one vector similarity, determining a vector similarity arrangement order of the at least one candidate routing node; Selecting candidate routing nodes that meet the vector similarity arrangement order condition and reach the routing node quantity threshold from the basic routing node set; The obtained at least one candidate routing node is used as at least one candidate routing node that satisfies the attribute feature similarity condition and reaches the routing node quantity condition.

7. A routing network expression device, characterized in that: include: A determination module, configured to determine a basic routing node set corresponding to a target routing node based on a node identification interval to which a node identification of the target routing node belongs in a routing network; wherein at least one basic routing node included in the basic routing node set is a routing node of the same node level; A screening module, configured to screen out candidate routing nodes from the basic routing node set that satisfy the attribute feature similarity condition with the target routing node and the routing node quantity condition, and save at least one candidate routing node to the initial associated routing node set; A modification module, configured to perform multiple iterations of modification on the initial associated routing node set until a target associated routing node set that satisfies the associated routing node condition is obtained; wherein, during one round of iterations of modification, the following operations are performed: based on routing node association information of at least one candidate routing node currently included in the initial associated routing node set, a subset of associated routing nodes adjacent to a node level corresponding to the routing node association information is obtained; and based on the attribute feature similarity between at least one candidate routing node included in the associated routing node subset and the target routing node, the initial associated routing node set is modified to obtain a modified initial associated routing node set; The processing module is used to use at least one candidate routing node included in the target associated routing node set as an associated routing node of the target routing node.

8. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Citation Information

Patent Citations

  • Display method and device, electronic equipment and storage medium

    CN111078967A

  • Information processing method, graph convolutional neural network training method and electronic equipment

    CN114399028A