KR20230094229A - Apparatus and method for generating subgraph for recommending terminal group corresponding to service request - Google Patents

Abstract

The present invention relates to a subgraph generating method for terminal recommendation in response to a service request, comprising the steps of: expressing edge resource information of an edge by a line graph, wherein the edge is made of a terminal; building a subgraph by inputting the edge resource information based on an artificial intelligence-based edge analysis model and by deriving relationship information about the corresponding edge; and evaluating the subgraph by determining, based on the subgraph, whether a starting terminal and at least one terminal connected from the starting terminal can provide a service.

Description

Apparatus and method for generating subgraphs for recommending terminal clusters corresponding to service requests

The present disclosure relates to an apparatus and method for generating a subgraph for recommending a terminal cluster corresponding to a service request. For example, the present disclosure relates to an apparatus and method for generating a subgraph for recommending a terminal cluster corresponding to a service request in a cellular network.

A technique for selecting and recommending a terminal suitable for providing a corresponding service in response to a service request of a specific terminal within a cellular network may be applied in various ways. For example, in the prior art, a method of providing a corresponding service by selecting a terminal having the best resource information from a list of provider terminals capable of providing the corresponding service has been mainly applied.

In addition, a method of presenting a recommendation ranking of provider terminals based on records capable of evaluating provider terminals, such as the number of times each provider terminal is selected and satisfaction (preference) with respect to the provider terminal, may be applied. In this case, various recommendation algorithms may be applied. can As another example, a method of recommending a terminal having resource information similar to the requester terminal or the selected terminal may be considered.

However, in the case of the above-mentioned provider terminal recommendation method, since it is evaluated whether the corresponding terminal (node) is suitable for providing a specific service by utilizing the location or resource information of each terminal (node), the terminal (node) participating in the network The higher the number, the lower the efficiency.

In addition, the method of recommending a provider terminal based on the similarity between terminals also has a limitation in that a lot of time and calculation are required because a new algorithm must be applied after resource recommendation through evaluation. In order to make a hierarchical recommendation between candidate terminals, there is a problem that a complicated additional algorithm such as a descending order sorting algorithm must be applied.

The background technology of the present application is disclosed in Korean Patent Registration No. 10-2208094.

The present invention is intended to solve the problems of the prior art described above, and responds to a service request capable of recommending a provider terminal suitable for a predetermined service in consideration of an edge defining a relationship between terminals, rather than considering only resource information of individual terminals. An object of the present invention is to provide an apparatus and method for generating a subgraph for recommending a terminal cluster.

The present invention is to solve the above-mentioned problems of the prior art, and provides a terminal cluster in response to a service request that can provide distributed services from one or more provider terminals having small resources to a single requester terminal requiring a large number of resources. An object of the present invention is to provide a device and method for generating a subgraph for recommendation.

However, the technical problem to be achieved by the embodiments of the present application is not limited to the technical problems described above, and other technical problems may exist.

As a technical means for achieving the above technical problem, a method for generating a subgraph for recommending a terminal corresponding to a service request according to an embodiment of the present invention includes expressing edge resource information of an edge composed of terminals in a line graph; Based on an artificial intelligence-based edge analysis model, constructing a sub-graph by deriving relation information about a corresponding edge with the edge resource information as an input, and a starting terminal and at least one terminal connected from the starting terminal based on the sub-graph It may include evaluating the sub-graph by determining whether the service of can be provided.

According to an embodiment of the present application, in the constructing of the subgraph, whether each edge is active may be determined based on the relationship information, and the subgraph may be constructed based on the activeness.

According to an embodiment of the present invention, the constructing of the subgraph comprises: constructing the subgraph including an activated edge, and expressing the activated edge to extend from the start terminal to the end terminal; At least some of the terminals among the plurality of terminals in the network may be hierarchically constructed.

According to an embodiment of the present application, in the step of evaluating the subgraph, it is possible to determine whether the service can be provided by summing resource information that can be provided through at least one terminal connected from the starting terminal.

According to an embodiment of the present application, the line graph may be a graph expressed by utilizing the terminals in the network to include nodes generated based on the edge resource information.

According to an embodiment of the present application, resource information expressed as an average value of a plurality of resource information of the start terminal and the end terminal, resource information replaced by a maximum value of the plurality of resource information, and the plurality of resource information It may include at least one resource information among resource information expressed as a subtraction result value.

According to an embodiment of the present application, the resource information may include at least one of storage capacity information, computing capability information, memory information, and sensor mounting information of each terminal.

According to an embodiment of the present application, the relationship information may be information related to the validity of the edge including a possibility of providing a service corresponding to the service request when the start terminal and the end terminal are connected.

According to an embodiment of the present application, the edge analysis model may be an artificial intelligence model learned to output relationship information about a corresponding edge when the edge resource information is input.

According to an embodiment of the present invention, the subgraph may be generated in consideration of resource information required by a corresponding service included in the service request.

An apparatus for generating a subgraph for recommending a terminal corresponding to a service request according to an embodiment of the present invention provides resource information of each terminal in a network and edge resource information of an edge composed of a pair of terminals based on the resource information as a line graph. Line graph generation unit expressed as; a sub-graph construction unit that constructs a sub-graph by deriving relational information about a corresponding edge based on an artificial intelligence-based edge analysis model based on the edge resource information as an input; and an evaluation unit that evaluates the subgraph by determining whether a starting terminal and at least one terminal connected from the starting terminal can provide a service based on the subgraph.

According to an embodiment of the present application, the sub-graph building unit may determine whether each edge is active based on the relationship information, and build a sub-graph based on the active state.

According to an embodiment of the present invention, the subgraph construction unit constructs the subgraph including an activated edge and expressing the activated edge to extend from the start terminal to the end terminal, at least some of a plurality of terminals in the network. It can be hierarchically built for the terminals of.

According to an embodiment of the present application, the evaluator may determine whether the service can be provided by summing up resource information that can be provided through all terminals between the start terminal and the end terminal.

The above-described problem solving means are merely exemplary and should not be construed as intended to limit the present disclosure. In addition to the exemplary embodiments described above, additional embodiments may exist in the drawings and detailed description of the invention.

According to the above-described problem solving means of the present application, it is possible to recommend a terminal corresponding to a service request capable of recommending a provider terminal suitable for a predetermined service in consideration of an edge defining a relationship between terminals, rather than considering only resource information of individual terminals. It is possible to provide an apparatus and method for generating a subgraph for

According to the above-described problem solving means of the present application, the resource information of each terminal is changed to edge resource information between terminals, the relationship between terminals in the network is learned through a learning neural network, and the learned relationship corresponds to a specific service requested. You can build a subgraph.

According to the above-described problem solving means of the present application, a recommendation for a service providing terminal can be made based on a subgraph that is hierarchically generated for a plurality of terminals based on edge information, so that even if there are a plurality of terminals in the network, efficient evaluation can be made.

According to the above-mentioned problem solving means of the present application, an updated subgraph can be provided more quickly in response to a change in state in the network.

According to the foregoing problem solving means of the present invention, it is possible to provide a distributed service from one or more provider terminals having few resources to a single requestor terminal requiring many resources.

However, the effects obtainable herein are not limited to the effects described above, and other effects may exist.

1 is a schematic configuration diagram of a service providing system including a subgraph generating apparatus for recommending a terminal cluster corresponding to a service request according to an embodiment of the present invention.
2 is a conceptual diagram for explaining the relationship between a requester terminal and a provider terminal for a predetermined service.
3 is a conceptual diagram for explaining the entire process of constructing a subgraph.
4 is a conceptual diagram for explaining the definition of an edge and a line graph in a network.
5 is a diagram exemplarily illustrating target values set to perform supervised learning based on information on whether a predetermined service is provided through a start terminal and an end terminal.
6 is a diagram showing a subgraph including a requestor terminal and a provider terminal according to whether an edge is active.
7 is a diagram showing a hierarchical structure of subgraphs.
8 is a diagram visually illustrating a process of summing up resource information of each terminal to determine whether a service can be provided.
9 is a schematic configuration diagram of an apparatus for generating a subgraph for recommending a terminal corresponding to a service request according to an embodiment of the present invention.
10 is an operation flowchart of a method for generating a subgraph for recommending a terminal corresponding to a service request according to an embodiment of the present invention.

Hereinafter, embodiments of the present application will be described in detail so that those skilled in the art can easily practice with reference to the accompanying drawings. However, the present disclosure may be implemented in many different forms and is not limited to the embodiments described herein. And in order to clearly describe the present application in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the present specification, when a part is said to be “connected” to another part, it is not only “directly connected”, but also “electrically connected” or “indirectly connected” with another element in between. "Including cases where

Throughout the present specification, when a member is referred to as being “on,” “above,” “on top of,” “below,” “below,” or “below” another member, this means that a member is located in relation to another member. This includes not only the case of contact but also the case of another member between the two members.

Throughout the present specification, when a certain component is said to "include", it means that it may further include other components without excluding other components unless otherwise stated.

The present disclosure relates to an apparatus and method for generating a subgraph for recommending a terminal cluster corresponding to a service request. For example, the present disclosure relates to an apparatus and method for generating a subgraph for recommending a terminal cluster corresponding to a service request in a cellular network.

1 is a schematic configuration diagram of a service providing system including a subgraph generating device for recommending a terminal cluster corresponding to a service request according to an embodiment of the present invention.

Referring to FIG. 1 , the service providing system 10 according to an embodiment of the present invention includes a sub-graph generation device 100 (hereinafter referred to as 'sub-graph') for recommending a terminal cluster corresponding to a service request according to an embodiment of the present invention. It may include a graph generating device 100'), a requester terminal 1 and a provider terminal 2. Also, referring to FIG. 1 , a plurality of provider terminals 2 may be provided.

The subgraph generator 100 , the requester terminal 1 and the provider terminal 2 may communicate with each other through the network 20 . The network 20 means a connection structure capable of exchanging information between nodes such as terminals and servers, and an example of such a network 20 includes a 3rd Generation Partnership Project (3GPP) network, LTE ( Long Term Evolution (Long Term Evolution) network, 5G network, WIMAX (World Interoperability for Microwave Access) network, Internet (Internet), LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), wifi network, Bluetooth network, satellite broadcasting network, analog broadcasting network, DMB (Digital Multimedia Broadcasting) network, etc. are included, but are not limited thereto. Meanwhile, the network 20 in the description of the embodiment of the present application may mainly mean a cellular network.

The requestor terminal 1 and the provider terminal 2 are, for example, a smartphone, a smart pad, a tablet PC, and the like, and PCS (Personal Communication System), GSM (Global System for Mobile communication), PDC (Personal Digital Cellular), Personal Handyphone System (PHS), Personal Digital Assistant (PDA), International Mobile Telecommunication (IMT)-2000, Code Division Multiple Access (CDMA)-2000, W-Code Division Multiple Access (W-CDMA), Wibro (Wireless Broadband Internet) It can be any kind of wireless communication device such as a terminal.

2 is a conceptual diagram for explaining a relationship between a requester terminal and a provider terminal for a predetermined service.

Referring to FIG. 2, in the description of the embodiment of the present application, the requester terminal 1 requests that the corresponding service be provided when a predetermined service can be provided through a connection (connection) between two terminals. It may refer to a user terminal requesting a search for the provider terminal 2 for providing the corresponding service. In addition, the provider terminal 2 is at least one of a plurality of terminals participating in the same network as the requester terminal 1 (eg, a cellular network) and can provide a predetermined service to the requester terminal 1. It may refer to a user terminal that possesses available resources and utilizes these resources to provide a specific service through connection (connection) with the requestor terminal 1.

In other words, in the description of the embodiment of the present application, the requestor terminal 1 requests a predetermined service from the provider terminal 2 (service request), and the provider terminal 2 sends the requested service to the requester terminal 1 can be provided to (service provision). In addition, the subgraph generating apparatus 100 of the present invention is based on relationship information about an edge defined between a plurality of provider terminals 2 in the network 20 including the requester terminal 1 for a predetermined service. Thus, a subgraph including information on the provider terminal 2 suitable for providing the corresponding service to the requester terminal 1 may be generated and provided.

Hereinafter, with reference to FIGS. 3 to 6 , specific functions and operations of the apparatus 100 for generating a subgraph will be described in detail.

3 is a conceptual diagram for explaining the entire process of constructing a subgraph.

Referring to FIG. 3, the subgraph generation apparatus 100 derives edge relationship information from terminal resource information of terminals (eg, a requester terminal 1 and a plurality of provider terminals 2, etc.) in the network 20. and builds a learning neural network to determine whether an edge is active in the subgraph that is finally output based on the derived relationship information, and creates (discovers) a subgraph for a given service based on the built learning neural network. It can be done ('Step 1: Subgraph discovery' in FIG. 3). Also, although not shown in FIG. 3 , the sub-graph generating apparatus 100 may perform re-learning of the learning neural network according to circumstances.

In addition, referring to FIG. 3, the subgraph generating apparatus 100 aggregates resource information of the requester terminal 1 and one or a plurality of provider terminals based on the generated (discovered) subgraph, and aggregates the resource information. By comparing with the service resource requirements of the service requested by the requester terminal 1, it is possible to determine whether the service can be provided, and to evaluate the subgraph according to whether the service can be provided ('Step 2: Subgraph evaluation' in FIG. 3). ).

Specifically, the subgraph generating apparatus 100 may calculate edge resource information for each edge composed of a pair of terminals based on resource information of each terminal in the network 20 . In other words, the subgraph generating apparatus 100 may define a plurality of edges defined between a pair of terminals among a plurality of terminals in the network 20 and calculate edge resource information corresponding to the defined edges.

In addition, 'subgraph' refers to a subgraph consisting of a subset of vertices (nodes) and edges (edges) belonging to the graph for a graph generated to include terminals participating in the network 20, Depending on the embodiment of the present application, it may be otherwise referred to as a partial graph or the like.

4 is a conceptual diagram for explaining the definition of an edge and a line graph in a network.

Referring to (a) of FIG. 4 , an edge E in the present application may be represented on a subgraph in the form of a line connecting a pair of nodes N corresponding to each terminal in the network 20 .

In this regard, the subgraph generating apparatus 100 does not individually utilize resource information of each terminal (node), but derives a resource relationship between terminals and uses the derived resource relationship between terminals to at least one terminal. A subgraph reflecting information on whether a predetermined service can be provided through connection can be created. In addition, for this purpose, the subgraph generating apparatus 100 may derive edge resource information from resource information of a pair of terminals (nodes) related to each edge E displayed in the form of a line on the subgraph. there is.

In addition, according to an embodiment of the present application, one 'edge' in the present application may be determined to include a start terminal and an end terminal, and a direction for each edge may be defined. In this regard, the sub-graph generating apparatus 100 may calculate edge resource information by subtracting each resource information of the starting terminal from each of at least one type of resource information of the end terminal, but is not limited thereto. Edge resource information may be calculated in various ways.

Specifically, according to an embodiment of the present application, the edge resource information is resource information expressed as an average value of a plurality of resource information of a start terminal and an end terminal, resource information replaced by a maximum value of a plurality of resource information, and a plurality of resources. It may include resource information expressed as a result value of subtraction between pieces of information, and the edge resource information may include resource information generated based on the similarity of each resource information of the start terminal and the end terminal and the size relationship of each resource information. It may be resource information generated based on comparison, but is not limited thereto. In other words, the edge resource information is generated (calculated) in various ways that can define the relationship (correlation) between the resource information of both terminals based on a plurality of types of resource information included in the start terminal and the end terminal. it could be

In addition, according to an embodiment of the present application, it is assumed that all terminals (eg, a plurality of candidate terminals that can be selected as the provider terminal 2) in the network 20 can connect (access) to each other, and A relationship between the terminals may be defined, and according to this, (a) of FIG. 4 may represent a network structure of a cellular network composed of four terminals (A, B, C, and D).

In addition, according to an embodiment of the present application, the resource information of the terminal may include at least one of storage capacity information, computing capability information, memory information, and sensor mounting information of each terminal.

As a specific example, the storage capacity among the resource information may refer to the capacity of a storage device mounted in a terminal. In addition, among the resource information, computational capability information may mean information on computational performance of an computational processing unit mountable in a terminal, such as a CPU or a GPU. Also, the memory information may include RAM information. In addition, the sensor mounting information may indicate whether a specific type of IoT (Internet-of-Things) sensor, such as a temperature sensor, is installed in a corresponding terminal. However, the above-described resource information is not limited to the above-described type, and various resource information may be considered according to an embodiment of the present invention.

In addition, according to an embodiment of the present application, the subgraph generating apparatus 100 defines resource information of each terminal associated with a predetermined service K based on Equation 1 below, and a pair of terminals based on Equation 2 below It is possible to calculate edge resource information for an edge defined between

[Equation 1]

Here, f ₁ , f ₂ , etc. may correspond to each type of resource information, and accordingly, n in Equation 1 may indicate the number of types of resources.

[Equation 2]

Here, Eji is edge resource information defined between the start terminal (i) and the end terminal (j), and Fi and Fj may be resource information of the start terminal (i) and the end terminal (j), respectively.

As such, the subgraph generating apparatus 100 calculates edge resource information for an edge defined between a pair of terminals based on the resource information of each pair of terminals, and provides a hierarchical structure in which directionality between a plurality of terminals is given. To represent the subgraph, edge resource information may be derived by subtracting the resource information of the starting terminal (i) from the resource information of the terminal terminal (j) for each type component of a resource corresponding to each other.

[Equation 3]

Also, referring to Equation 3, edge resource information according to an embodiment of the present disclosure may be derived using various equations including a subtraction method using an additional processing method. For example, the edge resource information is obtained by synthesizing (combining) the resource information of the type components of the terminal terminal (j), synthesizing the resource information of each type component of the starting terminal (i), and combining the combined resource information of the terminal terminal (j). Edge resource information may be derived by subtracting the resource information of the initiating terminal (i) synthesized from the resource information.

In addition, according to an embodiment of the present invention, the subgraph generation apparatus 100 considers the type of resource information required by the service requested by the requester terminal 1 so that the subgraph can be generated, Resource information and edge resource information of each terminal may be derived in consideration of the type of related resource information.

For example, when the subgraph generating apparatus 100 determines that a specific type of sensor-mounted information is irrelevant to the provision of the corresponding service in consideration of the type of resource required for the requested service, the resource information and edge resource information of each terminal When calculating information, resource information of a corresponding type may be excluded from being considered.

In addition, resource information has different categories depending on the type of resource, which can reduce or delay the learning effect of an artificial intelligence-based edge analysis model to be described later. The graph generating apparatus 100 may process edge resource information. Such processing may be applied when a method of final subtraction is used when calculating edge resource information, but is not limited thereto.

Specifically, the subgraph generating apparatus 100 according to an embodiment of the present application sets the value corresponding to the resource information to 0 when the difference between each resource information has a value of 0 or a negative number through the above-described subtraction method. Edge resource information may be processed by substituting 1 for a value corresponding to the corresponding resource information when it has a positive value, but is not limited thereto.

According to an embodiment of the present application, the subgraph generating apparatus 100 may recommend a provider terminal suitable for a predetermined service in consideration of an edge defining a relationship between terminals, rather than considering only resource information of individual terminals.

In addition, according to an embodiment of the present application, the subgraph generating apparatus 100 may express edge resource information of an edge composed of a pair of terminals in a line graph based on resource information and resource information of each terminal in the network. Specifically, the sub-graph generating apparatus 100 may generate a node N' based on edge resource information and generate a line graph by utilizing terminals in the network to include the created node N'. In other words, the sub-graph generating apparatus 100 may express the relationship of resource information in the form of a line graph.

Referring to (b) of FIG. 4, for example, the subgraph generating apparatus 100 generates a node AB based on the resource information of device A and device B in FIG. A node CD may be generated based on resource information of terminal D.

Therefore, according to an embodiment of the present application, the sub-graph generation apparatus 100 processes edge resource information and expresses it in the form of a line graph, thereby facilitating learning of an artificial intelligence-based edge analysis model to be described later. .

In addition, the sub-graph generating apparatus 100 may train an artificial intelligence-based edge analysis model. The edge analysis model may be an artificial intelligence model trained to output relation information about a corresponding edge when edge resource information is input.

Specifically, the sub-graph generating apparatus 100 is performed based on edge resource information collected for at least one edge and provision information of a predetermined service through a start terminal and an end terminal corresponding to the at least one edge. An edge analysis model can be trained based on supervised learning. In other words, the sub-graph generating apparatus 100 may build an edge analysis model (learning neural network in FIG. 3 ) based on a supervised learning method.

Specifically, the subgraph generating apparatus 100 corresponds to a first target value corresponding to a state in which a predetermined service can be provided by a start terminal and an end terminal, and a state in which a predetermined service cannot be provided by a start terminal but can be provided by an end terminal. The edge analysis model may be trained through supervised learning based on a second target value that corresponds to a second target value and a third target value corresponding to a state in which a predetermined service cannot be provided in an end terminal.

In this regard, the subgraph generating apparatus 100 derives the relationship information digitized with a higher value as the possibility of provision through connection (connection) between the start terminal and the end terminal for the service increases, so that the first target value and the second target value are derived. Among the 2 target values and the 3 target values, the first target value is set to the largest value, the second target value is set to be larger, and the third target value is set to the smallest value (in other words, first target value > second target value > third target value).

In addition, according to an embodiment of the present application, the sub-graph generation apparatus 100 calculates an error between the output value of the edge analysis model and the above-described target value in the process of learning the edge analysis model, and calculates the error as the edge analysis model (learning neural network). ), the edge analysis model can be trained.

5 is a diagram exemplarily illustrating target values set to perform supervised learning based on information on whether a predetermined service is provided through a start terminal and an end terminal.

Referring to FIG. 5, the aforementioned first target value may be set to '1', the second target value to '0.5', and the third target value to '0', but are not limited thereto, The subgraph generating apparatus 100 may operate to determine a specific numerical value for each target value based on the type and characteristics of a service to generate a subgraph and the number of terminals participating in the network 20 .

In addition, in order to build the edge analysis model disclosed herein, various supervised learning algorithm models previously known or developed in the future may be applied.

In addition, the subgraph generating apparatus 100 is learned (constructed) through the above-described process, and each edge defined between a plurality of provider terminals 2 within a specific network 20 including the requester terminal 1 When resource information is input, relational information for each edge in the network 20 may be derived based on an artificial intelligence-based edge analysis model (learning neural network) that outputs relational information for each inputted edge. In other words, the sub-graph generation apparatus 100 may derive relationship information on a corresponding edge based on an artificial intelligence-based edge analysis model by taking edge resource information as an input.

More specifically, the relationship information for each edge derived by the edge analysis model (learning neural network) is information related to the effectiveness of the edge, including the possibility of providing a service corresponding to a service request when a start terminal and an end terminal are connected. can

In other words, the edge analysis model takes edge resource information defined for a predetermined edge as an input, and when a start terminal and an end terminal included in the edge are connected, a digitized relationship indicating the possibility of effectively providing the corresponding service Information can be provided as output.

More specifically, the edge analysis model (learning neural network) effectively provides a service when an end terminal is connected to a starting terminal in relation to a corresponding edge based on input edge resource information through supervised learning based on the target value described in detail above. The higher the probability to provide, the closer the value to the first target value (eg, 1) may be output.

In addition, the sub-graph generating apparatus 100 may determine whether to activate each edge based on relationship information derived by an edge analysis model. According to an embodiment of the present application, the sub-graph generating apparatus 100 deactivates an edge in which the digitized relationship information derived by the edge analysis model is less than a preset threshold, and the digitized relationship information derived by the edge analysis model. It is possible to determine for each edge whether or not to activate an edge that is equal to or greater than the aforementioned threshold value.

For example, the threshold value, which is a criterion for determining whether each edge is active, is set based on the type of requested service, the number of terminals (nodes) participating in the network 20, etc. A threshold value may be determined so that the terminal recommendation or terminal cluster recommendation corresponding to the state of the network 20 in which the requestor terminal 1 participates is provided.

According to an embodiment of the present invention, the sub-graph generating apparatus 100 changes resource information of each terminal into edge resource information between terminals, learns a relationship between terminals in the network through a learning neural network, and makes a request through the learned relationship. A subgraph corresponding to a specific service can be constructed.

In detail, the subgraph generating apparatus 100 may build a subgraph based on whether the edge resource is active or not determined based on edge resource information. In addition, according to an embodiment of the present application, the subgraph generated by the subgraph generation apparatus 100 may be generated in consideration of resource information required by a corresponding service included in a service request as described above.

According to an embodiment of the present application, the subgraph generating apparatus 100 generates a subgraph including activated edges based on the determined activation status, and extends each activated edge from the start terminal of the corresponding edge to the end terminal. The sub-graph represented to go can be hierarchically constructed for at least some of the terminals among the plurality of terminals in the network 20 .

On the other hand, since the sub-graph generating device 100 generates a sub-graph by assuming a situation in which a service required to create a target value is requested based on a set of rules through a learned edge analysis model, the sub-graph generating device 100 It can be understood that the subgraph constructed by ) is created (constructed) based on the requester terminal 1 requesting the corresponding service to the network 20.

In addition, according to an embodiment of the present invention, the subgraph generating apparatus 100 may transmit the generated subgraph to the requester terminal 1 so that the subgraph generated through the requester terminal 1 is displayed (visualized). there is.

In addition, according to an embodiment of the present invention, when at least one of a change in a service request rule and a change in the environment of the network 20 occurs, the subgraph generating apparatus 100 reflects the change to update and generate the subgraph. The above-described sub-graph generation process may be performed again.

According to an embodiment of the present application, the sub-graph generation apparatus 100 modularizes the process of calculating the above-described edge resource information and the process of determining the relationship information and activation between edges based on the edge resource information, and sets a preset update period. It is possible to operate to repeatedly perform the sub-graph creation process every time.

In addition, changes in the environment of the network 20 may cause, for example, some nodes (terminals) previously participating in the network 20 to depart from the network 20 or a new node (terminal) to the network 20. If added, etc. may be included.

In this regard, since the subgraph generation apparatus 100 disclosed herein defines edge resource information between edges defined for a pair of terminals and related relation information for a combination of all terminals, a node (terminal) leaves Alternatively, in response to the addition, it is necessary to re-determine the edge resource information, relationship information, and activation status only for the edge associated with the node (terminal) that has been left or added, so the network including the departure or addition of the node (terminal) (20 ), it is possible to quickly regenerate a subgraph reflecting changes in the network 20 by actively responding to changes. In other words, the subgraph generation apparatus 100 can more quickly provide an updated subgraph in response to a state change in the network.

According to an embodiment of the present invention, the subgraph generation apparatus 100 may evaluate the subgraph by determining whether a starting terminal and one or more terminals connected to the starting terminal can provide a service based on the subgraph.

6 to 8 are diagrams for explaining evaluation of a subgraph by determining whether a starting terminal and one or a plurality of terminals connected to the starting terminal can provide a service based on the subgraph according to an embodiment of the present invention.

6 is a diagram illustrating a subgraph including a requester terminal and a provider terminal.

Referring to FIG. 6, in the sub-graph, a start terminal may mean a requester terminal 1, and terminals connected through an edge in a hierarchical structure from the requester terminal 1 become an activated edge in the above-described activation determination process. It may mean an active terminal 3 that is a connected terminal, and a terminal other than the requester terminal 1 and the active terminal 3 (terminals marked in gray in FIG. 6) may mean inactive terminals.

7 is a diagram showing a hierarchical structure of subgraphs. 7 may be a hierarchical representation of the subgraphs of FIG. 6 .

Referring to FIG. 7 , A is a requester terminal 1, and B and C may mean terminals (provider terminal 2) capable of providing at least a portion of resource information corresponding to the service requested by A. In addition, the terminal 3 between B and C is an activated terminal 3, but may mean a terminal that cannot provide resource information corresponding to the service requested by A.

According to an embodiment of the present application, the subgraph generating apparatus 100 may determine whether service can be provided by adding up resource information that can be provided through one or a plurality of terminals connected from the requester terminal 1, that is, the initiating terminal. there is. In other words, referring to FIG. 7 , when the resource information that can be provided by terminals B and C, which can provide the service requested by A, the starting terminal, is summed up, and the summed resource information satisfies the service resource requirements of the requested service. , the sub-graph generating apparatus 100 may determine that the service can be provided.

8 is a diagram visually illustrating a process of adding up resource information of each terminal to determine whether a service can be provided.

A is the requester terminal 1, and B and C represent terminals capable of providing resource information corresponding to the service requested by A (provider terminal 2).

As an example with reference to FIG. 8, A includes only resource information of types a, b, c, and f among a, b, c, d, e, and f, which are types of resource information necessary to satisfy a service. It includes the amount of resource information as shown in FIG. 8, B includes resource information of types a, b, c, and e, and C includes resource information of types a, b, c, and d as shown in FIG. At this time, since the result of summing (combining) all the resource information that can be provided by A, B, and C terminals can satisfy all service resource requirements (type and amount of resource information) corresponding to the service requested by A, the present application The sub-graph generation apparatus 100 according to an embodiment of may determine that the service can be provided with respect to the corresponding service.

In other words, the subgraph generating apparatus 100 may receive service resource requirements corresponding to the requested service from a plurality of provider terminals. That is, the subgraph generator 100 may provide a distributed service from one or more provider terminals having few resources to a single requester terminal requiring many resources.

In addition, according to an embodiment of the present invention, a result of summing (synthesizing) resource information of the requester terminal 1 and one or a plurality of provider terminals 2 is the service resource request corresponding to the service requested by the requester terminal 1. If the conditions are not satisfied, the sub-graph generation device 100 may determine that the service cannot be provided with respect to the corresponding service, but is not limited thereto.

In other words, the subgraph generating apparatus 100 aggregates resource information of the requester terminal 1 and one or a plurality of provider terminals based on the generated (discovered) subgraph, and the aggregated resource information is combined with the requestor terminal 1 ), it is possible to determine whether the service can be provided by comparing with the service resource requirements of the requested service.

Also, the subgraph generating apparatus 100 may evaluate the subgraph based on a result of determining whether the service can be provided. Specifically, based on the result of determining whether the service can be provided, the apparatus 100 for generating a new subgraph may provide feedback on a process of determining whether each edge is active using each relationship information when generating a new subgraph. there is.

According to an embodiment of the present application, the subgraph generation apparatus 100 can recommend a service providing terminal based on a subgraph that is hierarchically generated for a plurality of terminals based on edge information, so that multiple terminals in the network can be recommended. Efficient evaluation can be performed even if there are terminals of

According to an embodiment of the present invention, the subgraph generation apparatus 100 repeatedly performs the above-described subgraph creation (discovery) process (step 1 in FIG. 3) and subgraph evaluation process (step 2 in FIG. 3), so that the requester A suitable subgraph corresponding to the service requested by the terminal 1 may be created (discovered).

9 is a schematic configuration diagram of an apparatus for generating a subgraph for recommending a terminal corresponding to a service request according to an embodiment of the present invention.

Referring to FIG. 9 , the subgraph generating apparatus 100 may include a line graph generating unit 110 , a subgraph constructing unit 120 and an evaluation unit 130 .

According to an embodiment of the present application, the line graph generator 110 may express resource information of each terminal in the network and edge resource information of an edge composed of a pair of terminals in a line graph based on the resource information.

According to an embodiment of the present application, the resource information may include at least one of storage capacity information, computing capability information, memory information, and sensor mounting information of each terminal.

According to an embodiment of the present application, the edge resource information is resource information expressed as an average value of a plurality of resource information of a start terminal and an end terminal, resource information replaced by a maximum value of a plurality of resource information, and information between a plurality of resource information. It may include at least one resource information among resource information expressed as a result value of subtraction.

According to an embodiment of the present application, a line graph may be a graph expressed using terminals in a network to include nodes generated based on edge resource information.

The sub-graph builder 120 may construct a sub-graph by deriving relation information about a corresponding edge based on edge resource information as an input based on an artificial intelligence-based edge analysis model.

According to an embodiment of the present application, the relationship information may be information related to the validity of an edge including a possibility of providing a service corresponding to a service request when a start terminal and an end terminal are connected.

According to an embodiment of the present application, the edge analysis model may be an artificial intelligence model learned to output relational information about a corresponding edge when edge resource information is input.

Specifically, the sub-graph construction unit 120. It may be to determine whether each edge is active based on relationship information, and build a subgraph based on whether it is active.

In addition, the subgraph builder 120 hierarchically constructs a subgraph including activated edges and expressing the activated edges to extend from the terminal to the terminal terminal for at least some terminals among a plurality of terminals in the network. it may be

The evaluator 130 may evaluate the subgraph by determining whether a starting terminal and one or more terminals connected to the starting terminal can provide services based on the subgraph.

Specifically, the evaluator 130 may determine whether a service can be provided by summing up resource information that can be provided through one or a plurality of terminals connected from the starting terminal.

Hereinafter, based on the details described above, the operation flow of the present application will be briefly reviewed.

10 is an operation flowchart of a method for generating a subgraph for recommending a terminal corresponding to a service request according to an embodiment of the present invention.

The method of generating a subgraph for recommending a terminal corresponding to a service request shown in FIG. 10 may be performed by the apparatus 100 for generating a subgraph for recommending a terminal corresponding to a service request described above. Therefore, even if the contents are omitted below, the description of the subgraph generating apparatus 100 for recommending a terminal corresponding to a service request is equally applicable to the description of the method of generating a subgraph for recommending a terminal corresponding to a service request. can

Referring to FIG. 10, in step S11, the line graph generator 110 may express resource information of each terminal in the network and edge resource information of an edge composed of a pair of terminals based on the resource information as a line graph. .

Next, in step S12, the sub-graph building unit 120 may build a sub-graph by deriving relation information about a corresponding edge based on edge resource information as an input based on an artificial intelligence-based edge analysis model.

Specifically, in step S12, the sub-graph building unit 120 It may be to determine whether each edge is active based on relationship information, and build a subgraph based on whether it is active.

In addition, in step S12, the subgraph builder 120 constructs a subgraph including an activated edge and extending the activated edge from the terminal to the end terminal for at least some terminals among a plurality of terminals in the network. It may be built adversarially.

Next, in step S13, the evaluator 130 may evaluate the subgraph by determining whether a starting terminal and one or more terminals connected to the starting terminal can provide services based on the subgraph.

Specifically, in step S13, the evaluator 130 may determine whether the service can be provided by adding up resource information that can be provided through one or a plurality of terminals connected from the starting terminal.

In the foregoing description, steps S610 to S650 may be further divided into additional steps or combined into fewer steps, depending on an embodiment of the present invention. Also, some steps may be omitted if necessary, and the order of steps may be changed.

A method for generating a subgraph for recommending a terminal corresponding to a service request according to an embodiment of the present disclosure may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the medium may be those specially designed and configured for the present invention or those known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. - includes hardware devices specially configured to store and execute program instructions, such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine language codes such as those produced by a compiler. The hardware devices described above may be configured to act as one or more software modules to perform the operations of the present invention, and vice versa.

In addition, the above-described method for generating a subgraph for recommending a terminal corresponding to a service request may be implemented in the form of a computer program or application stored in a recording medium and executed by a computer.

The above description of the present application is for illustrative purposes, and those skilled in the art will understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present application. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

The scope of the present application is indicated by the following claims rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts thereof should be construed as being included in the scope of the present application.

10: service delivery system
110: Sub-graph generating device for recommending a terminal in response to a service request
110: edge definition unit
120: model learning unit
130: edge evaluation unit
140: sub graph construction unit
20: Network
1: requester terminal
2: provider terminal
E: edge
N: node

Claims (14)

In the subgraph generation method for recommending a terminal corresponding to a service request,
Expressing resource information of each terminal in the network and edge resource information of an edge composed of a pair of terminals in a line graph based on the resource information;
constructing a sub-graph by deriving relational information about a corresponding edge based on an artificial intelligence-based edge analysis model with the edge resource information as an input; and
Evaluating the subgraph by determining whether a starting terminal and at least one terminal connected from the starting terminal can provide a service based on the subgraph;
Including, sub-graph generation method. According to claim 1,
The step of constructing the subgraph,
Wherein determining whether each edge is active based on the relationship information and constructing a subgraph based on the active state. According to claim 2,
The step of constructing the subgraph,
Creating a subgraph that includes an activated edge and hierarchically builds the subgraph expressing the activated edge to extend from the start terminal to the end terminal for at least some terminals among a plurality of terminals in the network method. According to claim 1,
In the step of evaluating the subgraph,
and determining whether the service can be provided by summing up resource information that can be provided through at least one terminal connected from the starting terminal. According to claim 1,
The line graph is
The method of generating a subgraph, which is a graph expressed by utilizing the terminals in the network to include a node generated based on the edge resource information. According to claim 5,
The edge resource information,
Resource information expressed as an average value of a plurality of resource information of the start terminal and the end terminal, resource information replaced with a maximum value of the plurality of resource information, and resource information expressed as a value resulting from subtraction between the plurality of resource information A method for generating a subgraph, comprising at least one resource information of According to claim 1,
The resource information,
A method for generating a sub-graph comprising at least one of storage capacity information, computational capability information, memory information, and sensor mounting information of each terminal. According to claim 1,
The relationship information is
The sub-graph generation method of claim 1 , wherein the information related to the validity of the edge including a possibility of providing a service corresponding to the service request when the start terminal and the end terminal are connected. According to claim 1,
The edge analysis model,
A method of generating a sub-graph, which is an artificial intelligence model learned to output relation information about a corresponding edge when the edge resource information is input. According to claim 1,
The subgraph,
The method of generating a subgraph, which is generated in consideration of resource information required by a corresponding service included in the service request. An apparatus for generating a subgraph for recommending a terminal corresponding to a service request,
A line graph generating unit that expresses resource information of each terminal in the network and edge resource information of an edge composed of a pair of terminals in a line graph based on the resource information;
a sub-graph building unit that constructs a sub-graph by deriving relational information about a corresponding edge based on an artificial intelligence-based edge analysis model based on the edge resource information as an input; and
an evaluation unit that evaluates the subgraph by determining whether a starting terminal and at least one terminal connected from the starting terminal can provide a service based on the subgraph;
Including, a sub-graph generating device. According to claim 1,
The sub-graph building unit,
The apparatus for generating a sub-graph, wherein whether each edge is active is determined based on the relationship information, and a sub-graph is constructed based on the active state. According to claim 2,
The sub-graph building unit,
The subgraph including an activated edge and expressing the activated edge to extend from the start terminal to the end terminal is hierarchically constructed for at least some terminals among a plurality of terminals in the network. generating device. According to claim 1,
The evaluation unit,
and determining whether the service can be provided by adding up resource information that can be provided through all terminals between the start terminal and the end terminal.

Images