# US 12,216,629 B2 · Data processing method and apparatus, computerreadable medium, and electronic device

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,216,629 B2
- Document ID: US12216629B2
- Status: Granted
- Kind: B2
- Application number: 17/964,778
- Filed: 2022-10-12
- Published: 2023-02-02
- Granted: 2025-02-04
- Earliest claimed: 2020-12-31
- Assignees: Tencent Technology (Shenzhen) Company Limited
- Applicants: Tencent Technology (Shenzhen) Company Limited
- Inventors: Xiaosen Li; Jie Xu; Wen Ouyang; Yangyu Tao; Pin Xiao
- Web page: https://patentgrep.com/patent/US12216629B2

## CPC

- G06F 16/9024: {Graphs; Linked lists (G06F16/9027 takes precedence)}
- G06F 16/22: Indexing; Data structures therefor; Storage structures
- G06F 18/211: Selection of the most significant subset of features
- G06F 18/2323: based on graph theory, e.g. minimum spanning trees [MST] or graph cuts
- G06Q 10/48: using social graphs
- G06Q 30/0201: Market modelling; Market analysis; Collecting market data
- H04L 41/082: the condition being updates or upgrades of network functionality
- H04L 41/122: of virtualised topologies, e.g. software-defined networks [SDN] or network function virtualisation [NFV]
- H04L 41/0806: for initial configuration or provisioning, e.g. plug-and-play
- H04L 41/0895: Configuration of virtualised networks or elements, e.g. virtualised network function or OpenFlow elements
- H04L 41/145: {involving simulating, designing, planning or modelling of a network}
- H04L 41/16: using machine learning or artificial intelligence
- H04L 41/22: comprising specially adapted graphical user interfaces [GUI]
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

This application belongs to the technical field of artificial intelligence, and particularly relates to a data processing method, a data processing apparatus, a computer-readable medium, and an electronic device. The method includes acquiring a relationship graph network, the relationship graph network comprising nodes for representing interacting objects and edges for representing interactive relationships between multiple interacting objects; performing core degree mining on the relationship graph network through a device cluster comprising multiple computing devices, and iteratively updating node core degrees of all or some of the nodes in the relationship graph network; pruning the relationship graph network according to the node core degrees to remove some of the nodes and edges in the relationship graph network; and compressing the device cluster to remove some of the computing devices in the device cluster, when a network scale of the relationship graph network satisfies a network compression condition.

## Claims

### Claim 1

A computer implemented data processing method, the method comprising:
acquiring a relationship graph network, the relationship graph network comprising nodes for representing interacting objects and edges for representing interactive relationships between multiple interacting objects;
partitioning the relationship graph network to obtain a first number of partition graph networks, wherein a partition graph network includes part of the nodes and edges in the relationship graph network;
performing, based on the first number of partition graph networks, core degree mining on the relationship graph network in a distributed computing mode through a device cluster comprising multiple computing devices, and iteratively updating node core degrees of all or some of the nodes in the relationship graph network, each computing device in the device cluster being configured to perform computing tasks of the core degree mining on the relationship graph network;
pruning the relationship graph network according to the node core degrees to remove some of the nodes and edges in the relationship graph network; and
compressing the device cluster to remove some of the computing devices in the device cluster, when a network scale of the relationship graph network corresponding to the first number of the partitioned graph networks meets a network compression condition, comprising:
repartitioning, according to the network scale, the relationship graph network to obtain a second number of partition graph networks, the second number being smaller than the first number;
invoking one or more computing devices in the multiple computing devices based on the second number of partition graph networks; and
in response to the network scale being reduced to satisfy a specified condition, transforming from the distributed computing mode to a standalone computing mode by: selecting a single computing device from the device cluster as a target device for standalone computing of the relationship graph network, and removing the other computing devices except the target device from the device cluster, wherein the target device has a computing capability to perform the computing tasks of the core degree mining on the relationship graph network with the reduced network scale.

### Claim 2 (depends on 1)

The computer implemented data processing method according to claim 1, wherein the performing core degree mining on the relationship graph network through a device cluster comprising multiple computing devices and iteratively updating node core degrees of the respective nodes in the relationship graph network comprises:
allocating the first number of partition graph networks to the device cluster comprising the multiple computing devices, and determining a computing device for core degree mining of the first number of partition graph networks; and
performing core degree mining on the first number of partition graph networks through the allocated computing device, and iteratively updating the node core degree of each node in the first number of partition graph networks.

### Claim 3 (depends on 2)

The computer implemented data processing method according to claim 2, wherein the performing core degree mining on the first number of partition graph networks and updating the node core degree of each node in the first number of partition graph networks iteratively comprises:
selecting a computing node for core degree mining in a current iteration from the first number of partition graph networks, and identifying a neighbor node adjacent to the computing node;
acquiring current node core degrees of the computing node and the neighbor node in the current iteration;
determining a temporary node core degree of the computing node according to the current node core degree of the neighbor node, determining the temporary node core degree of the computing node is less than the current node core degree of the computing node, and labeling the computing node as an active node; and
updating the current node core degree of the active node according to the temporary node core degree, and determining the active node and a neighbor node adjacent to the active node as computing nodes for core degree mining in a next iteration.

### Claim 4 (depends on 3)

The computer implemented data processing method according to claim 3, wherein the determining a temporary node core degree of the computing node according to the current node core degree of the neighbor node comprises:
determining an h-index of the computing node according to the current node core degree of the neighbor node, and determining the h-index as the temporary node core degree of the computing node, the h-index representing that current node core degrees of at most h of all neighbor nodes of the computing node are greater than or equal to h.

### Claim 5 (depends on 4)

The computer implemented data processing method according to claim 4, wherein the determining an h-index of the computing node according to the current node core degree of the neighbor node comprises:
sequencing all the neighbor nodes of the computing node from high to low current node core degrees, and allocating an arrangement sequence number to each neighbor node from 0;
comparing the arrangement sequence number and current node core degree of each neighbor node, and screening out the neighbor node whose arrangement sequence number is greater than or equal to the current node core degree according to a comparison result; and
determining the current node core degree of the neighbor node with the minimum arrangement sequence number in the screened-out neighbor node as the h-index of the computing node.

### Claim 6 (depends on 3)

The computer implemented data processing method according to claim 3, wherein the selecting a computing node for core degree mining in a current iteration from the first number of partition graph networks comprises:
reading a node identifier of a node to be updated from a first storage space, the node to be updated comprising an active node whose node core degree is updated in a previous iteration and a neighbor node adjacent to the active node; and
selecting the computing node for core degree mining in the current iteration from the first number of partition graph networks according to the node identifier of the node to be updated.

### Claim 7 (depends on 6)

The computer implemented data processing method according to claim 6, further comprising:
writing an updated current node core degree of the active node to a second storage space, the second storage space storing the node core degrees of all the nodes in the relationship graph network;
acquiring a node identifier of the active node and a node identifier of the neighbor node of the active node, and writing the acquired node identifiers to a third storage space, the third storage space storing a node identifier of a computing node for core degree mining in a next iteration; and
overwriting, after core degree mining of all partition graph networks in the current iteration, data in the first storage space with data in the third storage space, and resetting the third storage space.

### Claim 8 (depends on 3)

The computer implemented data processing method according to claim 3, wherein the acquiring current node core degrees of the computing node and the neighbor node in the current iteration comprises:
reading the current node core degrees of the computing node and the neighbor node in the current iteration from a second storage space, the second storage space storing the node core degrees of all the nodes in the relationship graph network.

### Claim 9 (depends on 1)

The computer implemented data processing method according to claim 1, wherein the performing a pruning process on the relationship graph network according to the node core degrees to remove some of the nodes and edges in the relationship graph network comprises:
acquiring a minimum core degree of an active node in a current iteration and a minimum core degree of the active node in a previous iteration;
identifying a convergent node in the relationship graph network according to the minimum core degree of the active node in the previous iteration when the minimum core degree of the active node in the current iteration is greater than that of the active node in the previous iteration, the convergent node being a node whose node core degree is less than or equal to the minimum core degree of the active node in the previous iteration; and
removing the convergent node and an edge connected with the convergent node from the relationship graph network.

### Claim 10 (depends on 1)

The computer implemented data processing method according to claim 1, wherein the network compression condition comprises that a number of the edges in the relationship graph network is less than a number threshold.

### Claim 11 (depends on 1)

The computer implemented data processing method according to claim 1, further comprising:
acquiring, for each node in the relationship graph network, a node number of neighbor nodes adjacent to the node; and
initializing, for each node, the node core degree of the node according to the node number of the neighbor nodes adjacent to the node.

### Claim 12 (depends on 1)

The method according to claim 1, wherein partitioning the relationship graph network comprises:
selecting multiple partitioning center points from the relationship graph network according to a preset partition number;
performing a clustering process on one or more nodes in the relationship graph network to obtain partitioning center points as clustering centers for allocating the one or more nodes to a respective closest partitioning center point; and
partitioning the relationship graph network into one or more partition graph networks according to a clustering result of the nodes.

### Claim 13 (depends on 12)

The method according to claim 12, wherein the partitioning center points include a node selected according to a preset rule.

### Claim 14 (depends on 1)

The method according to claim 1, further comprising:
in the single computing mode, processing the whole relationship graph network serially by centralized computing on the target computing device.

### Claim 15

A non-transitory computer-readable storage medium, storing a computer program, the computer program, when executed by a processor, causing the processor to perform:
acquiring a relationship graph network, the relationship graph network comprising nodes for representing interacting objects and edges for representing interactive relationships between multiple interacting objects;
partitioning the relationship graph network to obtain a first number of partition graph networks, wherein a partition graph network includes part of the nodes and edges in the relationship graph network;
performing, based on the first number of partition graph networks, core degree mining on the relationship graph network in a distributed computing mode through a device cluster comprising multiple computing devices, and iteratively updating node core degrees of all or some of the nodes in the relationship graph network, each computing device in the device cluster being configured to perform computing tasks of the core degree mining on the relationship graph network;
pruning the relationship graph network according to the node core degrees to remove some of the nodes and edges in the relationship graph network; and
compressing the device cluster to remove some of the computing devices in the device cluster, when a network scale of the relationship graph network corresponding to the first number of the partitioned graph networks meets a network compression condition, comprising:
repartitioning, according to the network scale, the relationship graph network to obtain a second number of partition graph networks, the second number being smaller than the first number;
invoking one or more computing devices in the multiple computing devices based on the second number of partition graph networks; and
in response to the network scale being reduced to satisfy a specified condition, transforming from a distributed computing mode to a standalone computing mode by: selecting a single computing device from the device cluster as a target device for standalone computing of the relationship graph network, and removing the other computing devices except the target device from the device cluster, wherein the target device has a computing capability to perform the computing tasks of the core degree mining on the relationship graph network with the reduced network scale.

### Claim 16 (depends on 15)

The computer-readable storage medium according to claim 15, wherein the performing core degree mining on the relationship graph network through a device cluster comprising multiple computing devices and iteratively updating node core degrees of the respective nodes in the relationship graph network comprises:
allocating the first number of partition graph networks to the device cluster comprising the multiple computing devices, and determining a computing device for core degree mining of the first number of partition graph networks; and
performing core degree mining on the first number of partition graph networks through the allocated computing device, and iteratively updating the node core degree of each node in the first number of partition graph networks.

### Claim 17 (depends on 16)

The computer-readable storage medium according to claim 16, wherein the performing core degree mining on the first number of partition graph networks and updating the node core degree of each node in the first number of partition graph networks iteratively comprises:
selecting a computing node for core degree mining in a current iteration from the first number of partition graph networks, and identifying a neighbor node adjacent to the computing node;
acquiring current node core degrees of the computing node and the neighbor node in the current iteration;
determining a temporary node core degree of the computing node according to the current node core degree of the neighbor node, determining the temporary node core degree of the computing node is less than the current node core degree of the computing node, and labeling the computing node as an active node; and
updating the current node core degree of the active node according to the temporary node core degree, and determining the active node and a neighbor node adjacent to the active node as computing nodes for core degree mining in a next iteration.

### Claim 18 (depends on 17)

The computer-readable storage medium according to claim 17, wherein the determining a temporary node core degree of the computing node according to the current node core degree of the neighbor node comprises:
determining an h-index of the computing node according to the current node core degree of the neighbor node, and determining the h-index as the temporary node core degree of the computing node, the h-index representing that current node core degrees of at most h of all neighbor nodes of the computing node are greater than or equal to h.

### Claim 19

An electronic device, comprising:
a processor; and
a memory, configured to store executable instructions of the processor,
the processor being configured to execute the executable instructions and perform:
acquiring a relationship graph network, the relationship graph network comprising nodes for representing interacting objects and edges for representing interactive relationships between multiple interacting objects;
partitioning the relationship graph network to obtain a first number of partition graph networks, wherein a partition graph network includes part of the nodes and edges in the relationship graph network;
performing, based on the first number of partition graph networks, core degree mining on the relationship graph network in a distributed computing mode through a device cluster comprising multiple computing devices, and iteratively updating node core degrees of all or some of the nodes in the relationship graph network, each computing device in the device cluster being configured to perform computing tasks of the core degree mining on the relationship graph network;
pruning the relationship graph network according to the node core degrees to remove some of the nodes and edges in the relationship graph network; and
compressing the device cluster to remove some of the computing devices in the device corresponding to the first number of the partitioned graph networks meets a network compression condition, comprising:
repartitioning, according to the network scale, the relationship graph network to obtain a second number of partition graph networks, the second number being smaller than the first number;
invoking one or more computing devices in the multiple computing devices based on the second number of partition graph networks; and
in response to the network scale being reduced to satisfy a specified condition, transforming from a distributed computing mode to a standalone computing mode by: selecting a single computing device from the device cluster as a target device for standalone computing of the relationship graph network, and removing the other computing devices except the target device from the device cluster, wherein the target device has a computing capability to perform the computing tasks of the core degree mining on the relationship graph network with the reduced network scale.