Training method and device of graph representation system
A map and characterization technology, applied in the computer field, to achieve the effect of reducing the occupied resources, reducing resources, and broadening the scope of use
Active Publication Date: 2021-03-26
ALIPAY (HANGZHOU) INFORMATION TECH CO LTD
View PDF5 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
There is no such plan yet
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0034] Multiple embodiments disclosed in this specification will be described below in conjunction with the accompanying drawings.
[0035] The embodiment of this specification discloses a training method for an atlas characterization system. The atlas characterization system obtained through training can realize the absolute characterization of atlases, and the absolute characterization vectors of atlases can be used repeatedly. For example, see Figure 1B , through the absolute graph representation space designed in the embodiment of this specification, when calculating the graph representation vector of the query graph, it only needs to be represented once, and can be used to calculate the similarity between it and three different candidate graphs.
[0036] For easy understanding, figure 2 A schematic diagram showing a training architecture of a graph representation system according to an embodiment, such as figure 2 As shown, for the relationship graph pairs (that is, ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The embodiment of the invention provides a training method of a graph representation system. The graph representation system comprises a graph neural network, a plurality of node-level learning vectors, a plurality of intermediate-level learning vectors and a plurality of graph-level learning vectors. The method comprises the steps of obtaining a first training sample, the first training sample comprises two corresponding relation maps and similarity label values, and any first relation map comprises a plurality of object nodes; then, performing graph embedding processing on the first relationship graph by utilizing a graph neural network to obtain a plurality of object node embedding vectors; then, based on the plurality of object node embedding vectors, sequentially utilizing a node similarity memory component, a node graph similarity memory component and a graph similarity memory component to extract information of different scales so as to obtain an absolute graph representation vector of the first relation graph; and then, calculating a similarity prediction value between the two map representation vectors corresponding to the two relationship maps, and training the map representation system in combination with the similarity label value.
Description
technical field [0001] The embodiment of this specification relates to the field of computer technology, and in particular to a training method for a graph representation system. Background technique [0002] A relational network graph (or relational graph, relational graph, graph) is a description of the relationship between entities in the real world, and is currently widely used in various computer information processing. Generally, a relational network graph includes a set of nodes and a set of edges. Nodes represent entities in the real world, and edges represent connections between entities in the real world. For example, in a social network, people are entities, and relationships or links between people are edges. [0003] In some cases, a graph representation of the relational network graph is required. For example, in the retrieval scenario, it is necessary to characterize the query map input by the user and the map in the candidate library, and then use the repre...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06F16/28G06F16/36G06K9/62G06N3/04G06N3/08
CPCG06F16/288G06F16/367G06N3/084G06N3/045G06F18/22Y02D10/00
Inventor 熊涛马博群刘杰石磊磊漆远
Owner ALIPAY (HANGZHOU) INFORMATION TECH CO LTD