Graph convolution collaborative filtering recommendation method fusing social relation
The invention provides a graph convolution collaborative filtering recommendation method fusing social relations. The method comprises the following steps: S1, randomly initializing an embedding matrix of nodes, and querying to respectively obtain initialized embedding of a user u and an article i;...
Saved in:
Main Authors | , , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
11.01.2022
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention provides a graph convolution collaborative filtering recommendation method fusing social relations. The method comprises the following steps: S1, randomly initializing an embedding matrix of nodes, and querying to respectively obtain initialized embedding of a user u and an article i; S2, after initial embedding of the nodes is obtained, aggregating and updating node embedding through a semantic aggregation layer; firstly, introducing first-order semantic aggregation into a semantic aggregation layer, then expanding the first-order semantic aggregation to each layer, and achieving high-order semantic aggregation; S3, after the semantic aggregation embedding vector of the social embedding propagation layer and the semantic aggregation embedding vector of the interactive embedding propagation layer are obtained respectively, fusing user embedding vectors of the social embedding propagation layer and the interactive embedding propagation layer; performing weighted summation fusion on each order of |
---|---|
Bibliography: | Application Number: CN202111235558 |