Personalized comment recommendation method based on link prediction model of graph bidirectional aggregation network

Patent metadata

Field	Value
Publication number	US-11748426-B2
Application number	US-202318105891-A
Country	US
Kind code	B2
Filing date	Feb 6, 2023
Priority date	Jun 18, 2021
Publication date	Sep 5, 2023
Grant date	Sep 5, 2023

Abstract

A personalized comment recommendation method based on a link prediction model of a graph bidirectional aggregation network. In a user-comment bipartite graph, comment features are aggregated into a user feature. A social network is used to fuse a neighbor feature of a user to obtain an embedding representation of the user. The embedding representation of the user is aggregated into a comment after an original feature of the user is removed, and the embedding representation of the user is adjusted based on a difference before and after comment aggregation. On this basis, a forwarding network is used to calculate a score of an edge based on an inner product of user node features at both ends of the edge, and finally make a recommendation based on the score. Furthermore, a recommendation system converts a comment recommendation task into a link prediction task between users in a small range.

First claim

What is claimed is:

1. A personalized comment recommendation method based on a link prediction model of a graph bidirectional aggregation network, comprising the following steps:
a) building a user-comment bipartite graph, a social network, and a forwarding network based on specific comment content, a user concern relationship, and a comment forwarding relationship;
b) converting a comment made by a user into an N-dimensional original comment feature h com , and initializing an original feature h user of the user;
c) randomly deleting a comment node from the user-comment bipartite graph, and calculating an aggregated feature aggh user i of an i th user according to a formula aggh user i =SELU(W agg ·concat(h user i , aggregate({h com i , ∀j∈N com (i)}))), wherein an aggregated feature of all users is represented by aggh user , SELU(·) represents an activation function, W agg represents a weight of a feature extraction part, concat(·) represents a splicing function, h user i represents a user feature of an i th node, h com j represents a comment feature of a j th node, N com (i) represents a quantity of comment nodes associated with an i th user node, and aggregate represents an aggregation function, namely,

aggregate
=

1

N
com

(
i
)

·

∑

j
=
0

N
com

(
i
)

⁢
Re
⁢

LU
⁡
(

w
·

h
com
j

+
b

)

,

wherein ReLU(·) represents an activation function, w represents a weight in the aggregate function, and b represents an offset in the aggregate function;
d) calculating a fused neighbor node feature aggh user i , of the i th user according to a formula aggh user i =SELU(Σ j=0 N(i) α ij ·W agg ·h user j ), wherein α ij represents an attention coefficient of associating the i th user and a j th comment edge, and h user j represents a user feature of the j th node;
e) calculating a new comment feature aggh com i of the i th user according to a formula aggh com i =SELU(W agg ·aggregate({aggh user j −h user j , ∀j∈N user (i)})), wherein new comment features of all the users are represented by aggh com , aggh user j represents the fused neighbor node feature of the j th user, and N user (i) represents a quantity of user nodes associated with an i th comment node;
f) calculating a loss function value loss agg of a feature aggregation part according to a formula loss agg =smooth L 1 (h com , aggh com ), wherein smooth L 1 (·) represents a loss function, h com represents comment features of all nodes; and establishing a model of the feature extraction part;
g) aggregating neighbor features by using a GraphSAGE algorithm according to a formula preh=Graph SAGE (g retweet , aggh user ) to obtain a user feature preh, wherein g retweet represents a graph structure of the forwarding network;
h) calculating a score score of each edge in a positive-sample forwarding network in a form of an inner product according to a formula score=preh u ·preh v , wherein preh u represents a user node feature on a left side of a positive-sample edge, and preh v represents a user node feature on a right side of the positive-sample edge; and calculating a score score′ of each edge in a negative-sample forwarding network in the form of an inner product according to a formula score′=preh u ′·preh v ′, wherein preh u ′ represents a user node feature on a left side of a negative-sample edge, and preh v ′ represents a node feature on a right side of the negative-sample edge;
i) calculating a loss function value loss pre of a link prediction part according to a formula loss pre =max(0, M−score+score′), wherein M represents a boundary value, and setting M=1 to establish a model of the link prediction part;
j) calculating a total loss Loss according to a formula Loss=loss agg ×agg+loss pre ×pre, wherein agg represents the weight of the feature extraction part, and pre represents a weight of the link prediction part;
k) repeating steps c) to j) for no less than N times to complete model training and optimization; and
l) generating a recommendation list in descending order based on the score score in step h), and making a recommendation based on the list.

2. The personalized comment recommendation method based on the link prediction model of the graph bidirectional aggregation network according to claim 1 , wherein in step a), a negative sample of the forwarding network is constructed by using a negative sampling algorithm, wherein K represents a parameter in negative sampling and K=5.

3. The personalized comment recommendation method based on the link prediction model of the graph bidirectional aggregation network according to claim 1 , wherein in step b), the comment made by the user is converted into a 64-dimensional original comment feature h com by using a Doc2Vec model.

4. The personalized comment recommendation method based on the link prediction model of the graph bidirectional aggregation network according to claim 1 , wherein in step c), the comment node is randomly deleted from the user-comment bipartite graph based on a probability of 60%.

5. The personalized comment recommendation method based on the link prediction model of the graph bidirectional aggregation network according to claim 1 , wherein in step d), α ij is calculated by using an attention weight calculation method in a Graph Attention Network (GAT) algorithm.

6. The personalized comment recommendation method based on the link prediction model of the graph bidirectional aggregation network according to claim 1 , wherein in step j), agg=1, and pre=2.

7. The personalized comment recommendation method based on the link prediction model of the graph bidirectional aggregation network according to claim 1 , wherein in step k), a value of N is 100.

Assignees

• Shandong Artificial Intelligence Inst
• Univ Qilu Technology
• Shandong Computer Science Ct Nat Supercomputer Ct Jinan

Inventors

• Shu Minglei
• Wang Muchen
• Wang Yinglong
• Li Zhao
• Gao Tianlei

Classifications

• G06Q10/40

  Business processes related to social networking or social networking services · CPC title

• G06Q10/42

  Determination of affinities or common interests between users · CPC title

• G06Q10/48

  using social graphs · CPC title

• G06F16/9024 Primary

  Graphs; Linked lists (G06F16/9027 takes precedence) · CPC title

• Y02D10/00

  Energy efficient computing, e.g. low power processors, power management or thermal management · CPC title