Unlock instant, AI-driven research and patent intelligence for your innovation.

Heterogeneous information network enhanced academic paper recommendation method

A heterogeneous information network and recommendation method technology, which is applied in neural learning methods, biological neural network models, digital data information retrieval, etc., can solve problems such as sparse interactive data, and achieve the effect of improving accuracy

Active Publication Date: 2022-05-20
NINGBO UNIV
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The present invention uses heterogeneous information network to solve the problem of sparse interactive data, which can improve the accuracy of recommendation

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Heterogeneous information network enhanced academic paper recommendation method
  • Heterogeneous information network enhanced academic paper recommendation method
  • Heterogeneous information network enhanced academic paper recommendation method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0060] Embodiment: A method for recommending academic papers enhanced by a heterogeneous information network, the process is as follows figure 1 shown.

[0061] Step 1. Build a heterogeneous information network.

[0062] Such as figure 2 For a heterogeneous information network built on a citeulike dataset, figure 2 The part (a) in the figure represents the node type, the part (b) represents the heterogeneous information network, the part (c) represents the meta-path, and the part (d) represents the meta-path neighbors. The network contains three types of nodes: user U, paper P and label T, 3 kinds of relationships: user-paper interaction relationship, inter-paper citation relationship, and paper label inclusion relationship. The citeulike data set is a public data set suitable for the field of paper recommendation. Three files, users.dat, citations.dat and item-tag.dat are selected as the original data, among which users.dat is the user's historical click paper record, ci...

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

No PUM Login to View More

Abstract

The invention discloses a heterogeneous information network enhanced academic paper recommendation method, which comprises the following steps of: 1, constructing a heterogeneous information network which comprises three types of nodes, namely a user node, a paper node and a label node, the three relations include the interaction relation between the user and the papers, the reference relation between the papers and the affiliation relation between the papers and the labels; 2, learning interaction characteristics of the user and the paper by using a matrix decomposition algorithm; 3, inputting the interaction features into the heterogeneous graph attention network, and learning the high-order features of the papers in the heterogeneous information network; 4, utilizing outer product calculation to fuse the features obtained by learning in the steps 2 and 3; and step 5, inputting the features fused in the step 4 into a depth recommendation model to predict scores. According to the method, the problem of sparse interaction data is solved by utilizing the heterogeneous information network, and the recommendation accuracy can be improved.

Description

technical field [0001] The invention relates to a method for recommending academic papers, in particular to a method for recommending academic papers enhanced by a heterogeneous information network. Background technique [0002] With the explosive growth of the number of academic publications and the rapid iteration of knowledge, it is difficult for researchers to easily find academic papers that meet their needs, and they are facing an increasingly serious problem of information overload in papers. The Academic Paper Recommendation System (Academic Paper Recommendation, APR) accurately recommends papers to scientific researchers and is becoming an indispensable tool for scientific researchers. Collaborative Filtering (Collaboration Filtering, CF) is widely used in recommendation systems. It predicts users' personalized preferences by exploring users' historical interactions. However, when the interaction matrix is ​​very sparse, CF cannot produce robust performance; in rece...

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
no application Login to View More
IPC IPC(8): G06F16/335G06N3/04G06N3/08
CPCG06F16/335G06N3/08G06N3/045
Inventor 刘柏嵩吴俊超沈小烽张雪垣王冰源
Owner NINGBO UNIV