Reinforced negative sampling over knowledge graph for recommendation
Properly handling missing data is a fundamental challenge in recommendation. Most present works perform negative sampling from unobserved data to supply the training of recommender models with negative signals. Nevertheless, existing negative sampling strategies, either static or adaptive ones, are...
Saved in:
Main Authors: | , , , , , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2020
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/7459 https://ink.library.smu.edu.sg/context/sis_research/article/8462/viewcontent/3366423.3380098.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Singapore Management University |
Language: | English |
id |
sg-smu-ink.sis_research-8462 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-84622022-10-20T07:16:08Z Reinforced negative sampling over knowledge graph for recommendation WANG, Xiang XU, Yaokun HE, Xiangnan CAO, Yixin WANG, Meng CHUA, Tat-Seng Properly handling missing data is a fundamental challenge in recommendation. Most present works perform negative sampling from unobserved data to supply the training of recommender models with negative signals. Nevertheless, existing negative sampling strategies, either static or adaptive ones, are insufficient to yield high-quality negative samples — both informative to model training and reflective of user real needs. In this work, we hypothesize that item knowledge graph (KG), which provides rich relations among items and KG entities, could be useful to infer informative and factual negative samples. Towards this end, we develop a new negative sampling model, Knowledge Graph Policy Network (KGPolicy), which works as a reinforcement learning agent to explore high-quality negatives. Specifically, by conducting our designed exploration operations, it navigates from the target positive interaction, adaptively receives knowledgeaware negative signals, and ultimately yields a potential negative item to train the recommender. We tested on a matrix factorization (MF) model equipped with KGPolicy, and it achieves significant improvements over both state-of-the-art sampling methods like DNS [39] and IRGAN [30], and KG-enhanced recommender models like KGAT [32]. Further analyses from different angles provide insights of knowledge-aware sampling. We release the code 2020-04-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/7459 info:doi/10.1145/3366423.3380098 https://ink.library.smu.edu.sg/context/sis_research/article/8462/viewcontent/3366423.3380098.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Recommendation Knowledge Graph Negative Sampling Databases and Information Systems Graphics and Human Computer Interfaces |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
Recommendation Knowledge Graph Negative Sampling Databases and Information Systems Graphics and Human Computer Interfaces |
spellingShingle |
Recommendation Knowledge Graph Negative Sampling Databases and Information Systems Graphics and Human Computer Interfaces WANG, Xiang XU, Yaokun HE, Xiangnan CAO, Yixin WANG, Meng CHUA, Tat-Seng Reinforced negative sampling over knowledge graph for recommendation |
description |
Properly handling missing data is a fundamental challenge in recommendation. Most present works perform negative sampling from unobserved data to supply the training of recommender models with negative signals. Nevertheless, existing negative sampling strategies, either static or adaptive ones, are insufficient to yield high-quality negative samples — both informative to model training and reflective of user real needs. In this work, we hypothesize that item knowledge graph (KG), which provides rich relations among items and KG entities, could be useful to infer informative and factual negative samples. Towards this end, we develop a new negative sampling model, Knowledge Graph Policy Network (KGPolicy), which works as a reinforcement learning agent to explore high-quality negatives. Specifically, by conducting our designed exploration operations, it navigates from the target positive interaction, adaptively receives knowledgeaware negative signals, and ultimately yields a potential negative item to train the recommender. We tested on a matrix factorization (MF) model equipped with KGPolicy, and it achieves significant improvements over both state-of-the-art sampling methods like DNS [39] and IRGAN [30], and KG-enhanced recommender models like KGAT [32]. Further analyses from different angles provide insights of knowledge-aware sampling. We release the code |
format |
text |
author |
WANG, Xiang XU, Yaokun HE, Xiangnan CAO, Yixin WANG, Meng CHUA, Tat-Seng |
author_facet |
WANG, Xiang XU, Yaokun HE, Xiangnan CAO, Yixin WANG, Meng CHUA, Tat-Seng |
author_sort |
WANG, Xiang |
title |
Reinforced negative sampling over knowledge graph for recommendation |
title_short |
Reinforced negative sampling over knowledge graph for recommendation |
title_full |
Reinforced negative sampling over knowledge graph for recommendation |
title_fullStr |
Reinforced negative sampling over knowledge graph for recommendation |
title_full_unstemmed |
Reinforced negative sampling over knowledge graph for recommendation |
title_sort |
reinforced negative sampling over knowledge graph for recommendation |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2020 |
url |
https://ink.library.smu.edu.sg/sis_research/7459 https://ink.library.smu.edu.sg/context/sis_research/article/8462/viewcontent/3366423.3380098.pdf |
_version_ |
1770576342327230464 |