Graph-Based Feature Crossing to Enhance Recommender Systems
In recommendation tasks, most existing models that learn users’ preferences from user–item interactions ignore the relationships between items. Additionally, ensuring that the crossed features capture both global graph structures and local context is non-trivial, requiring innovative techniques for...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/13/2/302 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832588079527886848 |
|---|---|
| author | Congyu Cai Hong Chen Yunxuan Liu Daoquan Chen Xiuze Zhou Yuanguo Lin |
| author_facet | Congyu Cai Hong Chen Yunxuan Liu Daoquan Chen Xiuze Zhou Yuanguo Lin |
| author_sort | Congyu Cai |
| collection | DOAJ |
| description | In recommendation tasks, most existing models that learn users’ preferences from user–item interactions ignore the relationships between items. Additionally, ensuring that the crossed features capture both global graph structures and local context is non-trivial, requiring innovative techniques for multi-scale representation learning. To overcome these difficulties, we develop a novel neural network, CoGraph, which uses a graph to build the relations between items. The item co-occurrence pattern assumes that certain items consistently appear in pairs in users’ viewing or consumption logs. First, to learn relationships between items, a graph whose distance is measured by Normalised Point-Wise Mutual Information (NPMI) is applied to link items for the co-occurrence pattern. Then, to learn as many useful features as possible for higher recommendation quality, a Convolutional Neural Network (CNN) and the Transformer model are used to parallelly learn local and global feature interactions. Finally, a series of comprehensive experiments were conducted on several public data sets to show the performance of our model. It provides valuable insights into the capability of our model in recommendation tasks and offers a viable pathway for the public data operation. |
| format | Article |
| id | doaj-art-a3c8cb6330e24eb3a46e9c1b7bf8ced9 |
| institution | Kabale University |
| issn | 2227-7390 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj-art-a3c8cb6330e24eb3a46e9c1b7bf8ced92025-01-24T13:40:06ZengMDPI AGMathematics2227-73902025-01-0113230210.3390/math13020302Graph-Based Feature Crossing to Enhance Recommender SystemsCongyu Cai0Hong Chen1Yunxuan Liu2Daoquan Chen3Xiuze Zhou4Yuanguo Lin5School of Marxism, Jimei University, Xiamen 361021, ChinaInformation Hub, The Hong Kong University of Science and Technology (Guangzhou), Guangzhou 511453, ChinaSchool of Computer Engineering, Jimei University, Xiamen 361021, ChinaSchool of Intelligent Transportation, Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053, ChinaInformation Hub, The Hong Kong University of Science and Technology (Guangzhou), Guangzhou 511453, ChinaSchool of Computer Engineering, Jimei University, Xiamen 361021, ChinaIn recommendation tasks, most existing models that learn users’ preferences from user–item interactions ignore the relationships between items. Additionally, ensuring that the crossed features capture both global graph structures and local context is non-trivial, requiring innovative techniques for multi-scale representation learning. To overcome these difficulties, we develop a novel neural network, CoGraph, which uses a graph to build the relations between items. The item co-occurrence pattern assumes that certain items consistently appear in pairs in users’ viewing or consumption logs. First, to learn relationships between items, a graph whose distance is measured by Normalised Point-Wise Mutual Information (NPMI) is applied to link items for the co-occurrence pattern. Then, to learn as many useful features as possible for higher recommendation quality, a Convolutional Neural Network (CNN) and the Transformer model are used to parallelly learn local and global feature interactions. Finally, a series of comprehensive experiments were conducted on several public data sets to show the performance of our model. It provides valuable insights into the capability of our model in recommendation tasks and offers a viable pathway for the public data operation.https://www.mdpi.com/2227-7390/13/2/302recommender systemsconvolutional neural networkgraph neural networkco-occurrence patterntransformer model |
| spellingShingle | Congyu Cai Hong Chen Yunxuan Liu Daoquan Chen Xiuze Zhou Yuanguo Lin Graph-Based Feature Crossing to Enhance Recommender Systems Mathematics recommender systems convolutional neural network graph neural network co-occurrence pattern transformer model |
| title | Graph-Based Feature Crossing to Enhance Recommender Systems |
| title_full | Graph-Based Feature Crossing to Enhance Recommender Systems |
| title_fullStr | Graph-Based Feature Crossing to Enhance Recommender Systems |
| title_full_unstemmed | Graph-Based Feature Crossing to Enhance Recommender Systems |
| title_short | Graph-Based Feature Crossing to Enhance Recommender Systems |
| title_sort | graph based feature crossing to enhance recommender systems |
| topic | recommender systems convolutional neural network graph neural network co-occurrence pattern transformer model |
| url | https://www.mdpi.com/2227-7390/13/2/302 |
| work_keys_str_mv | AT congyucai graphbasedfeaturecrossingtoenhancerecommendersystems AT hongchen graphbasedfeaturecrossingtoenhancerecommendersystems AT yunxuanliu graphbasedfeaturecrossingtoenhancerecommendersystems AT daoquanchen graphbasedfeaturecrossingtoenhancerecommendersystems AT xiuzezhou graphbasedfeaturecrossingtoenhancerecommendersystems AT yuanguolin graphbasedfeaturecrossingtoenhancerecommendersystems |