OCPHN: Outfit Compatibility Prediction with Hypergraph Networks
With the rapid development of the online shopping, the pursuit of outfit compatibility has become a basic requirement for an increasing number of customers. The existing work on outfit compatibility prediction largely focuses on modeling pairwise item compatibility without considering modeling the w...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/10/20/3913 |
| _version_ | 1797471675162296320 |
|---|---|
| author | Zhuo Li Jian Li Tongtong Wang Xiaolin Gong Yinwei Wei Peng Luo |
| author_facet | Zhuo Li Jian Li Tongtong Wang Xiaolin Gong Yinwei Wei Peng Luo |
| author_sort | Zhuo Li |
| collection | DOAJ |
| description | With the rapid development of the online shopping, the pursuit of outfit compatibility has become a basic requirement for an increasing number of customers. The existing work on outfit compatibility prediction largely focuses on modeling pairwise item compatibility without considering modeling the whole outfit directly. To address the problem, in this paper, we propose a novel hypergraph-based compatibility modeling scheme named OCPHN, which is able to better model complex relationships among outfits. In OCPHN, we represent the outfit as a hypergraph, where each hypernode represents a category and each hyperedge represents the interactions between multiple categories (i.e., they appear in the same outfit). To better predict outfit compatibility, the hypergraph is transformed into a simple graph, and the message propagation mechanism in the graph convolution network is used to aggregate the neighbours’ information on the node and update the node representations. Furthermore, with learned node representations, an attention mechanism is introduced to compute the outfit compatibility score. Using a benchmark dataset, the experimental results show that the proposed method is an improvement over the strangest baselines in terms of accuracy by about 3% and 1% in the fill-in-the-blank and compatibility prediction tasks, respectively. |
| first_indexed | 2024-03-09T19:52:29Z |
| format | Article |
| id | doaj.art-c05f38416cf54da0930e72b7d4c8cd87 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-09T19:52:29Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-c05f38416cf54da0930e72b7d4c8cd872023-11-24T01:08:26ZengMDPI AGMathematics2227-73902022-10-011020391310.3390/math10203913OCPHN: Outfit Compatibility Prediction with Hypergraph NetworksZhuo Li0Jian Li1Tongtong Wang2Xiaolin Gong3Yinwei Wei4Peng Luo5The School of Microelectronics, Tianjin University, Tianjin 300072, ChinaThe School of Microelectronics, Tianjin University, Tianjin 300072, ChinaThe School of Microelectronics, Tianjin University, Tianjin 300072, ChinaThe School of Microelectronics, Tianjin University, Tianjin 300072, ChinaThe School of Computing, National University of Singapore, Singapore 37580, SingaporeThe State Grid Hebei Electric Power Research Institute, Shijiazhuang 050021, ChinaWith the rapid development of the online shopping, the pursuit of outfit compatibility has become a basic requirement for an increasing number of customers. The existing work on outfit compatibility prediction largely focuses on modeling pairwise item compatibility without considering modeling the whole outfit directly. To address the problem, in this paper, we propose a novel hypergraph-based compatibility modeling scheme named OCPHN, which is able to better model complex relationships among outfits. In OCPHN, we represent the outfit as a hypergraph, where each hypernode represents a category and each hyperedge represents the interactions between multiple categories (i.e., they appear in the same outfit). To better predict outfit compatibility, the hypergraph is transformed into a simple graph, and the message propagation mechanism in the graph convolution network is used to aggregate the neighbours’ information on the node and update the node representations. Furthermore, with learned node representations, an attention mechanism is introduced to compute the outfit compatibility score. Using a benchmark dataset, the experimental results show that the proposed method is an improvement over the strangest baselines in terms of accuracy by about 3% and 1% in the fill-in-the-blank and compatibility prediction tasks, respectively.https://www.mdpi.com/2227-7390/10/20/3913outfit compatibilityhypergraph networkgraph convolution networkattention mechanism |
| spellingShingle | Zhuo Li Jian Li Tongtong Wang Xiaolin Gong Yinwei Wei Peng Luo OCPHN: Outfit Compatibility Prediction with Hypergraph Networks Mathematics outfit compatibility hypergraph network graph convolution network attention mechanism |
| title | OCPHN: Outfit Compatibility Prediction with Hypergraph Networks |
| title_full | OCPHN: Outfit Compatibility Prediction with Hypergraph Networks |
| title_fullStr | OCPHN: Outfit Compatibility Prediction with Hypergraph Networks |
| title_full_unstemmed | OCPHN: Outfit Compatibility Prediction with Hypergraph Networks |
| title_short | OCPHN: Outfit Compatibility Prediction with Hypergraph Networks |
| title_sort | ocphn outfit compatibility prediction with hypergraph networks |
| topic | outfit compatibility hypergraph network graph convolution network attention mechanism |
| url | https://www.mdpi.com/2227-7390/10/20/3913 |
| work_keys_str_mv | AT zhuoli ocphnoutfitcompatibilitypredictionwithhypergraphnetworks AT jianli ocphnoutfitcompatibilitypredictionwithhypergraphnetworks AT tongtongwang ocphnoutfitcompatibilitypredictionwithhypergraphnetworks AT xiaolingong ocphnoutfitcompatibilitypredictionwithhypergraphnetworks AT yinweiwei ocphnoutfitcompatibilitypredictionwithhypergraphnetworks AT pengluo ocphnoutfitcompatibilitypredictionwithhypergraphnetworks |