Bundle recommendation aims to provide users with convenient one-stop solutions by recommending bundles of related items that cater to their diverse needs. However, previous research has neglected the interaction betwe...Bundle recommendation aims to provide users with convenient one-stop solutions by recommending bundles of related items that cater to their diverse needs. However, previous research has neglected the interaction between bundle and item views and relied on simplistic methods for predicting user-bundle relationships. To address this limitation, we propose Hybrid Contrastive Learning for Bundle Recommendation (HCLBR). Our approach integrates unsupervised and supervised contrastive learning to enrich user and bundle representations, promoting diversity. By leveraging interconnected views of user-item and user-bundle nodes, HCLBR enhances representation learning for robust recommendations. Evaluation on four public datasets demonstrates the superior performance of HCLBR over state-of-the-art baselines. Our findings highlight the significance of leveraging contrastive learning and interconnected views in bundle recommendation, providing valuable insights for marketing strategies and recommendation system design.展开更多
In the recommendation system,bundle recommendation is a prevalent sales strategy in which a combination of diverse,related,or complementary products is suggested to consumers.Recent methodologies frequently utilize gr...In the recommendation system,bundle recommendation is a prevalent sales strategy in which a combination of diverse,related,or complementary products is suggested to consumers.Recent methodologies frequently utilize graph neural networks to capture information from user-bundle,user-item,and bundle-item interactions,deriving corresponding feature representations.However,these approaches often emphasize the distinctions among these three interaction types or treat them uniformly,neglecting the varying importance within one type of interaction and failing to consider the acquisition of information at varying granularities from different types of interactions.In this study,we employ a graph attention mechanism to process user-bundle interaction information,and optimize it using an association enhancement method to extract and construct coarse-grained information representations for users and bundles.By analyzing interactions between users and items,as well as between bundles and items,we identify disparities in item popularity and update the items’feature representations,facilitating the acquisition of fine-grained information representations for users and bundles.By merging this information,we achieve more comprehensive representations of user intent and bundle characteristics.Extensive experiments on two real-world datasets convincingly demonstrate that our approach significantly advances the task of bundle recommendation,outperforming state-of-the-art methods.展开更多
Bundle recommendation offers users more holistic insights by recommending multiple compatible items at once.However,the intricate correlations between items,varied user preferences,and the pronounced data sparsity in ...Bundle recommendation offers users more holistic insights by recommending multiple compatible items at once.However,the intricate correlations between items,varied user preferences,and the pronounced data sparsity in combinations present significant challenges for bundle recommendation algorithms.Furthermore,current bundle recommendation methods fail to identify mismatched items within a given set,a process termed as‘‘outlier item detection’’.These outlier items are those with the weakest correlations within a bundle.Identifying them can aid users in refining their item combinations.While the correlation among items can predict the detection of such outliers,the adaptability of combinations might not be adequately responsive to shifts in individual items during the learning phase.This limitation can hinder the algorithm’s performance.To tackle these challenges,we introduce an encoder–decoder architecture tailored for outlier item detection.The encoder learns potential item correlations through a self-attention mechanism.Concurrently,the decoder garners efficient inference frameworks by directly assessing item anomalies.We have validated the efficacy and efficiency of our proposed algorithm using real-world datasets.展开更多
文摘Bundle recommendation aims to provide users with convenient one-stop solutions by recommending bundles of related items that cater to their diverse needs. However, previous research has neglected the interaction between bundle and item views and relied on simplistic methods for predicting user-bundle relationships. To address this limitation, we propose Hybrid Contrastive Learning for Bundle Recommendation (HCLBR). Our approach integrates unsupervised and supervised contrastive learning to enrich user and bundle representations, promoting diversity. By leveraging interconnected views of user-item and user-bundle nodes, HCLBR enhances representation learning for robust recommendations. Evaluation on four public datasets demonstrates the superior performance of HCLBR over state-of-the-art baselines. Our findings highlight the significance of leveraging contrastive learning and interconnected views in bundle recommendation, providing valuable insights for marketing strategies and recommendation system design.
基金supported by the Open Project of Xiangjiang Laboratory(No.23XJ03006)the Open Project of Anhui Provincial Key Laboratory of Multimodal Cognitive Computation,Anhui University(No.MMC202408)+3 种基金the Fundamental Research Funds for the Central Universities,JLU(No.93K172024K17)the Fundamental Research Funds for the Central Universities(No.21623402)the Science and Technology Program of Guangzhou,China(No.2024A04j6317)the Hubei Key Laboratory of Intelligent Robot(Wuhan Institute of Technology)(No.HBIR 202302).
文摘In the recommendation system,bundle recommendation is a prevalent sales strategy in which a combination of diverse,related,or complementary products is suggested to consumers.Recent methodologies frequently utilize graph neural networks to capture information from user-bundle,user-item,and bundle-item interactions,deriving corresponding feature representations.However,these approaches often emphasize the distinctions among these three interaction types or treat them uniformly,neglecting the varying importance within one type of interaction and failing to consider the acquisition of information at varying granularities from different types of interactions.In this study,we employ a graph attention mechanism to process user-bundle interaction information,and optimize it using an association enhancement method to extract and construct coarse-grained information representations for users and bundles.By analyzing interactions between users and items,as well as between bundles and items,we identify disparities in item popularity and update the items’feature representations,facilitating the acquisition of fine-grained information representations for users and bundles.By merging this information,we achieve more comprehensive representations of user intent and bundle characteristics.Extensive experiments on two real-world datasets convincingly demonstrate that our approach significantly advances the task of bundle recommendation,outperforming state-of-the-art methods.
基金supported in part by the Guangxi Key Laboratory of Trusted Software,China(KX202037)the Project of Guangxi Science and Technology,China(GuiKeAD 20297054)the Guangxi Natural Science Foundation Project,China(2020GXNSFBA297108)。
文摘Bundle recommendation offers users more holistic insights by recommending multiple compatible items at once.However,the intricate correlations between items,varied user preferences,and the pronounced data sparsity in combinations present significant challenges for bundle recommendation algorithms.Furthermore,current bundle recommendation methods fail to identify mismatched items within a given set,a process termed as‘‘outlier item detection’’.These outlier items are those with the weakest correlations within a bundle.Identifying them can aid users in refining their item combinations.While the correlation among items can predict the detection of such outliers,the adaptability of combinations might not be adequately responsive to shifts in individual items during the learning phase.This limitation can hinder the algorithm’s performance.To tackle these challenges,we introduce an encoder–decoder architecture tailored for outlier item detection.The encoder learns potential item correlations through a self-attention mechanism.Concurrently,the decoder garners efficient inference frameworks by directly assessing item anomalies.We have validated the efficacy and efficiency of our proposed algorithm using real-world datasets.
