Personalized Transfer of User Preferences for Cross-domain Recommendation

Zhu, Yongchun; Tang, Zhenwei; Liu, Yudan; Zhuang, Fuzhen; Xie, Rong; Zhang, Xu; Lin, Leyu; He, Qing

doi:10.1145/3488560.3498392

Cited by 100 publications

(38 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Following previous works, we adopt their selected crossdomain recommendation datasets [6,7,10,37,38,39], and the preprocessing settings [6,7] to build our CDR scenarios. Specifically, we conduct experiments on the large scale public Amazon [40] datasets 3 .…”

Section: A Datasetsmentioning

confidence: 99%

“…Following this paradigm, SSCDR [6] exploits multi-hop information of users to improve the robustness of the mapping function, and DCDIR [48] further introduces meta-path [18] information of item knowledge graph [60,61] to fuse the external knowledge. Recent TMCDR [7] and PTUPCDR [39] follow the MAML [62] framework to learn a meta network to substitute the mapping function for the better recommendation, which can still be regarded as a particular form of the embedding-and-mapping paradigm. Besides, motivated by the success of VAE [25] framework in collaborative filtering, CDVAE [9], AlignVAE [63] and SA-VAE [8] are proposed based on Bayesian-VAE to learn the mapping function.…”

Section: ) Cross-domain Recommendation To Overlapping Usersmentioning

confidence: 99%

See 1 more Smart Citation

Cross-Domain Recommendation to Cold-Start Users via Variational Information Bottleneck

Cao¹,

Sheng²,

Xu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Recommender systems have been widely deployed in many real-world applications, but usually suffer from the longstanding user cold-start problem. As a promising way, Cross-Domain Recommendation (CDR) has attracted a surge of interest, which aims to transfer the user preferences observed in the source domain to make recommendations in the target domain. Previous CDR approaches mostly achieve the goal by following the Embedding and Mapping (EMCDR) idea which attempts to learn a mapping function to transfer the pre-trained user representations (embeddings) from the source domain into the target domain. However, they pre-train the user/item representations independently for each domain, ignoring to consider both domain interactions simultaneously. Therefore, the biased pre-trained representations inevitably involve the domain-specific information which may lead to negative impact to transfer information across domains. In this work, we consider a key point of the CDR task: what information needs to be shared across domains? To achieve the above idea, this paper utilizes the information bottleneck (IB) principle, and proposes a novel approach termed as CDRIB to enforce the representations encoding the domainshared information. To derive the unbiased representations, we devise two IB regularizers to model the cross-domain/in-domain user-item interactions simultaneously and thereby CDRIB could consider both domain interactions jointly for de-biasing. With an additional contrastive information regularizer, CDRIB can also capture cross-domain user-user correlations. In this way, those regularizers encourage the representations to encode the domain-shared information, which has the capability to make recommendations in both domains directly. To the best of our knowledge, this paper is the first work to capture the domain-shared information for cold-start users via variational information bottleneck. Empirical experiments illustrate that CDRIB outperforms the state-of-the-art approaches on four realworld cross-domain datasets, demonstrating the effectiveness of adopting the information bottleneck for CDR.

show abstract

Section: A Datasetsmentioning

confidence: 99%

Section: ) Cross-domain Recommendation To Overlapping Usersmentioning

confidence: 99%

Cross-Domain Recommendation to Cold-Start Users via Variational Information Bottleneck

Cao¹,

Sheng²,

Xu³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…It is widely verified that users having similar historical behaviors are likely to share the same preferences [36]. This could be an effective supplement to the user modeling, especially when we can only learn very little from the historical and current sessions for cold-start [20,[42][43][44] users. However, the user preferences often evolve over dialogues dynamically [37].…”

Section: Temporal Look-alike User Selectormentioning

confidence: 99%

User-Centric Conversational Recommendation with Multi-Aspect User Modeling

Li,

Xie,

Zhu

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Conversational recommender systems (CRS) aim to provide highquality recommendations in conversations. However, most conventional CRS models mainly focus on the dialogue understanding of the current session, ignoring other rich multi-aspect information of the central subjects (i.e., users) in recommendation. In this work, we highlight that the user's historical dialogue sessions and look-alike users are essential sources of user preferences besides the current dialogue session in CRS. To systematically model the multi-aspect information, we propose a User-Centric Conversational Recommendation (UCCR) model, which returns to the essence of user preference learning in CRS tasks. Specifically, we propose a historical session learner to capture users' multi-view preferences from knowledge, semantic, and consuming views as supplements to the current preference signals. A multi-view preference mapper is conducted to learn the intrinsic correlations among different views in current and historical sessions via self-supervised objectives. We also design a temporal look-alike user selector to understand users via their similar users. The learned multi-aspect multi-view user preferences are then used for the recommendation and dialogue generation. In experiments, we conduct comprehensive evaluations on both Chinese and English CRS datasets. The significant improvements over competitive models in both recommendation and dialogue generation verify the superiority of UCCR. The source code will be available on https://github.com/lisk123/UCCR.

show abstract

“…The work can be grouped into three clusters, e.g. metric-based [26], model-based [37], and optimizationbased [6] approaches. Recently, meta-learning has also been applied to recommender systems [21,14,36,38,35,29] and natural language processing [5].…”

Section: Related Workmentioning

confidence: 99%

Customized Conversational Recommender Systems

Li¹,

Zhu²,

Xie³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Conversational recommender systems (CRS) aim to capture user's current intentions and provide recommendations through real-time multi-turn conversational interactions. As a human-machine interactive system, it is essential for CRS to improve the user experience. However, most CRS methods neglect the importance of user experience. In this paper, we propose two key points for CRS to improve the user experience: (1) Speaking like a human, human can speak with different styles according to the current dialogue context. (2) Identifying fine-grained intentions, even for the same utterance, different users have diverse finegrained intentions, which are related to users' inherent preference. Based on the observations, we propose a novel CRS model, coined Customized Conversational Recommender System (CCRS), which customizes CRS model for users from three perspectives. For human-like dialogue services, we propose multi-style dialogue response generator which selects context-aware speaking style for utterance generation. To provide personalized recommendations, we extract user's current fine-grained intentions from dialogue context with the guidance of user's inherent preferences. Finally, to customize the model parameters for each user, we train the model from the meta-learning perspective. Extensive experiments and a series of analyses have shown the superiority of our CCRS on both the recommendation and dialogue services.

show abstract

Personalized Transfer of User Preferences for Cross-domain Recommendation

Cited by 100 publications

References 27 publications

Cross-Domain Recommendation to Cold-Start Users via Variational Information Bottleneck

Cross-Domain Recommendation to Cold-Start Users via Variational Information Bottleneck

User-Centric Conversational Recommendation with Multi-Aspect User Modeling

Customized Conversational Recommender Systems

Contact Info

Product

Resources

About