While recommendation plays an increasingly critical role in our living, study, work, and entertainment, the recommendations we receive are often for irrelevant, duplicate, or uninteresting products and services. A critical reason for such bad recommendations lies in the intrinsic assumption that recommended users and items are independent and identically distributed (IID) in existing theories and systems. Another phenomenon is that, while tremendous efforts have been made to model specific aspects of users or items, the overall user and item characteristics and their non-IIDness have been overlooked. In this paper, the non-IID nature and characteristics of recommendation are discussed, followed by the non-IID theoretical framework in order to build a deep and comprehensive understanding of the intrinsic nature of recommendation problems, from the perspective of both couplings and heterogeneity. This non-IID recommendation research triggers the paradigm shift from IID to non-IID recommendation research and can hopefully deliver informed, relevant, personalized, and actionable recommendations. It creates exciting new directions and fundamental solutions to address various complexities including cold-start, sparse data-based, cross-domain, group-based, and shilling attack-related issues.
Do Recommender Systems Promote Local Music? A Reproducibility Study Using Music Streaming Data
CNRS（法国国家科学研究中心）；LIPN（巴黎第十三大学计算机科学实验室）；USPN（巴黎第十三大学）；LTCI（法国电信学院）；Télécom Paris（巴黎电信学院）；Deezer Research（Deezer研究院）；PACTE（政治与公共行动研究中心）；Sciences Po Grenoble（格勒诺布尔政治学院）
Recommender systems research has experienced different stages such as from user preference understanding to content analysis. Typical recommendation algorithms were built on the following bases: (1) assuming users and items are IID, namely independent and identically distributed, and (2) focusing on specific aspects such as user preferences or contents. In reality, complex recommendation tasks involve and request (1) personalized outcomes to tailor heterogeneous subjective preferences; and (2) explicit and implicit objective coupling relationships between users, items, and ratings to be consid…
Dual Test-time Training for Out-of-distribution Recommender System
National University of Defense Technology（国防科技大学）；Hong Kong University of Science and Technology（香港科技大学）；The Hong Kong Polytechnic University（香港理工大学）；City University of Hong Kong（香港城市大学）；University of Science and Technology of China（中国科学技术大学）
Recommendation algorithms are known to suffer from popularity bias; a few popular items are recommended frequently while the majority of other items are ignored. These recommendations are then consumed by the users, their reaction will be logged and added to the system: what is generally known as a feedback loop. In this paper, we propose a method for simulating the users interaction with the recommenders in an offline setting and study the impact of feedback loop on the popularity bias amplification of several recommendation algorithms. We then show how this bias amplification leads to severa…
Do Not Wait: Learning Re-Ranking Model Without User Feedback At Serving Time in E-Commerce
Alibaba Group（阿里巴巴集团）；Gaoling School of Artificial Intelligence, Renmin University of China（中国人民大学高瓴人工智能学院）；School of Information, Renmin University of China（中国人民大学信息学院）
Recently, embedding techniques have achieved impressive success in recommender systems. However, the embedding techniques are data demanding and suffer from the cold-start problem. Especially, for the cold-start item which only has limited interactions, it is hard to train a reasonable item ID embedding, called cold ID embedding, which is a major challenge for the embedding techniques. The cold item ID embedding has two main problems: (1) A gap is existing between the cold ID embedding and the deep model. (2) Cold ID embedding would be seriously affected by noisy interaction. However, most exi…
Owing to its nature of scalability and privacy by design, federated learning (FL) has received increasing interest in decentralized deep learning. FL has also facilitated recent research on upscaling and privatizing personalized recommendation services, using on-device data to learn recommender models locally. These models are then aggregated globally to obtain a more performant model, while maintaining data privacy. Typically, federated recommender systems (FRSs) do not consider the lack of resources and data availability at the end-devices. In addition, they assume that the interaction data …
Recommender systems are a valuable way to engage users in a system, increase participation and show them resources they may not have found otherwise. One significant challenge is that user interests may change over time and certain items have an inherently temporal aspect. As a result, a recommender system should try and take into account the time-dependant user-item relationships. However, temporal aspects of a user profile may not always be explicitly available and so we may need to infer this information from available resources. Product reviews on sites, such as Amazon, represent a valuabl…
Business success in e-commerce depends on customer perceived value. A customer with high perceived value buys, returns, and recommends items. The perceived value is at risk whenever the information load harms users' shopping experience. In e-grocery, shoppers face an overwhelming number of items, the majority of which is irrelevant for the shopper. Recommender systems (RS) enable businesses to master information overload (IO) by providing users with an item ranking by relevance. Prior work proposes RS with short personalized rankings (top-k). Given large order sizes and high user heterogeneity…
This paper is an extended version of [Burashnikova et al., 2021, arXiv: 2012.06910], where we proposed a theoretically supported sequential strategy for training a large-scale Recommender System (RS) over implicit feedback, mainly in the form of clicks. The proposed approach consists in minimizing pairwise ranking loss over blocks of consecutive items constituted by a sequence of non-clicked items followed by a clicked one for each user. We present two variants of this strategy where model parameters are updated using either the momentum method or a gradient-based approach. To prevent updating…
Recommender systems easily face the issue of user preference shifts. User representations will become out-of-date and lead to inappropriate recommendations if user preference has shifted over time. To solve the issue, existing work focuses on learning robust representations or predicting the shifting pattern. There lacks a comprehensive view to discover the underlying reasons for user preference shifts. To understand the preference shift, we abstract a causal graph to describe the generation procedure of user interaction sequences. Assuming user preference is stable within a short period, we a…
