In recent years, two-sided marketplaces have emerged as viable business models in many real-world applications. In particular, we have moved from the social network paradigm to a network with two distinct types of participants representing the supply and demand of a specific good. Examples of industries include but are not limited to accommodation (Airbnb, Booking.com), video content (YouTube, Instagram, TikTok), ridesharing (Uber, Lyft), online shops (Etsy, Ebay, Facebook Marketplace), music (Spotify, Amazon), app stores (Apple App Store, Google App Store) or job sites (LinkedIn). The traditional research in most of these industries focused on satisfying the demand. OTAs would sell hotel accommodation, TV networks would broadcast their own content, or taxi companies would own their own vehicle fleet. In modern examples like Airbnb, YouTube, Instagram, or Uber, the platforms operate by outsourcing the service they provide to their users, whether they are hosts, content creators or drivers, and have to develop their models considering their needs and goals.

It has become increasingly clear that recommender systems that overly focus on short-term engagement prevents users from exploring diverse interests, ultimately hurting long-term user experience. To tackle this challenge, numerous diversification algorithms have been proposed as the final stage of recommender systems. These algorithms typically rely on measures of item similarity, aiming to maximize the dissimilarity across items in the final set of recommendations. However, in this work, we demonstrate the benefits of going beyond item-level similarities by utilizing higher-level user understanding-specifically, user intents that persist across multiple interactions or recommendation sessions-in diversification. Our approach is motivated by the observation that user behaviors on online platforms are largely driven by their underlying intents. Therefore, final recommendations should ensure that a diverse set of user intents is accurately represented. While user intent has primarily been studied in the context of search, it is less clear how to incorporate real-time dynamic intent predictions in recommender systems.

In recent years, two-sided marketplaces have emerged as viable business models in many real-world applications. In particular, we have moved from the social network paradigm to a network with two distinct types of participants representing the supply and demand of a specific good. Examples of industries include but are not limited to accommodation (Airbnb, Booking.com), video content (YouTube, Instagram, TikTok), ridesharing (Uber, Lyft), online shops (Etsy, Ebay, Facebook Marketplace), music (Spotify, Amazon), app stores (Apple App Store, Google App Store) or job sites (LinkedIn). The traditional research in most of these industries focused on satisfying the demand. OTAs would sell hotel accommodation, TV networks would broadcast their own content, or taxi companies would own their own vehicle fleet. In modern examples like Airbnb, YouTube, Instagram, or Uber, the platforms operate by outsourcing the service they provide to their users, whether they are hosts, content creators or drivers, and have to develop their models considering their needs and goals.

Exposure bias and its induced feedback loop effect are well-known problems in recommender systems. Exploration is believed to be the key to break such feedback loops. While classical contextual bandit algorithms such as Upper-Confidence-Bound and Thompson Sampling have been successful in addressing the exploration-exploitation trade-off in the single-task settings with one clear reward signal, modern recommender systems often leverage multiple rich sources of feedback such as clicks, likes, dislikes, shares, satisfaction survey responses, and employ multi-task learning in practice. It is unclear how one can incorporate exploration in the multi-task setup with different objectives. In this paper, we study an efficient bandit algorithm tailored to multi-task recommender systems, named Multi-task Neural Linear Bandit (mtNLB). In particular, we investigate efficient feature embeddings in the multi-task setups that could be used as contextual features in the Neural Linear Bandit, a contextual bandit algorithm that nicely combines the representation power from DNN and simplicity in uncertainty calculation from linear models. We further study cost-effective approximations of the uncertainty estimate and principled ways to incorporate uncertainty into the multi-task scoring of items. To showcase the efficacy of our proposed method, we conduct live experiments on a large-scale commercial recommendation platform that serves billions of users. We evaluate the quality of the uncertainty estimate and demonstrate its ability to improve exploration across the different dimensions of the reward signals in comparison to baseline approaches.

Traditional recommendation systems are subject to a strong feedback loop by learning from and reinforcing past user-item interactions, which in turn limits the discovery of novel user interests. To address this, we introduce a hybrid hierarchical framework combining Large Language Models (LLMs) and classic recommendation models for user interest exploration. The framework controls the interfacing between the LLMs and the classic recommendation models through “interest clusters”, the granularity of which can be explicitly determined by algorithm designers. It recommends the next novel interests by first representing “interest clusters” using language, and employs a fine-tuned LLM to generate novel interest descriptions that are strictly within these predefined clusters. At the low level, it grounds these generated interests to an item-level policy by restricting classic recommendation models, in this case a transformer-based sequence recommender to return items that fall within the novel clusters generated at the high level. We showcase the efficacy of this approach on an industrial-scale commercial platform serving billions of users. Live experiments show a significant increase in both exploration of novel interests and overall user enjoyment of the platform.

Effective exploration is believed to positively influence the long-term user experience on recommendation platforms. Determining its exact benefits, however, has been challenging. Regular A/B tests on exploration often measure neutral or even negative engagement metrics while failing to capture its long-term benefits. We here introduce new experiment designs to formally quantify the long-term value of exploration by examining its effects on content corpus, and connecting content corpus growth to the long-term user experience from real-world experiments. Once established the values of exploration, we investigate the Neural Linear Bandit algorithm as a general framework to introduce exploration into any deep learning based ranking systems. We conduct live experiments on one of the largest short-form video recommendation platforms that serves billions of users to validate the new experiment designs, quantify the long-term values of exploration, and to verify the effectiveness of the adopted neural linear bandit algorithm for exploration.

Recommendation system serves as a conduit connecting users to an incredibly large, diverse and ever growing collection of contents. In practice, missing information on fresh contents needs to be filled in order for them to be exposed and discovered by their audience. In this context, we are delighted to share our success stories in building a dedicated fresh content recommendation stack on a large commercial platform and also shed a light on the utilization of Large Language Models (LLMs) for fresh content recommendations within an industrial framework. To nominate fresh contents, we built a multi-funnel nomination system that combines (i) a two-tower model with strong generalization power for coverage, and (ii) a sequence model with near real-time update on user feedback for relevance, which effectively balances between coverage and relevance. Beyond that, by harnessing the reasoning and generalization capabilities of LLMs, we are presented with exciting prospects to enhance recommendation systems. We share our initial efforts on employing LLMs as data augmenters to bridge the knowledge gap on cold-start items during the training phase. This innovative approach circumvents the costly generation process during inference, presenting a model-agnostic, forward-looking solution for fresh content recommendation.

Recommendation system serves as a conduit connecting users to an incredibly large, diverse and ever growing collection of contents. In practice, missing information on fresh (and tail) contents needs to be filled in order for them to be exposed and discovered by their audience. We here share our success stories in building a dedicated fresh content recommendation stack on a large commercial platform. To nominate fresh contents, we built a multi-funnel nomination system that combines (i) a two-tower model with strong generalization power for coverage, and (ii) a sequence model with near real-time update on user feedback for relevance. The multi-funnel setup effectively balances between coverage and relevance. An in-depth study uncovers the relationship between user activity level and their proximity toward fresh contents, which further motivates a contextual multi-funnel setup. Nominated fresh candidates are then scored and ranked by systems considering prediction uncertainty to further bootstrap content with less exposure. We evaluate the benefits of the dedicated fresh content recommendation stack, and the multi-funnel nomination system in particular, through user corpus co-diverted live experiments. We conduct multiple rounds of live experiments on a commercial platform serving billion of users demonstrating efficacy of our proposed methods.

Share on Towards Companion Recommenders Assisting Users' Long-Term Journeys Authors: Konstantina Christakopoulou Google DeepMind, Google, USA Google DeepMind, Google, USA 0000-0002-1650-1796View Profile , Minmin Chen Google DeepMind, Google, USA Google DeepMind, Google, USA 0000-0002-7342-9022View Profile Authors Info & Claims RecSys '23: Proceedings of the 17th ACM Conference on Recommender SystemsSeptember 2023Pages 1039–1041https://doi.org/10.1145/3604915.3610241Published:14 September 2023Publication History 0citation175DownloadsMetricsTotal Citations0Total Downloads175Last 12 Months175Last 6 weeks175 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

The paradigm of framing recommendations as (sequential) decision-making processes has gained significant interest. To achieve long-term user satisfaction, these interactive systems need to strike a balance between exploitation (recommending high-reward items) and exploration (exploring uncertain regions for potentially better items). Classical bandit algorithms like Upper-Confidence-Bound and Thompson Sampling, and their contextual extensions with linear payoffs have exhibited strong theoretical guarantees and empirical success in managing the exploration-exploitation trade-off. Building efficient exploration-based systems for deep neural network powered real-world, large-scale industrial recommender systems remains under studied. In addition, these systems are often multi-stage, multi-objective and response time sensitive. In this talk, we share our experience in addressing these challenges in building exploration based industrial recommender systems. Specifically, we adopt the Neural Linear Bandit algorithm, which effectively combines the representation power of deep neural networks, with the simplicity of linear bandits to incorporate exploration in DNN based recommender systems. We introduce exploration capability to both the nomination and ranking stage of the industrial recommender system. In the context of the ranking stage, we delve into the extension of this algorithm to accommodate the multi-task setup, enabling exploration in systems with multiple objectives. Moving on to the nomination stage, we will address the development of efficient bandit algorithms tailored to factorized bi-linear models. These algorithms play a crucial role in facilitating maximum inner product search, which is commonly employed in large-scale retrieval systems. We validate our algorithms and present findings from real-world live experiments.

The last decade has witnessed many successes of deep learning-based models for industry-scale recommender systems. These models are typically trained offline in a batch manner. While being effective in capturing users’ past interactions with recommendation platforms, batch learning suffers from long model-update latency and is vulnerable to system biases, making it hard to adapt to distribution shift and explore new items or user interests. Although online learning-based approaches (e.g., multi-armed bandits) have demonstrated promising theoretical results in tackling these challenges, their practical real-time implementation in large-scale recommender systems remains limited. First, the scalability of online approaches in servicing a massive online traffic while ensuring timely updates of bandit parameters poses a significant challenge. Additionally, exploring uncertainty in recommender systems can easily result in unfavorable user experience, highlighting the need for devising intricate strategies that effectively balance the trade-off between exploitation and exploration. In this paper, we introduce Online Matching: a scalable closed-loop bandit system learning from users’ direct feedback on items in real time. We present a hybrid offline + online approach for constructing this system, accompanied by a comprehensive exposition of the end-to-end system architecture. We propose Diag-LinUCB – a novel extension of the LinUCB algorithm – to enable distributed updates of bandits parameter in a scalable and timely manner. We conduct live experiments in YouTube and show that Online Matching is able to enhance the capabilities of fresh content discovery and item exploration in the present platform.

Over the years we have seen recommender systems shifting focus from optimizing short-term engagement toward improving long-term user experience on the platforms. While defining good long-term user experience is still an active research area, we focus on one specific aspect of improved long-term user experience here, which is user revisiting the platform. These long term outcomes however are much harder to optimize due to the sparsity in observing these events and low signal-to-noise ratio (weak connection) between these long-term outcomes and a single recommendation. To address these challenges, we propose to establish the association between these long-term outcomes and a set of more immediate term user behavior signals that can serve as surrogates for optimization.

Industrial recommendation platforms are increasingly concerned with how to make recommendations that cause users to enjoy their long term experience on the platform. Reinforcement learning emerged naturally as an appealing approach for its promise in 1) combating feedback loop effect resulted from myopic system behaviors; and 2) sequential planning to optimize long term outcome. Scaling RL algorithms to production recommender systems serving billions of users and contents, however remain challenging. Sample inefficiency and instability of online RL hinder its widespread adoption in production. Offline RL enables usage of off-policy data and batch learning. It on the other hand faces significant challenges in learning due to the distribution shift.

In the era of increasing choices, recommender systems are becoming indispensable in helping users navigate the million or billion pieces of content on recommendation platforms. Most of the recommender systems are powered by ML models trained on a large amount of user-item interaction data. Such a setup however induces a strong feedback loop that creates the rich gets richer phenomenon where head contents are getting more and more exposure while tail and fresh contents are not discovered. At the same time, it pigeonholes users to contents they are already familiar with. We believe exploration is key to break away from the feedback loop and to optimize long term user experience on recommendation platforms.

Users who come to recommendation platforms are heterogeneous in activity levels. There usually exists a group of core users who visit the platform regularly and consume a large body of content upon each visit, while others are casual users who tend to visit the platform occasionally and consume less each time. As a result, consumption activities from core users often dominate the training data used for learning. As core users can exhibit different activity patterns from casual users, recommender systems trained on historical user activity data usually achieve much worse performance on casual users than core users. To bridge the gap, we propose a model-agnostic framework L2Aug to improve recommendations for casual users through data augmentation, without sacrificing core user experience. L2Aug is powered by a data augmentor that learns to generate augmented interaction sequences, in order to fine-tune and optimize the performance of the recommendation system for casual users. On four real-world public datasets, L2Aug outperforms other treatment methods and achieves the best sequential recommendation performance for both casual and core users. We also test L2Aug in an online simulation environment with real-time feedback to further validate its efficacy, and showcase its flexibility in supporting different augmentation actions.

In the era of increasing choices, recommender systems are becoming indispensable in helping users navigate the million or billion pieces of content on recommendation platforms. As the focus of these systems shifts from attracting short-term user attention toward optimizing long term user experience on these platforms, reinforcement learning (and bandits) have emerged as appealing techniques to power these systems [5, 9, 26, 27]. The exploration-exploitation tradeoff, being the foundation of bandits and RL research, has been extensively studied [1, 2, 4, 6, 8, 10, 11, 18, 20, 21, 22, 23]. An agent is incentivized to exploit to maximize its return, i.e., by repeating actions taken in the past that produced high rewards. On the other hand, the agent needs to explore previously unseen actions in order to discover potentially better ones. Exploration has been shown to be extremely useful in solving tasks of long horizons or sparse reward in many RL applications [2, 14, 15, 16, 19]. While effective exploration is believed to positively influence the user experience on the platform, the exact value of exploration in recommender systems has not been well established.

Reinforcement Learning (RL) has been sought after to bring next-generation recommender systems to further improve user experience on recommendation platforms. While the exploration-exploitation tradeoff is the foundation of RL research, the value of exploration in (RL-based) recommender systems is less well understood. Exploration, commonly seen as a tool to reduce model uncertainty in regions of sparse user interaction/feedback, is believed to cost user experience in the short term, while the indirect benefit of better model quality arrives at a later time. We focus on another aspect of exploration, which we refer to as user exploration to help discover new user interests, and argue it can improve user experience even in the more imminent term.

Reinforcement Learning (RL) techniques have been sought after as the next-generation tools to further advance the field of recommendation research. Different from classic applications of RL, recommender agents, especially those deployed on commercial recommendation platforms, have to operate in extremely large state and action spaces, serving a dynamic user base in the order of billions, and a long-tail item corpus in the order of millions or billions. The (positive) user feedback available to train such agents is extremely scarce in retrospect. Improving the sample efficiency of RL algorithms is thus of paramount importance when developing RL agents for recommender systems. In this work, we present a general framework to augment the training of model-free RL agents with auxiliary tasks for improved sample efficiency. More specifically, we opt to add additional tasks that predict users' immediate responses (positive or negative) toward recommendations, i.e., user response modeling, to enhance the learning of the state and action representations for the recommender agents. We also introduce a tool based on gradient correlation analysis to guide the model design. We showcase the efficacy of our method in offline experiments, learning and evaluating agent policies over hundreds of millions of user trajectories. We also conduct live experiments on an industrial recommendation platform serving billions of users and tens of millions of items to verify its benefit.

Most existing recommender systems focus primarily on matching users (content consumers) to content which maximizes user satisfaction on the platform. It is increasingly obvious, however, that content providers have a critical influence on user satisfaction through content creation, largely determining the content pool available for recommendation. A natural question thus arises: can we design recommenders taking into account the long-term utility of both users and content providers? By doing so, we hope to sustain more content providers and a more diverse content pool for long-term user satisfaction. Understanding the full impact of recommendations on both user and content provider groups is challenging. This paper aims to serve as a research investigation of one approach toward building a content provider aware recommender, and evaluating its impact in a simulated setup.

Improving user satisfaction is at the forefront of industrial recommender systems. While significant progress has been made by utilizing logged implicit data of user-item interactions (i.e., clicks, dwell/watch time, and other user engagement signals), there has been a recent surge of interest in measuring and modeling user satisfaction, as provided by orthogonal data sources. Such data sources typically originate from responses to user satisfaction surveys, which explicitly ask users to rate their experience with the system and/or specific items they have consumed in the recent past. This data can be valuable for measuring and modeling the degree to which a user has had a satisfactory experience on the recommendation platform, since what users do (engagement) does not always align with what users say they want (satisfaction as measured by surveys).

Many real-world recommender systems need to be highly scalable: matching millions of items with billions of users, with milliseconds latency. The scalability requirement has led to widely used two-stage recommender systems, consisting of efficient candidate generation model(s) in the first stage and a more powerful ranking model in the second stage.

Characterizing temporal dependence patterns is a critical step in understanding the statistical properties of sequential data. Long Range Dependence (LRD) --- referring to long-range correlations decaying as a power law rather than exponentially w.r.t. distance --- demands a different set of tools for modeling the underlying dynamics of the sequential data. While it has been widely conjectured that LRD is present in language modeling and sequential recommendation, the amount of LRD in the corresponding sequential datasets has not yet been quantified in a scalable and model-independent manner. We propose a principled estimation procedure of LRD in sequential datasets based on established LRD theory for real-valued time series and apply it to sequences of symbols with million-item-scale dictionaries. In our measurements, the procedure estimates reliably the LRD in the behavior of users as they write Wikipedia articles and as they interact with YouTube. We further show that measuring LRD better informs modeling decisions in particular for RNNs whose ability to capture LRD is still an active area of research. The quantitative measure informs new Evolutive Recurrent Neural Networks (EvolutiveRNNs) designs, leading to state-of-the-art results on language understanding and sequential recommendation tasks at a fraction of the computational cost.

Aggregation is a common technique in data-driven applications for handling issues like privacy, scalability and reliability in a vast range of domains including healthcare, sensor networks and web applications. However, despite the ubiquitousness, extending machine learning methods to the aggregation context is unfortunately not well-studied. In this work, we consider the problem of learning individual level predictive models when the target variables used for training are only available as aggregates. In particular, this problem is a critical bottleneck in designing effective bidding strategies in the context of online advertising where ground-truth cost-per-click (CPC) data is aggregated before being released to advertisers. We introduce a novel learning framework that can use aggregates computed at varying levels of granularity for building individual-level predictive models. We generalise our modelling and algorithmic framework to handle data from diverse domains, and extend our techniques to cover arbitrary aggregation paradigms like sliding windows and overlapping/non-uniform aggregation. We show empirical evidence for the efficacy of our techniques with experiments on both synthetic data and real data from the online advertising domain as well as healthcare to demonstrate the wider applicability of our framework.

Industrial recommender systems deal with extremely large action spaces -- many millions of items to recommend. Moreover, they need to serve billions of users, who are unique at any point in time, making a complex user state space. Luckily, huge quantities of logged implicit feedback (e.g., user clicks, dwell time) are available for learning. Learning from the logged feedback is however subject to biases caused by only observing feedback on recommendations selected by the previous versions of the recommender. In this work, we present a general recipe of addressing such biases in a production top-K recommender system at Youtube, built with a policy-gradient-based algorithm, i.e. REINFORCE. The contributions of the paper are: (1) scaling REINFORCE to a production recommender system with an action space on the orders of millions; (2) applying off-policy correction to address data biases in learning from logged feedback collected from multiple behavior policies; (3) proposing a novel top-K off-policy correction to account for our policy recommending multiple items at a time; (4) showcasing the value of exploration. We demonstrate the efficacy of our approaches through a series of simulations and multiple live experiments on Youtube.

Understanding temporal dynamics has proved to be highly valuable for accurate recommendation. Sequential recommenders have been successful in modeling the dynamics of users and items over time. However, while different model architectures excel at capturing various temporal ranges or dynamics, distinct application contexts require adapting to diverse behaviors.

Many techniques to utilize side information of users and/or items as inputs to recommenders to improve recommendation, especially on cold-start items/users, have been developed over the years. In this work, we test the approach of utilizing item side information, specifically categorical attributes, in the output of recommendation models either through multi-task learning or hierarchical classification. We first demonstrate the efficacy of these approaches for both matrix factorization and neural networks with a medium-size real-word data set. We then show that they improve a neural-network based production model in an industrial-scale recommender system. We demonstrate the robustness of the hierarchical classification approach by introducing noise in building the hierarchy. Lastly, we investigate the generalizability of hierarchical classification on a simulated dataset by building two user models in which we can fully control the generative process of user-item interactions.

In text mining, information retrieval, and machine learning, text documents are commonly represented through variants of sparse Bag of Words (sBoW) vectors (e.g. TF-IDF [1]). Although simple and intuitive, sBoW style representations suffer from their inherent over-sparsity and fail to capture word-level synonymy and polysemy. Especially when labeled data is limited (e.g. in document classification), or the text documents are short (e.g. emails or abstracts), many features are rarely observed within the training corpus. This leads to overfitting and reduced generalization accuracy. In this paper we propose Dense Cohort of Terms (dCoT), an unsupervised algorithm to learn improved sBoW document features. dCoT explicitly models absent words by removing and reconstructing random sub-sets of words in the unlabeled corpus. With this approach, dCoT learns to reconstruct frequent words from co-occurring infrequent words and maps the high dimensional sparse sBoW vectors into a low-dimensional dense representation. We show that the feature removal can be marginalized out and that the reconstruction can be solved for in closed-form. We demonstrate empirically, on several benchmark datasets, that dCoT features significantly improve the classification accuracy across several document classification tasks.

Topical classification of user queries is critical for general-purpose web search systems. It is also a challenging task, due to the sparsity of query terms and the lack of labeled queries. On the other hand, search contexts embedded in query sessions and unlabeled queries free on the web have not been fully utilized in most query classification systems. In this work, we leverage these information to improve query classification accuracy.

Conditional random fields(CRFs) are a class of undirected graphical models which have been widely used for classifying and labeling sequence data. The training of CRFs is typically formulated as an unconstrained optimization problem that maximizes the conditional likelihood. However, maximum likelihood training is prone to overfitting. To address this issue, we propose a novel constrained nonlinear optimization formulation in which the prediction accuracy of cross-validation sets are included as constraints. Instead of requiring multiple passes of training, the constrained formulation allows the cross-validation be handled in one pass of constrained optimization.