Field-aware Factorization Machines (FFMs) have emerged as a powerful model for click-through rate prediction, particularly excelling in capturing complex feature interactions. In this work, we present an in-depth analysis of our in-house, Rust-based Deep FFM implementation, and detail its deployment on a CPU-only, multi-data-center scale. We overview key optimizations devised for both training and inference, demonstrated by previously unpublished benchmark results in efficient model search and online training. Further, we detail an in-house weight quantization that resulted in more than an order of magnitude reduction in bandwidth footprint related to weight transfers across data-centres. We disclose the engine and associated techniques under an open-source license to contribute to the broader machine learning community. This paper showcases one of the first successful CPU-only deployments of Deep FFMs at such scale, marking a significant stride in practical, low-footprint click-through rate prediction methodologies.
The modeling of users' behaviors is crucial in modern recommendation systems. A lot of research focuses on modeling users' lifelong sequences, which can be extremely long and sometimes exceed thousands of items. These models use the target item to search for the most relevant items from the historical sequence. However, training lifelong sequences in click through rate (CTR) prediction or personalized search ranking (PSR) is extremely difficult due to the insufficient learning problem of ID embedding, especially when the IDs in the lifelong sequence features do not exist in the samples of training dataset. Additionally, existing target attention mechanisms struggle to learn the multi-modal representations of items in the sequence well. The distribution of multi-modal embedding (text, image and attributes) output of user's interacted items are not properly aligned and there exist divergence across modalities. We also observe that users' search query sequences and item browsing sequences can fully depict users' intents and benefit from each other. To address these challenges, we propose a unified lifelong multi-modal sequence model called SEMINAR-Search Enhanced Multi-Modal Interest Network and Approximate Retrieval. Specifically, a network called Pretraining Search Unit (PSU) learns the lifelong sequences of multi-modal query-item pairs in a pretraining-finetuning manner with multiple objectives: multi-modal alignment, next query-item pair prediction, query-item relevance prediction, etc. After pretraining, the downstream model restores the pretrained embedding as initialization and finetunes the network. To accelerate the online retrieval speed of multi-modal embedding, we propose a multi-modal codebook-based product quantization strategy to approximate the exact attention calculati
This paper investigates the one-epoch overfitting phenomenon in Click-Through Rate (CTR) models, where performance notably declines at the start of the second epoch. Despite extensive research, the efficacy of multi-epoch training over the conventional one-epoch approach remains unclear. We identify the overfitting of the embedding layer, caused by high-dimensional data sparsity, as the primary issue. To address this, we introduce a novel and simple Multi-Epoch learning with Data Augmentation (MEDA) framework, suitable for both non-continual and continual learning scenarios, which can be seamlessly integrated into existing deep CTR models and may have potential applications to handle the "forgetting or overfitting" dilemma in the retraining and the well-known catastrophic forgetting problems. MEDA minimizes overfitting by reducing the dependency of the embedding layer on subsequent training data or the Multi-Layer Perceptron (MLP) layers, and achieves data augmentation through training the MLP with varied embedding spaces. Our findings confirm that pre-trained MLP layers can adapt to new embedding spaces, enhancing performance without overfitting. This adaptability underscores the MLP layers' role in learning a matching function focused on the relative relationships among embeddings rather than their absolute positions. To our knowledge, MEDA represents the first multi-epoch training strategy tailored for deep CTR prediction models. We conduct extensive experiments on several public and business datasets, and the effectiveness of data augmentation and superiority over conventional single-epoch training are fully demonstrated. Besides, MEDA has exhibited significant benefits in a real-world online advertising system.
Click-through rate (CTR) prediction is a crucial task in online advertising to recommend products that users are likely to be interested in. To identify the best-performing models, rigorous model evaluation is necessary. Offline experimentation plays a significant role in selecting models for live user-item interactions, despite the value of online experimentation like A/B testing, which has its own limitations and risks. Often, the correlation between offline performance metrics and actual online model performance is inadequate. One main reason for this discrepancy is the common practice of using random splits to create training, validation, and test datasets in CTR prediction. In contrast, real-world CTR prediction follows a temporal order. Therefore, the methodology used in offline evaluation, particularly the data splitting strategy, is crucial. This study aims to address the inconsistency between current offline evaluation methods and real-world use cases, by focusing on data splitting strategies. To examine the impact of different data split strategies on offline performance, we conduct extensive experiments using both random and temporal splits on a large open benchmark dataset, Criteo.
Extreme Learning Machines (ELM) provide a fast alternative to traditional gradient-based learning in neural networks, offering rapid training and robust generalization capabilities. Its theoretical basis shows its universal approximation capability. We explore the application of ELMs for the task of Click-Through Rate (CTR) prediction, which is largely unexplored by ELMs due to the high dimensionality of the problem. We introduce an ELM-based model enhanced with embedding layers to improve the performance on CTR tasks, which is a novel addition to the field. Experimental results on benchmark datasets, including Avazu and Criteo, demonstrate that our proposed ELM with embeddings achieves competitive F1 results while significantly reducing training time compared to state-of-the-art models such as Masknet. Our findings show that ELMs can be useful for CTR prediction, especially when fast training is needed.
Click-through-rate (CTR) prediction has an essential impact on improving user experience and revenue in e-commerce search. With the development of deep learning, graph-based methods are well exploited to utilize graph structure extracted from user behaviors and other information to help embedding learning. However, most of the previous graph-based methods mainly focus on recommendation scenarios, and therefore their graph structures highly depend on item's sequential information from user behaviors, ignoring query's sequential signal and query-item correlation. In this paper, we propose a new approach named Light-weight End-to-End Graph Interest Network (EGIN) to effectively mine users' search interests and tackle previous challenges. (i) EGIN utilizes query and item's correlation and sequential information from the search system to build a heterogeneous graph for better CTR prediction in e-commerce search. (ii) EGIN's graph embedding learning shares the same training input and is jointly trained with CTR prediction, making the end-to-end framework effortless to deploy in large-scale search systems. The proposed EGIN is composed of three parts: query-item heterogeneous graph, light-weight graph sampling, and multi-interest network. The query-item heterogeneous graph captures correlation and sequential information of query and item efficiently by the proposed light-weight graph sampling. The multi-interest network is well designed to utilize graph embedding to capture various similarity relationships between query and item to enhance the final CTR prediction. We conduct extensive experiments on both public and industrial datasets to demonstrate the effectiveness of the proposed EGIN. At the same time, the training cost of graph learning is relatively low compared with the main CTR prediction task, ensuring efficiency in practical applications.
In this paper, we introduce Star+, a novel multi-domain model for click-through rate (CTR) prediction inspired by the Star model. Traditional single-domain approaches and existing multi-task learning techniques face challenges in multi-domain environments due to their inability to capture domain-specific data distributions and complex inter-domain relationships. Star+ addresses these limitations by enhancing the interaction between shared and domain-specific information through various fusion strategies, such as add, adaptive add, concatenation, and gating fusions, to find the optimal balance between domain-specific and shared information. We also investigate the impact of different normalization techniques, including layer normalization, batch normalization, and partition normalization, on the performance of our model. Our extensive experiments on both industrial and public datasets demonstrate that Star+ significantly improves prediction accuracy and efficiency. This work contributes to the advancement of recommendation systems by providing a robust, scalable, and adaptive solution for multi-domain environments.
Click-Through Rate (CTR) prediction has become an essential task in digital industries, such as digital advertising or online shopping. Many deep learning-based methods have been implemented and have become state-of-the-art models in the domain. To further improve the performance of CTR models, Knowledge Distillation based approaches have been widely used. However, most of the current CTR prediction models do not have much complex architectures, so it's hard to call one of them 'cumbersome' and the other one 'tiny'. On the other hand, the performance gap is also not very large between complex and simple models. So, distilling knowledge from one model to the other could not be worth the effort. Under these considerations, Mutual Learning could be a better approach, since all the models could be improved mutually. In this paper, we showed how useful the mutual learning algorithm could be when it is between equals. In our experiments on the Criteo and Avazu datasets, the mutual learning algorithm improved the performance of the model by up to 0.66% relative improvement.
This paper proposes new methods to enhance click-through rate (CTR) prediction models using the Deep Interest Network (DIN) model, specifically applied to the advertising system of Alibaba's Taobao platform. Unlike traditional deep learning approaches, this research focuses on localized user behavior activation for tailored ad targeting by leveraging extensive user behavior data. Compared to traditional models, this method demonstrates superior ability to handle diverse and dynamic user data, thereby improving the efficiency of ad systems and increasing revenue.
This paper introduces a novel approach for click-through rate (CTR) prediction within industrial recommender systems, addressing the inherent challenges of numerical imbalance and geometric asymmetry. These challenges stem from imbalanced datasets, where positive (click) instances occur less frequently than negatives (non-clicks), and geometrically asymmetric distributions, where positive samples exhibit visually coherent patterns while negatives demonstrate greater diversity. To address these challenges, we have used a deep neural network classifier that uses the polyhedral conic functions. This classifier is similar to the one-class classifiers in spirit and it returns compact polyhedral acceptance regions to separate the positive class samples from the negative samples that have diverse distributions. Extensive experiments have been conducted to test the proposed approach using state-of-the-art (SOTA) CTR prediction models on four public datasets, namely Criteo, Avazu, MovieLens and Frappe. The experimental evaluations highlight the superiority of our proposed approach over Binary Cross Entropy (BCE) Loss, which is widely used in CTR prediction tasks.
Position bias, i.e., users' preference of an item is affected by its placing position, is well studied in the recommender system literature. However, most existing methods ignore the widely coupled ranking bias, which is also related to the placing position of the item. Using both synthetic and industrial datasets, we first show how this widely coexisted ranking bias deteriorates the performance of the existing position bias estimation methods. To mitigate the position bias with the presence of the ranking bias, we propose a novel position bias estimation method, namely gradient interpolation, which fuses two estimation methods using a fusing weight. We further propose an adaptive method to automatically determine the optimal fusing weight. Extensive experiments on both synthetic and industrial datasets demonstrate the superior performance of the proposed methods.
With the applications of recommendation systems rapidly expanding, an increasing number of studies have focused on every aspect of recommender systems with different data inputs, models, and task settings. Therefore, a flexible library is needed to help researchers implement the experimental strategies they require. Existing open libraries for recommendation scenarios have enabled reproducing various recommendation methods and provided standard implementations. However, these libraries often impose certain restrictions on data and seldom support the same model to perform different tasks and input formats, limiting users from customized explorations. To fill the gap, we propose ReChorus2.0, a modular and task-flexible library for recommendation researchers. Based on ReChorus, we upgrade the supported input formats, models, and training&evaluation strategies to help realize more recommendation tasks with more data types. The main contributions of ReChorus2.0 include: (1) Realization of complex and practical tasks, including reranking and CTR prediction tasks; (2) Inclusion of various context-aware and rerank recommenders; (3) Extension of existing and new models to support different tasks with the same models; (4) Support of highly-customized input with impression logs, negative items, or click labels, as well as user, item, and situation contexts. To summarize, ReChorus2.0 serves as a comprehensive and flexible library better aligning with the practical problems in the recommendation scenario and catering to more diverse research needs. The implementation and detailed tutorials of ReChorus2.0 can be found at https://github.com/THUwangcy/ReChorus.
Deep Click-Through Rate (CTR) prediction models play an important role in modern industrial recommendation scenarios. However, high memory overhead and computational costs limit their deployment in resource-constrained environments. Low-rank approximation is an effective method for computer vision and natural language processing models, but its application in compressing CTR prediction models has been less explored. Due to the limited memory and computing resources, compression of CTR prediction models often confronts three fundamental challenges, i.e., (1). How to reduce the model sizes to adapt to edge devices? (2). How to speed up CTR prediction model inference? (3). How to retain the capabilities of original models after compression? Previous low-rank compression research mostly uses tensor decomposition, which can achieve a high parameter compression ratio, but brings in AUC degradation and additional computing overhead. To address these challenges, we propose a unified low-rank decomposition framework for compressing CTR prediction models. We find that even with the most classic matrix decomposition SVD method, our framework can achieve better performance than the original model. To further improve the effectiveness of our framework, we locally compress the output features instead of compressing the model weights. Our unified low-rank compression framework can be applied to embedding tables and MLP layers in various CTR prediction models. Extensive experiments on two academic datasets and one real industrial benchmark demonstrate that, with 3-5x model size reduction, our compressed models can achieve both faster inference and higher AUC than the uncompressed original models. Our code is at https://github.com/yuhao318/Atomic_Feature_Mimicking.
Sequential recommendation aims to estimate how a user's interests evolve over time via uncovering valuable patterns from user behavior history. Many previous sequential models have solely relied on users' historical information to model the evolution of their interests, neglecting the crucial role that future information plays in accurately capturing these dynamics. However, effectively incorporating future information in sequential modeling is non-trivial since it is impossible to make the current-step prediction for any target user by leveraging his future data. In this paper, we propose a novel framework of sequential recommendation called Look into the Future (LIFT), which builds and leverages the contexts of sequential recommendation. In LIFT, the context of a target user's interaction is represented based on i) his own past behaviors and ii) the past and future behaviors of the retrieved similar interactions from other users. As such, the learned context will be more informative and effective in predicting the target user's behaviors in sequential recommendation without temporal data leakage. Furthermore, in order to exploit the intrinsic information embedded within the context itself, we introduce an innovative pretraining methodology incorporating behavior masking. In our extensive experiments on five real-world datasets, LIFT achieves significant performance improvement on click-through rate prediction and rating prediction tasks in sequential recommendation over strong baselines, demonstrating that retrieving and leveraging relevant contexts from the global user pool greatly benefits sequential recommendation. The experiment code is provided at https://anonymous.4open.science/r/LIFT-277C/Readme.md.
Click-Through Rate (CTR) prediction holds a paramount position in recommender systems. The prevailing ID-based paradigm underperforms in cold-start scenarios due to the skewed distribution of feature frequency. Additionally, the utilization of a single modality fails to exploit the knowledge contained within textual features. Recent efforts have sought to mitigate these challenges by integrating Pre-trained Language Models (PLMs). They design hard prompts to structure raw features into text for each interaction and then apply PLMs for text processing. With external knowledge and reasoning capabilities, PLMs extract valuable information even in cases of sparse interactions. Nevertheless, compared to ID-based models, pure text modeling degrades the efficacy of collaborative filtering, as well as feature scalability and efficiency during both training and inference. To address these issues, we propose \textbf{C}ost-\textbf{E}fficient \textbf{L}anguage Model \textbf{A}lignment (\textbf{CELA}) for CTR prediction. CELA incorporates textual features and language models while preserving the collaborative filtering capabilities of ID-based models. This model-agnostic framework can be equipped with plug-and-play textual features, with item-level alignment enhancing the utilization of external information while maintaining training and inference efficiency. Through extensive offline experiments, CELA demonstrates superior performance compared to state-of-the-art methods. Furthermore, an online A/B test conducted on an industrial App recommender system showcases its practical effectiveness, solidifying the potential for real-world applications of CELA.
Most current click-through rate prediction(CTR)models create explicit or implicit high-order feature crosses through Hadamard product or inner product, with little attention to the importance of feature crossing; only few models are either limited to the second-order explicit feature crossing, implicitly to high-order feature crossing, or can learn the importance of high-order explicit feature crossing but fail to provide good interpretability for the model. This paper proposes a new model, FiiNet (Multiple Order Feature Interaction Importance Neural Networks). The model first uses the selective kernel network (SKNet) to explicitly construct multi-order feature crosses. It dynamically learns the importance of feature interaction combinations in a fine grained manner, increasing the attention weight of important feature cross combinations and reducing the weight of featureless crosses. To verify that the FiiNet model can dynamically learn the importance of feature interaction combinations in a fine-grained manner and improve the model's recommendation performance and interpretability, this paper compares it with many click-through rate prediction models on two real datasets, proving that the FiiNet model incorporating the selective kernel network can effectively improve the recommendation effect and provide better interpretability. FiiNet model implementations are available in PyTorch.
Effective feature interaction modeling is critical for enhancing the accuracy of click-through rate (CTR) prediction in industrial recommender systems. Most of the current deep CTR models resort to building complex network architectures to better capture intricate feature interactions or user behaviors. However, we identify two limitations in these models: (1) the samples given to the model are undifferentiated, which may lead the model to learn a larger number of easy samples in a single-minded manner while ignoring a smaller number of hard samples, thus reducing the model's generalization ability; (2) differentiated feature interaction encoders are designed to capture different interactions information but receive consistent supervision signals, thereby limiting the effectiveness of the encoder. To bridge the identified gaps, this paper introduces a novel CTR prediction framework by integrating the plug-and-play Twin Focus (TF) Loss, Sample Selection Embedding Module (SSEM), and Dynamic Fusion Module (DFM), named the Twin Focus Framework for CTR (TF4CTR). Specifically, the framework employs the SSEM at the bottom of the model to differentiate between samples, thereby assigning a more suitable encoder for each sample. Meanwhile, the TF Loss provides tailored supervision signals to both simple and complex encoders. Moreover, the DFM dynamically fuses the feature interaction information captured by the encoders, resulting in more accurate predictions. Experiments on five real-world datasets confirm the effectiveness and compatibility of the framework, demonstrating its capacity to enhance various representative baselines in a model-agnostic manner. To facilitate reproducible research, our open-sourced code and detailed running logs will be made available at: https://github.com/salmon1802/TF4CTR.
Click-through rate (CTR) prediction is crucial for personalized online services. Sample-level retrieval-based models, such as RIM, have demonstrated remarkable performance. However, they face challenges including inference inefficiency and high resource consumption due to the retrieval process, which hinder their practical application in industrial settings. To address this, we propose a universal plug-and-play \underline{r}etrieval-\underline{o}riented \underline{k}nowledge (\textbf{\name}) framework that bypasses the real retrieval process. The framework features a knowledge base that preserves and imitates the retrieved \& aggregated representations using a decomposition-reconstruction paradigm. Knowledge distillation and contrastive learning optimize the knowledge base, enabling the integration of retrieval-enhanced representations with various CTR models. Experiments on three large-scale datasets demonstrate \name's exceptional compatibility and performance, with the neural knowledge base serving as an effective surrogate for the retrieval pool. \name surpasses the teacher model while maintaining superior inference efficiency and demonstrates the feasibility of distilling knowledge from non-parametric methods using a parametric approach. These results highlight \name's strong potential for real-world applications and its ability to transform retrieval-based methods into practical solutions. Our implementation code is available to support reproducibility in \url{https://github.com/HSLiu-Initial/ROK.git}.
Click-through rate (CTR) prediction tasks play a pivotal role in real-world applications, particularly in recommendation systems and online advertising. A significant research branch in this domain focuses on user behavior modeling. Current research predominantly centers on modeling co-occurrence relationships between the target item and items previously interacted with by users in their historical data. However, this focus neglects the intricate modeling of user behavior patterns. In reality, the abundance of user interaction records encompasses diverse behavior patterns, indicative of a spectrum of habitual paradigms. These patterns harbor substantial potential to significantly enhance CTR prediction performance. To harness the informational potential within user behavior patterns, we extend Target Attention (TA) to Target Pattern Attention (TPA) to model pattern-level dependencies. Furthermore, three critical challenges demand attention: the inclusion of unrelated items within behavior patterns, data sparsity in behavior patterns, and computational complexity arising from numerous patterns. To address these challenges, we introduce the Deep Pattern Network (DPN), designed to comprehensively leverage information from user behavior patterns. DPN efficiently retrieves target-related user behavior patterns using a target-aware attention mechanism. Additionally, it contributes to refining user behavior patterns through a pre-training paradigm based on self-supervised learning while promoting dependency learning within sparse patterns. Our comprehensive experiments, conducted across three public datasets, substantiate the superior performance and broad compatibility of DPN.
Click-through rate (CTR) prediction plays an indispensable role in online platforms. Numerous models have been proposed to capture users' shifting preferences by leveraging user behavior sequences. However, these historical sequences often suffer from severe homogeneity and scarcity compared to the extensive item pool. Relying solely on such sequences for user representations is inherently restrictive, as user interests extend beyond the scope of items they have previously engaged with. To address this challenge, we propose a data-driven approach to enrich user representations. We recognize user profiling and recall items as two ideal data sources within the cross-stage framework, encompassing the u2u (user-to-user) and i2i (item-to-item) aspects respectively. In this paper, we propose a novel architecture named Recall-Augmented Ranking (RAR). RAR consists of two key sub-modules, which synergistically gather information from a vast pool of look-alike users and recall items, resulting in enriched user representations. Notably, RAR is orthogonal to many existing CTR models, allowing for consistent performance improvements in a plug-and-play manner. Extensive experiments are conducted, which verify the efficacy and compatibility of RAR against the SOTA methods.