Shuai Zhang

Yixin Chen, Shuai Zhang, Boran Han et al. · amazon-science

In this work, we introduce Context-Aware MultiModal Learner (CaMML), for tuning large multimodal models (LMMs). CaMML, a lightweight module, is crafted to seamlessly integrate multimodal contextual samples into large models, thereby empowering the model to derive knowledge from analogous, domain-specific, up-to-date information and make grounded inferences. Importantly, CaMML is highly scalable and can efficiently handle lengthy multimodal context examples owing to its hierarchical design. Based on CaMML, we have developed two multimodal models, CaMML-7B and CaMML-13B, that have shown exceptional performance across an array of benchmark datasets for multimodal tasks. Remarkably, CaMML-13B achieves the state-of-the-art performance on over ten widely recognized multimodal benchmark datasets, surpassing LLaVA-1.5 (13B) with a noticeable margin, without integration of any external resources. Moreover, we have conducted extensive ablative studies to inspect the inner workings of CaMML and performed qualitative analyses to showcase its effectiveness in handling real-world challenging cases. Code and models are available at: https://github.com/amazon-science/camml.

Shuai Zhang, Yi Tay, Lina Yao et al.

Deep learning based recommender systems have been extensively explored in recent years. However, the large number of models proposed each year poses a big challenge for both researchers and practitioners in reproducing the results for further comparisons. Although a portion of papers provides source code, they adopted different programming languages or different deep learning packages, which also raises the bar in grasping the ideas. To alleviate this problem, we released the open source project: \textbf{DeepRec}. In this toolkit, we have implemented a number of deep learning based recommendation algorithms using Python and the widely used deep learning package - Tensorflow. Three major recommendation scenarios: rating prediction, top-N recommendation (item ranking) and sequential recommendation, were considered. Meanwhile, DeepRec maintains good modularity and extensibility to easily incorporate new models into the framework. It is distributed under the terms of the GNU General Public License. The source code is available at github: \url{https://github.com/cheungdaven/DeepRec}

Yashar Deldjoo, Nikhil Mehta, Maheswaran Sathiamoorthy et al. · amazon-science

Recommender systems powered by generative models (Gen-RecSys) extend beyond classical item ranking by producing open-ended content, which simultaneously unlocks richer user experiences and introduces new risks. On one hand, these systems can enhance personalization and appeal through dynamic explanations and multi-turn dialogues. On the other hand, they might venture into unknown territory-hallucinating nonexistent items, amplifying bias, or leaking private information. Traditional accuracy metrics cannot fully capture these challenges, as they fail to measure factual correctness, content safety, or alignment with user intent. This paper makes two main contributions. First, we categorize the evaluation challenges of Gen-RecSys into two groups: (i) existing concerns that are exacerbated by generative outputs (e.g., bias, privacy) and (ii) entirely new risks (e.g., item hallucinations, contradictory explanations). Second, we propose a holistic evaluation approach that includes scenario-based assessments and multi-metric checks-incorporating relevance, factual grounding, bias detection, and policy compliance. Our goal is to provide a guiding framework so researchers and practitioners can thoroughly assess Gen-RecSys, ensuring effective personalization and responsible deployment.

Dyah Adila, Shuai Zhang, Boran Han et al. · amazon-science

The integration of contextual information has significantly enhanced the performance of large language models (LLMs) on knowledge-intensive tasks. However, existing methods often overlook a critical challenge: the credibility of context documents can vary widely, potentially leading to the propagation of unreliable information. In this paper, we introduce CrEst, a novel weakly supervised framework for assessing the credibility of context documents during LLM inference--without requiring manual annotations. Our approach is grounded in the insight that credible documents tend to exhibit higher semantic coherence with other credible documents, enabling automated credibility estimation through inter-document agreement. To incorporate credibility into LLM inference, we propose two integration strategies: a black-box approach for models without access to internal weights or activations, and a white-box method that directly modifies attention mechanisms. Extensive experiments across three model architectures and five datasets demonstrate that CrEst consistently outperforms strong baselines, achieving up to a 26.86% improvement in accuracy and a 3.49% increase in F1 score. Further analysis shows that CrEst maintains robust performance even under high-noise conditions.

Lichao Zhang, Jia Yu, Shuai Zhang et al.

Large Language Models (LLMs) have significantly advanced user-bot interactions, enabling more complex and coherent dialogues. However, the prevalent text-only modality might not fully exploit the potential for effective user engagement. This paper explores the impact of multi-modal interactions, which incorporate images and audio alongside text, on user engagement in chatbot conversations. We conduct a comprehensive analysis using a diverse set of chatbots and real-user interaction data, employing metrics such as retention rate and conversation length to evaluate user engagement. Our findings reveal a significant enhancement in user engagement with multi-modal interactions compared to text-only dialogues. Notably, the incorporation of a third modality significantly amplifies engagement beyond the benefits observed with just two modalities. These results suggest that multi-modal interactions optimize cognitive processing and facilitate richer information comprehension. This study underscores the importance of multi-modality in chatbot design, offering valuable insights for creating more engaging and immersive AI communication experiences and informing the broader AI community about the benefits of multi-modal interactions in enhancing user engagement.

Dyah Adila, Shuai Zhang, Boran Han et al.

The question-answering (QA) capabilities of foundation models are highly sensitive to prompt variations, rendering their performance susceptible to superficial, non-meaning-altering changes. This vulnerability often stems from the model's preference or bias towards specific input characteristics, such as option position or superficial image features in multi-modal settings. We propose to rectify this bias directly in the model's internal representation. Our approach, SteerFair, finds the bias direction in the model's representation space and steers activation values away from it during inference. Specifically, we exploit the observation that bias often adheres to simple association rules, such as the spurious association between the first option and correctness likelihood. Next, we construct demonstrations of these rules from unlabeled samples and use them to identify the bias directions. We empirically show that SteerFair significantly reduces instruction-tuned model performance variance across prompt modifications on three benchmark tasks. Remarkably, our approach surpasses a supervised baseline with 100 labels by an average of 10.86% accuracy points and 12.95 score points and matches the performance with 500 labels.

Zixiang Ding, Guoqing Jiang, Shuai Zhang et al.

In this paper, we propose Stochastic Knowledge Distillation (SKD) to obtain compact BERT-style language model dubbed SKDBERT. In each iteration, SKD samples a teacher model from a pre-defined teacher ensemble, which consists of multiple teacher models with multi-level capacities, to transfer knowledge into student model in an one-to-one manner. Sampling distribution plays an important role in SKD. We heuristically present three types of sampling distributions to assign appropriate probabilities for multi-level teacher models. SKD has two advantages: 1) it can preserve the diversities of multi-level teacher models via stochastically sampling single teacher model in each iteration, and 2) it can also improve the efficacy of knowledge distillation via multi-level teacher models when large capacity gap exists between the teacher model and the student model. Experimental results on GLUE benchmark show that SKDBERT reduces the size of a BERT$_{\rm BASE}$ model by 40% while retaining 99.5% performances of language understanding and being 100% faster.

Yongliang Shen, Xinyin Ma, Zeqi Tan et al.

Named entity recognition (NER) is a well-studied task in natural language processing. Traditional NER research only deals with flat entities and ignores nested entities. The span-based methods treat entity recognition as a span classification task. Although these methods have the innate ability to handle nested NER, they suffer from high computational cost, ignorance of boundary information, under-utilization of the spans that partially match with entities, and difficulties in long entity recognition. To tackle these issues, we propose a two-stage entity identifier. First we generate span proposals by filtering and boundary regression on the seed spans to locate the entities, and then label the boundary-adjusted span proposals with the corresponding categories. Our method effectively utilizes the boundary information of entities and partially matched spans during training. Through boundary regression, entities of any length can be covered theoretically, which improves the ability to recognize long entities. In addition, many low-quality seed spans are filtered out in the first stage, which reduces the time complexity of inference. Experiments on nested NER datasets demonstrate that our proposed method outperforms previous state-of-the-art models.

Shuai Zhang, Yi Tay, Wenqi Jiang et al.

Learning embedding spaces of suitable geometry is critical for representation learning. In order for learned representations to be effective and efficient, it is ideal that the geometric inductive bias aligns well with the underlying structure of the data. In this paper, we propose Switch Spaces, a data-driven approach for learning representations in product space. Specifically, product spaces (or manifolds) are spaces of mixed curvature, i.e., a combination of multiple euclidean and non-euclidean (hyperbolic, spherical) manifolds. To this end, we introduce sparse gating mechanisms that learn to choose, combine and switch spaces, allowing them to be switchable depending on the input data with specialization. Additionally, the proposed method is also efficient and has a constant computational complexity regardless of the model size. Experiments on knowledge graph completion and item recommendations show that the proposed switch space achieves new state-of-the-art performances, outperforming pure product spaces and recently proposed task-specific models.

Aston Zhang, Yi Tay, Shuai Zhang et al.

Recent works have demonstrated reasonable success of representation learning in hypercomplex space. Specifically, "fully-connected layers with Quaternions" (4D hypercomplex numbers), which replace real-valued matrix multiplications in fully-connected layers with Hamilton products of Quaternions, both enjoy parameter savings with only 1/4 learnable parameters and achieve comparable performance in various applications. However, one key caveat is that hypercomplex space only exists at very few predefined dimensions (4D, 8D, and 16D). This restricts the flexibility of models that leverage hypercomplex multiplications. To this end, we propose parameterizing hypercomplex multiplications, allowing models to learn multiplication rules from data regardless of whether such rules are predefined. As a result, our method not only subsumes the Hamilton product, but also learns to operate on any arbitrary nD hypercomplex space, providing more architectural flexibility using arbitrarily $1/n$ learnable parameters compared with the fully-connected layer counterpart. Experiments of applications to the LSTM and Transformer models on natural language inference, machine translation, text style transfer, and subject verb agreement demonstrate architectural flexibility and effectiveness of the proposed approach.

Ye Bai, Jiangyan Yi, Jianhua Tao et al.

Attention-based encoder-decoder (AED) models have achieved promising performance in speech recognition. However, because the decoder predicts text tokens (such as characters or words) in an autoregressive manner, it is difficult for an AED model to predict all tokens in parallel. This makes the inference speed relatively slow. We believe that because the encoder already captures the whole speech utterance, which has the token-level relationship implicitly, we can predict a token without explicitly autoregressive language modeling. When the prediction of a token does not rely on other tokens, the parallel prediction of all tokens in the sequence is realizable. Based on this idea, we propose a non-autoregressive speech recognition model called LASO (Listen Attentively, and Spell Once). The model consists of an encoder, a decoder, and a position dependent summarizer (PDS). The three modules are based on basic attention blocks. The encoder extracts high-level representations from the speech. The PDS uses positional encodings corresponding to tokens to convert the acoustic representations into token-level representations. The decoder further captures token-level relationships with the self-attention mechanism. At last, the probability distribution on the vocabulary is computed for each token position. Therefore, speech recognition is re-formulated as a position-wise classification problem. Further, we propose a cross-modal transfer learning method to refine semantics from a large-scale pre-trained language model BERT for improving the performance.

Susie Xi Rao, Shuai Zhang, Zhichao Han et al.

Massive account registration has raised concerns on risk management in e-commerce companies, especially when registration increases rapidly within a short time frame. To monitor these registrations constantly and minimize the potential loss they might incur, detecting massive registration and predicting their riskiness are necessary. In this paper, we propose a Dynamic Heterogeneous Graph Neural Network framework to capture suspicious massive registrations (DHGReg). We first construct a dynamic heterogeneous graph from the registration data, which is composed of a structural subgraph and a temporal subgraph. Then, we design an efficient architecture to predict suspicious/benign accounts. Our proposed model outperforms the baseline models and is computationally efficient in processing a dynamic heterogeneous graph constructed from a real-world dataset. In practice, the DHGReg framework would benefit the detection of suspicious registration behaviors at an early stage.

Shuai Zhang, Huoyu Liu, Aston Zhang et al.

Modeling user interests is crucial in real-world recommender systems. In this paper, we present a new user interest representation model for personalized recommendation. Specifically, the key novelty behind our model is that it explicitly models user interests as a hypercuboid instead of a point in the space. In our approach, the recommendation score is learned by calculating a compositional distance between the user hypercuboid and the item. This helps to alleviate the potential geometric inflexibility of existing collaborative filtering approaches, enabling a greater extent of modeling capability. Furthermore, we present two variants of hypercuboids to enhance the capability in capturing the diversities of user interests. A neural architecture is also proposed to facilitate user hypercuboid learning by capturing the activity sequences (e.g., buy and rate) of users. We demonstrate the effectiveness of our proposed model via extensive experiments on both public and commercial datasets. Empirical results show that our approach achieves very promising results, outperforming existing state-of-the-art.

Shuai Zhang, Peng Zhang, Xindian Ma et al.

Transformer has been widely-used in many Natural Language Processing (NLP) tasks and the scaled dot-product attention between tokens is a core module of Transformer. This attention is a token-wise design and its complexity is quadratic to the length of sequence, limiting its application potential for long sequence tasks. In this paper, we propose a dimension-wise attention mechanism based on which a novel language modeling approach (namely TensorCoder) can be developed. The dimension-wise attention can reduce the attention complexity from the original $O(N^2d)$ to $O(Nd^2)$, where $N$ is the length of the sequence and $d$ is the dimensionality of head. We verify TensorCoder on two tasks including masked language modeling and neural machine translation. Compared with the original Transformer, TensorCoder not only greatly reduces the calculation of the original model but also obtains improved performance on masked language modeling task (in PTB dataset) and comparable performance on machine translation tasks.

Zhengkun Tian, Jiangyan Yi, Ye Bai et al.

For most of the attention-based sequence-to-sequence models, the decoder predicts the output sequence conditioned on the entire input sequence processed by the encoder. The asynchronous problem between the encoding and decoding makes these models difficult to be applied for online speech recognition. In this paper, we propose a model named synchronous transformer to address this problem, which can predict the output sequence chunk by chunk. Once a fixed-length chunk of the input sequence is processed by the encoder, the decoder begins to predict symbols immediately. During training, a forward-backward algorithm is introduced to optimize all the possible alignment paths. Our model is evaluated on a Mandarin dataset AISHELL-1. The experiments show that the synchronous transformer is able to perform encoding and decoding synchronously, and achieves a character error rate of 8.91% on the test set.

Xindian Ma, Peng Zhang, Shuai Zhang et al.

Latest development of neural models has connected the encoder and decoder through a self-attention mechanism. In particular, Transformer, which is solely based on self-attention, has led to breakthroughs in Natural Language Processing (NLP) tasks. However, the multi-head attention mechanism, as a key component of Transformer, limits the effective deployment of the model to a resource-limited setting. In this paper, based on the ideas of tensor decomposition and parameters sharing, we propose a novel self-attention model (namely Multi-linear attention) with Block-Term Tensor Decomposition (BTD). We test and verify the proposed attention method on three language modeling tasks (i.e., PTB, WikiText-103 and One-billion) and a neural machine translation task (i.e., WMT-2016 English-German). Multi-linear attention can not only largely compress the model parameters but also obtain performance improvements, compared with a number of language modeling approaches, such as Transformer, Transformer-XL, and Transformer with tensor train decomposition.

Yi Tay, Aston Zhang, Luu Anh Tuan et al.

Many state-of-the-art neural models for NLP are heavily parameterized and thus memory inefficient. This paper proposes a series of lightweight and memory efficient neural architectures for a potpourri of natural language processing (NLP) tasks. To this end, our models exploit computation using Quaternion algebra and hypercomplex spaces, enabling not only expressive inter-component interactions but also significantly ($75\%$) reduced parameter size due to lesser degrees of freedom in the Hamilton product. We propose Quaternion variants of models, giving rise to new architectures such as the Quaternion attention Model and Quaternion Transformer. Extensive experiments on a battery of NLP tasks demonstrates the utility of proposed Quaternion-inspired models, enabling up to $75\%$ reduction in parameter size without significant loss in performance.

Shuai Zhang, Lina Yao, Lucas Vinh Tran et al.

This paper proposes Quaternion Collaborative Filtering (QCF), a novel representation learning method for recommendation. Our proposed QCF relies on and exploits computation with Quaternion algebra, benefiting from the expressiveness and rich representation learning capability of Hamilton products. Quaternion representations, based on hypercomplex numbers, enable rich inter-latent dependencies between imaginary components. This encourages intricate relations to be captured when learning user-item interactions, serving as a strong inductive bias as compared with the real-space inner product. All in all, we conduct extensive experiments on six real-world datasets, demonstrating the effectiveness of Quaternion algebra in recommender systems. The results exhibit that QCF outperforms a wide spectrum of strong neural baselines on all datasets. Ablative experiments confirm the effectiveness of Hamilton-based composition over multi-embedding composition in real space.

Shuai Zhang, Yi Tay, Lina Yao et al.

In this work, we move beyond the traditional complex-valued representations, introducing more expressive hypercomplex representations to model entities and relations for knowledge graph embeddings. More specifically, quaternion embeddings, hypercomplex-valued embeddings with three imaginary components, are utilized to represent entities. Relations are modelled as rotations in the quaternion space. The advantages of the proposed approach are: (1) Latent inter-dependencies (between all components) are aptly captured with Hamilton product, encouraging a more compact interaction between entities and relations; (2) Quaternions enable expressive rotation in four-dimensional space and have more degree of freedom than rotation in complex plane; (3) The proposed framework is a generalization of ComplEx on hypercomplex space while offering better geometrical interpretations, concurrently satisfying the key desiderata of relational representation learning (i.e., modeling symmetry, anti-symmetry and inversion). Experimental results demonstrate that our method achieves state-of-the-art performance on four well-established knowledge graph completion benchmarks.

Lucas Vinh Tran, Yi Tay, Shuai Zhang et al.

This paper investigates the notion of learning user and item representations in non-Euclidean space. Specifically, we study the connection between metric learning in hyperbolic space and collaborative filtering by exploring Mobius gyrovector spaces where the formalism of the spaces could be utilized to generalize the most common Euclidean vector operations. Overall, this work aims to bridge the gap between Euclidean and hyperbolic geometry in recommender systems through metric learning approach. We propose HyperML (Hyperbolic Metric Learning), a conceptually simple but highly effective model for boosting the performance. Via a series of extensive experiments, we show that our proposed HyperML not only outperforms their Euclidean counterparts, but also achieves state-of-the-art performance on multiple benchmark datasets, demonstrating the effectiveness of personalized recommendation in hyperbolic geometry.

Shuai Zhang, Yi Tay, Lina Yao et al.

In this paper, we propose a novel sequence-aware recommendation model. Our model utilizes self-attention mechanism to infer the item-item relationship from user's historical interactions. With self-attention, it is able to estimate the relative weights of each item in user interaction trajectories to learn better representations for user's transient interests. The model is finally trained in a metric learning framework, taking both short-term and long-term intentions into consideration. Experiments on a wide range of datasets on different domains demonstrate that our approach outperforms the state-of-the-art by a wide margin.

Chaoran Huang, Lina Yao, Xianzhi Wang et al.

Knowledge acquisition and exchange are generally crucial yet costly for both businesses and individuals, especially when the knowledge concerns various areas. Question Answering Communities offer an opportunity for sharing knowledge at a low cost, where communities users, many of whom are domain experts, can potentially provide high-quality solutions to a given problem. In this paper, we propose a framework for finding experts across multiple collaborative networks. We employ the recent techniques of tree-guided learning (via tensor decomposition), and matrix factorization to explore user expertise from past voted posts. Tensor decomposition enables to leverage the latent expertise of users, and the posts and related tags help identify the related areas. The final result is an expertise score for every user on every knowledge area. We experiment on Stack Exchange Networks, a set of question answering websites on different topics with a huge group of users and posts. Experiments show our proposed approach produces steady and premium outputs.

Shuai Zhang, Lina Yao, Aixin Sun et al.

Modeling user-item interaction patterns is an important task for personalized recommendations. Many recommender systems are based on the assumption that there exists a linear relationship between users and items while neglecting the intricacy and non-linearity of real-life historical interactions. In this paper, we propose a neural network based recommendation model (NeuRec) that untangles the complexity of user-item interactions and establishes an integrated network to combine non-linear transformation with latent factors. We further design two variants of NeuRec: user-based NeuRec and item-based NeuRec, by concentrating on different aspects of the interaction matrix. Extensive experiments on four real-world datasets demonstrated their superior performances on personalized ranking task.

Xiang Zhang, Lina Yao, Shuai Zhang et al.

A Brain-Computer Interface (BCI) acquires brain signals, analyzes and translates them into commands that are relayed to actuation devices for carrying out desired actions. With the widespread connectivity of everyday devices realized by the advent of the Internet of Things (IoT), BCI can empower individuals to directly control objects such as smart home appliances or assistive robots, directly via their thoughts. However, realization of this vision is faced with a number of challenges, most importantly being the issue of accurately interpreting the intent of the individual from the raw brain signals that are often of low fidelity and subject to noise. Moreover, pre-processing brain signals and the subsequent feature engineering are both time-consuming and highly reliant on human domain expertise. To address the aforementioned issues, in this paper, we propose a unified deep learning based framework that enables effective human-thing cognitive interactivity in order to bridge individuals and IoT objects. We design a reinforcement learning based Selective Attention Mechanism (SAM) to discover the distinctive features from the input brain signals. In addition, we propose a modified Long Short-Term Memory (LSTM) to distinguish the inter-dimensional information forwarded from the SAM. To evaluate the efficiency of the proposed framework, we conduct extensive real-world experiments and demonstrate that our model outperforms a number of competitive state-of-the-art baselines. Two practical real-time human-thing cognitive interaction applications are presented to validate the feasibility of our approach.

Shuai Zhang, Lina Yao, Yi Tay et al.

In the past decade, matrix factorization has been extensively researched and has become one of the most popular techniques for personalized recommendations. Nevertheless, the dot product adopted in matrix factorization based recommender models does not satisfy the inequality property, which may limit their expressiveness and lead to sub-optimal solutions. To overcome this problem, we propose a novel recommender technique dubbed as {\em Metric Factorization}. We assume that users and items can be placed in a low dimensional space and their explicit closeness can be measured using Euclidean distance which satisfies the inequality property. To demonstrate its effectiveness, we further designed two variants of metric factorization with one for rating estimation and the other for personalized item ranking. Extensive experiments on a number of real-world datasets show that our approach outperforms existing state-of-the-art by a large margin on both rating prediction and item ranking tasks.

Shuai Zhang, Lina Yao, Aixin Sun et al.

With the ever-growing volume of online information, recommender systems have been an effective strategy to overcome such information overload. The utility of recommender systems cannot be overstated, given its widespread adoption in many web applications, along with its potential impact to ameliorate many problems related to over-choice. In recent years, deep learning has garnered considerable interest in many research fields such as computer vision and natural language processing, owing not only to stellar performance but also the attractive property of learning feature representations from scratch. The influence of deep learning is also pervasive, recently demonstrating its effectiveness when applied to information retrieval and recommender systems research. Evidently, the field of deep learning in recommender system is flourishing. This article aims to provide a comprehensive review of recent research efforts on deep learning based recommender systems. More concretely, we provide and devise a taxonomy of deep learning based recommendation models, along with providing a comprehensive summary of the state-of-the-art. Finally, we expand on current trends and provide new perspectives pertaining to this new exciting development of the field.

Shuai Zhang, Lina Yao, Xiwei Xu et al.

In this paper, we present a novel structure, Semi-AutoEncoder, based on AutoEncoder. We generalize it into a hybrid collaborative filtering model for rating prediction as well as personalized top-n recommendations. Experimental results on two real-world datasets demonstrate its state-of-the-art performances.

Shuai Zhang, Lina Yao

Recommender systems have been actively and extensively studied over past decades. In the meanwhile, the boom of Big Data is driving fundamental changes in the development of recommender systems. In this paper, we propose a dynamic intention-aware recommender system to better facilitate users to find desirable products and services. Compare to prior work, our proposal possesses the following advantages: (1) it takes user intentions and demands into account through intention mining techniques. By unearthing user intentions from the historical user-item interactions, and various user digital traces harvested from social media and Internet of Things, it is capable of delivering more satisfactory recommendations by leveraging rich online and offline user data; (2) it embraces the benefits of embedding heterogeneous source information and shared representations of multiple domains to provide accurate and effective recommendations comprehensively; (3) it recommends products or services proactively and timely by capturing the dynamic influences, which can significantly reduce user involvements and efforts.