Search

This paper presents the solution of our team APEX in the Meta KDD CUP 2024: CRAG Comprehensive RAG Benchmark Challenge. The CRAG benchmark addresses the limitations of existing QA benchmarks in evaluating the diverse and dynamic challenges faced by Retrieval-Augmented Generation (RAG) systems. It provides a more comprehensive assessment of RAG performance and contributes to advancing research in this field. We propose a routing-based domain and dynamic adaptive RAG pipeline, which performs specific processing for the diverse and dynamic nature of the question in all three stages: retrieval, augmentation, and generation. Our method achieved superior performance on CRAG and ranked 2nd for Task 2&3 on the final competition leaderboard. Our implementation is available at this link: https://github.com/USTCAGI/CRAG-in-KDD-Cup2024.

Retrieval-augmented generation (RAG) is increasingly recognized as an effective approach for mitigating the hallucination of large language models (LLMs) through the integration of external knowledge. While numerous efforts, most studies focus on a single type of externeal knowledge source. However, in real-world applications, most situations involve diverse knowledge from various sources, yet this area has been less explored. The main dilemma is the lack of a suitable dataset containing multiple knowledge sources and pre-exploration of the associated issues. To address these challenges, we standardize a benchmark dataset that combines structured and unstructured knowledge across diverse and complementary domains. Based on this dataset, we further develop a plug-and-play RAG framework, PruningRAG, whose main characteristic is to employ multi-granularity pruning strategies for optimizing the integration of relevant information and minimizing misleading context. Building upon the standardized dataset and PruningRAG, we also report a series of experimental results, as well as insightful findings. Our dataset and code are publicly available\footnote{https://github.com/USTCAGI/PruningRAG}, with the aim of advancing future research in the RAG community., Comment: 10 pages, 11 figures

Graph transformers are a recent advancement in machine learning, offering a new class of neural network models for graph-structured data. The synergy between transformers and graph learning demonstrates strong performance and versatility across various graph-related tasks. This survey provides an in-depth review of recent progress and challenges in graph transformer research. We begin with foundational concepts of graphs and transformers. We then explore design perspectives of graph transformers, focusing on how they integrate graph inductive biases and graph attention mechanisms into the transformer architecture. Furthermore, we propose a taxonomy classifying graph transformers based on depth, scalability, and pre-training strategies, summarizing key principles for effective development of graph transformer models. Beyond technical analysis, we discuss the applications of graph transformer models for node-level, edge-level, and graph-level tasks, exploring their potential in other application scenarios as well. Finally, we identify remaining challenges in the field, such as scalability and efficiency, generalization and robustness, interpretability and explainability, dynamic and complex graphs, as well as data quality and diversity, charting future directions for graph transformer research., Comment: 23 pages, 4 figures

The training of large models, involving fine-tuning, faces the scarcity of high-quality data. Compared to the solutions based on centralized data centers, updating large models in the Internet of Things (IoT) faces challenges in coordinating knowledge from distributed clients by using their private and heterogeneous data. To tackle such a challenge, we propose KOALA (Federated Knowledge Transfer Fine-tuning Large Server Model with Resource-Constrained IoT Clients) to impel the training of large models in IoT. Since the resources obtained by IoT clients are limited and restricted, it is infeasible to locally execute large models and also update them in a privacy-preserving manner. Therefore, we leverage federated learning and knowledge distillation to update large models through collaboration with their small models, which can run locally at IoT clients to process their private data separately and enable large-small model knowledge transfer through iterative learning between the server and clients. Moreover, to support clients with similar or different computing capacities, KOALA is designed with two kinds of large-small model joint learning modes, namely to be homogeneous or heterogeneous. Experimental results demonstrate that compared to the conventional approach, our method can not only achieve similar training performance but also significantly reduce the need for local storage and computing power resources.

Quantum conferencing enables multiple nodes within a quantum network to share a secure group key for private message broadcasting. The key rate, however, is limited by the repeaterless capacity to distribute multiparticle entangled states across the network. Currently, in the finite-size regime, no feasible schemes utilizing existing experimental techniques can overcome the fundamental rate-distance limit of quantum conferencing in quantum networks without repeaters. Here, we propose a practical, multi-field scheme that breaks this limit, involving virtually establishing Greenberger-Horne-Zeilinger states through post-measurement coincidence matching. This proposal features a measurement-device-independent characteristic and can directly scale to support any number of users. Simulations show that the fundamental limitation on the group key rate can be overcome in a reasonable running time of sending $10^{14}$ pulses. We predict that it offers an efficient design for long-distance broadcast communication in future quantum networks., Comment: 20 pages, 6 figures

Quantum conference key agreement facilitates secure communication among multiple parties through multipartite entanglement and is anticipated to be an important cryptographic primitive for future quantum networks. However, the experimental complexity and low efficiency associated with the synchronous detection of multipartite entangled states have significantly hindered their practical application. In this work, we propose a measurement-device-independent conference key agreement protocol that utilizes asynchronous Greenberger-Horne-Zeilinger state measurement.This approach achieves a linear scaling of the conference key rate among multiple parties, exhibiting performance similar to that of the single-repeater scheme in quantum networks. The asynchronous measurement strategy bypasses the need for complex global phase locking technologies, concurrently extending the intercity transmission distance with composable security in the finite key regime. Additionally, our work also showcases the advantages of the asynchronous pairing concept in multiparty quantum entanglement., Comment: 21 pages, 7 figures

Music recommender systems play a critical role in music streaming platforms by providing users with music that they are likely to enjoy. Recent studies have shown that user emotions can influence users' preferences for music moods. However, existing emotion-aware music recommender systems (EMRSs) explicitly or implicitly assume that users' actual emotional states expressed through identical emotional words are homogeneous. They also assume that users' music mood preferences are homogeneous under the same emotional state. In this article, we propose four types of heterogeneity that an EMRS should account for: emotion heterogeneity across users, emotion heterogeneity within a user, music mood preference heterogeneity across users, and music mood preference heterogeneity within a user. We further propose a Heterogeneity-aware Deep Bayesian Network (HDBN) to model these assumptions. The HDBN mimics a user's decision process of choosing music with four components: personalized prior user emotion distribution modeling, posterior user emotion distribution modeling, user grouping, and Bayesian neural network-based music mood preference prediction. We constructed two datasets, called EmoMusicLJ and EmoMusicLJ-small, to validate our method. Extensive experiments demonstrate that our method significantly outperforms baseline approaches on metrics of HR, Precision, NDCG, and MRR. Ablation studies and case studies further validate the effectiveness of our HDBN. The source code and datasets are available at https://github.com/jingrk/HDBN., Comment: 43 pages, 20 figures

COVID-19 as a global pandemic causes a massive disruption to social stability that threatens human life and the economy. Policymakers and all elements of society must deliver measurable actions based on the pandemic's severity to minimize the detrimental impact of COVID-19. A proper forecasting system is arguably important to provide an early signal of the risk of COVID-19 infection so that the authorities are ready to protect the people from the worst. However, making a good forecasting model for infection risks in different cities or regions is not an easy task, because it has a lot of influential factors that are difficult to be identified manually. To address the current limitations, we propose a deep graph learning model, called PANDORA, to predict the infection risks of COVID-19, by considering all essential factors and integrating them into a geographical network. The framework uses geographical position relations and transportation frequency as higher-order structural properties formulated by higher-order network structures (i.e., network motifs). Moreover, four significant node attributes (i.e., multiple features of a particular area, including climate, medical condition, economy, and human mobility) are also considered. We propose three different aggregators to better aggregate node attributes and structural features, namely, Hadamard, Summation, and Connection. Experimental results over real data show that PANDORA outperforms the baseline method with higher accuracy and faster convergence speed, no matter which aggregator is chosen. We believe that PANDORA using deep graph learning provides a promising approach to get superior performance in infection risk level forecasting and help humans battle the COVID-19 crisis.

Scientific collaboration is a significant behavior in knowledge creation and idea exchange. To tackle large and complex research questions, a trend of team formation has been observed in recent decades. In this study, we focus on recognizing collaborative teams and exploring inner patterns using scholarly big graph data. We propose a collaborative team recognition (CORE) model with a "core + extension" team structure to recognize collaborative teams in large academic networks. In CORE, we combine an effective evaluation index called the collaboration intensity index with a series of structural features to recognize collaborative teams in which members are in close collaboration relationships. Then, CORE is used to guide the core team members to their extension members. CORE can also serve as the foundation for team-based research. The simulation results indicate that CORE reveals inner patterns of scientific collaboration: senior scholars have broad collaborative relationships and fixed collaboration patterns, which are the underlying mechanisms of team assembly. The experimental results demonstrate that CORE is promising compared with state-of-the-art methods.

Federated recommender systems have been crucially enhanced through data sharing and continuous model updates, attributed to the pervasive connectivity and distributed computing capabilities of Internet of Things (IoT) devices. Given the sensitivity of IoT data, transparent data processing in data sharing and model updates is paramount. However, existing methods fall short in tracing the flow of shared data and the evolution of model updates. Consequently, data sharing is vulnerable to exploitation by malicious entities, raising significant data privacy concerns, while excluding data sharing will result in sub-optimal recommendations. To mitigate these concerns, we present LIBERATE, a privacy-traceable federated recommender system. We design a blockchain-based traceability mechanism, ensuring data privacy during data sharing and model updates. We further enhance privacy protection by incorporating local differential privacy in user-server communication. Extensive evaluations with the real-world dataset corroborate LIBERATE's capabilities in ensuring data privacy during data sharing and model update while maintaining efficiency and performance. Results underscore blockchain-based traceability mechanism as a promising solution for privacy-preserving in federated recommender systems.

Anomaly detection (such as telecom fraud detection and medical image detection) has attracted the increasing attention of people. The complex interaction between multiple entities widely exists in the network, which can reflect specific human behavior patterns. Such patterns can be modeled by higher-order network structures, thus benefiting anomaly detection on attributed networks. However, due to the lack of an effective mechanism in most existing graph learning methods, these complex interaction patterns fail to be applied in detecting anomalies, hindering the progress of anomaly detection to some extent. In order to address the aforementioned issue, we present a higher-order structure based anomaly detection (GUIDE) method. We exploit attribute autoencoder and structure autoencoder to reconstruct node attributes and higher-order structures, respectively. Moreover, we design a graph attention layer to evaluate the significance of neighbors to nodes through their higher-order structure differences. Finally, we leverage node attribute and higher-order structure reconstruction errors to find anomalies. Extensive experiments on five real-world datasets (i.e., ACM, Citation, Cora, DBLP, and Pubmed) are implemented to verify the effectiveness of GUIDE. Experimental results in terms of ROC-AUC, PR-AUC, and Recall@K show that GUIDE significantly outperforms the state-of-art methods.

Fairness-aware Graph Neural Networks (GNNs) often face a challenging trade-off, where prioritizing fairness may require compromising utility. In this work, we re-examine fairness through the lens of spectral graph theory, aiming to reconcile fairness and utility within the framework of spectral graph learning. We explore the correlation between sensitive features and spectrum in GNNs, using theoretical analysis to delineate the similarity between original sensitive features and those after convolution under different spectra. Our analysis reveals a reduction in the impact of similarity when the eigenvectors associated with the largest magnitude eigenvalue exhibit directional similarity. Based on these theoretical insights, we propose FUGNN, a novel spectral graph learning approach that harmonizes the conflict between fairness and utility. FUGNN ensures algorithmic fairness and utility by truncating the spectrum and optimizing eigenvector distribution during the encoding process. The fairness-aware eigenvector selection reduces the impact of convolution on sensitive features while concurrently minimizing the sacrifice of utility. FUGNN further optimizes the distribution of eigenvectors through a transformer architecture. By incorporating the optimized spectrum into the graph convolution network, FUGNN effectively learns node representations. Experiments on six real-world datasets demonstrate the superiority of FUGNN over baseline methods. The codes are available at https://github.com/yushuowiki/FUGNN., Comment: Accepted in SIGKDD 2024

The design of Graph Transformers (GTs) generally neglects considerations for fairness, resulting in biased outcomes against certain sensitive subgroups. Since GTs encode graph information without relying on message-passing mechanisms, conventional fairness-aware graph learning methods cannot be directly applicable to address these issues. To tackle this challenge, we propose FairGT, a Fairness-aware Graph Transformer explicitly crafted to mitigate fairness concerns inherent in GTs. FairGT incorporates a meticulous structural feature selection strategy and a multi-hop node feature integration method, ensuring independence of sensitive features and bolstering fairness considerations. These fairness-aware graph information encodings seamlessly integrate into the Transformer framework for downstream tasks. We also prove that the proposed fair structural topology encoding with adjacency matrix eigenvector selection and multi-hop integration are theoretically effective. Empirical evaluations conducted across five real-world datasets demonstrate FairGT's superiority in fairness metrics over existing graph transformers, graph neural networks, and state-of-the-art fairness-aware graph learning approaches.

Knowledge graphs (KGs) have garnered significant attention for their vast potential across diverse domains. However, the issue of outdated facts poses a challenge to KGs, affecting their overall quality as real-world information evolves. Existing solutions for outdated fact detection often rely on manual recognition. In response, this paper presents DEAN (Deep outdatEd fAct detectioN), a novel deep learning-based framework designed to identify outdated facts within KGs. DEAN distinguishes itself by capturing implicit structural information among facts through comprehensive modeling of both entities and relations. To effectively uncover latent out-of-date information, DEAN employs a contrastive approach based on a pre-defined Relations-to-Nodes (R2N) graph, weighted by the number of entities. Experimental results demonstrate the effectiveness and superiority of DEAN over state-of-the-art baseline methods., Comment: 10 pages, 6 figures

While Large Language Models (LLMs) dominate tasks like natural language processing and computer vision, harnessing their power for spatial-temporal forecasting remains challenging. The disparity between sequential text and complex spatial-temporal data hinders this application. To address this issue, this paper introduces STG-LLM, an innovative approach empowering LLMs for spatial-temporal forecasting. We tackle the data mismatch by proposing: 1) STG-Tokenizer: This spatial-temporal graph tokenizer transforms intricate graph data into concise tokens capturing both spatial and temporal relationships; 2) STG-Adapter: This minimalistic adapter, consisting of linear encoding and decoding layers, bridges the gap between tokenized data and LLM comprehension. By fine-tuning only a small set of parameters, it can effectively grasp the semantics of tokens generated by STG-Tokenizer, while preserving the original natural language understanding capabilities of LLMs. Extensive experiments on diverse spatial-temporal benchmark datasets show that STG-LLM successfully unlocks LLM potential for spatial-temporal forecasting. Remarkably, our approach achieves competitive performance on par with dedicated SOTA methods.

We study the radiative decay of $X(3872)$ based on the assumption that $X(3872)$ is regarded as a $c\overline{c}$ charmonium with quantum number $J^{PC}=1^{++}$ ($J,\,P,\,C$ represent the spin, parity and charge conjugation, respectively). The form factors of $X(3872)$ transitions to $J/\psi\gamma$ and $\psi'\gamma$ ($\psi'$ denotes $\psi(2S)$ throughout the paper) are calculated in the framework of the covariant light-front quark model. The phenomenological wave function of a meson depends on the parameter $\beta$, whose inverse essentially describes the confinement scale. In the present work, the parameters $\beta$ for the vector $J/\psi$ and $\psi'$ mesons will be determined through their decay constants, which are obtained from the experimental values of their partial decay widths to the electron-positron pair. For $X(3872)$, we determined the value of $\beta$ by the decay width of $X(3872)\rightarrow \psi'\gamma$. Then, we examined the width of $X(3872)\to J/\psi\gamma$ in a manner of parameter-free prediction and compared it with the experimental value. As a result, an inconsistency or contradiction occurs between the widths of $X(3872)\to J/\psi\gamma$ and $X(3872)\to \psi'\gamma$. We thus conclude that $X(3872)$ cannot be a pure $c\overline c$ resonance and that other components are necessary in its wave function., Comment: two columu, 8 pages, more clarifications and references added, accepted for publication in Phys. Letts. B

Multivariate time series anomaly detection (MTAD) plays a vital role in a wide variety of real-world application domains. Over the past few years, MTAD has attracted rapidly increasing attention from both academia and industry. Many deep learning and graph learning models have been developed for effective anomaly detection in multivariate time series data, which enable advanced applications such as smart surveillance and risk management with unprecedented capabilities. Nevertheless, MTAD is facing critical challenges deriving from the dependencies among sensors and variables, which often change over time. To address this issue, we propose a coupled attention-based neural network framework (CAN) for anomaly detection in multivariate time series data featuring dynamic variable relationships. We combine adaptive graph learning methods with graph attention to generate a global-local graph that can represent both global correlations and dynamic local correlations among sensors. To capture inter-sensor relationships and temporal dependencies, a convolutional neural network based on the global-local graph is integrated with a temporal self-attention module to construct a coupled attention module. In addition, we develop a multilevel encoder-decoder architecture that accommodates reconstruction and prediction tasks to better characterize multivariate time series data. Extensive experiments on real-world datasets have been conducted to evaluate the performance of the proposed CAN approach, and the results show that CAN significantly outperforms state-of-the-art baselines.

In the field of quantitative trading, it is common practice to transform raw historical stock data into indicative signals for the market trend. Such signals are called alpha factors. Alphas in formula forms are more interpretable and thus favored by practitioners concerned with risk. In practice, a set of formulaic alphas is often used together for better modeling precision, so we need to find synergistic formulaic alpha sets that work well together. However, most traditional alpha generators mine alphas one by one separately, overlooking the fact that the alphas would be combined later. In this paper, we propose a new alpha-mining framework that prioritizes mining a synergistic set of alphas, i.e., it directly uses the performance of the downstream combination model to optimize the alpha generator. Our framework also leverages the strong exploratory capabilities of reinforcement learning~(RL) to better explore the vast search space of formulaic alphas. The contribution to the combination models' performance is assigned to be the return used in the RL process, driving the alpha generator to find better alphas that improve upon the current set. Experimental evaluations on real-world stock market data demonstrate both the effectiveness and the efficiency of our framework for stock trend forecasting. The investment simulation results show that our framework is able to achieve higher returns compared to previous approaches., Comment: Accepted by KDD '23, ADS track

In this paper, a modified version of conservative Physics-informed Neural Networks (cPINN for short) is provided to construct the weak solutions of Riemann problem for the hyperbolic scalar conservation laws in non-conservative form. To demonstrate the results, we use the model of generalized Buckley-Leverett equation (GBL equation for short) with discontinuous porosity in porous media. By inventing a new unknown, the GBL equation is transformed into a two-by-two resonant hyperbolic conservation laws in conservative form. The modified method of cPINN is invented to overcome the difficulties due to the discontinuity of the porosity and the appearance of the critical states (near vacuum) in the Riemann data. We experiment with our idea by using a deep learning algorithm to solve the GBL equation in both conservative and non-conservative forms, as well as the cases of critical and non-critical states. This method provides a combination of two different neural networks and corresponding loss functions, one is for the two-by-two resonant hyperbolic system, and the other is for the scalar conservation law with a discontinuous perturbation term in the non-convex flux. The technique of re-scaling to the unknowns is adopted to avoid the oscillation of the Riemann solutions in the cases of critical Riemann data. The solutions constructed by the modified cPINN match the exact solutions constructed by the theoretical analysis for hyperbolic conservation laws. In addition, the solutions are identical in both conservative and non-conservative cases. Finally, we compare the performance of the modified cPINN with numerical method called WENO5. Whereas WENO5 struggles with the highly oscillation of approximate solutions for the Riemann problems of GBL equation in non-conservative form, cPINN works admirably., Comment: 23 pages, 26 figures

The new variant of measurement-device-independent quantum key distribution (MDI-QKD), called asynchronous MDI-QKD or mode-pairing MDI-QKD, offers similar repeater-like rate-loss scaling but has the advantage of simple technology implementation by exploiting an innovative post-measurement pairing technique. We herein present an evaluation of the practical aspects of decoy-state asynchronous MDI-QKD. To determine its effectiveness, we analyze the optimal method of decoy-state calculation and examine the impact of asymmetrical channels and multi-user networks. Our simulations show that, under realistic conditions, aynchronous MDI-QKD can furnish the highest key rate with MDI security as compared to other QKD protocols over distances ranging from 50 km to 480 km. At fiber distances of 50 km and 100 km, the key rates attain 6.02 Mbps and 2.29 Mbps respectively, which are sufficient to facilitate real-time one-time-pad video encryption. Our findings indicate that experimental implementation of asynchronous MDI-QKD in intercity networks can be both practical and efficient., Comment: 15 pages, 4 figures

Quantum theory has shown its superiority in enhancing machine learning. However, facilitating quantum theory to enhance graph learning is in its infancy. This survey investigates the current advances in quantum graph learning (QGL) from three perspectives, i.e., underlying theories, methods, and prospects. We first look at QGL and discuss the mutualism of quantum theory and graph learning, the specificity of graph-structured data, and the bottleneck of graph learning, respectively. A new taxonomy of QGL is presented, i.e., quantum computing on graphs, quantum graph representation, and quantum circuits for graph neural networks. Pitfall traps are then highlighted and explained. This survey aims to provide a brief but insightful introduction to this emerging field, along with a detailed discussion of frontiers and outlook yet to be investigated.

Secure key rate (SKR) of point-point quantum key distribution (QKD) is fundamentally bounded by the rate-loss limit. Recent breakthrough of twin-field (TF) QKD can overcome this limit and enables long distance quantum communication, but its implementation necessitates complex global phase tracking and requires strong phase references which not only add to noise but also reduce the duty cycle for quantum transmission. Here, we resolve these shortcomings, and importantly achieve even higher SKRs than TF-QKD, via implementing an innovative but simpler measurement-device-independent QKD which realizes repeater-like communication through asynchronous coincidence pairing. Over 413 and 508 km optical fibers, we achieve finite-size SKRs of 590.61 and 42.64 bit/s, which are respectively 1.80 and 4.08 times of their corresponding absolute rate limits. Significantly, the SKR at 306 km exceeds 5 kbit/s and meets the bitrate requirement for live one-time-pad encryption of voice communication. Our work will bring forward economical and efficient intercity quantum-secure networks., Comment: 29 pages, 10 figures, 3 tables

In the framework of the factorization approach we calculate the branching fractions of 100 two-body nonleptonic decay channels in total, including 44 channels of the charm meson decays and 56 channels of the bottom meson decays. For charm meson decays, we test and confirm the previous observation that taking the limit for the number of colors $N\to\infty$ significantly improves theoretical predictions. For bottom meson decays, the penguin contributions are included in addition. As an essential input, we employ the weak decay form factors obtained in the framework of the relativistic quark model based on the quasi-potential approach. These form factors have well been tested by calculating observables in the semileptonic $D$ and $B$ meson decays and confronting obtained results with experimental data. In general, the predictions for the nonleptonic decay branching fractions are acceptable. However, for a quantitative calculation it is necessary to account for a more subtle effects of the final-state interaction., Comment: 24 pages, 5 tables, accepted for publication in Frontiers of Physics

Byzantine agreement, the underlying core of blockchain, aims to make every node in a decentralized network reach consensus. Classical Byzantine agreements unavoidably face two major problems. One is $1/3$ fault-tolerance bound, which means that the system to tolerate $f$ malicious players requires at least $3f+1$ players. The other is the security loopholes from its classical cryptography methods. Here, we propose a Byzantine agreement framework with unconditional security to break this bound with nearly $1/2$ fault tolerance due to multiparty correlation provided by quantum digital signatures. \textcolor{black}{It is intriguing that quantum entanglement is not necessary to break the $1/3$ fault-tolerance bound, and we show that weaker correlation, such as asymmetric relationship of quantum digital signature, can also work.} Our work strictly obeys two Byzantine conditions and can be extended to any number of players without requirements for multiparticle entanglement. We experimentally demonstrate three-party and five-party consensus for a digital ledger. Our work indicates the quantum advantage in terms of consensus problems and suggests an important avenue for quantum blockchain and quantum consensus networks., Comment: 21 pages, 7 figures. All comments are welcome!