arXiv subjects

Xiaofeng Meng

Publications and source records attributed to Xiaofeng Meng.

15 recordsLinked to original sources

arXiv2026

From Passive Response to Proactive Correction: Enhancing LLM Robustness Against Input Fact Perturbations

Large language models (LLMs) frequently produce confident yet factually incorrect responses when user inputs contain misleading premises, a phenomenon we attribute to fact perturbations in the input. Existing approaches to hallucination mitigation typically assume reliable user inputs, overlooking how such factual errors can actively mislead model reasoning. To address this vulnerability, we propose DEDUCE, a three-stage framework that transforms LLMs from passive responders into proactive error correctors. DEDUCE operates in three stages: (1) detect errors through fine-grained fact extraction and verification; (2) devise correction strategies via multi perspective deliberation; and (3) correct misconceptions while delivering reliable answers. We also present MisFactQA, a dataset containing factual errors of varying degrees, and propose new metrics for evaluating model robustness. Experiments on TruthfulQA, FalseQA, and our MisFactQA benchmark demonstrate that DEDUCE significantly improves both accuracy and error correction capability. Consistent gains across Qwen, LLaMA, and Gemma families confirm its effectiveness and scalability.

cs.CL↗

arXiv2026

From Time Series Analysis to Question Answering: A Survey in the LLM Era

Recently, Large Language Models (LLMs) have introduced a novel paradigm in Time Series Analysis (TSA), leveraging strong language capabilities to support tasks such as forecasting and anomaly detection. However, these analysis tasks cannot adequately cover temporal language tasks, such as interpretation and captioning. A fundamental gap remains between TSA and LLMs: LLMs are pre-trained to optimize natural language relevance for question answering rather than objectives specialized for TSA. To bridge this gap, TSA is evolving toward Time Series Question Answering (TSQA), shifting from expert-driven and task-specific analysis to user-driven and task-unified question answering. TSQA depends on flexible exploration rather than predefined TSA pipelines. In this survey, we first propose a taxonomy that reflects the evolution from TSA to TSQA, driven by a shift from external to internal alignment. We then organize existing literature into three alignment paradigms: Injective Alignment, Bridging Alignment, and Internal Alignment, and provide practical guidance for flexible, economical, and generalizable selection of alignment paradigms. We finally analyze datasets across domains and characteristics, identify challenges, and highlight future research directions.

cs.LG↗

arXiv2025

KV-Auditor: Auditing Local Differential Privacy for Correlated Key-Value Estimation

To protect privacy for data-collection-based services, local differential privacy (LDP) is widely adopted due to its rigorous theoretical bound on privacy loss. However, mistakes in complex theoretical analysis or subtle implementation errors may undermine its practical guarantee. To address this, auditing is crucial to confirm that LDP protocols truly protect user data. However, existing auditing methods, though, mainly target machine learning and federated learning tasks based on centralized differentially privacy (DP), with limited attention to LDP. Moreover, the few studies on LDP auditing focus solely on simple frequency estimation task for discrete data, leaving correlated key-value data - which requires both discrete frequency estimation for keys and continuous mean estimation for values - unexplored. To bridge this gap, we propose KV-Auditor, a framework for auditing LDP-based key-value estimation mechanisms by estimating their empirical privacy lower bounds. Rather than traditional LDP auditing methods that relies on binary output predictions, KV-Auditor estimates this lower bound by analyzing unbounded output distributions, supporting continuous data. Specifically, we classify state-of-the-art LDP key-value mechanisms into interactive and non-interactive types. For non-interactive mechanisms, we propose horizontal KV-Auditor for small domains with sufficient samples and vertical KV-Auditor for large domains with limited samples. For interactive mechanisms, we design a segmentation strategy to capture incremental privacy leakage across iterations. Finally, we perform extensive experiments to validate the effectiveness of our approach, offering insights for optimizing LDP-based key-value estimators.

cs.CR↗

arXiv2025

Intra-Query Runtime Elasticity for Cloud-Native Data Analysis

We propose the concept of Intra-Query Runtime Elasticity (IQRE) for cloud-native data analysis. IQRE enables a cloud-native OLAP engine to dynamically adjust a query's Degree of Parallelism (DOP) during execution. This capability allows users to utilize cloud computing resources more cost-effectively. We present Accordion, the first IQRE query engine. Accordion can adjust the parallelism of a query at any point during query execution without pausing data processing. It features a user-friendly interface and an auto-tuner backed by a "what-if" service to allow users to adjust the DOP according to their query latency constraints. The design of Accordion follows the execution model in Presto, an open-source distributed SQL query engine developed at Meta. We present the implementation of Accordion and demonstrate its ease of use, showcasing how it enables users to minimize compute resource consumption while meeting their query time constraints.

cs.DB↗

arXiv2024

Research on the Spatial Data Intelligent Foundation Model

This report focuses on spatial data intelligent large models, delving into the principles, methods, and cutting-edge applications of these models. It provides an in-depth discussion on the definition, development history, current status, and trends of spatial data intelligent large models, as well as the challenges they face. The report systematically elucidates the key technologies of spatial data intelligent large models and their applications in urban environments, aerospace remote sensing, geography, transportation, and other scenarios. Additionally, it summarizes the latest application cases of spatial data intelligent large models in themes such as urban development, multimodal systems, remote sensing, smart transportation, and resource environments. Finally, the report concludes with an overview and outlook on the development prospects of spatial data intelligent large models.

cs.AI↗

arXiv2024

PriPL-Tree: Accurate Range Query for Arbitrary Distribution under Local Differential Privacy

Answering range queries in the context of Local Differential Privacy (LDP) is a widely studied problem in Online Analytical Processing (OLAP). Existing LDP solutions all assume a uniform data distribution within each domain partition, which may not align with real-world scenarios where data distribution is varied, resulting in inaccurate estimates. To address this problem, we introduce PriPL-Tree, a novel data structure that combines hierarchical tree structures with piecewise linear (PL) functions to answer range queries for arbitrary distributions. PriPL-Tree precisely models the underlying data distribution with a few line segments, leading to more accurate results for range queries. Furthermore, we extend it to multi-dimensional cases with novel data-aware adaptive grids. These grids leverage the insights from marginal distributions obtained through PriPL-Trees to partition the grids adaptively, adapting the density of underlying distributions. Our extensive experiments on both real and synthetic datasets demonstrate the effectiveness and superiority of PriPL-Tree over state-of-the-art solutions in answering range queries across arbitrary data distributions.

cs.CR↗

arXiv2024

Cross-silo Federated Learning with Record-level Personalized Differential Privacy

Federated learning (FL) enhanced by differential privacy has emerged as a popular approach to better safeguard the privacy of client-side data by protecting clients' contributions during the training process. Existing solutions typically assume a uniform privacy budget for all records and provide one-size-fits-all solutions that may not be adequate to meet each record's privacy requirement. In this paper, we explore the uncharted territory of cross-silo FL with record-level personalized differential privacy. We devise a novel framework named \textit{rPDP-FL}, employing a two-stage hybrid sampling scheme with both uniform client-level sampling and non-uniform record-level sampling to accommodate varying privacy requirements. A critical and non-trivial problem is how to determine the ideal per-record sampling probability $q$ given the personalized privacy budget $\varepsilon$. We introduce a versatile solution named \textit{Simulation-CurveFitting}, allowing us to uncover a significant insight into the nonlinear correlation between $q$ and $\varepsilon$ and derive an elegant mathematical model to tackle the problem. Our evaluation demonstrates that our solution can provide significant performance gains over the baselines that do not consider personalized privacy preservation.

cs.CR↗

arXiv2024

From Chaos to Clarity: Time Series Anomaly Detection in Astronomical Observations

With the development of astronomical facilities, large-scale time series data observed by these facilities is being collected. Analyzing anomalies in these astronomical observations is crucial for uncovering potential celestial events and physical phenomena, thus advancing the scientific research process. However, existing time series anomaly detection methods fall short in tackling the unique characteristics of astronomical observations where each star is inherently independent but interfered by random concurrent noise, resulting in a high rate of false alarms. To overcome the challenges, we propose AERO, a novel two-stage framework tailored for unsupervised anomaly detection in astronomical observations. In the first stage, we employ a Transformer-based encoder-decoder architecture to learn the normal temporal patterns on each variate (i.e., star) in alignment with the characteristic of variate independence. In the second stage, we enhance the graph neural network with a window-wise graph structure learning to tackle the occurrence of concurrent noise characterized by spatial and temporal randomness. In this way, AERO is not only capable of distinguishing normal temporal patterns from potential anomalies but also effectively differentiating concurrent noise, thus decreasing the number of false alarms. We conducted extensive experiments on three synthetic datasets and three real-world datasets. The results demonstrate that AERO outperforms the compared baselines. Notably, compared to the state-of-the-art model, AERO improves the F1-score by up to 8.76% and 2.63% on synthetic and real-world datasets respectively.

cs.LG↗

arXiv2022

Overview and Introduction to Development of Non-Ergodic Earthquake Ground-Motion Models

This paper provides an overview and introduction to the development of non-ergodic ground-motion models, GMMs. It is intended for a reader who is familiar with the standard approach for developing ergodic GMMs. It starts with a brief summary of the development of ergodic GMMs and then describes different methods that are used in the development of non-ergodic GMMs with an emphasis on Gaussian Process (GP) regression, as that is currently the method preferred by most researchers contributing to this special issue. Non-ergodic modeling requires the definition of locations for the source and site characterizing the systematic source and site effects; the non-ergodic domain is divided into cells for describing the systematic path effects. Modeling the cell-specific anelastic attenuation as a GP and considerations on constraints for extrapolation of the non-ergodic GMMs are also discussed. An updated unifying notation for non-ergodic GMMs is also presented, which has been adopted by the authors of this issue.

stat.AP↗

arXiv2022

Lower Bound for the Simplicial Volume of Closed Manifolds Covered by $\mathbb{H}^{2}\times\mathbb{H}^{2}\times\mathbb{H}^{2}$

We estimate the upper bound for the $\ell^{\infty}$-norm of the volume form on $\mathbb{H}^2\times\mathbb{H}^2\times\mathbb{H}^2$ seen as a class in $H_{c}^{6}(\mathrm{PSL}_{2}\mathbb{R}\times\mathrm{PSL}_{2}\mathbb{R}\times\mathrm{PSL}_{2}\mathbb{R};\mathbb{R})$. This gives the lower bound for the simplicial volume of closed Riemennian manifolds covered by $\mathbb{H}^{2}\times\mathbb{H}^{2}\times\mathbb{H}^{2}$. The proof of these facts yields an algorithm to compute the lower bound of closed Riemannian manifolds covered by $\big(\mathbb{H}^2\big)^n$.

math.DG↗

arXiv2021

Simplicial volume of fiber bundles with nonpositively curved fibers

We prove the simplicial volume of the total space of a smooth fiber bundle with fiber being an oriented closed connected (occ) manifold of nonpositive curvature and negative Ricci curvature over an occ manifold with a closed universal covering is zero. Furthermore, if the fiber is an occ negatively curved manifold with dimension more than $2$, the simplicial volume of the total space is zero if and only if the simplicial volume of the base space is zero.

math.GT↗

arXiv2018

A Frequency Scaling based Performance Indicator Framework for Big Data Systems

It is important for big data systems to identify their performance bottleneck. However, the popular indicators such as resource utilizations, are often misleading and incomparable with each other. In this paper, a novel indicator framework which can directly compare the impact of different indicators with each other is proposed to identify and analyze the performance bottleneck efficiently. A methodology which can construct the indicator from the performance change with the CPU frequency scaling is described. Spark is used as an example of a big data system and two typical SQL benchmarks are used as the workloads to evaluate the proposed method. Experimental results show that the proposed method is accurate compared with the resource utilization method and easy to implement compared with some white-box method. Meanwhile, the analysis with our indicators lead to some interesting findings and valuable performance optimization suggestions for big data systems.

cs.DB↗

arXiv2018

Data Management in Time-Domain Astronomy: Requirements and Challenges

In time-domain astronomy, we need to use the relational database to manage star catalog data. With the development of sky survey technology, the size of star catalog data is larger, and the speed of data generation is faster. So, in this paper, we make a systematic and comprehensive introduction to process the data in time-domain astronomy, and valuable research questions are detailed. Then, we list candidate systems usually used in astronomy and point out the advantages and disadvantages of these systems. In addition, we present the key techniques needed to deal with astronomical data. Finally, we summarize the challenges faced by the design of our database prototype.

cs.DB↗

arXiv2018

AstroServ: Distributed Database for Serving Large-Scale Full Life-Cycle Astronomical Data

In time-domain astronomy, STLF (Short-Timescale and Large Field-of-view) sky survey is the latest way of sky observation. Compared to traditional sky survey who can only find astronomical phenomena, STLF sky survey can even reveal how short astronomical phenomena evolve. The difference does not only lead the new survey data but also the new analysis style. It requires that database behind STLF sky survey should support continuous analysis on data streaming, real-time analysis on short-term data and complex analysis on long-term historical data. In addition, both insertion and query latencies have strict requirements to ensure that scientific phenomena can be discovered. However, the existing databases cannot support our scenario. In this paper, we propose AstroServ, a distributed system for analysis and management of large-scale and full life-cycle astronomical data. AstroServ's core components include three data service layers and a query engine. Each data service layer serves for a specific time period of data and query engine can provide the uniform analysis interface on different data. In addition, we also provide many applications including interactive analysis interface and data mining tool to help scientists efficiently use data. The experimental results show that AstroServ can meet the strict performance requirements and the good recognition accuracy.

cs.DB↗

arXiv2018

Cloud based Real-Time and Low Latency Scientific Event Analysis

Astronomy is well recognized as big data driven science. As the novel observation infrastructures are developed, the sky survey cycles have been shortened from a few days to a few seconds, causing data processing pressure to shift from offline to online. However, existing scientific databases focus on offline analysis of long-term historical data, not real-time and low latency analysis of large-scale newly arriving data. In this paper, a cloud based method is proposed to efficiently analyze scientific events on large-scale newly arriving data. The solution is implemented as a highly efficient system, namely Aserv. A set of compact data store and index structures are proposed to describe the proposed scientific events and a typical analysis pattern is formulized as a set of query operations. Domain aware filter, accuracy aware data partition, highly efficient index and frequently used statistical data designs are four key methods to optimize the performance of Aserv. Experimental results under the typical cloud environment show that the presented optimization mechanism can meet the low latency demand for both large data insertion and scientific event analysis. Aserv can insert 3.5 million rows of data within 3 seconds and perform the heaviest query on 6.7 billion rows of data also within 3 seconds. Furthermore, a performance model is given to help Aserv choose the right cloud resource setup to meet the guaranteed real-time performance requirement.

cs.DB↗