arXiv subjects

Ying Qiao

Publications and source records attributed to Ying Qiao.

6 recordsLinked to original sources

arXiv2026

Semi-Implicit Pairwise Descent for Nonlocal Continuum Mechanics

We propose Semi-Implicit Pairwise Descent (SIPD), a unified nonlocal pairwise framework for simulating large-scale hyperelastic materials involving complex contact and friction. By reformulating the Finite Element Method (FEM) equations of motion into a pairwise force representation from a nonlocal perspective, our approach avoids costly Hessian computations, leading to a reduction in per-iteration computational overhead. Furthermore, we propose an analytical projection strategy for projecting our Hessian-free coefficient matrices to positive semi-definiteness. And we treat contact and friction as a unified anisotropic elastic energy, allowing for a seamless integration into the elastic solver framework. We mathematically prove that our method is unconditionally stable and numerically convergent.Experimental results demonstrate that SIPD achieves real-time performance for million-scale simulations even under intricate contact and friction conditions.

cs.GR↗

arXiv2025

Supervised Fine Tuning of Large Language Models for Domain Specific Knowledge Graph Construction:A Case Study on Hunan's Historical Celebrities

Large language models and knowledge graphs offer strong potential for advancing research on historical culture by supporting the extraction, analysis, and interpretation of cultural heritage. Using Hunan's modern historical celebrities shaped by Huxiang culture as a case study, pre-trained large models can help researchers efficiently extract key information, including biographical attributes, life events, and social relationships, from textual sources and construct structured knowledge graphs. However, systematic data resources for Hunan's historical celebrities remain limited, and general-purpose models often underperform in domain knowledge extraction and structured output generation in such low-resource settings. To address these issues, this study proposes a supervised fine-tuning approach for enhancing domain-specific information extraction. First, we design a fine-grained, schema-guided instruction template tailored to the Hunan historical celebrities domain and build an instruction-tuning dataset to mitigate the lack of domain-specific training corpora. Second, we apply parameter-efficient instruction fine-tuning to four publicly available large language models - Qwen2.5-7B, Qwen3-8B, DeepSeek-R1-Distill-Qwen-7B, and Llama-3.1-8B-Instruct - and develop evaluation criteria for assessing their extraction performance. Experimental results show that all models exhibit substantial performance gains after fine-tuning. Among them, Qwen3-8B achieves the strongest results, reaching a score of 89.3866 with 100 samples and 50 training iterations. This study provides new insights into fine-tuning vertical large language models for regional historical and cultural domains and highlights their potential for cost-effective applications in cultural heritage knowledge extraction and knowledge graph construction.

cs.CL↗

arXiv2024

A Framework for QoS of Integration Testing in Satellite Edge Clouds

The diversification of satellite communication services imposes varied requirements on network service quality, making quality of service (QoS) testing for microservices running on satellites more complex. Existing testing tools have limitations, potentially offering only single-functionality testing, thus failing to meet the requirements of QoS testing for edge cloud services in mobile satellite scenarios. In this paper, we propose a framework for integrating quality of service testing in satellite edge clouds. More precisely, the framework can integrate changes in satellite network topology, create and manage satellite edge cloud cluster testing environments on heterogeneous edge devices, customize experiments for users, support deployment and scaling of various integrated testing tools, and publish and visualize test results. Our experimental results validate the framework's ability to test key service quality metrics in a satellite edge cloud cluster.

cs.SE↗

arXiv2020

KRED: Knowledge-Aware Document Representation for News Recommendations

News articles usually contain knowledge entities such as celebrities or organizations. Important entities in articles carry key messages and help to understand the content in a more direct way. An industrial news recommender system contains various key applications, such as personalized recommendation, item-to-item recommendation, news category classification, news popularity prediction and local news detection. We find that incorporating knowledge entities for better document understanding benefits these applications consistently. However, existing document understanding models either represent news articles without considering knowledge entities (e.g., BERT) or rely on a specific type of text encoding model (e.g., DKN) so that the generalization ability and efficiency is compromised. In this paper, we propose KRED, which is a fast and effective model to enhance arbitrary document representation with a knowledge graph. KRED first enriches entities' embeddings by attentively aggregating information from their neighborhood in the knowledge graph. Then a context embedding layer is applied to annotate the dynamic context of different entities such as frequency, category and position. Finally, an information distillation layer aggregates the entity embeddings under the guidance of the original document representation and transforms the document vector into a new one. We advocate to optimize the model with a multi-task framework, so that different news recommendation applications can be united and useful information can be shared across different tasks. Experiments on a real-world Microsoft News dataset demonstrate that KRED greatly benefits a variety of news recommendation applications.

cs.IR↗

arXiv2020

XGLUE: A New Benchmark Dataset for Cross-lingual Pre-training, Understanding and Generation

In this paper, we introduce XGLUE, a new benchmark dataset that can be used to train large-scale cross-lingual pre-trained models using multilingual and bilingual corpora and evaluate their performance across a diverse set of cross-lingual tasks. Comparing to GLUE(Wang et al., 2019), which is labeled in English for natural language understanding tasks only, XGLUE has two main advantages: (1) it provides 11 diversified tasks that cover both natural language understanding and generation scenarios; (2) for each task, it provides labeled data in multiple languages. We extend a recent cross-lingual pre-trained model Unicoder(Huang et al., 2019) to cover both understanding and generation tasks, which is evaluated on XGLUE as a strong baseline. We also evaluate the base versions (12-layer) of Multilingual BERT, XLM and XLM-R for comparison.

cs.CL↗

arXiv2009

CliqueStream: an efficient and fault-resilient live streaming network on a clustered peer-to-peer overlay

Several overlay-based live multimedia streaming platforms have been proposed in the recent peer-to-peer streaming literature. In most of the cases, the overlay neighbors are chosen randomly for robustness of the overlay. However, this causes nodes that are distant in terms of proximity in the underlying physical network to become neighbors, and thus data travels unnecessary distances before reaching the destination. For efficiency of bulk data transmission like multimedia streaming, the overlay neighborhood should resemble the proximity in the underlying network. In this paper, we exploit the proximity and redundancy properties of a recently proposed clique-based clustered overlay network, named eQuus, to build efficient as well as robust overlays for multimedia stream dissemination. To combine the efficiency of content pushing over tree structured overlays and the robustness of data-driven mesh overlays, higher capacity stable nodes are organized in tree structure to carry the long haul traffic and less stable nodes with intermittent presence are organized in localized meshes. The overlay construction and fault-recovery procedures are explained in details. Simulation study demonstrates the good locality properties of the platform. The outage time and control overhead induced by the failure recovery mechanism are minimal as demonstrated by the analysis.

cs.NI↗