Qiyu Liu

Qiyu Liu

Auto-generated profile, based on publicly available information

Qiyu Liu

Southwest University, Chongqing, China

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 January 2025 in The VLDB Journal
Verified
Journal Article
Open access
PDF file
QLQiyu Liu
Qiyu Liu
MLMaocheng Li
Maocheng Li
YZYuxiang Zeng
Yuxiang Zeng
YSYanyan Shen
Yanyan Shen
LCLei Chen
Lei Chen
Efficient indexing is fundamental to managing and analyzing multi-dimensional data. A growing trend is to directly learn the storage layout of multi-dimensional data using simple machine learning models, leading to the concept of Learned Index. Compared to conventional indexing methods that have been used for decades (e.g., kd-tree and R-tree variants), learned indexes have demonstrated empirical advantages in both space and time efficiency on modern architectures. However, there is a lack of comprehensive evaluation across existing multi-dimensional learned indexes under a standardized benchmark, making it challenging to identify the most suitable index for specific data types and query patterns. This gap also hinders the widespread adoption of learned indexes in practical applications. In this paper, we present the first in-depth empirical study to answer the question: how good are multi-dimensional learned indexes? We evaluate ten recently published indexes under a unified experimental framework, which includes standardized implementations, datasets, query workloads, and evaluation metrics. We thoroughly investigate the evaluation results and discuss the findings that may provide insights for future learned index design.

30 May 2024
Verified
Conference Paper
Open access
YYYunfeng Yu
Yunfeng Yu
LLLonglong Lin
Longlong Lin
QLQiyu Liu
Qiyu Liu
ZWZeli Wang
Zeli Wang
XOXi Ou
Xi Ou
Tao JiaTao Jia
Tao Jia
Graph Neural Networks (GNNs) have achieved remarkable performance in various applications, including social media analysis, computer vision, and natural language processing. Decoupled GNNs are a ubiquitous framework because of their high efficiency. However, existing decoupled GNNs suffer from the following several defects. (1) Their studies on GNN feature propagation are isolated, with each study emphasizing a user-specified propagation matrix. (2) They still have high computation costs to achieve provable performance on massive graphs with millions of nodes and billions of edges. (3) Their feature propagation steps are uniform, which makes it difficult for them to escape the dilemmas of over-smoothing. In this paper, we propose GSD-GNN, a Generalized and Scalable Decoupled GNN framework based on the spectral graph theory, which offers the following advantages. Firstly, through minor parameter adjustments, it can degenerate into most existing Decoupled GNNs, such as APPNP, GDC, SGC, etc. Secondly, it efficiently computes an arbitrary propagation matrix with near-linear time complexity and theoretical guarantees. Thirdly, it customizes the adaptive feature propagation mechanism for each node to mitigate the over-smoothing dilemma. Finally, extensive experiments on massive graphs demonstrate that the proposed GSD-GNN indeed is effective, scalable, and flexible.

31 January 2025 in The VLDB Journal
Verified
Journal Article
Open access
PDF file
JPJingshu Peng
Jingshu Peng
QLQiyu Liu
Qiyu Liu
ZCZhao Chen
Zhao Chen
YSYingxia Shao
Yingxia Shao
YSYanyan Shen
Yanyan Shen
JCJiannong Cao
Jiannong Cao
Graph neural networks (GNNs) have emerged due to their success at modeling graph data. Yet, it is challenging for GNNs to efficiently scale to large graphs. Thus, distributed GNNs come into play. To avoid communication caused by expensive data movement between workers, we propose Sancus and its advanced version Sancus"Equation missing", the staleness and quantization-aware communication-avoiding decentralized GNN system. By introducing a set of novel bounded embedding staleness metrics and adaptively skipping broadcasts, Sancus abstracts decentralized GNN processing as sequential matrix multiplication and uses historical embeddings via cache. To further mitigate the communication volume, Sancus"Equation missing" conducts quantization-aware communication on embeddings to reduce the size of broadcast messages. Theoretically, we show bounded approximation errors of embeddings and gradients with a known fastest convergence guarantee. Empirically, we evaluate Sancus and Sancus"Equation missing" with common GNN models via different system setups on large-scale benchmark datasets. Compared to SOTA works, Sancus"Equation missing" can avoid up to $$86\%$$ 86 % communication with $$3.0\times $$ 3.0 × faster throughput on average without accuracy loss.

14 December 2024 in arXiv
Verified
Preprint
Open access
PDF file
QLQiyu Liu
Qiyu Liu
SHSiyuan Han
Siyuan Han
JLJianwei Liao
Jianwei Liao
JLJin Li
Jin Li
JPJingshu Peng
Jingshu Peng
LCLei Chen
Lei Chen
Compressing integer keys is a fundamental operation among multiple communities, such as database management (DB), information retrieval (IR), and high-performance computing (HPC). Recent advances in \emph{learned indexes} have inspired the development of \emph{learned compressors}, which leverage simple yet compact machine learning (ML) models to compress large-scale sorted keys. The core idea behind learned compressors is to \emph{losslessly} encode sorted keys by approximating them with \emph{error-bounded} ML models (e.g., piecewise linear functions) and using a \emph{residual array} to guarantee accurate key reconstruction. While the concept of learned compressors remains in its early stages of exploration, our benchmark results demonstrate that an SIMD-optimized learned compressor can significantly outperform state-of-the-art CPU-based compressors. Drawing on our preliminary experiments, this vision paper explores the potential of learned data compression to enhance critical areas in DBMS and related domains. Furthermore, we outline the key technical challenges that existing systems must address when integrating this emerging methodology.

9 May 2024 in arXiv
Verified
Preprint
Open access
PDF file
QLQiyu Liu
Qiyu Liu
MLMaocheng Li
Maocheng Li
YZYuxiang Zeng
Yuxiang Zeng
YSYanyan Shen
Yanyan Shen
LCLei Chen
Lei Chen
Efficient indexing is fundamental for multi-dimensional data management and analytics. An emerging tendency is to directly learn the storage layout of multi-dimensional data by simple machine learning models, yielding the concept of Learned Index. Compared with the conventional indices used for decades (e.g., kd-tree and R-tree variants), learned indices are empirically shown to be both space- and time-efficient on modern architectures. However, there lacks a comprehensive evaluation of existing multi-dimensional learned indices under a unified benchmark, which makes it difficult to decide the suitable index for specific data and queries and further prevents the deployment of learned indices in real application scenarios. In this paper, we present the first in-depth empirical study to answer the question of how good multi-dimensional learned indices are. Six recently published indices are evaluated under a unified experimental configuration including index implementation, datasets, query workloads, and evaluation metrics. We thoroughly investigate the evaluation results and discuss the findings that may provide insights for future learned index design.

15 December 2025 in ACM Transactions on Architecture and Code Optimization
Verified
Journal Article
Open access
SZShiyu Zhong
Shiyu Zhong
JWJiaojiao Wu
Jiaojiao Wu
XGXinyu Guo
Xinyu Guo
FYFan Yang
Fan Yang
AYAobo Yang
Aobo Yang
JLJianwei Liao
Jianwei Liao
Solid State Disks (SSDs) provide high degrees of channel-level parallelism, and sub-requests of a large request are forwarded to different flash channels to maximize parallelism and utilization. The scheduler, however, may experience differential wait times since their individual queues of channels are not coordinated and keep balanced workloads at all times. Since a large request is regarded as a complete only when its last sub-request is completed, some of its sub-requests that are completed earlier inevitably wait for this last one that is termed as intra-request lag , thus affecting I/O responsiveness. This article proposes intra-request lag-aware cache management to maximize channel-level read parallelism, thus improving user experience, as well as balance I/O workloads over all SSD channels. Once a part of sub-request(s) belonging to the same large read requests are blocked on the congested channel(s), indicating intra-request lag(s) appear, we preferentially load the lag-corresponding data, instead of all relevant data of read request, into the data cache. Consequently, it can respond quickly and balance the workload over channels if the read is hit in the near future. Besides, in order to specifically buffer the data of small read requests or the data of intra-request lags of large read request to improve read responsiveness, we present a cache partition model for splitting the data cache into the read cache and the write cache. Trace-driven experiments show that our proposal can reduce the long-tail latency of I/O requests by 26.2 % at the 99.9-th percentile, and the overall I/O time by 35.0 % on average, in contrast to existing approaches.