arXiv subjects

Guoliang Li

Publications and source records attributed to Guoliang Li.

At least 127 records · Page 7Linked to original sources

arXiv2023

Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs

Text-to-SQL parsing, which aims at converting natural language instructions into executable SQLs, has gained increasing attention in recent years. In particular, Codex and ChatGPT have shown impressive results in this task. However, most of the prevalent benchmarks, i.e., Spider, and WikiSQL, focus on database schema with few rows of database contents leaving the gap between academic study and real-world applications. To mitigate this gap, we present Bird, a big benchmark for large-scale database grounded in text-to-SQL tasks, containing 12,751 pairs of text-to-SQL data and 95 databases with a total size of 33.4 GB, spanning 37 professional domains. Our emphasis on database values highlights the new challenges of dirty database contents, external knowledge between NL questions and database contents, and SQL efficiency, particularly in the context of massive databases. To solve these problems, text-to-SQL models must feature database value comprehension in addition to semantic parsing. The experimental results demonstrate the significance of database values in generating accurate text-to-SQLs for big databases. Furthermore, even the most effective text-to-SQL models, i.e. ChatGPT, only achieves 40.08% in execution accuracy, which is still far from the human result of 92.96%, proving that challenges still stand. Besides, we also provide an efficiency analysis to offer insights into generating text-to-efficient-SQLs that are beneficial to industries. We believe that BIRD will contribute to advancing real-world applications of text-to-SQL research. The leaderboard and source code are available: https://bird-bench.github.io/.

cs.CL↗

arXiv2023

Towards Transaction as a Service

This paper argues for decoupling transaction processing from existing two-layer cloud-native databases and making transaction processing as an independent service. By building a transaction as a service (TaaS) layer, the transaction processing can be independently scaled for high resource utilization and can be independently upgraded for development agility. Accordingly, we architect an execution-transaction-storage three-layer cloud-native database. By connecting to TaaS, 1) the AP engines can be empowered with ACID TP capability, 2) multiple standalone TP engine instances can be incorporated to support multi-master distributed TP for horizontal scalability, 3) multiple execution engines with different data models can be integrated to support multi-model transactions, and 4) high performance TP is achieved through extensive TaaS optimizations and consistent evolution. Cloud-native databases deserve better architecture: we believe that TaaS provides a path forward to better cloud-native databases.

cs.DB↗

arXiv2023

Constraints on the ghost-free parity-violating gravity from Laser-ranged Satellites

This paper explores the evolutionary behavior of the Earth-satellite binary system within the framework of the ghost-free parity-violating gravity and the corresponding discussion on the parity-violating effect from the laser-ranged satellites. For this purpose, we start our study with the Parameterized Post-Newtonian (PPN) metric of this gravity theory to study the orbital evolution of the satellites in which the spatial-time sector of the spacetime is modified due to the parity violation. With this modified PPN metric, we calculate the effects of the parity-violating sector of metrics on the time evolution of the orbital elements for an Earth-satellite binary system. We find that among the five orbital elements, the parity violation has no effect on the semi-latus rectum, while the eccentricity and ascending node are affected only in a periodic manner. These three orbital elements are the same as the results of general relativity and are also consistent with the observations of the present experiment. In particular, parity violation produces non-zero corrections to the eccentricity and pericenter, which will accumulate with the evolution of time, indicating that the parity violation of gravity produces observable secular effects. The observational constraint on the parity-violating effect is derived by confronting the theoretical prediction with the observation by the LAGEOS II pericenter advance, giving a constraint on the parity-violating parameter space from the satellite experiments.

gr-qc↗

arXiv2023

SkyMath: Technical Report

Large language models (LLMs) have shown great potential to solve varieties of natural language processing (NLP) tasks, including mathematical reasoning. In this work, we present SkyMath, a large language model for mathematics with 13 billion parameters. By applying self-compare fine-tuning, we have enhanced mathematical reasoning abilities of Skywork-13B-Base remarkably. On GSM8K, SkyMath outperforms all known open-source models of similar size and has established a new SOTA performance.

cs.CL↗

arXiv2023

Survey of Vector Database Management Systems

There are now over 20 commercial vector database management systems (VDBMSs), all produced within the past five years. But embedding-based retrieval has been studied for over ten years, and similarity search a staggering half century and more. Driving this shift from algorithms to systems are new data intensive applications, notably large language models, that demand vast stores of unstructured data coupled with reliable, secure, fast, and scalable query processing capability. A variety of new data management techniques now exist for addressing these needs, however there is no comprehensive survey to thoroughly review these techniques and systems. We start by identifying five main obstacles to vector data management, namely vagueness of semantic similarity, large size of vectors, high cost of similarity comparison, lack of natural partitioning that can be used for indexing, and difficulty of efficiently answering hybrid queries that require both attributes and vectors. Overcoming these obstacles has led to new approaches to query processing, storage and indexing, and query optimization and execution. For query processing, a variety of similarity scores and query types are now well understood; for storage and indexing, techniques include vector compression, namely quantization, and partitioning based on randomization, learning partitioning, and navigable partitioning; for query optimization and execution, we describe new operators for hybrid queries, as well as techniques for plan enumeration, plan selection, and hardware accelerated execution. These techniques lead to a variety of VDBMSs across a spectrum of design and runtime characteristics, including native systems specialized for vectors and extended systems that incorporate vector capabilities into existing systems. We then discuss benchmarks, and finally we outline research challenges and point the direction for future work.

cs.DB↗

arXiv2023

Sciences with the 2.5-meter Wide Field Survey Telescope (WFST)

The Wide Field Survey Telescope (WFST) is a dedicated photometric surveying facility being built jointly by the University of Science and Technology of China and the Purple Mountain Observatory. It is equipped with a 2.5-meter diameter primary mirror, an active optics system, and a mosaic CCD camera with 0.73 gigapixels on the primary focal plane for high-quality image capture over an FOV of 6.5-square-degree. It is anticipated that WFST will be set up at the Lenghu site in the summer of 2023 and begin to observe the northern sky in four optical bands (u, g, r, and i) with a range of cadences, from hourly/daily in the Deep High-Cadence Survey (DHS) program to semiweekly in the Wide-Field Survey (WFS) program, three months later. During a photometric night, a nominal 30 s exposure in the WFS program will reach a depth of 22.27, 23.32, 22.84, and 22.31 (AB magnitudes) in these four bands, respectively, allowing for the detection of a tremendous amount of transients in the low-z universe and a systematic investigation of the variability of Galactic and extragalactic objects. In the DHS program, intranight 90 s exposures as deep as 23 (u) and 24 mag (g), in combination with target of opportunity follow-ups, will provide a unique opportunity to explore energetic transients in demand for high sensitivities, including the electromagnetic counterparts of gravitational wave events, supernovae within a few hours of their explosions, tidal disruption events and fast, luminous optical transients even beyond a redshift of unity. In addition, the final 6-year co-added images, anticipated to reach g=25.8 mag in WFS or 1.5 mags deeper in DHS, will be of fundamental importance to general Galactic and extragalactic science. The highly uniform legacy surveys of WFST will serve as an indispensable complement to those of LSST that monitor the southern sky.

astro-ph.IM↗

arXiv2023

HybPSF: Hybrid PSF reconstruction for the observed JWST NIRCam image

The James Webb Space Telescope (JWST) ushers in a new era of astronomical observation and discovery, offering unprecedented precision in a variety of measurements such as photometry, astrometry, morphology, and shear measurement. Accurate point spread function (PSF) models are crucial for many of these measurements. In this paper, we introduce a hybrid PSF construction method called HybPSF for JWST NIRCam imaging data. HybPSF combines the WebbPSF software, which simulates the PSF for JWST, with observed data to produce more accurate and reliable PSF models. We apply this method to the SMACS J0723 imaging data and construct supplementary structures from residuals obtained by subtracting the WebbPSF PSF model from the data. Our results show that HybPSF significantly reduces discrepancies between the PSF model and the data compared to WebbPSF. Specifically, the PSF shape parameter ellipticity and size comparisons indicate that HybPSF improves precision by a factor of approximately 10 for \$R^2\$ and \$50\%\$ for \$e\$. This improvement has important implications for astronomical measurements using JWST NIRCam imaging data.

astro-ph.IM↗

arXiv2023

Theoretical Analysis of Random Scattering Induced by Microlensing

Theoretical investigations into the deflection angle caused by microlenses offer a direct path to uncovering principles of the cosmological microlensing effect. This work specifically concentrates on the the probability density function (PDF) of the light deflection angle induced by microlenses. We have made several significant improvements to the widely used formula from Katz et al. First, we update the coefficient from 3.05 to 1.454, resulting in a better fit between the theoretical PDF and our simulation results. Second, we developed an elegant fitting formula for the PDF that can replace its integral representation within a certain accuracy, which is numerically divergent unless arbitrary upper limits are chosen. Third, to facilitate further theoretical work in this area, we have identified a more suitable Gaussian approximation for the fitting formula.

astro-ph.GA↗

arXiv2023

LLM As DBA

Database administrators (DBAs) play a crucial role in managing, maintaining and optimizing a database system to ensure data availability, performance, and reliability. However, it is hard and tedious for DBAs to manage a large number of database instances (e.g., millions of instances on the cloud databases). Recently large language models (LLMs) have shown great potential to understand valuable documents and accordingly generate reasonable answers. Thus, we propose D-Bot, a LLM-based database administrator that can continuously acquire database maintenance experience from textual sources, and provide reasonable, well-founded, in-time diagnosis and optimization advice for target databases. This paper presents a revolutionary LLM-centric framework for database maintenance, including (i) database maintenance knowledge detection from documents and tools, (ii) tree of thought reasoning for root cause analysis, and (iii) collaborative diagnosis among multiple LLMs. Our preliminary experimental results that D-Bot can efficiently and effectively diagnose the root causes and our code is available at github.com/TsinghuaDatabaseGroup/DB-GPT.

cs.DB↗

arXiv2023

From Large Language Models to Databases and Back: A discussion on research and education

This discussion was conducted at a recent panel at the 28th International Conference on Database Systems for Advanced Applications (DASFAA 2023), held April 17-20, 2023 in Tianjin, China. The title of the panel was "What does LLM (ChatGPT) Bring to Data Science Research and Education? Pros and Cons". It was moderated by Lei Chen and Xiaochun Yang. The discussion raised several questions on how large language models (LLMs) and database research and education can help each other and the potential risks of LLMs.

cs.DB↗

arXiv2023

Microlensing bias on the detection of strong lensing gravitational wave

Identifying strong lensing gravitational wave (SLGW) events is of utmost importance in astrophysics as we approach the historic first detection of SLGW amidst the growing number of gravitational wave (GW) events. Currently, one crucial method for identifying SLGW signals involves assessing the overlap of parameters between two GWs. However, the distribution of discrete matter, such as stars and sub-halos, within the strong lensing galaxy can imprint a wave optical (WO) effect on the SLGW waveform. These frequency dependent imprints introduce biases in parameter estimation and impact SLGW identification. In this study, we assess the influence of the stellar microlensing field embedded in a strong lensing galaxy. Our finding demonstrate that the WO effect reduces the detection efficiency of SLGW by $5\%\sim 50\%$ for various false alarm probabilities per pair (${\rm FAP}_{\rm per~pair}$). Specifically, at an ${\rm FAP}_{\rm per~pair}$ of $10^{-5}$, the detection efficiency decreases from $\sim 10\%$ to $\sim 5\%$. Consequently, the presence of the microlensing field can result in missing half of the strong lensing candidates. Additionally, the microlensing WO effect introduces a noticeable bias in intrinsic parameters, particularly for chirp mass and mass ratio. However, it has tiny influence on extrinsic parameters. Considering all parameters, $\sim 30\%$ of events exhibit a $1σ$ parameter bias, $\sim 12\%$ exhibit a $2σ$ parameter bias, and $\sim 5\%$ exhibit a $3σ$ parameter bias.

astro-ph.CO↗

arXiv2023

DoF Upper Bound on Three-User MISO Broadcast Channel with Two Antennas and Mixed CSIT

With delayed and imperfect current channel state information at the transmitter (CSIT), namely mixed CSIT, the sum degrees-of-freedom (sum-DoF) in the two-user multiple-input multiple-output (MIMO) broadcast channel and the $K$-user multiple-input single-output (MISO) broadcast channel with not-less-than-$K$ transmit antennas have been obtained. However, the case of the three-user broadcast channel with two transmit antennas and mixed CSIT is still unexplored. In this paper, we investigate the sum-DoF upper bound of three-user MISO broadcast channel with two transmit antennas and mixed CSIT. By exploiting genie-aided signaling and extremal inequalities, we derive the sum-DoF upper bound as $(1-α)3/2 + 9α/4$, which is at most $12.5\%$ larger than the expected sum-DoF, given by $(1-α)3/2 + 2α$. This indicates that the gap may mitigate by better bounding the imperfect current CSIT counterpart.

cs.IT↗

arXiv2023

Interleaving Pre-Trained Language Models and Large Language Models for Zero-Shot NL2SQL Generation

Zero-shot NL2SQL is crucial in achieving natural language to SQL that is adaptive to new environments (e.g., new databases, new linguistic phenomena or SQL structures) with zero annotated NL2SQL samples from such environments. Existing approaches either fine-tune pre-trained language models (PLMs) based on annotated data or use prompts to guide fixed large language models (LLMs) such as ChatGPT. PLMs can perform well in schema alignment but struggle to achieve complex reasoning, while LLMs is superior in complex reasoning tasks but cannot achieve precise schema alignment. In this paper, we propose a ZeroNL2SQL framework that combines the complementary advantages of PLMs and LLMs for supporting zero-shot NL2SQL. ZeroNL2SQL first uses PLMs to generate an SQL sketch via schema alignment, then uses LLMs to fill the missing information via complex reasoning. Moreover, in order to better align the generated SQL queries with values in the given database instances, we design a predicate calibration method to guide the LLM in completing the SQL sketches based on the database instances and select the optimal SQL query via an execution-based strategy. Comprehensive experiments show that ZeroNL2SQL can achieve the best zero-shot NL2SQL performance on real-world benchmarks. Specifically, ZeroNL2SQL outperforms the state-of-the-art PLM-based methods by 3.2% to 13% and exceeds LLM-based methods by 10% to 20% on execution accuracy.

cs.CL↗

arXiv2023

RITA: Group Attention is All You Need for Timeseries Analytics

Timeseries analytics is of great importance in many real-world applications. Recently, the Transformer model, popular in natural language processing, has been leveraged to learn high quality feature embeddings from timeseries, core to the performance of various timeseries analytics tasks. However, the quadratic time and space complexities limit Transformers' scalability, especially for long timeseries. To address these issues, we develop a timeseries analytics tool, RITA, which uses a novel attention mechanism, named group attention, to address this scalability issue. Group attention dynamically clusters the objects based on their similarity into a small number of groups and approximately computes the attention at the coarse group granularity. It thus significantly reduces the time and space complexity, yet provides a theoretical guarantee on the quality of the computed attention. The dynamic scheduler of RITA continuously adapts the number of groups and the batch size in the training process, ensuring group attention always uses the fewest groups needed to meet the approximation quality requirement. Extensive experiments on various timeseries datasets and analytics tasks demonstrate that RITA outperforms the state-of-the-art in accuracy and is significantly faster -- with speedups of up to 63X.

cs.LG↗

arXiv2023

iCOIL: Scenario Aware Autonomous Parking Via Integrated Constrained Optimization and Imitation Learning

Autonomous parking (AP) is an emering technique to navigate an intelligent vehicle to a parking space without any human intervention. Existing AP methods based on mathematical optimization or machine learning may lead to potential failures due to either excessive execution time or lack of generalization. To fill this gap, this paper proposes an integrated constrained optimization and imitation learning (iCOIL) approach to achieve efficient and reliable AP. The iCOIL method has two candidate working modes, i.e., CO and IL, and adopts a hybrid scenario analysis (HSA) model to determine the better mode under various scenarios. We implement and verify iCOIL on the Macao Car Racing Metaverse (MoCAM) platform. Results show that iCOIL properly adapts to different scenarios during the entire AP procedure, and achieves significantly larger success rates than other benchmarks.

cs.RO↗

arXiv2023

Local Group Dwarf Galaxy Detection Limit in the CSST survey

We predict the dwarf galaxy detection limits for the upcoming Chinese Space Station Telescope (CSST) survey that will cover 17,500 deg$^{2}$ of the sky with a wide field of view of 1.1 deg$^2$. The point-source depth reaches 26.3 mag in the $g$ band and 25.9 mag in the $i$ band. Constructing mock survey data based on the designed photometric bands, we estimate the recovery rate of artificial dwarf galaxies from mock point-source photometric catalogues. The detection of these artificial dwarf galaxies is strongly dependent on their distance, magnitude and size, in agreement with searches in current surveys. We expect CSST to enable the detection of dwarf galaxies with $M_V = -3.0$ and $μ_{250} = 32.0$ mag/arcsec$^2$ (surface-brightness limit for a system of half-light radius $r_{\rm h}$ = 250 pc at 400 kpc, and $M_V = -4.9$ and $μ_{250} = 30.5$ mag/arcsec$^2$ around the Andromeda galaxy. Beyond the Local Group, the CSST survey will achieve $M_V = -5.8$, and $μ_{250}$ = 29.7 mag/arcsec$^2$ in the distance range of 1--2 Mpc, opening up an exciting discovery space for faint field dwarf galaxies. With its optical bands, wide survey footprint, and space resolution, CSST will undoubtedly expand our knowledge of low-mass dwarf galaxies to an unprecedented volume.

astro-ph.GA↗

arXiv2023

GeoGauss: Strongly Consistent and Light-Coordinated OLTP for Geo-Replicated SQL Database

Multinational enterprises conduct global business that has a demand for geo-distributed transactional databases. Existing state-of-the-art databases adopt a sharded master-follower replication architecture. However, the single-master serving mode incurs massive cross-region writes from clients, and the sharded architecture requires multiple round-trip acknowledgments (e.g., 2PC) to ensure atomicity for cross-shard transactions. These limitations drive us to seek yet another design choice. In this paper, we propose a strongly consistent OLTP database GeoGauss with full replica multi-master architecture. To efficiently merge the updates from different master nodes, we propose a multi-master OCC that unifies data replication and concurrent transaction processing. By leveraging an epoch-based delta state merge rule and the optimistic asynchronous execution, GeoGauss ensures strong consistency with light-coordinated protocol and allows more concurrency with weak isolation, which are sufficient to meet our needs. Our geo-distributed experimental results show that GeoGauss achieves 7.06X higher throughput and 17.41X lower latency than the state-of-the-art geo-distributed database CockroachDB on the TPC-C benchmark.

cs.DB↗

arXiv2023

The spherical Fast Multipole Method (sFMM) for Gravitational Lensing Simulation

In this paper, we present a spherical Fast Multipole Method (sFMM) for ray tracing simulation of gravitational lensing (GL) on a curved sky. The sFMM is a non-trivial extension of the Fast Multiple Method (FMM) to sphere $\mathbb S^2$, and it can accurately solve the Poisson equation with time complexity of $O(N)\log(N)$, where $N$ is the number of particles. It is found that the time complexity of the sFMM is near $O(N)$ and the computational accuracy can reach $10^{-10}$ in our test. In addition, compared with the Fast Spherical Harmonic Transform (FSHT), the sFMM is not only faster but more accurate, as it has the ability to reserve high-frequency components of the density field. These merits make the sFMM an optimum method to simulate the gravitational lensing on a curved sky, which is the case for upcoming large-area sky surveys, such as the Vera Rubin Observatory and the China Space Station Telescope.

astro-ph.IM↗