检索结果-内蒙古大学图书馆

Multi-pattern retrieval-augmented framework for text-to-sql with Poincaré-Skeleton retrieval and meta-instruction reasoning

引用

INFORMATION PROCESSING & MANAGEMENT 2025年第3期62卷

作者： Guo, Chunxi Tian, Zhiliang Tang, Jintao Li, Shasha Wang, Ting Natl Univ Def Technol Coll Comp 109 Deya Rd Changsha 410073 Hunan Peoples R China

text-to-sql transforms natural language text into sql queries, a task complicated by sql's complex syntax and the need for specialized knowledge. Retrieval from the sql-generated case repository assists large language models (LLMs) by providing relevant examples. However, complex queries often involve complicated sql syntax, which can confuse LLMs because they are designed for natural language rather than sql. In this paper, we propose a multi- pattern retrieval-augmented framework for sql generation, which dynamically selects relevant examples based on the query and reasoning patterns. To retrieve similar query patterns, we construct question skeletons in the Poincar & eacute;model, which better distinguishes entities aligned with a question's needs. To provide tailored examples of reasoning patterns for each logical step, especially for complex problems, we design meta-instruction-based retrieval repositories for multi-category chain-of-thought fragments. To mitigate biases from initial retrievals, we implement a revision strategy that leverages LLMs to interact with databases, enabling LLMs to self-correct errors during sql generation. Experiments on four benchmarks show that our method outperforms strong baseline models, which increases execution accuracy by 13% to 23.1% with the same LLM. Ablation studies provide insights into the framework's performance sensitivity to different components and strategies. Further analysis reveals a correlation between the framework's effectiveness and LLMs' quality and stability, and substantial performance gains from initial iteration modifications are the most significant.

关键词： Large language model text-to-sql Prompt learning Retrieval augmented

来源：评论

学校读者我要写书评

暂无评论

A Question-Aware Few-Shot text-to-sql Neural Model for Industrial Databases

引用

INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS 2025年第1期2025卷

作者： Li, Ren Chen, Yu Zhang, Hongyi Yang, Jianxi Xiao, Qiao Jiang, Shixin Chongqing Jiaotong Univ Sch Informat Sci & Engn Chongqing 400074 Peoples R China Chongqing Jiaotong Univ Sch Traff & Transportat Chongqing 400074 Peoples R China

Intelligent question answering over industrial databases is a challenging task due to the multicolumn context and complex questions. The existing methods need to be improved in terms of sql generation accuracy. In this paper, we propose a question-aware few-shot text-to-sql approach based on the SDCUP pretrained model. Specifically, an attention-based filtering approach is proposed to reduce the redundant information from multiple columns in the industrial database scenario. We further propose an operator semantics enhancement method to improve the ability of identifying complex conditions in queries. Experimental results on the industrial benchmarks in the fields of electric energy and structural inspection show that the proposed model outperforms the baseline models across all few-shot settings.

关键词： few-shot industrial databases question answering text-to-sql

来源：评论

学校读者我要写书评

暂无评论

Finetuning LLMs for text-to-sql with Two-Stage Progressive Learning 13th

Finetuning LLMs for Text-to-SQL with Two-Stage Progressive L...

引用

13th International Conference on Natural Language Processing and Chinese Computing

作者： Ling, Xiao Liu, Jialin Liu, Jindu Wu, Jianhua Liu, Jie Nankai Univ Coll Artificial Intelligence Engn Res Ctr Trusted Behav Intelligence Natl Key Lab Intelligent Tracking & Forecasting I Tianjin Peoples R China

ISBN: (纸本)9789819794331;9789819794348

With the widespread usage of large language model (LLMs), LLM-based method has become the mainstream approach for text-to-sql tasks, achieving leading performance on text-to-sql leaderboards. However, generating complex sql queries correctly has always been a main challenge. Current LLM-based models primarily utilize prompting-based methods on large scale closed-source LLMs (e.g., GPT-4 and ChatGPT), which may cause concerns of usage costs and data privacy. For fine-tuning based methods, it is difficult to generate complex sql accurately in only one fine-tuning step. Focusing on this, we propose TSPsql, a Two-Stage Progressive learning method for text-to-sql. TSPsql decomposes text-to-sql task into two stages: sql elements generation auxiliary task, and sql query generation main task. The two tasks are progressively fine-tuned on a single model, effectively reducing the difficulty of sql generation and improving accuracy. TSP-sql achieves state-of-the-art performance among open-source fine-tuning based methods on Spider dev set, and surpasses most of the methods based on large scale closed-source LLMs.

关键词： text-to-sql Large Languages Models Progressive Learning

来源：评论

学校读者我要写书评

暂无评论

LLM-Based text-to-sql for Real-World Databases

引用

SN Computer Science 2025年第2期6卷 1-13页

作者： Nascimento, Eduardo R. García, Grettel Izquierdo, Yenier T. Feijó, Lucas Coelho, Gustavo M. C. de Oliveira, Aiko R. Lemos, Melissa Garcia, Robinson L. S. Leme, Luiz A. P. Paes Casanova, Marco A. Instituto Tecgraf PUC-Rio RJ Rio de Janeiro 22451-900 Brazil Petrobras RJ Rio de Janeiro 20031-912 Brazil Depto. Informática PUC-Rio RJ Rio de Janeiro 22451-900 Brazil Instituto de Computação UFF RJ Niterói 24210-310 Brazil

text-to-sql refers to the task defined as “given a relational databaseDand a natural language sentenceSthat describes a question on D, generate an sql query QoverDthat expressesS”. Several LLM-based text-to-sql tools, that is, text-to-sql tools that explore Large Language Models (LLMs), emerged that outperformed previous approaches on well-known benchmarks. This article first shows that the performance of a selected set of LLM-based text-to-sql tools is, however, significantly less when run on two challenging databases with a large number of tables, columns, and foreign keys. A closer analysis reveals that one of the problems lie in that the relational schema is an inappropriate specification of the database from the point of view of the LLM. The article then introduces database specifications based on LLM-friendly views, that are close to the language of the users’ questions and that eliminate frequently used joins, and LLM-friendly data descriptions of the database values. The article proceeds to show that the use of a set of LLM-friendly views and data samples considerably improves the performance of a text-to-sql prompt strategy over a real-world database. This result suggests that real-world databases require rethinking how schema specifications should be passed to the LLM to recover state-of-the-art performance. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2025.

关键词： GPT Large language models Natural language Relational databases text-to-sql

来源：评论

学校读者我要写书评

暂无评论

Small, Medium, and Large Language Models for text-to-sql 43rd

Small, Medium, and Large Language Models for Text-to-SQL

引用

43rd International Conference on Conceptual Modeling

作者： Oliveira, Aiko Nascimento, Eduardo Pinheiro, Joao Avila, Caio Viktor S. Coelho, Gustavo Feijo, Lucas Izquierdo, Yenier Garcia, Grettel Paes Leme, Luiz Andre P. Lemos, Melissa Casanova, Marco A. Pontificia Univ Catolica Rio de Janeiro Tecgraf Inst BR-22451900 Rio de Janeiro RJ Brazil Pontificia Univ Catolica Rio de Janeiro Dept Informat BR-22451900 Rio de Janeiro RJ Brazil Univ Fed Fluminense Inst Comp BR-24210310 Niteroi RJ Brazil Univ Fed Ceara UFC BR-60440900 Fortaleza Ceara Brazil

ISBN: (纸本)9783031758713;9783031758720

This paper investigates how the model size affects the ability of a Generative AI Language Model, or briefly a GLM, to support the text-to-sql task for databases with large schemas typical of real-world applications. The paper first introduces a text-to-sql framework that combines a prompt strategy and a Retrieval-Augmented Generation (RAG) technique, leaving as flexibilization points the GLM and the database. Then, it describes a benchmark based on an open-source database featuring a schema much larger than the schemas of most of the databases in familiar text-to-sql benchmarks. The paper proceeds with experiments to assess the performance of the text-to-sql framework instantiated with the benchmark database and GLMs of different sizes. The paper concludes with recommendations to help select which GLM size is appropriate for a text-to-sql scenario, characterized by the difficulty of the expected NL questions and the data privacy requirements, among other characteristics.

关键词： text-to-sql Generative AI Language Model Retrieval-Augmented Generation Prompt engineering Real-World Databases

来源：评论

学校读者我要写书评

暂无评论

大语言模型时代text-to-sql更准确的评估指标

引用

电脑知识与技术 2025年第1期21卷 76-78,88页

作者：蒋鹏西南石油大学计算机科学学院四川成都610000

大型语言模型(LLM)已成为推进text-to-sql任务的强大工具。研究发现,基于LLM的模型在不同评估指标下,其性能表现与经过微调的模型存在显著差异。因此,文章分析了测试套件执行准确度(EXE)和精确集匹配准确度(ESM)在评估基于LLM的text-to-... 详细信息

大型语言模型(LLM)已成为推进text-to-sql任务的强大工具。研究发现,基于LLM的模型在不同评估指标下,其性能表现与经过微调的模型存在显著差异。因此,文章分析了测试套件执行准确度(EXE)和精确集匹配准确度(ESM)在评估基于LLM的text-to-sql模型时的不足,并提出了改进指标EESM(Enhanced Exact Set Matching)。实验结果表明,EXE和ESM分别存在高达13.2%和10.8%的假阳性和假阴性率,而EESM的假阳性率和假阴性率分别仅为0.2%和1.8%,表明EESM能够提供更准确的评估。

关键词： EESM 增强的精确集匹配准确度测试套件执行准确度精确集匹配准确度 text-to-sql

来源：评论

学校读者我要写书评

暂无评论

text-to-sql: A methodical review of challenges and models

引用

TURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCES 2024年第3期32卷 403-419页

作者： Kanburoglu, Ali Bugra Tek, F. Boray Isik Univ Dept Comp Engn Istanbul Turkiye Istanbul Tech Univ Dept Artificial Intelligence & Data Engn Istanbul Turkiye

This survey focuses on text -to -sql, automated translation of natural language queries into sql queries. Initially, we describe the problem and its main challenges. Then, by following the PRISMA systematic review methodology, we survey the existing text -to -sql review papers in the literature. We apply the same method to extract proposed text -to -sql models and classify them with respect to used evaluation metrics and benchmarks. We highlight the accuracies achieved by various models on text -to -sql datasets and discuss execution -guided evaluation strategies. We present insights into model training times and implementations of different models. We also explore the availability of text -to -sql datasets in non-English languages. Additionally, we focus on large language model (LLM) based approaches for the text -to -sql task, where we examine LLM-based studies in the literature and subsequently evaluate the LLMs on the cross -domain Spider dataset. Finally, we conclude with a discussion of future directions for text -to -sql research, identifying potential areas of improvement and advancements in this field.

关键词： text-to-sql large language model natural language processing deep learning

来源：评论

学校读者我要写书评

暂无评论

Fine-tuning text-to-sql models with reinforcement-learning training objectives

Natural Language Processing Journal

引用

Natural Language Processing Journal 2025年 10卷

作者： Xuan-Bang Nguyen Xuan-Hieu Phan Massimo Piccardi University of Engineering and Technology Vietnam National University Hanoi Viet Nam FPT Technology Research Institute FPT University Hanoi Viet Nam Faculty of Engineering and Information Technology University of Technology Sydney Broadway NSW 2007 Australia

text-to-sql is an important natural language processing task that helps users automatically convert natural language queries into formal sql code. While transformer-based models have pushed text-to-sql to unprecedented accuracy levels in recent years, such performance is confined to models of very large size that can only be run in specialised clouds. For this reason, in this paper we explore the use of reinforcement learning to improve the performance of models of more conservative size, which can fit within standard user hardware. As reinforcement learning reward, we propose a novel function which better aligns with the text-to-sql evaluation metrics, applied in conjunction with two strong policy gradient algorithms, REINFORCE and RELAX. Our experimental results over the popular Spider benchmark show that the proposed approach has been able to outperform a conventionally-trained T5 Small baseline by 6.6 pp (percentage points) of exact-set-match accuracy and 4.6 pp of execution accuracy, and a T5 Base baseline by 2.0 pp and 1.9 pp, respectively. The proposed model has also achieved a remarkable comparative performance against ChatGPT instances.

关键词： text-to-sql Reinforcement learning Reward functions Policy-gradient algorithms Fine-tuning

来源：评论

学校读者我要写书评

暂无评论

UniSAr: a unified structure-aware autoregressive language model for text-to-sql semantic parsing

引用

INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS 2023年第12期14卷 4361-4376页

作者： Dou, Longxu Gao, Yan Pan, Mingyang Wang, Dingzirui Che, Wanxiang Lou, Jian-Guang Zhan, Dechen Harbin Inst Technol Harbin 150001 Heilongjiang Peoples R China Microsoft Res Asia Beijing 100089 Peoples R China

Existing text-to-sql semantic parsers are typically designed for particular settings such as handling queries that span multiple tables, domains, or turns which makes them ineffective when applied to different settings. We present UniSAr (Unified Structure-Aware Autoregressive Language Model), which benefits from directly using an off-the-shelf language model architecture and demonstrates consistently high performance under different settings. Specifically, UniSAr extends existing autoregressive language models to incorporate two non-invasive extensions to make them structure-aware: (1) adding structure mark to encode database schema, conversation context, and their relationships;(2) constrained decoding to decode well-structured sql for a given database schema. On seven well-known text-to-sql datasets covering multi-domain, multi-table, and multi-turn, UniSAr demonstrates highly comparable or better performance to the most advanced specifically-designed text-to-sql models.

关键词： text-to-sql Semantic parsing Natural language interfaces to databases Natural language processing Constrained decoding

来源：评论

学校读者我要写书评

暂无评论

A survey on deep learning approaches for text-to-sql

引用

VLDB JOURNAL 2023年第4期32卷 905-936页

作者： Katsogiannis-Meimarakis, George Koutrika, Georgia Athena Res Ctr Athens Greece

To bridge the gap between users and data, numerous text-to-sql systems have been developed that allow users to pose natural language questions over relational databases. Recently, novel text-to-sql systems are adopting deep learning methods with very promising results. At the same time, several challenges remain open making this area an active and flourishing field of research and development. To make real progress in building text-to-sql systems, we need to de-mystify what has been done, understand how and when each approach can be used, and, finally, identify the research challenges ahead of us. The purpose of this survey is to present a detailed taxonomy of neural text-to-sql systems that will enable a deeper study of all the parts of such a system. This taxonomy will allow us to make a better comparison between different approaches, as well as highlight specific challenges in each step of the process, thus enabling researchers to better strategise their quest towards the "holy grail" of database accessibility.

关键词： text-to-sql Deep learning Natural language processing Natural language interface for databases

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：