检索结果-内蒙古大学图书馆

计算机应用 2024年第9期44卷 2689-2695页

作者：吴相岚肖洋刘梦莹刘明铭南开大学软件学院天津300457

为优化基于异构图编码器的text-to-sql生成效果,提出SELsql模型。首先,模型采用端到端的学习框架,使用双曲空间下的庞加莱距离度量替代欧氏距离度量,以此优化使用探针技术从预训练语言模型中构建的语义增强的模式链接图;其次,利用K头加... 详细信息

为优化基于异构图编码器的text-to-sql生成效果,提出SELsql模型。首先,模型采用端到端的学习框架,使用双曲空间下的庞加莱距离度量替代欧氏距离度量,以此优化使用探针技术从预训练语言模型中构建的语义增强的模式链接图;其次,利用K头加权的余弦相似度以及图正则化方法学习相似度度量图使得初始模式链接图在训练中迭代优化;最后,使用改良的关系图注意力网络(RGAT)图编码器以及多头注意力机制对两个模块的联合语义模式链接图进行编码,并且使用基于语法的神经语义解码器和预定义的结构化语言进行结构化查询语言(sql)语句解码。在Spider数据集上的实验结果表明,使用ELECTRA-large预训练模型时,SELsql模型比最佳基线模型的准确率提升了2.5个百分点,对于复杂sql语句生成的提升效果很大。

关键词：模式链接图结构学习预训练语言模型 text-to-sql 异构图

来源：评论

学校读者我要写书评

暂无评论

text-to-sql转换关键技术研究

Text-to-SQL转换关键技术研究

引用

作者：余伟国防科技大学

学位级别：博士

语义解析,旨在将非结构化的自然语言表达转换为特定的结构化表示,是自然语言处理和人工智能领域内挑战难度较大的任务之一。text-to-sql转换是语义解析的一个子领域,在近年来备受学术界和工业界关注。text-to-sql转换的目标是根据已知... 详细信息

语义解析,旨在将非结构化的自然语言表达转换为特定的结构化表示,是自然语言处理和人工智能领域内挑战难度较大的任务之一。text-to-sql转换是语义解析的一个子领域,在近年来备受学术界和工业界关注。text-to-sql转换的目标是根据已知的关系数据库或表格,将给定的自然语言文本转换为对应的sql查询。该任务需综合多领域的知识,解决其特有的领域需求、复杂的转换语境、多样化的上下文约束等方面带来的诸多挑战。尽管任务复杂,但text-to-sql及其相关技术无论在学术研究上还是实际应用中都具有重大且深远的意义。它不仅能为复杂的应用程序提供数据支撑,还能为用户、应用程序和数据库三方搭建自然语言的桥梁。近年来,随着大量text-to-sql转换数据集的发布以及深度学习技术的飞速发展,text-to-sql转换相关研究也取得了长足的进步。本文立足text-to-sql转换任务,针对能不能转、怎么转和转错了怎么办等问题所面临的挑战,从可回答性分类、单轮text-to-sql转换、多轮text-to-sql转换及text-to-sql错误校正等方面开展若干技术研究。本文的主要贡献总结如下:第一,针对text-to-sql可回答性分类所面临的结构化信息丢失及独立编码间的信息交流问题,本文构建了问题—数据库模式图结构和提出了条件层归一化方法。本文提出的模型对sql数据的结构信息进行了良好的建模,同时通过条件层归一化调制的预训练模型编码更适合text-to-sql可回答性分类任务。在Triage sql数据集上的实验结果表明,我们模型的性能显著优于其它基线方法,在召回率、准确率和F1值上比现有Sota分别高出6.04%、7.88%和8.35%。第二,针对单轮text-to-sql转换所面临的挑战,我们探索性地利用案例推理技术进行text-to-sql转换,并发现在转换过程中存在相似度度量偏差问题。针对所存在的问题,本文提出一种混合的问题相似度度量方法。该方法综合地利用文本余弦相似度和分类概率来进行案例相似度度量。在单轮text-to-sql转换基准数据集Wiki sql上测试,我们的模型比现有的Sota提升了1.2%的聚合函数预测准确率。第三,针对多轮text-to-sql转换所面临的多源编码困难,以及缺乏建模多轮上下文的方法,本文提出了一种异构图融合的交互建模机制。该机制将多轮text-to-sql转换任务的所有数据均转换为图,并利用异构图聚合将多轮交互聚合为统一的表示。在Spar C数据集上的实验表明,我们的模型比现有基线高出0.8%的问题准确度。同时,在Co sql数据集上的实验显示,我们的模型比现有基线高出0.8%的问题准确度和0.4%的交互准确度。第四,针对text-to-sql错误校正所面临的上下文-sql建模、编辑节点选择、中间编辑过程利用等问题,本文提出一种迭代图编辑模型以充分利用sql的结构化信息,并且对迭代过程中的临时编辑状态进行了合理利用。在SPLASH数据集上的实验表明,我们模型的性能显著优于其它基线方法,在校正准确率、编辑距离下降率、编辑距离增长率和完成度上比现有的最优模型分别高出4.87%、7.92%、4.24%和7.59%。

关键词：语义解析 text-to-sql 可回答分类案例推理交互建模 sql校正图编辑

来源：评论

学校读者我要写书评

暂无评论

text-to-sql中的解码方法研究

Text-to-SQL中的解码方法研究

引用

作者：潘名扬哈尔滨工业大学

学位级别：硕士

text-to-sql是自然语言处理领域的一个重要任务,该任务将自然语言描述或问题转换为对一个具体的数据库的sql查询。现如今,数据库技术有着广泛的应用,互联网上绝大多数的信息数据都是保存在数据库中的,text-to-sql技术可以帮助更多非专... 详细信息

text-to-sql是自然语言处理领域的一个重要任务,该任务将自然语言描述或问题转换为对一个具体的数据库的sql查询。现如今,数据库技术有着广泛的应用,互联网上绝大多数的信息数据都是保存在数据库中的,text-to-sql技术可以帮助更多非专业人士快速地从数据库中获取信息。text-to-sql中的一个重要问题是:如何准确快速地解码出sql语句。因此本文针对text-to-sql中的解码问题进行研究,主要研究内容分为三个部分:基于转移系统的sql解码方法,根据sql固有的语法结构,通过构建转移系统的方式,在sql的解码过程中构建部分抽象语法树,限制解码空间,确保生成的sql符合语法规则。同时对现有的转移系统进行改进,缩减文法序列长度,提升解码的时间效率。基于预训练序列到序列语言模型的sql解码方法使用预训练语言模型完成text-to-sql的解码任务,无需增加额外的结构,只需在输入序列中构建相应的特征。同时该模型应用于国家电网调控AI创新大赛中,同时增加了表格检索和数据增强两个额外的部分,以适应该比赛提供的数据集特点,最终在比赛中夺得第二赛道冠军。基于模板检索的非自回归sql解码方法以提升sql的解码速度为目标,利用非自回归模型的时间优势,加速sql解码。同时使用模板检索的方式为非自回归模型增加更多的编码信息,以弥补非自回归模型的不足之处。

关键词： text-to-sql 预训练模型序列到序列生成非自回归模型语义解析

来源：评论

学校读者我要写书评

暂无评论

Bridging the gap between text-to-sql research and real-world applications: A unified all-in-one framework for text-to-sql

引用

KNOWLEDGE-BASED SYSTEMS 2024年 306卷

作者： Han, Mirae Park, Seongsik Kim, Harksoo Kim, Seulgi Konkuk Univ Dept Artificial Intelligence 120 Neungdong Ro Seoul 05029 South Korea Konkuk Univ Dept Comp Sci & Engn 120 Neungdong Ro Seoul 05029 South Korea

Existing text-to-sql research assumes the availability of gold table when generating sql queries. It is possible to effectively generate complex and difficult queries by leveraging information from the gold table. However, in real-world scenarios, determining which of the numerous tables in a database should be referenced is challenging. Therefore, existing models reveal a gap in achieving the core objective of practicality in text-to- sql research. In response, we propose a practical framework that can effectively convert user questions into queries, even in scenarios where reference tables are not provided. By adding a phase to find tables, it can generate queries using only information from questions, mitigating the limitations that arise when restricting reference tables to a single one. We demonstrate that our methods are suitable for practical use in text-to-sql systems by achieving performances comparable to those of existing models with simple structures.

关键词： Semantic parsing text-to-sql Table selection Natural language processing

来源：评论

学校读者我要写书评

暂无评论

CodeS: Towards Building Open-source Language Models for text-to-sql

引用

Proceedings of the ACM on Management of Data 2024年第3期2卷 1-28页

作者： Haoyang Li Jing Zhang Hanbing Liu Ju Fan Xiaokang Zhang Jun Zhu Renjie Wei Hongyan Pan Cuiping Li Hong Chen Renmin University of China Beijing China BEIJING AI-FINANCE TECHNOLOGIES CO. LTD Beijing China

Language models have shown promising performance on the task of translating natural language questions into sql queries (text-to-sql). However, most of the state-of-the-art (SOTA) approaches rely on powerful yet closed-source large language models (LLMs), such as ChatGPT and GPT-4, which may have the limitations of unclear model architectures, data privacy risks, and expensive inference overheads. To address the limitations, we introduce CodeS, a series of pre-trained language models with parameters ranging from 1B to 15B, specifically designed for the text-to-sql task. CodeS is a fully open-source language model, which achieves superior accuracy with much smaller parameter sizes. This paper studies the research challenges in building CodeS. To enhance the sql generation abilities of CodeS, we adopt an incremental pre-training approach using a specifically curated sql-centric corpus. Based on this, we address the challenges of schema linking and rapid domain adaptation through strategic prompt construction and a bi-directional data augmentation technique. We conduct comprehensive evaluations on multiple datasets, including the widely used Spider benchmark, the newly released BIRD benchmark, robustness-diagnostic benchmarks such as Spider-DK, Spider-Syn, Spider-Realistic, and ***, as well as two real-world datasets created for financial and academic applications. The experimental results show that our CodeS achieves new SOTA accuracy and robustness on nearly all challenging text-to-sql benchmarks.

关键词： language model natural language interface for databases text-to-sql

来源：评论

学校读者我要写书评

暂无评论

Few-shot text-to-sql Translation using Structure and Content Prompt Learning

引用

Proceedings of the ACM on Management of Data 2023年第2期1卷 1-28页

作者： Zihui Gu Ju Fan Nan Tang Lei Cao Bowen Jia Sam Madden Xiaoyong Du Renmin University of China Beijing China QCRI & HKUST (GZ) Doha Qatar MIT CSAIL & University of Arizona Boston MA USA MIT CSAIL Boston MA USA

A common problem with adopting text-to-sql translation in database systems is poor generalization. Specifically, when there is limited training data on new datasets, existing few-shot text-to-sql techniques, even with carefully designed textual prompts on pre-trained language models (PLMs), tend to be ineffective. In this paper, we present a divide-and-conquer framework to better support few-shot text-to-sql translation, which divides text-to-sql translation into two stages (or sub-tasks), such that each sub-task is simpler to be tackled. The first stage, called the structure stage, steers a PLM to generate an sql structure (including sql commands such as SELECT, FROM, WHERE and sql operators such as <", ?>") with placeholders for missing identifiers. The second stage, called the content stage, guides a PLM to populate the placeholders in the generated sql structure with concrete values (including sql identifies such as table names, column names, and constant values). We propose a hybrid prompt strategy that combines learnable vectors and fixed vectors (i.e., word embeddings of textual prompts), such that the hybrid prompt can learn contextual information to better guide PLMs for prediction in both stages. In addition, we design keyword constrained decoding to ensure the validity of generated sql structures, and structure guided decoding to guarantee the model to fill correct content. Extensive experiments, by comparing with ten state-of-the-art text-to-sql solutions at the time of writing, show that SC-Prompt significantly outperforms them in the few-shot scenario. In particular, on the widely-adopted Spider dataset, given less than 500 labeled training examples (5% of the official training set), SC-Prompt outperforms the previous SOTA methods by around 5% on accuracy.

关键词： pre-trained language model prompt learning text-to-sql

来源：评论

学校读者我要写书评

暂无评论

Measuring text-to-sql Semantic Parsing Model on the Question Generalizability 22

Measuring Text-to-SQL Semantic Parsing Model on the Question...

引用

Proceedings of the 2022 6th International Conference on Natural Language Processing and Information Retrieval

作者： Thanakrit Julavanich Akiko Aizawa Department of Computer Science Graduate School of Information Science and Technology The University of Tokyo Japan Aizawa Laboratory National Institute of Informatics Japan and Department of Computer Science Graduate School of Information Science and Technology The University of Tokyo Japan

ISBN: (纸本)9781450397629

One of the challenges in NLP tasks, such as text-to-sql semantic parsing, is generalization. In the text-to-sql task, having separate training and testing data can measure one aspect of the generalization: how well the model generalizes to unseen databases. Other aspects, however, remain unaccounted for. We propose a new dataset and a more challenging and thorough evaluation process that focuses on the two challenges of generalizing the text-to-sql model: database content references and question patterns. We create SPIDER-QG, an augmented dataset that employs three techniques, to assess generalizability. First, we replace the set of values in the existing test set with other values from the same column in the same database. Second, we use the synonym of each value as a replacement instead. Third, we generate new questions for the existing sql query by back-translating the original question. Our evaluation setup demonstrates the generalization challenges and struggles of the current models.

关键词： datasets model generalizability text-to-sql

来源：评论

学校读者我要写书评

暂无评论

基于text-to-sql问答系统的设计与实现

基于Text-to-SQL问答系统的设计与实现

引用

作者：宁泽楠西北民族大学

学位级别：硕士

随着人工智能技术的发展,将自然语言转化为结构化查询语言的任务得到了学术界的广泛关注,因为它向非专业用户提供了数据库查询的接口,大大降低了用户查询数据的学习成本,不但可以提高查询效率,而且也能提高人们的生活效率,具有较高的研... 详细信息

随着人工智能技术的发展,将自然语言转化为结构化查询语言的任务得到了学术界的广泛关注,因为它向非专业用户提供了数据库查询的接口,大大降低了用户查询数据的学习成本,不但可以提高查询效率,而且也能提高人们的生活效率,具有较高的研究价值。目前,中文的自然语言转化为结构化查询语言任务面临诸多挑战,第一,在中文数据集的研究方面,其sql语句的预测准确率较低,还存在很大的优化空间;第二,中文的表达中存在众多的同义词,如何准确的做到自然语言问句同数据库列名之间的映射,是该任务面临的主要挑战。本文以CSpider中文数据集为研究对象,该数据集将sql语句分为简单、中等、困难三种预测难度。本文根据中文字和词的不同表达形式,分为基于词的text-to-sql的研究和基于字的text-to-sql的研究。主要工作如下:(1)构建了融合词性和依存句法特征的词序列语义增强模型基于词的研究方面,本文对原始数据采用了新的分词方法,分词效果显示更贴近用户日常生活的语义理解。同时在词向量方面,为了增强模型对语义的理解,在原有词向量中加入了词性特征向量和依存句法特征向量,实验结果表明,加入词性特征后对实验的KEYWORD和AND\OR子任务提升明显,对于中等难度的sql语句来说,提升了15%和7.8%。对困难难度的sql语句来说,提升了12.2%和13.8%。(2)构建了基于BERT的字序列模型基于字的研究方面,本文利用BERT模型作为编码器编码,训练完成后的字的向量同时也包含词的语义特征。除此之外,为了增强自然语言问句与数据库表的列名之间的语义关系,将自然语言问句和数据库列名做拼接组成句对送到BERT多语言模型中来生成向量,实验结果表明在将数据库列名作为BERT模型的输入序列后,对于该任务的各项子任务来说,其准确率、召回率和F1值都上升了5%左右,对于WHERE子任务上升了10%左右。最终对于sql语句的整体准确率,其简单难度、中等难度和困难难度分别提高了3.6%、4.4%、1.2%。(3)搭建了基于text-to-sql的问答系统原型以CSpider数据集作为系统的数据库,搭建基于text-to-sql技术的系统,为用户提供航空公司信息、高校信息、图书信息等多个领域的数据转换接口,同时包含辅助用户提问和相关信息反馈的功能,来提升用户使用体验。

关键词： text-to-sql CSpider 语义增强模型字序列模型 BERT

来源：评论

学校读者我要写书评

暂无评论

Demonstration of a Multi-agent Framework for text to sql Applications with Large Language Models 24

Demonstration of a Multi-agent Framework for Text to SQL App...

引用

33rd ACM International Conference on Information and Knowledge Management (CIKM)

作者： Shen, Chen Wang, Jin Rahman, Sajjadur Kandogan, Eser Megagon Labs Mountain View CA 94041 USA

ISBN: (纸本)9798400704369

The text-to-sql problem aims at developing natural language query interfaces for relational database systems by converting the text input into executable sql queries. Recently, using Large Language Models (LLM) has emerged as a new paradigm for the textto-sql problem. To this end, the LLM needs to understand not only user input but also information from the database. In this demo, we present multi-agent sql (Magesql), an LLM based text-to-sql approach that tackles the task by orchestrating multiple agents in a pipeline. We will showcase a user-friendly interface to demonstrate the inner workings of our approach that allows users to add and modify the agents with different functionalities, customize prompts, and see their impact on specific examples. Through several use cases, we will demonstrate how to (i) construct a text-to-sql pipeline with multiple agents;(ii) generate prompts for LLM with various templates and strategies;and (iii) monitor the results of natural language queries and perform debugging.

关键词： text-to-sql Large Language Model multi-agent system

来源：评论

学校读者我要写书评

暂无评论

SEOSS-Queries - a software engineering dataset for text-to-sql and question answering tasks

引用

DATA IN BRIEF 2022年 42卷 108211页

作者： Tomova, Mihaela Todorova Hofmann, Martin Maeder, Patrick Tech Univ Ilmenau D-98693 Ilmenau Germany Friedrich Schiller Univ Fac Biol Sci D-07745 Jena Germany

Stakeholders of software development projects have various information needs for making rational decisions during their daily work. Satisfying these needs requires substantial knowledge of where and how the relevant information is stored and consumes valuable time that is often not available. Easing the need for this knowledge is an ideal text-to-sql benchmark problem, a field where public datasets are scarce and needed. We propose the SEOSSQueries dataset consisting of natural language utterances and accompanying sql queries extracted from previous studies, software projects, issue tracking tools, and through expert surveys to cover a large variety of information need perspectives. Our dataset consists of 1,162 English utterances translating into 166 sql queries;each query has four precise utterances and three more general ones. Furthermore, the dataset contains 393,086 labeled utterances extracted from issue tracker comments. We provide pre-trained sqlNet and Ratsql baseline models for benchmark comparisons, a replication package facilitating a seamless application, and discuss various other tasks that may be solved and evaluated using the dataset. The whole dataset with paraphrased natural language utterances and sql queries is hosted at ***/s/75ed49ef01ac2f83b3e2. (C) 2022 The Authors. Published by Elsevier Inc.

关键词： Software and systems requirement engineering text-to-sql Dataset Question answering Natural language processing

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：