基于 LLM 的异构数据 Q&A 系统和基准测试

Achille Fokoue, Srideepika Jayaraman, Elham Khabiri, Jeffrey O. Kephart, Yingjie Li, Dhruv Shah, Youssef Drissi, Fenno F. Heath III, Anu Bhamidipaty, Fateh A. Tipu, Robert J. Baseman
{"title":"基于 LLM 的异构数据 Q&A 系统和基准测试","authors":"Achille Fokoue, Srideepika Jayaraman, Elham Khabiri, Jeffrey O. Kephart, Yingjie Li, Dhruv Shah, Youssef Drissi, Fenno F. Heath III, Anu Bhamidipaty, Fateh A. Tipu, Robert J. Baseman","doi":"arxiv-2409.05735","DOIUrl":null,"url":null,"abstract":"In many industrial settings, users wish to ask questions whose answers may be\nfound in structured data sources such as a spreadsheets, databases, APIs, or\ncombinations thereof. Often, the user doesn't know how to identify or access\nthe right data source. This problem is compounded even further if multiple (and\npotentially siloed) data sources must be assembled to derive the answer.\nRecently, various Text-to-SQL applications that leverage Large Language Models\n(LLMs) have addressed some of these problems by enabling users to ask questions\nin natural language. However, these applications remain impractical in\nrealistic industrial settings because they fail to cope with the data source\nheterogeneity that typifies such environments. In this paper, we address\nheterogeneity by introducing the siwarex platform, which enables seamless\nnatural language access to both databases and APIs. To demonstrate the\neffectiveness of siwarex, we extend the popular Spider dataset and benchmark by\nreplacing some of its tables by data retrieval APIs. We find that siwarex does\na good job of coping with data source heterogeneity. Our modified Spider\nbenchmark will soon be available to the research community","PeriodicalId":501123,"journal":{"name":"arXiv - CS - Databases","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A System and Benchmark for LLM-based Q\\\\&A on Heterogeneous Data\",\"authors\":\"Achille Fokoue, Srideepika Jayaraman, Elham Khabiri, Jeffrey O. Kephart, Yingjie Li, Dhruv Shah, Youssef Drissi, Fenno F. Heath III, Anu Bhamidipaty, Fateh A. Tipu, Robert J. Baseman\",\"doi\":\"arxiv-2409.05735\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"In many industrial settings, users wish to ask questions whose answers may be\\nfound in structured data sources such as a spreadsheets, databases, APIs, or\\ncombinations thereof. Often, the user doesn't know how to identify or access\\nthe right data source. This problem is compounded even further if multiple (and\\npotentially siloed) data sources must be assembled to derive the answer.\\nRecently, various Text-to-SQL applications that leverage Large Language Models\\n(LLMs) have addressed some of these problems by enabling users to ask questions\\nin natural language. However, these applications remain impractical in\\nrealistic industrial settings because they fail to cope with the data source\\nheterogeneity that typifies such environments. In this paper, we address\\nheterogeneity by introducing the siwarex platform, which enables seamless\\nnatural language access to both databases and APIs. To demonstrate the\\neffectiveness of siwarex, we extend the popular Spider dataset and benchmark by\\nreplacing some of its tables by data retrieval APIs. We find that siwarex does\\na good job of coping with data source heterogeneity. Our modified Spider\\nbenchmark will soon be available to the research community\",\"PeriodicalId\":501123,\"journal\":{\"name\":\"arXiv - CS - Databases\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Databases\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.05735\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Databases","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.05735","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

在许多工业环境中,用户希望提出问题,而问题的答案可以在结构化数据源中找到,如电子表格、数据库、API 或它们的组合。通常情况下,用户不知道如何识别或访问正确的数据源。最近,各种利用大型语言模型(LLM)的文本到 SQL 应用程序通过让用户使用自然语言提问,解决了其中的一些问题。然而,这些应用在现实的工业环境中仍然不切实际,因为它们无法应对数据源的异构性,而这种异构性正是此类环境的典型特征。在本文中,我们通过引入 siwarex 平台来解决异构问题,该平台可实现对数据库和应用程序接口的无缝自然语言访问。为了证明 siwarex 的有效性,我们扩展了流行的 Spider 数据集和基准,用数据检索 API 代替了其中的一些表格。我们发现 siwarex 能够很好地应对数据源的异构性。我们修改后的 Spider 基准将很快提供给研究界。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
A System and Benchmark for LLM-based Q\&A on Heterogeneous Data
In many industrial settings, users wish to ask questions whose answers may be found in structured data sources such as a spreadsheets, databases, APIs, or combinations thereof. Often, the user doesn't know how to identify or access the right data source. This problem is compounded even further if multiple (and potentially siloed) data sources must be assembled to derive the answer. Recently, various Text-to-SQL applications that leverage Large Language Models (LLMs) have addressed some of these problems by enabling users to ask questions in natural language. However, these applications remain impractical in realistic industrial settings because they fail to cope with the data source heterogeneity that typifies such environments. In this paper, we address heterogeneity by introducing the siwarex platform, which enables seamless natural language access to both databases and APIs. To demonstrate the effectiveness of siwarex, we extend the popular Spider dataset and benchmark by replacing some of its tables by data retrieval APIs. We find that siwarex does a good job of coping with data source heterogeneity. Our modified Spider benchmark will soon be available to the research community
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
Development of Data Evaluation Benchmark for Data Wrangling Recommendation System Messy Code Makes Managing ML Pipelines Difficult? Just Let LLMs Rewrite the Code! Fast and Adaptive Bulk Loading of Multidimensional Points Matrix Profile for Anomaly Detection on Multidimensional Time Series Extending predictive process monitoring for collaborative processes
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1