arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.14941cs.CL
收藏

Text2Arch:一个用于从自然语言描述生成科学架构图的数据库

Text2Arch: A Dataset for Generating Scientific Architecture Diagrams from Natural Language Descriptions

IIT Roorkee, India(印度拉尔·克里希纳理工学院);Google, India(印度谷歌);Microsoft, India(印度微软)

Shivank Garg, Sankalp Mittal, Manish Gupta

AI总结:

本文提出Text2Arch数据库,通过语言模型生成高保真的架构图,解决了文本生成架构图的效率和歧义问题,并展示了其在多个应用中的优势。

摘要:

仅通过文本沟通复杂系统设计或科学过程效率低下且容易产生歧义。一个能够从文本中自动高语义保真生成科学架构图的系统在多个应用中都有用,如企业架构可视化、AI驱动的软件设计和教育内容创作。因此,本文专注于利用语言模型进行输入文本描述的语义理解,以生成可处理生成高保真架构图的中间代码。不幸的是,没有干净的大型开放访问数据集存在,这意味着缺乏任何有效的开放模型用于此任务。因此,我们贡献了一个全面的数据集,\system,包含科学架构图像、其对应文本描述以及相关的DOT代码表示。利用这一资源,我们微调了一套小型语言模型,并且使用GPT-4o进行上下文学习。通过广泛的实验,我们显示\system模型在现有基线模型如DiagramAgent上显著优于,并与基于上下文学习的GPT-4o生成表现相当。我们公开了代码、数据和模型。

英文摘要:

Communicating complex system designs or scientific processes through text alone is inefficient and prone to ambiguity. A system that automatically generates scientific architecture diagrams from text with high semantic fidelity can be useful in multiple applications like enterprise architecture visualization, AI-driven software design, and educational content creation. Hence, in this paper, we focus on leveraging language models to perform semantic understanding of the input text description to generate intermediate code that can be processed to generate high-fidelity architecture diagrams. Unfortunately, no clean large-scale open-access dataset exists, implying lack of any effective open models for this task. Hence, we contribute a comprehensive dataset, \system, comprising scientific architecture images, their corresponding textual descriptions, and associated DOT code representations. Leveraging this resource, we fine-tune a suite of small language models, and also perform in-context learning using GPT-4o. Through extensive experimentation, we show that \system{} models significantly outperform existing baseline models like DiagramAgent and perform at par with in-context learning-based generations from GPT-4o. We make the code, data and models publicly available.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑