大语言模型在标准信息服务中的应用实践与思考  

Application Practices and Reflections of Large Language Models in Standard Information Services

作　　者：吴永亮 李小龙 高俭 李宗鹤 刘炜 WU Yongliang;LI Xiaolong;GAO Jian;LI Zonghe;LIU Wei(China Astronautics Standards Institute,Beijing,100071,China)

机构地区：[1]中国航天标准化研究所,北京100071

出　　处：《航天标准化》2024年第4期15-18,42,共5页Aerospace Standardization

摘　　要：随着我国数字化、智能化高速发展,传统标准信息服务方法在面对复杂多变应用场景时显得力不从心,具备高质量标准数据和推理嵌入能力的标准信息服务需求日益凸显。随着深度学习和自然语言处理技术的突破,大语言模型技术为解决这一难题提供了新思路。探讨大语言模型在标准信息服务中的应用潜力,通过分析其核心技术和应用场景,揭示其如何提高服务效率与准确性。大语言模型能有效解析复杂的标准文本,支持自动化标准数据治理、知识敏捷动态更新、信息个性化推荐以及知识推理应用。同时,提出利用大语言模型实现增强检索的基本技术路径,并展望了未来的发展方向。With the rapid development of digitalization and intelligence in our country,traditional methods of standard information services seem powerless to cope with the complex and ever-changing application scenarios.The demand for standard information services that possess high-quality standard data and inference embedding capabilities is becoming increasingly prominent.Breakthroughs in deep learning and natural language processing technologies have provided new ideas for solving this problem through the large model technology.This article aims to explore the application potential of large language models in standard information services by analyzing their core technologies and application scenarios,revealing how they can improve the service efficiency and accuracy.Large language models can effectively parse complex standard texts,support automated standard data governance,agile dynamic updates of knowledge,personalized information recommendations,and knowledge reasoning applications.This article proposes a basic technical path for achieving enhanced retrieval using large language models and looks forward to future development directions.

关 键 词：大语言模型 大模型 标准信息服务 自然语言处理 人工智能 可信问答 

分 类 号：G63[文化科学—教育学]