大语言模型结合微信的智慧实验室探索  

Large language model combined with WeChat s smart laboratory exploration

作　　者：陈金英 张荣生 叶阿勇[2] CHEN Jinying;ZHANG Rongsheng;YE Ayong(College of Life Science,Fujian Normal University,Fuzhou 350117,China;College of Computer and Cyber Security,Fujian Normal University,Fuzhou 350117,China)

机构地区：[1]福建师范大学生命科学学院,福建福州350117 [2]福建师范大学计算机与网络空间安全学院,福建福州350117

出　　处：《实验室科学》2024年第3期125-130,134,共7页Laboratory Science

摘　　要：随着实验室物联化等需求发展,更为便捷的实验室智慧化也迫在眉睫。将大语言模型引入实验室管理,对人文感知、理解、分析、判断等能力提升具有重要作用。首先,分析了大语言模型的核心能力:语言到指令转换、多任务与自动化、知识响应、多模融合及信息过滤与汇总。其次,以微信公众号为基础,构建基于大语言模型的智慧化实验室方案,实现自然语言理解与执行、个性化学习计划制定、辅导、日志分析与反馈管理和实验报告自动评估等功能。最后,通过实验验证方案的有效性。With the rapid development of laboratory IoT and interconnection needs,more convenient laboratory intelligence is imminent.Introducing large language models into laboratory management plays an important role in improving humanistic perception,memory,understanding,analysis,judgment,sublimation and other abilities.First,the five core capabilities of large language models are analyzed:language-to-instruction conversion,multi-step tasks and automation,intelligent knowledge response,multi-modal fusion,and information filtering and aggregation.Secondly,using the WeChat public account as the basic platform,we build a smart laboratory solution based on large language models to realize the understanding and execution of users natural language,personalized learning plan formulation,experimental coaching,log analysis and user feedback management and experimental reports Automated assessment and other management functions.Finally,through experimental verification,the large language model can effectively optimize the experimental reservation and management process,improve the efficiency of experimental resource utilization,and manage user feedback and needs.

关 键 词：大语言模型 微信平台 智慧实验室 自然语言处理 聊天机器人技术 

分 类 号：G482[文化科学—教育学]