大语言模型背景下情报研究的数字化应对策略及实践场景  

Digital Strategies and Practice for Intelligence Research Under the Background of Large Language Model

作　　者：杨倩[1] 林鹤[1] YANG Qian;LIN He(Shanghai Library(Institute of Scientific and Technical Information of Shanghai),Shanghai 200031,China)

机构地区：[1]上海图书馆(上海科学技术情报研究所),上海200031

出　　处：《竞争情报》2023年第3期2-13,共12页Competitive Intelligence

摘　　要：ChatGPT、GPT-4系列模型的出现让人们看到了通用人工智能技术的快速发展,大语言模型将成为一种通用技术,对社会、经济、文化等方面产生重要影响。对情报领域来说,研究大语言模型对制定相关应对策略,推动情报研究在大语言模型背景下更好地发展和服务具有重要意义。在充分分析大语言模型对情报采集、加工、分析和服务4个环节影响的基础上,提出了实现情报工作全方位数据化,研发了基于大语言模型的情报新工具,制定了特定领域大语言模型的数字化应对策略,并就其在科技报告阅读与理解场景、科技查新场景、专业报告研究思路拓展及网络搜索场景中使用相关策略进行了实践。实践结果证明,上述应对策略在情报分析场景的关键环节能够有效辅助情报工作的开展,提高情报工作效率。The emergence of ChatGPT and GPT-4 series models has made people see the dawn of general artificial intelligence,and large language models will become a general-purpose technology,which will have an important impact on society,economy,culture and other aspects.Therefore,it is of great significance to study the impact of large language models on the intelligence field,formulate relevant coping strategies,and promote the better development and service of intelligence research in the context of large language models.Based on the full analysis of the impact of the big prediction model on the four links of intelligence collection,processing,analysis,and service,this paper proposes a digital response strategy for the digitization of the entire process of intelligence,the development of tools based on large language models,and the training of large language models in the field of intelligence.It also uses relevant strategies in practice in the scenarios of reading and comprehension of scientific and technological reports,scientific and technological novelty retrieval,and network search expansion.The practical scenario proves that the above countermeasures can effectively assist the development of intelligence work and improve the efficiency of intelligence work in the key links of intelligence analysis scenarios.

关 键 词：大语言模型 GPT-4 情报研究 科技查新 

分 类 号：G350[文化科学—情报学]