摘要
面向科技情报机构对于高效、实时、精准、智能科技情报服务方式和产品的需求,针对专利计量分析服务,开展专利计量情报问答系统的研究和实践。研究对专利计量情报问答系统的概念进行了界定,基于自然语言理解和自然语言生成技术,提出了以问句理解模块和答句生成模块为核心的专利计量情报问答系统实现技术方案。研究设计并实现了人工智能领域专利计量情报问答系统,能够应对同一问题的多种问法给出正确答案,对科技情报机构提升科技情报服务质量和用户体验具有积极意义。
Aiming at the demand of high-efficiency,real-time,accurate and intelligent S&T information service mode and products from S&T intelligence agencies,the research and application of intelligent Q&A system on patent bibliometrics is carried out for the service of patent bibliometrics.The research first defines the intelligence question-answering system on patent bibliometrics. And then based on natural language understanding and natural language generation technology, a technical proposal of the intelligent Q&A system on patent bibliometrics with question understanding module and answer generationmodule as the core is proposed. Finally, we design and implementation the intelligent Q&A system on patent bibliometrics in thefield of artificial intelligence. The system can answer various expression forms of the same question and give the correct answers.It is of positive significance to improve the service quality and user experience of scientific and technological intelligence.
作者
吕璐成
韩涛
王燕鹏
LYU Lucheng;HAN Tao;WANG Yanpeng(Development of Documentation and Information Centre,Chinese Academy of Sciences,Beijing 100190,China;School of economics and management,Chinese Academy of Sciences,Beijing 100190,China)
出处
《情报工程》
2019年第5期46-56,共11页
Technology Intelligence Engineering
基金
中国科学院文献情报新增能力建设专项“情报研究数据中心及服务体系建设”(院1680)
中国科学院青年创新促进会(2016157)
关键词
问答系统
智能情报服务
任务型对话系统
专利计量
Question-answering system
intelligent intelligence service
task-oriented dialogue system
patent bibliometrics
作者简介
吕璐成(1989-),博士研究生,助理研究员,研究方向:专利文本挖掘、智能情报方法与技术,E-mail:lucheng918@126.com;韩涛(1980-),博士,研究员,研究方向:智能情报方法与技术、科学计量与复杂性科学;王燕鹏(1992-),硕士,助理研究员,研究方向:文献计量与文本挖掘。
