一种基于知识树的科技前沿探测方法——以深度学习领域为例  

A Detection Method for Science and Technology Frontiers Based Knowledge Tree

作　　者：曾文[1] 闫甜甜 刘晓琳 张蕾 Zeng Wen

机构地区：[1]中国科学技术信息研究所,北京100038

出　　处：《情报理论与实践》2024年第3期158-162,共5页Information Studies:Theory & Application

基　　金：国家自然科学基金面上项目“基于开源情报的科技前沿多维度探测方法及模型研究”的成果,项目编号:72074201。

摘　　要：[目的/意义]当前全球科技竞争愈加剧烈,准确把握未来科技前沿是科技竞争取得胜利的关键,科技前沿探测是科技情报研究的热点问题之一。[方法/过程]文章提出一种基于知识树的科技前沿探测方法,通过科技前沿知识树反映领域科技前沿的发展信息,探索解决科技前沿探测过程中存在的信息冗余、探测结果不准确等问题的方法和途径,并以深度学习领域为例进行了实证研究。[结果/结论]研究结果表明,基于知识树的科技前沿探测方法可以在一定程度上提高科技前沿探测过程的效率,改善科技前沿探测结果的质量,可以为面向特定领域的科技前沿研究提供参考。[Purpose/significance] Currently,global technological competition is becoming increasingly fierce.Accurately grasping the future science and technology frontiers is the key to winning technological competition,and exploring the technological frontier is one of the hot topics in scientific and technological intelligence research.[Method/process] This artical proposes a technology frontier detection method based on a knowledge tree,which reflects the development information of the field's science and technology frontier,and explores the methods and paths to solve the problems such as information redundancy and inaccurate detection results in the process of frontier detection.Then,the empirical research in the field of deep learning is carried.[Result/conclusion] To a certain extent,the method proposes in this article can improve the efficiency of the detection process and the quality of detection results,and provides reference for research in science and technology fields.

关 键 词：知识树 科技前沿 科技情报 深度学习 

分 类 号：G350[文化科学—情报学]