网络化知识生产和演化智能分析模型构建及应用研究
周炫余1,2 陈 丽1* 郑勤华1 郭玉娟3 张思敏2
（1.北京师范大学 远程教育研究中心，北京 100875；2.湖南师范大学 基础教育大数据研究与应用重点实验室，湖南 长沙 410006；3.北京师范大学 系统科学学院，北京 100875）
【摘 要】互联网环境中出现了群智汇聚知识生产和知识动态演化现象，如何利用智能技术高效、精准地挖掘大规模网络数据中知识生产和演化的新规律，是“互联网+教育”研究领域面临的重要科学问题。基于此，论文以回归论知识观为哲学理论基础，以互联网知识发展框架为实践指导，采用人工智能技术构建了网络化知识生产和演化智能分析模型，并以国内cMOOC课程为应用案例分析网络化知识的演化规律。该模型首先对cMOOC课程的数据进行预处理，再基于prompt和TextCNN算法表征知识实体与属性，最后以周为单位进行时间切片探究知识实体演化规律。该智能分析模型得出的结论与现有学者的研究成果高度契合，有力证明了模型的准确性及有效性。研究不仅从数据的角度验证了回归论知识观的观点，也形成了一套基于智能技术分析群智涌现引起知识生产与演化的方法，为未来分析智能时代的大规模知识生产与演化规律提供了借鉴思路。
【关键词】知识生产；知识演化；智能分析；网络化知识；cMOOC
Construction and Application of Intelligent Analysis Model for Networked Knowledge Production and Evolution
ZHOU Xuanyu1,2, CHEN Li1*, ZHENG Qinhua1, GUO Yujuan3 and ZHANG Simin2
(1. Research Center of Distance Education, Beijing Normal University, Beijing 100875, China; 2. Key Laboratory of Big Data Research and Application for Basic Education, Hunan Normal University, Changsha 410006, China; 3. School of Systems Science, Beijing Normal University, Beijing 100875, China)
Abstract: The emergence of collective intelligence-driven knowledge production and dynamic knowledge evolution phenomena in internet environments poses a crucial scientific challenge for “Internet + Education” research: how to efficiently and accurately uncover novel patterns of knowledge production and evolution within large-scale network data using intelligent technologies. Based on this, this article employs the regressive theory of knowledge as its philosophical theoretical foundation and the internet knowledge development framework as practical guidance, constructing an intelligent analysis model for networked knowledge production and evolution through artificial intelligence technologies. Using domestic cMOOC courses as case studies, it analyzes the evolutionary patterns of networked knowledge. The model first preprocesses cMOOC course data, then represents knowledge entities and attributes using prompt and engineering and Text CNN algorithms, and finally investigates knowledge entity evolution through weekly time slicing. The conclusions derived from this intelligent analysis model demonstrate strong alignment with existing scholar research outcomes, strongly validating the model’s accuracy and effectiveness. The study not only empirically verifies the perspectives of regressive theory of knowledge from a data-driven approach, but also establishes a set of methodology for analyzing collective intelligence-emergent knowledge production and evolution through intelligent technologies, providing references for future investigations into large-scale knowledge production andevolution patterns in the intelligent era.
Keyword: knowledge production; knowledge evolution; intelligent analysis; networked knowledge; cMOOC
