统计与数据科学知识图谱构建与创新人才培养  

Building Knowledge Graphs for Talents Cultivating in Statistics and Data Science

作　　者：尹建鑫[1] 王晓军[1] YIN Jianxin;WANG Xiaojun

机构地区：[1]中国人民大学应用统计科学研究中心/统计学院,北京100872

出　　处：《中国人民大学教育学刊》2023年第2期69-79,共11页Renmin University of China Education Journal

摘　　要：统计学同时具有基础、应用与交叉三种学科属性。随着大数据、人工智能的兴起,“统计学”与“数据科学”成了近义词或同义词,在很多场合常常并称。由于数据形态、体量、质量发生丰富多样的变化,传统的数据获取、数据分析、统计推断、展示交流都需要适应大数据进行变革,从而要求统计学人才培养做出相应变革。随着研究前沿的不断推进,一方面,统计学理论将更加深入地融合概率论、最优化算法和复杂度理论;另一方面,统计学应用将更加深入地融合机器学习方法、大数据处理分析技术以及领域知识,并与经管、社科等各类学科进行交叉融合。在人才培养中,知识爆炸和教学容量有限的矛盾、知识广度和深度的矛盾以及学科领地认同与学科交叉边界逐渐模糊的矛盾等,对统计与数据科学创新人才培养提出了诸多挑战。面对挑战,应如何构建知识体系,使得教和学都能按图索骥?更重要的是,如何构建能力体系,使得学生拥有学习能力、自觉寻找路径、发现资源,实现自我完善、持续进步成长?本文在整理大量教材、前沿文献等资料的基础上,使用自然语言处理技术、大数据文本挖掘技术,结合专家判断,构建了统计学与数据科学知识图谱。在知识图谱的指引下,教学中很多问题,包括课程内容、先后修顺序、课程间衔接等问题都可以得到很好的重构,同时也为自主学习提供了方便。本文提出的构建知识体系、能力体系的方法模式同样适用于其他学科在面对大量新增周边领域知识的情况下更新构建知识体系与能力体系。特别适合拥抱数据科学新时代,在新工科、新文科、新医科、新农科建设中,为创新复合型人才培养提供参考。Statistics,as a subject,is both basic,applied,and interdisciplinary.And in the era of big data and AI,statistics has a synonym as“data science”.They are usually mentioned at the same time.As the form,volume,and property of the data are changing rapidly,traditional methods in data acquisition,analysis,statistical inference,and presentation have been changed accordingly.Hence the statistical education must adapt to this change.As the academic frontier updates rapidly,the statistical theory incorporates more and more theory from Probability,Optimization,and Complexity Theory.On the other hand,in application area,statistics intersects with machine learning algorithms,big data analytics,and domain knowledge,especially with economics and management science.In the process of talents training,the conflict between knowledge explosion and limited education time,the conflict between exploration and exploitation,as well as the conflict between subject self-awareness and blurring of the boundaries advocates many challenges.How to build the knowledge system?How to construct a knowledge graph both for the teaching and learning?More importantly,how to build the ability system for learners?If a student acquires the ability of self-learning,and know how to search for resources,then he can accomplish himself and keep lifelong learning.In this paper,based on a large corpus of textbooks and research papers,and utilizing the NLP technique and knowledge graph models,and combining with the expert knowledge,we build the knowledge graph for statistics and data science.Under the guidance of knowledge graph,the coverage of each course,the prerequisite relationship among courses can be re-constructed.At the same time,knowledge graph can also provide guidance for self-learning.The general method of learning knowledge graphs from domain educational corpus can be applied to other areas for the knowledge and ability system building.It is especially useful for training creative talents in new industrial,new liberal arts,new medicine,and new a

关 键 词：统计学 数据科学 知识体系 能力体系 

分 类 号：G642.3[文化科学—高等教育学]