Human-Machine Collaborated Data Wisdom Mechanisms： The Alchemy of Data Value for Smarter Education
Data value is increasingly emphasized in education， like other industries. However，the density of this value is so low that it cannot serve education purpose directly， let alone serve the smarter education. Data wisdom refers to a class of models for representing structural and/or functional relationships between data， information， knowledge ，and wisdom. The four levels of data wisdom provide a way for solving the above problems. But，the effective methods to make data evolve into wisdom through information and knowledge have not been found. Thus， we developed a kind of human-machine collaborated data wisdom mechanisms. It describes how to achieve the evolution of data into wisdom by using the understanding of both human and machine. Human understanding includes four layers ： know by practicing， know by sensing， know by constructing and know by critiquing. Machine understanding also includes four layers ： know by perceiving， know by describing， know by mining and know by learning. With the above-mentioned assumption， data wisdom mechanisms can be divided into three parts ： the mechanism of the relational organization of data， the mechanism of pattern recognition and interpretation of information and the mechanism of principle derivation of knowledge. The first part is for data evolution into information which has four steps ： a) introduce monitoring goals of teaching and students behavior，b) determine the relationship of monitored data， c) organize data based on determined relationship， d) and represent the meaning of organized data by using visual charts and dashboards. The second part is for information evolution into knowledge which also has four steps ： a ) extract the features of objects from organized data， b ) inquiry meaning from visual charts and dashboards， c ) mine information patterns from extracted objects’ features via inquired meaning，d) and explain and evaluate the mined information patterns. The third part is for knowledge evolution into wisdom which is four steps ： a) inquiry the reasons for the behavior and performance of students inspired by explained and evaluated information patterns， b ) evaluate the value of student behavior and performance based on inquired reasons， c) learn the service criteria from b) and the service decisions from d) by machines， and d) generate learning service decisions based on the insight obtained from b). We hope that the proposed data wisdom mechanisms can provide a feasible scheme of extracting value from data for the smarter education， and then it can further help to optimize the teaching and learning behaviors in smarter education.
|關鍵詞||数据智慧、智慧教育、数据价值、人机协同、精准决策、data wisdom、smarter education、data value、human-machine collaboration、precise decision mak-ing、CSSCI|