As big data and artificial intelligence become increasingly prevalent in the field of education, applications and systems such as learning analytics, adaptive testing, and intelligent learning platforms have penetrated the core aspects of elementary, secondary, and tertiary education. Powerful data-driven technologies are increasingly influencing educational policy and governance, integrating into everyday teaching practices and management, and changing classroom and extracurricular activities. Likewise, data on educational users is collected and controlled to a much greater extent than in other industries, and the trends of platformization and datafication in education warrant the attention of policy-makers, administrators, researchers, and educators.
It is undeniable that the vigorous research and development of intelligent technology and educational data and its effective application are the basis for promoting educational reform and developing future education, but we cannot focus only on the intelligent technology promoted by investors, media and technology companies. The potential and opportunities of innovation should pay more attention to and think about the problems and challenges that intelligent education innovation may bring. Technology companies assist and participate in educational technology innovation practices in primary and secondary schools and colleges and universities. During the process, educational platform owners can collect and analyze a large amount of behavioral data, while educational managers, teachers and students are users of data-driven educational platforms. Not really understanding how data is collected, processed and interpreted will inevitably lead to issues of privacy protection and inequality in the long run. At the same time, educational products and services produced only for commercial purposes and market orientation will inevitably violate the objective laws of education and teaching. For example, tech companies such as Google, Microsoft, Facebook (renamed “Metaverse” in 2021) have gained the unprecedented ability to grasp, predict and guide people’s behavior by extracting and using personal data on a large scale. Funding and technology go directly into education to develop core applications for classroom and out-of-school learning. Especially in the past three years, the global COVID-19 epidemic has led to the closure of campuses and the prevalence of large-scale remote online teaching, which has also turned the “platform in the classroom” into a “classroom on the platform”.
In this case, the data-driven platform will undoubtedly change the process and mode of education management and decision-making. First, tech companies reap huge commercial benefits from platform data, and the value of data formed by students, teachers, and administrators using the platform seems to far outweigh its own benefits; second, tech companies can predict, influence, and Control students’ behavior to maximize business interests, rather than focusing on students’ interest in learning and long-term development.
With the rapid development of artificial intelligence technology, the subject of learning is no longer limited to humans, and machines are also learning and influencing human behavior and decision-making in various forms. In intelligent education applications, on the one hand, students learn the required knowledge and skills through the platform (usually in a “personalized” and autonomous way); on the other hand, the platform builds models and algorithms through supervised or unsupervised learning and then analyzes student data. These data are mainly derived from the interaction between students and the platform and are mainly analyzed based on behaviorist education theory. Skinner, a behaviorist psychologist, established the operational behaviorist learning theory in the 1950s, proposed the program teaching theory teaching model, and designed a one-to-one “teaching machine” on this basis. . Skinner believes that human learning is a reinforcement process of operational responses. The key to successful teaching or training is to accurately analyze the reinforcement effect and design techniques that can manipulate this process to establish a specific reinforcement system. It has also become the basis for the development of modern educational technology.
In the era of data-driven intelligence, behaviorist teaching theory represented by intelligent learning machines is returning to classrooms and families in the name of personalized learning. This is contrary to the “student-centered” concept of future educational development generally advocated by the international education community, that is, it opposes the behaviorist teaching method that regards students as passive receptors, and advocates the construction of students’ initiative and initiative. ideology, and teaching modes and methods based on social learning.
Unlike other fields, educational teaching scenarios and models are very complex. At present, many artificial intelligence-based learning systems, such as intelligent question banks and question-making software, take the behaviorist teaching model as the concept, and build models and algorithms based on historical data. Such an intelligent system is very suitable for the traditional single evaluation standard, that is, the teaching goal is to improve student performance, but it is difficult to develop educational services and products that can change the existing educational model and meet future educational needs. In such intelligent systems, the power of teachers to provide more learning opportunities for students may be replaced by automatic feedback systems, and it is possible to scale behaviorist teaching methods, making technology a hindrance in educational teaching innovation. In an education system that emphasizes student development, we redefine and understand education, learning, and intelligence. Intelligent education aiming at meeting the needs of future social talents requires educational researchers, teachers, and computer technology developers to carry out interdisciplinary research and practice, go beyond traditional teaching models and narrow evaluation standards, and give full play to intelligent technology in future education. potential and role.
We need to cultivate students’ critical thinking about data-driven technologies and conduct relevant empirical research. We believe that students’ learning data is very rich and diverse, and data-driven technology needs to highlight the participation of students and teachers in the data construction process, so as to better integrate technological progress with the development of teaching practice, and promote intelligent technology To better serve to teach innovation and explore new educational theories and models.
