The application status and development trend of education big data

II) The meaning of educational big data

The meaning of educational big data needs to be analyzed from both data and technology levels. Among the big data definitions that are cited more, Wikipedia definitions and McKinsey definitions both emphasize the amount of big data that cannot be handled with common data tools; while the Gartner definition focuses on The characteristics and value of the data. In order to analyze the true meaning of educational big data, it is necessary to analyze the composition and characteristics of educational big data.

In terms of the composition of educational big data, online learning data is the first to bear the brunt. It can be said that the education of big data is widely concerned and has an inseparable relationship with the prevalence of online teaching and learning. In Schonberg's book, "With Big Data Peers - The Future of Learning and Education", the first big data education application case comes from online learning. With the increasing popularity of online teaching, all kinds of massive data recorded by the learning management system and various mobile devices have become an important source of analysis teaching process in the process of teaching and learning. These data include behavioral data for recording the learning process, evaluation data for recording learning results, and social network relationship data formed by learning. Expanded from these data, Education Big Data also contains various types of student personal information data, teaching management data and so on. It can be seen that the big data of education comes from the main body and process of education and teaching.

According to the different levels of the main body and the content of education and teaching activities, educational big data can be divided into four levels and six types. The four levels include individuals, schools, regions, and countries; six types include basic data, instructional data, research data, management data, service data, and public opinion data. The basic data includes basic information data of learners represented by demography; the teaching data includes the process, content and result data involved in the teaching process; the research data includes data obtained from various educational and experimental experiments and research projects; The data includes data recorded in various educational management systems, such as student's student status data, archival data and various statistical data, etc. The service data includes data recorded in various service systems related to education and teaching, for example, Various types of life and service services for teachers and students, book archives services, etc.; public opinion data includes education-related data in various public media, such as various educational news data, education-related data in social network systems such as Weibo.

From a feature perspective, the characteristics of big data are often summarized as 4V, including massive volume, Velocity, Variety, and Value. The characteristics of educational big data are different from and different from 4V: First, from the scale, the volume of educational big data has not reached the scale of retail and telecommunications, but it has exceeded the processing power of traditional data tools. Secondly, from the perspective of flow speed, education big data circulation speed is relatively slow, and it does not have the characteristics of fast flow of transaction data, search data or communication data. Correspondingly, the periodicity of education and teaching determines the typical periodicity of educational big data. Furthermore, from the perspective of data composition, unstructured data in educational big data, especially audio and video data, accounts for a large proportion. These data come from classroom videos, teaching resources, etc., which are different from the data recorded in traditional databases and have certain analytical complexity. At the same time, unlike the trading activities in the fields of e-commerce and other fields with clear steps, clear results and short cycle, education and teaching activities have higher process complexity. It is more difficult to find the law through education big data analysis. It can be seen that the characteristics of educational big data can be summarized as strong periodicity, high complexity and great value.

In summary, we can define educational big data as: service education subject and education process, high complexity data with strong periodicity and great educational value.

III) The application of educational big data

For the application of educational big data, researchers have put forward their own thinking from different angles. From the perspective of research paradigm, Professor Zhu Zhiting proposed the enlightenment of big data on educational technology research methods, and emphasized the adaptive learning supported by data. Zheng Yanlin and Liu Haimin believe that the application of educational big data is mainly to support education evaluation and education and teaching decision-making. Hu Yucheng and Wang Zulin summarized the application of big data as a scientific, complete, comprehensive and dynamic quality monitoring system that promotes the effectiveness of teaching through evaluation and prediction, and promotes educational decision-making based on changing educational forms and complex relationships. Yang Xianmin and other researchers believe that the application of educational big data can be divided into policy science, regional education balance, school education quality improvement, curriculum system and teaching effect optimization, individual individual development and other aspects.

The impact of big data on education is comprehensive. It can change the learning situation of individual learners, the depth of understanding of the laws of education, the way in which educational policies are formulated, and the structure of the entire education system. From the perspective of demand, the application of educational big data can be summarized into five levels, namely, learning, teaching, research, management and policy. Learning and teaching needs focus on adaptive learning; research needs focus on discovering the rules of education and teaching; management needs focus on fine management and scientific decision-making; policy requirements are derived from the design basis of the acquisition mechanism.

Education big data applications form a variety of products and services for different levels of demand. From adaptive teaching to dynamic tracking and evaluation, from management model building to data sharing portals, various types of applications outline the overall picture of big data impacting education.

We may wish to analyze the international application of educational big data with great influence from the three aspects of adaptive teaching, educational law discovery and precise management support through typical technologies, products and services in the international scope, with a view to the education big data of China. Development applications provide a reference.

1) Adaptive teaching support

Adaptive teaching and learning is the optimal state of teaching. The content, methods and processes in adaptive teaching can be customized according to the learner's situation, so that each learner is likely to get the most suitable development for himself. The realization of adaptive teaching needs to be based on a comprehensive analysis of the learner's individual characteristics and learning conditions. Big data provides the possibility to track and integrate these data and personalize support for students.

The most common adaptive teaching system comes from the field of online learning. The learner's learning process is fully documented in a variety of learning management systems and online learning platforms. The record of the learning process, combined with demographic data such as demographics and learning styles, can clearly characterize the learner's learning path and learner characteristics, and conduct diagnosis and recommendation based on effective recording of the learning process and comprehensive evaluation of the learning situation. Targeted teaching.

Adaptive learning support has almost become the "standard" for online learning, with varying degrees of adaptability in every commercial online learning platform. Content recommendation is a major form of adaptability. However, a truly effective adaptive teaching system needs to integrate three systems, namely knowledge system, behavior system and feature system. The knowledge system is used to describe the knowledge system; the behavioral system is used to record the learning, practice and feedback processes; the feature system is used to analyze the individual characteristics and learning traits of the students.

Currently, the most representative adaptive learning system is Knewton and Kehan Academy. This type of adaptive learning system focuses on student learning. The learning system attempts to play the role of a teacher, automating the recording, diagnosis and intervention of student learning. Another aspect of adaptive learning is support for teacher teaching. It should be noted that the current learning system still has many limitations. In the early stage of the development of MOOCs, whether teachers were replaced by online classrooms has become an eye-catching topic. However, as a key role in the educational process, teachers cannot disappear in a short period of time, but will use technology to achieve professional level improvement and role change. Big data will become a powerful assistant to teachers' teaching, helping teachers to better play their roles and better promote students' learning.

With the support of big data technology, teachers can monitor students' learning according to their own needs, and evaluate students automatically or semi-automatically through their own set standards. With the support of data, teachers can combine their own teaching experience to diagnose and intervene students. Under the training of teachers, big data tools will provide stronger support for teachers' teaching. Big data tools will be the best helper for teachers, not competitors.

Taking Masteryconnect as an example, Masteryconnect provides comprehensive data support for teachers' teaching. It provides data, analysis, presentation and data-based collaboration support from the daily work of teachers. Teachers can collect a variety of teaching data, including classroom observation data, answer sheet data, gauge data, and online test data. After collecting data, Masteryconnect can perform automated analysis and visualization. The results of the analysis can be shared with other teachers through the system, and teachers can communicate and collaborate on the basis of data. Masteryconnect provides comprehensive support for various formative evaluations. Teachers can build their own teaching content structure and design questionnaires, exercises, test papers and other assessment methods for each module and knowledge point. Assessments can be published to PCs and mobile devices, and students can choose the way they like to complete the assessment. At the same time, the results of the assessment can form a customized report and send it to parents.

Support for teaching and learning is two aspects of the application of big data in adaptive learning. Based on the data, the student's learning status is fully recorded, the learning system can push customized content, and the teacher can carry out more targeted teaching. It can be seen that data is changing the teaching process both online and offline.