2. Platform function: Strengthen the application of new technologies such as artificial intelligence
In terms of platform function, all platforms stepped up efforts to optimise the function design of the use of new technologies like artificial intelligence (AI), machine learning, and virtual reality/augmented reality (VR/AR) to improve the learning experience. For example, courses are chosen according to recommendations and the learning path is intelligently planned according to a learner’s learning objectives and experience in via a machine learning algorithm. The learners’ learning needs are identified through an AI chat or virtual learning buddies, increasing the sense of interaction as well as the intelligence of the system. An immersive learning environment is developed to help learners better understand complex concepts and complete practical projects.
Early in 2018, XuetangX released the smart learning assistant “Xiao Mu.” At the 2019 International Conference on Artificial Intelligence and Education, “Xiao Mu” appeared as a key case in Vice Education Minister Zhong Denghua’s speech, Intelligent Education Leading the Future: China's Understanding and Action. As an intelligent learning assistant, “Xiao Mu” can perform functions including reminding learners to make learning plans when choosing courses, giving reminders at different learning stages, and monitoring the learning process. Furthermore, “Xiao Mu” can interact with learners in the form of Q&A and even chat.
Coursera has also tried to decrease the drop-off rate of its courses using a machine learning algorithm. Through analysing a large number of learning logs, Coursera found that most learners drop out at the beginning of the courses and at the stage of grading work. They fall into three typical groups, including non-starters, dabblers, and explorers. With the help of a machine learning algorithm, Coursera is now able to identify learners who are at the edge of dropping off and deduce how and when to intervene. At present, these interventions are mainly achieved through information pushes. In Coursera's experiment, pushing individual pieces of information has increased the existing unit completing rate by 10% (Lockemer, 2019).
Another of Coursera’s prominent function improvements in 2019 is the release of Coursera Labs (Maggioncalda, 2019). Through Coursera Labs, learners are able to complete practical exercises or hands-on projects without unnecessary downloads, software installation, or configuration. This function can be widely used in computer programming and development courses. In order to further enhance the function of Coursera Labs, Coursera purchased the startup Rhyme Softworks, whose team is mainly engaged in the development of interactive projects based on browsers and cloud technology. Rhyme Softworks’ products allow users direct access to virtual environments using browsers, and developers can complete various kinds of tools and even nest relevant teaching videos and instructions needed by the projects in a built-in virtual environment.
In addition, Coursera has also further enhanced its notes function (Sun, 2019) to facilitate learning. Learners can now quickly add screenshots, mark them, and add their own thoughts during the course of watching teaching videos. They can directly highlight the lecture content provided by the teachers using text selection, which is then automatically recorded by the background system. An individual “note” button has been added to courses so learners can quickly view and retrieve all their notes.
3. Advance MOOC practice with investigations and studies
Analysis and research on innovative teaching methods and learning behaviours are also regarded as effective ways to improve the instruction quality and completion rate of MOOCs. Academic conferences such as Learning with MOOC5, Learning at Scale6, and European MOOCs Stakeholders Summit7 (EMOOCs) have been held once a year around the world. Coursera and edX also hold annual global partner meetings, when partners are invited to share their MOOC experiences and the results of academic research. The aforementioned meetings have also attracted the attention of a large number of MOOCs practitioners and researchers. The research from the EMOOCs sub-forum “Self-Regulated Learning and MOOCs” showed that more than half of learners are not fully prepared for learning via MOOCs. It also showed that learners with strong self-regulation abilities are more able to work according to their own initiative, more likely to repeatedly watch course materials, and tend to organise their learning process in a more flexible way. Therefore, learners can consciously improve their learning using a strategy of self-regulated learning (Winter, 2019). The platform and course development teams can also give better trainings and instructions on MOOC learning and include strategies like goal setting and learning planning in their teaching designs.
Most major MOOC platforms also have their own analytical research teams. These teams focus on improving the functions of the platform and offering references for the development of course products more suitable for the needs of customers through user investigation and learning data analysis. For example, edX conducted an investigation of 1,000 learners aged over 18 in 2019 in which more than one third of the informants indicated that they lack the skills needed for their job positions. Data skills, business skills, and other project management, leadership, and soft skills are the areas with the most shortages (Medros, 2019). With the goal of better understanding their learners, FutureLearn’s research team launched a questionnaire survey of 7,000 learners in 2018. The learners were divided into seven different types in three major categories, and the prototypes of each sub-type were described. The details are shown in Table 1 (Pickard, 2019b).