The effectiveness of our “digital Diamond Mandala Matrix (dDMM)” application on a real-time online class is shown based on the results of analysis of output by students. The dDMM application which is something like a mind map limiting words to nine plus a description field for explaining the whole of the dDMM data. We can easily view several dDMMs at a glance for comparing them with each other, so it is highly suitable for an online class where only one screen can be shared. In order to improve the effect on students’ learning, the class was offered in an active learning style, so that it consists of pre-learning, a real-time online class with a lecture, and discussions based on the pre-learning using Zoom Meeting, and students’ re-thinking based on the online class as post-learning. Students input data into the dDMM application as pre-learning dDMMs before participating in the online class, and then join the class consisting of a lecture and discussions based on the pre-learning dDMMs. They input data into the dDMM application as post-learning dDMMs after the online class. In order to investigate the efficacy of the dDMM application on real-time online class, we compared them statistically using the text data analysing tool KH Coder pre-learning and post-learning dDMMs, which are outputs by students. From the results, we found that post-learning dDMMs are improved from post-learning dDMMs due to the lecture and discussions in the online class. Therefore, the dDMM application is an effective tool in real-time online courses. We will improve the application by adding functions such as automatic grouping for supporting collaborative learning, with multilingual support.