Our challenge was to teach a very complex and technical knowledge to middle school students (grade 7 to 9). We need to set up a system such that it is easy for middle school students to follow and have fun learning it, but still understand the technical details.We choose a program called RapidMiner as our tool for teaching. The user interface allows students to explore machine learning, with no prior computer science knowledge or coding skills. We choose mangos as our target. Mango is a local and popular fruit in Thailand. When it is unripe, it is sour, green, and crispy. When it is ripe, it is sweet, yellow, and chewy. Its physical appearance also changes, from hard and green, to soft and yellow. There are many features of mango to observe, which is crucial for building the dataset.