The scoring problem for science teachers

Student-developed visualized modeling, in which student must show their cognition processing and understanding of scientific knowledge and concept through visualized presentation, denotes students' ability to express their mental schema of mechanism underlying. Student can present their mental model through multi visualized presentation, such as graphical medaling, flowchart, diagram, and pictures. Visualized modeling creates many meaningful opportunities for teachers to identify their students’ learning obstacles than multiple choice questions. However, the reality of typical timely-fashioned formative assessment restricts the options teachers have for applying student drawing in science teaching and learning

​

​Machine learning scoring is just here

This automatic analysis of visualized modeling web portal aims to use ensembled   machine learning (ML) algorithm models approach to automatically score student-developed visualized modeling  and to provide timely feedback. By uploading their models to the web portal and clicking the run button, users can receive a detailed report within one minute. The report can provide not only the score of students’ models but also specific learning guide. This portal is user-friendly because it requires limited programming knowledge for users. Also, users are not required to tune the parameters of algorithms. This project reduces the scoring burden for teachers and enhances the effectiveness of formative assessment with timely feedback, thus helping teachers’ instructional decision making and providing adaptive science learning guide for students in scientific modeling practice.