Research on Solving Text-Diagram Function Problems and Its Application in Tutoring System

The function learning, as a major focus of attention for the mathematics education research community for several decades, requires humanized tutoring. Recently, several intelligent tutoring systems for mathematics learning have been proposed, which are relying on expert design and hand-crafted worked examples. However, there exists a strong need for a humanized intelligent tutoring platform for function learning, because (i) the function problems being tutored should not be limited by the worked examples, and (ii) it must be able to handle step-wise tutoring and instant feedback in response to a learner's answer. In order to achieve this goal, a relationcentric solving algorithm to support the above needs is proposed, allowing students to provide unrestricted input to the ITSFL platform and providing humanized tutoring service. Results from several empirical studies suggest that the proposed platform can be used in function learning effectively.