Abstract
Constraint-Based student Modeling (CBM) is an important technique employed in intelligent tutoring systems to model student knowledge to provide relevant assistance. This paper introduces the Math Story Problem Tutor (MAST), a Web-based intelligent tutoring system for probability story problems, which is able to generate problems of different contexts, types and difficulty levels for self-paced learning. Constraints in MAST are specified at a low-level of granularity to allow fine-grained diagnosis of the student error. Furthermore, MAST extends CBM to address errors due to misunderstanding of the narrative story. It can locate and highlight keywords that may have been overlooked or misunderstood leading to an error. This is achieved by utilizing the roles and syntaxes of the problem segments and the semantic descriptions of the keywords that are defined through the Natural Language Generation (NLG) methods deployed in the story problem generation. MAST also integrates CBM with scaffolding questions and feedback to provide various forms of help and guidance to the student. This allows the student to discover and correct any errors in his/her solution. The tutoring effect of MAST has been evaluated empirically using different tests (paired samples t-test, normalized knowledge gain, Mann-Whitney U test, and power learning curves). The results demonstrated a positive effect of MAST on improving the exam scores, and the normalized knowledge gain. Additionally, keyword highlighting of errors integrated in MAST has been shown to speed up learning and improve reduction of average percentage of violated constraints along the learning curve. This suggests that using various forms of assistance can help speedup learning, and that well-designed tutoring systems providing relevant help can be superior to traditional approaches.