Design and implementation of semantic matching based automatic scoring system for C programming language

Abstract

With the development of education and computer, it has become an urgent demand for scoring programming questions automatically. This paper puts forward a semantic matching based automatic scoring method for C programming questions. The said method standardizes student programs and template programs, and then calculates their semantic similarity, so as to score student programs. Main idea: Firstly, convert student programs and template programs into an intermediate representation of programs (i.e. system dependence graph); secondly, carry out semantic equivalence conversion for the generated system dependence graph according to a series of standardization rules, so as to eliminate the diversification of program expression; thirdly, calculate the matching degree of the standardized system dependence graph, and score student programs according to the matching result and scoring rule. © 2010 Springer-Verlag Berlin Heidelberg.