AUTOMATIC MARKING OF NL TO FOL CONVERSIONS

摘要

In this paper, we present a web-based tool that automatically marks student answers in exercises related to converting Natural Language (NL) sentences into First Order Logic (FOL) formulas. It aims at helping the tutor in assessing student's performance in translating NL into FOL. It operates in two stages. In the first stage, the system checks student's answer (FOL sentence) in order to find and to recognize the errors made by using an error detection mechanism. In the second stage, it automatically marks the student's answer based on the types of the errors made. Each error type has a different contribution (weight) in the final mark. The final mark is composed of partial marks concerning different elements of a FOL sentence (atoms, connectives, quantifiers). Experimental results show that the automatic marking tool is quite successful.