Model Answer Generation for Word-Type Questions in Elementary Mathematics

摘要

There are several categories of word-type questions at elementary level Mathematics. These include addition, subtraction, multiplication, division and ratio. Addition and subtraction problems can be further divided based on their textual information. Those types are change type (join-separate type), compare type, and whole-part type. This paper presents a set of ensemble classifiers to automatically generate model answers for these three types of addition and subtraction problems. Currently, questions with one unknown variable are considered. In addition to the existing data sets, a new data set is created for the training and the evaluation purpose. Our results outperform the existing statistical approaches.