Arithmetic Word Problem Knowledge Acquisition and Representation for Automated Problem Solving

Yuhui Ma, Xiaojing Shang, and Kai Tan

Keywords

Knowledge acquisition, Knowledge representation, Knowledge engineering, Arithmetic word problem

Abstract

Many researches are focused on automated theorem proving and automated solving geometry problems, and some systems in this field have been developed successfully to help the students solve the problems. Arithmetic word problem is the major part of mathematical education at elementary schools, but it is very difficult for students to solve. However, most systems that help the students solve word problems can do nothing when the problem is dynamically submitted by students. In order to automatically solve the problems submitted by students in ITS, the knowledge of solving word problems must be acquired and represented. In this paper, we specify the knowledge of solving arithmetic word problems based on the researches of cognitive psychology and mathematical education. The knowledge of automated arithmetic word problem for ITS is classified into three types: linguistic knowledge, mathematical knowledge and strategy knowledge. The representation of it and the method of acquirement are described, too. An experiment is performed in order to verify the knowledge effectively. The experimental result has shown that the knowledge is self-contained and the method of acquirement and the representation are effective.

Important Links:



Go Back