Statistical graph classification in intelligent mathematics problem solving system for high school student

Abstract

In recent years, intelligent mathematics problem solving has aroused the interest of researchers. In the intelligent mathematics problem solving system related to high school, the classification of statistical graph is a key step. Consequently, the classification of statistical graphs has become an urgent problem to be solved. In this paper, a new method is proposed for statistical graphs classification. Firstly, the image features of statistical graphs are obtained by spatial pyramid matching using sparse coding (ScSPM). The extracted features are then fed into classifier: support vector machine (SVM). In this paper, a new statistical graph dataset was established to evaluate the proposed method. It contains 400 statistical graphs including line graphs, histograms, scatter plots, and pie charts. Experimental results on the established dataset demonstrate that the proposed statistical graphs classification method achieves better performance.