AUTOMATIC UNDERSTANDING AND FORMALIZATION OF NATURAL LANGUAGE GEOMETRY PROBLEMS USING SYNTAX-SEMANTICS MODELS

Automatic understanding of natural language problems is a long-standing challenge research problem in automatic solving. This paper models the understanding of geometry questions as a problem of relation extraction, instead of as the problem of semantic understanding of natural language; further it discovers that the entities and the geometric attribute pattern of elements can play an important role in relation extraction. Based on these ideas this paper proposes a syntax-semantics (S-2) model approach to understand geometry problem, targeting to produce a group of relations to represent the given geometry problem. The formalized geometric relations can then be transformed into the target system-native representations for manipulation to obtain geometric solutions. Experiments conducted on the test problem dataset show that 91.5% of questions can be correctly understood and solved, and the F-1 score in formalizing these problems is substantially high (0.990). The comparisons also demonstrate that the proposed method can achieve good performance against the state-of-the-art method. Integrating the automatic understanding method with different geometry systems will greatly enhance the efficiency and intelligence in automatic solving.