Automatic solver for mathematical word problems

Mathematical word problems (MWP) test critical aspects of reading comprehension in conjunction with generating a solution that agrees with the "story" in the problem. In this thesis we design and construct an MWP solver in a systematic manner, as a step toward enabling comprehension in mathematics. We do this by (a) identifying the discourse structure of MWPs that will enable comprehension in mathematics, and (b) utilizing the information in the discourse structure toward generating the solution in a systematic manner. We build a multistage software prototype that predicts the problem type, identifies the function of sentences in each problem, and extracts the necessary information from the question to generate the corresponding mathematical equation. Our prototype has an accuracy of 86% on a large corpus of MWPs of three problem types from the elementary grade mathematics curriculum