An Enhanced Relation-flow Algorithm for Solving Number Line Problems

Number line problems (NBLP) constitute an indispensable part of primary mathematics. It remains challenging for existing solving algorithms to solve NBLP due to the lack of specialized algorithms for understanding NBLP text and diagram. This paper presents an enhanced relation-flow algorithm for solving NBLP, utilizing customized syntax-semantics models and a diagram procedure to ensure corresponding problem text and diagram understanding performance. For NBLP text understanding, it uses a syntax-semantics method for extracting relations from the NBLP text by developing a group syntax-semantics model for NBLP text. For NBLP diagram understanding, it proposes a novel diagram procedure for acquiring relations from the NBLP diagram. Thus, a relation group is obtained from the problem text and diagram. Finally, a symbolic reasoner is proposed to infer the relation group and obtain the answer. Experimental results demonstrate the efficacy of the proposed method in solving the number line problem. To the best of our knowledge, this is the first algorithm capable of solving NBLP, making a significant contribution to expanding the scope of solvable problems.