Automatically solving math word problems,which involves comprehension,cognition,and reasoning,is a crucial issue in ar-tificial intelligence research.Existing math word problem solvers mainly work on word-level relationship extraction and the generation of expression solutions while lacking consideration of the clause-level relationship.To this end,inspired by the theory of two levels of process in comprehension,we propose a novel clause-level relationship-aware math solver(CLRSolver)to mimic the process of human comprehension from lower level to higher level.Specifically,in the lower-level processes,we split problems into clauses according to their natural division and learn their semantics.In the higher-level processes,following human's multi-view understanding of clause-level rela-tionships,we first apply a CNN-based module to learn the dependency relationships between clauses from word relevance in a local view.Then,we propose two novel relationship-aware mechanisms to learn dependency relationships from the clause semantics in a global view.Next,we enhance the representation of clauses based on the learned clause-level dependency relationships.In expression genera-tion,we develop a tree-based decoder to generate the mathematical expression.We conduct extensive experiments on two datasets,where the results demonstrate the superiority of our framework.

Chang-Yang Wu;Xin Lin;Zhen-Ya Huang;Yu Yin;Jia-Yu Liu;Qi Liu;Gang Zhou

Anhui Province Key Laboratory of Big Data Analysis and Application,School of Data Science,University of Science and Technology of China,Hefei 230026,China;Institute of Artificial Intelligence,Hefei Comprehensive National Science Center,Hefei 230088,China;Laboratory of Mathematical Engineering and Advanced Computing,Information Engineering University,Zhengzhou 450001,China

机器智能研究（英文）

[1]Chang-Yang Wu;Xin Lin;Zhen-Ya Huang;Yu Yin;Jia-Yu Liu;Qi Liu;Gang Zhou-.Clause-level Relationship-aware Math Word Problems Solver)[J].机器智能研究（英文）,2022(05):425-438

Clause,tificial,clause,CLRSolver,clauses

Relationship,aware,Math,Word,Problems,Automatically,solving,word,problems,which,involves,comprehension,cognition,reasoning,crucial,issue,intelligence,research,Existing,solvers,mainly,extraction,generation,expression,solutions,while,lacking,consideration,To,this,inspired,by,theory,two,levels,propose,novel,mimic,human,from,lower,higher,Specifically,processes,split,into,according,their,natural,division,semantics,In,following,multi,view,understanding,first,apply,module,dependency,relationships,between,relevance,local,Then,mechanisms,global,Next,enhance,representation,learned,develop,tree,decoder,generate,mathematical,We,conduct,extensive,experiments,datasets,where,results,demonstrate,superiority,our,framework

0.442996