跳到主要导航 跳到搜索 跳到主要内容

Enhancing Seq2seq Math Word Problem Solver with Entity Information and Math Knowledge

  • East China Normal University
  • Zhejiang University
  • Alibaba Group Holding Ltd.

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

Devising an automatic Math Word Problem (MWP) solver has emerged as an important task in recent years. Various applications such as online education and intelligent assistants are expecting better MWP solvers to process complex user queries that involve numerical reasoning. Current seq2seq MWP solvers encounter two critical challenges: ordinal indices without semantics and insufficient training data. In this work, we propose Entity Random Indexing to equip indices with semantics and design diverse representations of math expressions to augment training data. Experimental results show that our approach effectively enhances the seq2seq MWP solver, which outperforms strong baselines.

源语言英语
主期刊名Web Information Systems Engineering – WISE 2022 - 23rd International Conference, Proceedings
编辑Richard Chbeir, Helen Huang, Fabrizio Silvestri, Yannis Manolopoulos, Yanchun Zhang, Yanchun Zhang
出版商Springer Science and Business Media Deutschland GmbH
370-385
页数16
ISBN(印刷版)9783031208904
DOI
出版状态已出版 - 2022
活动23rd International Conference on Web Information Systems Engineering, WISE 2021 - Biarritz, 法国
期限: 1 11月 20223 11月 2022

出版系列

姓名Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
13724 LNCS
ISSN(印刷版)0302-9743
ISSN(电子版)1611-3349

会议

会议23rd International Conference on Web Information Systems Engineering, WISE 2021
国家/地区法国
Biarritz
时期1/11/223/11/22

指纹

探究 'Enhancing Seq2seq Math Word Problem Solver with Entity Information and Math Knowledge' 的科研主题。它们共同构成独一无二的指纹。

引用此