HMS: A Hierarchical Solver with Dependency-Enhanced Understanding for Math Word Problem

نویسندگان

چکیده

Automatically solving math word problems is a crucial task for exploring the intelligence levels of machines in general AI domain. It highly challenging since it requires not only natural language understanding but also mathematical expression inference. Existing solutions usually explore sequence-to-sequence models to generate expressions, where are simply encoded sequentially. However, such generally far from enough as similar humans and lead incorrect answers. To this end, paper, we propose novel Hierarchical Math Solver (HMS) make deep exploitation problems. In problem understanding, imitating human reading habits, hierarchical word-clause-problem encoder. Specifically, first split each into several clauses learn semantics local clause level global level. Then, dependency-based module enhance with dependency structure problem. Next, inference, tree-based decoder answer. decoder, apply attention mechanism context different levels, pointer-generator network guide model copy existing information infer extra knowledge. Extensive experimental results on two widely used datasets demonstrate that HMS achieves better answers more reasonable

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Explanation Generation for a Math Word Problem Solver

Background Machine Reading (MR) aims to make the knowledge contained in the text available in forms that machines can use them for automated processing. That is, machines will learn to read from a few examples and they will read to learn what they need in order to answer questions or perform some reasoning task [1]. Since a domain-independent MR system is difficult to build, the Math Word Probl...

متن کامل

A Meaning-based English Math Word Problem Solver with Understanding, Reasoning and Explanation

This paper presents a meaning-based statistical math word problem (MWP) solver with understanding, reasoning and explanation. It comprises a web user interface and pipelined modules for analysing the text, transforming both body and question parts into their logic forms, and then performing inference on them. The associated context of each quantity is represented with proposed role-tags (e.g., ...

متن کامل

A Tag-based English Math Word Problem Solver with Understanding, Reasoning and Explanation

This paper presents a tag-based statistical math word problem solver with understanding, reasoning, and explanation. It analyzes the text and transforms both body and question parts into their tag-based logic forms, and then performs inference on them. The proposed tag-based approach provides the flexibility for annotating an extracted math quantity with its associated syntactic and semantic in...

متن کامل

A Tag-Based Statistical English Math Word Problem Solver with Understanding, Reasoning and Explanation

This demonstration presents a tag-based statistical English math word problem (MWP) solver with understanding, reasoning, and explanation. It analyzes the text and transforms both body and question parts into their tag-based logic forms, and then performs inference on them. The proposed tag (e.g., Agent, Verb, etc.) provides the flexibility for annotating an extracted math quantity with its ass...

متن کامل

A Meaning-based Statistical English Math Word Problem Solver

We introduce MeSys, a meaning-based approach to solving English math word problems (MWPs) via understanding and reasoning in this paper. It first analyzes the text, transforms both body and question parts into their corresponding logic forms, and then performs inference on them. The associated context of each quantity is represented with proposed role-tags (e.g., nsubj, verb, etc.), which provi...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Proceedings of the ... AAAI Conference on Artificial Intelligence

سال: 2021

ISSN: ['2159-5399', '2374-3468']

DOI: https://doi.org/10.1609/aaai.v35i5.16547