Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Empowering Bengali Education with AI: Solving Bengali Math Word Problems through Transformer Models

Jalisha Jashim Era, Bidyarthi Paul, Tahmid Sattar Aothoi, Mirazur Rahman Zim, Faisal Muhammad Shah

TL;DR

This paper tackles Bengali math word problem solving by transforming Bengali descriptive text into mathematical equations using transformer models. It introduces the PatiGonit dataset of 10,000 Bengali MWPs and evaluates multiple transformer architectures (mT5, BanglaT5, mBART50, and a Basic Transformer) via fine-tuning and an equation solver to obtain final answers. The results show mT5 achieving the highest accuracy of 97.30%, with mBART50 and BanglaT5 close behind, underscoring the effectiveness of multilingual and Bengali-specific transformers in low-resource settings. The work provides a valuable resource and methodology for educational AI in Bengali, while also highlighting challenges in cross-linguistic adaptation and the need for richer multi-step problems.

Abstract

Mathematical word problems (MWPs) involve the task of converting textual descriptions into mathematical equations. This poses a significant challenge in natural language processing, particularly for low-resource languages such as Bengali. This paper addresses this challenge by developing an innovative approach to solving Bengali MWPs using transformer-based models, including Basic Transformer, mT5, BanglaT5, and mBART50. To support this effort, the "PatiGonit" dataset was introduced, containing 10,000 Bengali math problems, and these models were fine-tuned to translate the word problems into equations accurately. The evaluation revealed that the mT5 model achieved the highest accuracy of 97.30%, demonstrating the effectiveness of transformer models in this domain. This research marks a significant step forward in Bengali natural language processing, offering valuable methodologies and resources for educational AI tools. By improving math education, it also supports the development of advanced problem-solving skills for Bengali-speaking students.

Empowering Bengali Education with AI: Solving Bengali Math Word Problems through Transformer Models

TL;DR

This paper tackles Bengali math word problem solving by transforming Bengali descriptive text into mathematical equations using transformer models. It introduces the PatiGonit dataset of 10,000 Bengali MWPs and evaluates multiple transformer architectures (mT5, BanglaT5, mBART50, and a Basic Transformer) via fine-tuning and an equation solver to obtain final answers. The results show mT5 achieving the highest accuracy of 97.30%, with mBART50 and BanglaT5 close behind, underscoring the effectiveness of multilingual and Bengali-specific transformers in low-resource settings. The work provides a valuable resource and methodology for educational AI in Bengali, while also highlighting challenges in cross-linguistic adaptation and the need for richer multi-step problems.

Abstract

Mathematical word problems (MWPs) involve the task of converting textual descriptions into mathematical equations. This poses a significant challenge in natural language processing, particularly for low-resource languages such as Bengali. This paper addresses this challenge by developing an innovative approach to solving Bengali MWPs using transformer-based models, including Basic Transformer, mT5, BanglaT5, and mBART50. To support this effort, the "PatiGonit" dataset was introduced, containing 10,000 Bengali math problems, and these models were fine-tuned to translate the word problems into equations accurately. The evaluation revealed that the mT5 model achieved the highest accuracy of 97.30%, demonstrating the effectiveness of transformer models in this domain. This research marks a significant step forward in Bengali natural language processing, offering valuable methodologies and resources for educational AI tools. By improving math education, it also supports the development of advanced problem-solving skills for Bengali-speaking students.
Paper Structure (19 sections, 2 equations, 7 figures, 4 tables)

This paper contains 19 sections, 2 equations, 7 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Dataset
  4. Dataset Collection
  5. Dataset Annotation
  6. Dataset Pre-processing
  7. Train-Test-Validation Splits
  8. Methodology
  9. Basic Transformer
  10. mT5
  11. BanglaT5
  12. mBART-50
  13. Experiments
  14. Baseline setup
  15. Evaluation Metrics
  16. ...and 4 more sections

Figures (7)

  • Figure 1: Bengali Math Word Problem Equation generation and solver example
  • Figure 2: Sample of a simple equation
  • Figure 3: Sample of a complex equation
  • Figure 4: Some challenges of the Bengali Word Problems
  • Figure 5: Some challenges in solving the Bengali Math Word Problems
  • ...and 2 more figures