Automatic Classification of Mathematical Formulas with Natural Language Processing

In e-learning, for predicting a learner's ability, it is necessary to not only analyze learning records but also consider the links between the types of study problems. For analyzing mathematical problems automatically, a formula classifier was developed with Natural Language Processing, which...

Full description

Saved in:
Bibliographic Details
Main Authors: To, Yutaka, Ni, Baorong
Format: Conference Proceeding
Language:English
Subjects:
Analytical models
Data models
Data visualization
e-learning
Electronic learning
formula embeddings
machine learning
Mathematical models
Natural language processing
Semantics
text classification
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In e-learning, for predicting a learner's ability, it is necessary to not only analyze learning records but also consider the links between the types of study problems. For analyzing mathematical problems automatically, a formula classifier was developed with Natural Language Processing, which classifies formulas into mathematical operations including expansion, factorization, etc. It could classify 73% of formulas in test data into appropriate operations and visualize where this model was focused when classifying. From the visualization, this classifier can have the same judgement criteria as humans.
ISSN:2768-4857
DOI:10.1109/ICEAST58324.2023.10157296