Definition Extraction from Generic and Mathematical Domains with Deep Ensemble Learning

Natalia Vanetik,Marina Litvak

doi:10.3390/math9192502

Abstract

Definitions are extremely important for efficient learning of new materials. In particular, mathematical definitions are necessary for understanding mathematics-related areas. Automated extraction of definitions could be very useful for automated indexing educational materials, building taxonomies of relevant concepts, and more. For definitions that are contained within a single sentence, this problem can be viewed as a binary classification of sentences into definitions and non-definitions. In this paper, we focus on automatic detection of one-sentence definitions in mathematical and general texts. We experiment with different classification models arranged in an ensemble and applied to a sentence representation containing syntactic and semantic information, to classify sentences. Our ensemble model is applied to the data adjusted with oversampling. Our experiments demonstrate the superiority of our approach over state-of-the-art methods in both general and mathematical domains.

Highlights

Deep Ensemble Learning.Definitions play a very important role in scientific and educational literature because they define the major concepts that are operated inside the text
Despite mathematical and generic definitions being pretty similar in their linguistic style, supervised identification of mathematical definitions benefits from a training on a mathematical domain, as we previously showed in [1]
The results demonstrated the superiority of models with a Convolutional Neural Network (CNN) layer, which can be explained by the ability of CNN to learn features and reduce the number of free parameters in a high-dimensional sentence representation, allowing the network to be more accurate with fewer parameters

Summary

Introduction

Definitions play a very important role in scientific and educational literature because they define the major concepts that are operated inside the text. Despite mathematical and generic definitions being pretty similar in their linguistic style (see the example of two definitions below: the first, defining ASCII, is general, while the second defines mathematical object), supervised identification of mathematical definitions benefits from a training on a mathematical domain, as we previously showed in [1]. Academic Editors: Cornelia Caragea and Florentina Hristea. American Standard Code for Information Interchange, called ASCII, is a character encoding based on English alphabet.

Methods

Results

Discussion

Conclusion