Towards Math-Aware Automated Classification and Similarity Search of Scientific Publications: Methods of Mathematical Content Representations

TL;DR

This paper explores mathematical content representations for automated classification and similarity search in STEM documents, evaluating different methods on arXiv papers to improve performance using machine learning algorithms.

Contribution

It introduces structured math representations that outperform flat TeX tokens and assesses their impact on classification and similarity search tasks.

Findings

Structured math representations improve classification accuracy.

Weighted tokens slightly enhance search performance.

Math content representation influences machine learning effectiveness.

Abstract

In this paper, we investigate mathematical content representations suitable for the automated classification of and the similarity search in STEM documents using standard machine learning algorithms: the Latent Dirichlet Allocation (LDA) and the Latent Semantic Indexing (LSI). The methods are evaluated on a subset of arXiv.org papers with the Mathematics Subject Classification (MSC) as a reference classification and using the standard precision/recall/F1-measure metrics. The results give insight into how different math representations may influence the performance of the classification and similarity search tasks in STEM repositories. Non-surprisingly, machine learning methods are able to grab distributional semantics from textual tokens. A proper selection of weighted tokens representing math may improve the quality of the results slightly. A structured math representation that imitates successful text-processing techniques with math is shown to yield better results than flat TeX tokens.