Advancing Math Formula Search Using Diverse Structural and Symbolic Representations

Sumedh Vemuganti; Ayu Seiya; Nickvash Kani

doi:10.1007/978-3-031-88708-6_8

Advancing Math Formula Search Using Diverse Structural and Symbolic Representations

Sumedh Vemuganti
, Ayu Seiya
, Nickvash Kani

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We present a comprehensive analysis of search engines from past MathIR competitions (ARQMath, NTCIR) and investigate the semantic representations required for effective formula retrieval. After evaluating these search engines’ methodologies and performance, we uncover a diverse range of equation-parsing and feature-extraction strategies. Inspired by these approaches, we present an ensemble-based retrieval system, integrating symbolic, tree-based, and textual representations of formulas. A key innovation of our work is the Leaf-to-Leaf (L2L) feature extraction method, which captures structural dependencies in formula trees by identifying the shortest paths of operators between pairs of leaf nodes. Our system combines L2L features with symbol-based and LaTeX-based rankings, surpassing the top performing ARQMath-3 search engines in nDCG^′ and mAP^′ metrics, while matching or exceeding the top NTCIR-12 performers in PR^′@{15, 20} and P^′@{5, 10, 15, 20} measures. These results demonstrate the effectiveness of leveraging multiple feature spaces to enhance mathematical formula retrieval.

Original language	English (US)
Title of host publication	Advances in Information Retrieval - 47th European Conference on Information Retrieval, ECIR 2025, Proceedings
Editors	Claudia Hauff, Craig Macdonald, Dietmar Jannach, Gabriella Kazai, Franco Maria Nardini, Fabio Pinelli, Fabrizio Silvestri, Nicola Tonellotto
Publisher	Springer
Pages	116-131
Number of pages	16
ISBN (Print)	9783031887079
DOIs	https://doi.org/10.1007/978-3-031-88708-6_8
State	Published - 2025
Event	47th European Conference on Information Retrieval, ECIR 2025 - Lucca, Italy Duration: Apr 6 2025 → Apr 10 2025

Publication series

Name	Lecture Notes in Computer Science
Volume	15572 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	47th European Conference on Information Retrieval, ECIR 2025
Country/Territory	Italy
City	Lucca
Period	4/6/25 → 4/10/25

Keywords

mathematical information retrieval
query-by-expression
set-based retrieval

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Online availability

10.1007/978-3-031-88708-6_8

Library availability

Discover UIUC Full Text

Cite this

Vemuganti, S., Seiya, A., & Kani, N. (2025). Advancing Math Formula Search Using Diverse Structural and Symbolic Representations. In C. Hauff, C. Macdonald, D. Jannach, G. Kazai, F. M. Nardini, F. Pinelli, F. Silvestri, & N. Tonellotto (Eds.), Advances in Information Retrieval - 47th European Conference on Information Retrieval, ECIR 2025, Proceedings (pp. 116-131). (Lecture Notes in Computer Science; Vol. 15572 LNCS). Springer. https://doi.org/10.1007/978-3-031-88708-6_8

Advancing Math Formula Search Using Diverse Structural and Symbolic Representations. / Vemuganti, Sumedh; Seiya, Ayu; Kani, Nickvash.
Advances in Information Retrieval - 47th European Conference on Information Retrieval, ECIR 2025, Proceedings. ed. / Claudia Hauff; Craig Macdonald; Dietmar Jannach; Gabriella Kazai; Franco Maria Nardini; Fabio Pinelli; Fabrizio Silvestri; Nicola Tonellotto. Springer, 2025. p. 116-131 (Lecture Notes in Computer Science; Vol. 15572 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Vemuganti, S, Seiya, A & Kani, N 2025, Advancing Math Formula Search Using Diverse Structural and Symbolic Representations. in C Hauff, C Macdonald, D Jannach, G Kazai, FM Nardini, F Pinelli, F Silvestri & N Tonellotto (eds), Advances in Information Retrieval - 47th European Conference on Information Retrieval, ECIR 2025, Proceedings. Lecture Notes in Computer Science, vol. 15572 LNCS, Springer, pp. 116-131, 47th European Conference on Information Retrieval, ECIR 2025, Lucca, Italy, 4/6/25. https://doi.org/10.1007/978-3-031-88708-6_8

Vemuganti S, Seiya A, Kani N. Advancing Math Formula Search Using Diverse Structural and Symbolic Representations. In Hauff C, Macdonald C, Jannach D, Kazai G, Nardini FM, Pinelli F, Silvestri F, Tonellotto N, editors, Advances in Information Retrieval - 47th European Conference on Information Retrieval, ECIR 2025, Proceedings. Springer. 2025. p. 116-131. (Lecture Notes in Computer Science). doi: 10.1007/978-3-031-88708-6_8

Vemuganti, Sumedh ; Seiya, Ayu ; Kani, Nickvash. / Advancing Math Formula Search Using Diverse Structural and Symbolic Representations. Advances in Information Retrieval - 47th European Conference on Information Retrieval, ECIR 2025, Proceedings. editor / Claudia Hauff ; Craig Macdonald ; Dietmar Jannach ; Gabriella Kazai ; Franco Maria Nardini ; Fabio Pinelli ; Fabrizio Silvestri ; Nicola Tonellotto. Springer, 2025. pp. 116-131 (Lecture Notes in Computer Science).

@inproceedings{3859a2d82ca84aa6b8feaa2140a6b723,

title = "Advancing Math Formula Search Using Diverse Structural and Symbolic Representations",

abstract = "We present a comprehensive analysis of search engines from past MathIR competitions (ARQMath, NTCIR) and investigate the semantic representations required for effective formula retrieval. After evaluating these search engines{\textquoteright} methodologies and performance, we uncover a diverse range of equation-parsing and feature-extraction strategies. Inspired by these approaches, we present an ensemble-based retrieval system, integrating symbolic, tree-based, and textual representations of formulas. A key innovation of our work is the Leaf-to-Leaf (L2L) feature extraction method, which captures structural dependencies in formula trees by identifying the shortest paths of operators between pairs of leaf nodes. Our system combines L2L features with symbol-based and LaTeX-based rankings, surpassing the top performing ARQMath-3 search engines in nDCG′ and mAP′ metrics, while matching or exceeding the top NTCIR-12 performers in PR′@\{15, 20\} and P′@\{5, 10, 15, 20\} measures. These results demonstrate the effectiveness of leveraging multiple feature spaces to enhance mathematical formula retrieval.",

keywords = "mathematical information retrieval, query-by-expression, set-based retrieval",

author = "Sumedh Vemuganti and Ayu Seiya and Nickvash Kani",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.; 47th European Conference on Information Retrieval, ECIR 2025 ; Conference date: 06-04-2025 Through 10-04-2025",

year = "2025",

doi = "10.1007/978-3-031-88708-6\_8",

language = "English (US)",

isbn = "9783031887079",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "116--131",

editor = "Claudia Hauff and Craig Macdonald and Dietmar Jannach and Gabriella Kazai and Nardini, \{Franco Maria\} and Fabio Pinelli and Fabrizio Silvestri and Nicola Tonellotto",

booktitle = "Advances in Information Retrieval - 47th European Conference on Information Retrieval, ECIR 2025, Proceedings",

address = "Germany",

}

TY - GEN

T1 - Advancing Math Formula Search Using Diverse Structural and Symbolic Representations

AU - Vemuganti, Sumedh

AU - Seiya, Ayu

AU - Kani, Nickvash

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

PY - 2025

Y1 - 2025

N2 - We present a comprehensive analysis of search engines from past MathIR competitions (ARQMath, NTCIR) and investigate the semantic representations required for effective formula retrieval. After evaluating these search engines’ methodologies and performance, we uncover a diverse range of equation-parsing and feature-extraction strategies. Inspired by these approaches, we present an ensemble-based retrieval system, integrating symbolic, tree-based, and textual representations of formulas. A key innovation of our work is the Leaf-to-Leaf (L2L) feature extraction method, which captures structural dependencies in formula trees by identifying the shortest paths of operators between pairs of leaf nodes. Our system combines L2L features with symbol-based and LaTeX-based rankings, surpassing the top performing ARQMath-3 search engines in nDCG′ and mAP′ metrics, while matching or exceeding the top NTCIR-12 performers in PR′@{15, 20} and P′@{5, 10, 15, 20} measures. These results demonstrate the effectiveness of leveraging multiple feature spaces to enhance mathematical formula retrieval.

AB - We present a comprehensive analysis of search engines from past MathIR competitions (ARQMath, NTCIR) and investigate the semantic representations required for effective formula retrieval. After evaluating these search engines’ methodologies and performance, we uncover a diverse range of equation-parsing and feature-extraction strategies. Inspired by these approaches, we present an ensemble-based retrieval system, integrating symbolic, tree-based, and textual representations of formulas. A key innovation of our work is the Leaf-to-Leaf (L2L) feature extraction method, which captures structural dependencies in formula trees by identifying the shortest paths of operators between pairs of leaf nodes. Our system combines L2L features with symbol-based and LaTeX-based rankings, surpassing the top performing ARQMath-3 search engines in nDCG′ and mAP′ metrics, while matching or exceeding the top NTCIR-12 performers in PR′@{15, 20} and P′@{5, 10, 15, 20} measures. These results demonstrate the effectiveness of leveraging multiple feature spaces to enhance mathematical formula retrieval.

KW - mathematical information retrieval

KW - query-by-expression

KW - set-based retrieval

UR - https://www.scopus.com/pages/publications/105003103379

UR - https://www.scopus.com/pages/publications/105003103379#tab=citedBy

U2 - 10.1007/978-3-031-88708-6_8

DO - 10.1007/978-3-031-88708-6_8

M3 - Conference contribution

AN - SCOPUS:105003103379

SN - 9783031887079

T3 - Lecture Notes in Computer Science

SP - 116

EP - 131

BT - Advances in Information Retrieval - 47th European Conference on Information Retrieval, ECIR 2025, Proceedings

A2 - Hauff, Claudia

A2 - Macdonald, Craig

A2 - Jannach, Dietmar

A2 - Kazai, Gabriella

A2 - Nardini, Franco Maria

A2 - Pinelli, Fabio

A2 - Silvestri, Fabrizio

A2 - Tonellotto, Nicola

PB - Springer

T2 - 47th European Conference on Information Retrieval, ECIR 2025

Y2 - 6 April 2025 through 10 April 2025

ER -